ANNA FENKO - TU Delft

224
SENSORY DOMINANCE IN PRODUCT EXPERIENCE ANNA FENKO

Transcript of ANNA FENKO - TU Delft

Page 1: ANNA FENKO - TU Delft

             

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

    

ANNA FENKO         

Page 2: ANNA FENKO - TU Delft

  

Page 3: ANNA FENKO - TU Delft

 

    

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

PROEFSCHRIFT 

 

ter verkrijging van de graad van doctor 

aan de Technische Universiteit Delft, 

op gezag van de Rector Magnificus prof. ir. K.C.A.M.Luyben, 

voorzitter van het College voor Promoties, 

in het openbaar te verdedigen op donderdag 2 december 2010 om 12:30 uur 

 

 

DOOR ANNA FENKO Candidate of psychological sciences 

 Moscow State University, Russia 

geboren te Moskou, Rusland   

   

Page 4: ANNA FENKO - TU Delft

Dit proefschrift is goedgekeurd door de promotor: 

Prof. dr. P.P.M. Hekkert  

 

COPROMOTOR: 

Dr. ir. H.N.J. Schifferstein 

 

SAMENSTELLING PROMOTIECOMMISSIE: 

Rector Magnificus,     voorzitter  

Prof. dr. P.P.M. Hekkert,     Technische Universiteit Delft, promotor 

Dr. ir. H.N.J. Schifferstein,    Technische Universiteit Delft, copromotor 

Prof. dr. H. de Ridder,     Technische Universiteit Delft 

Prof. dr. G.J. Steen,     Vrije Universiteit Amsterdam  

Prof. dr. C. Spence,     University of Oxford 

Prof. dr. G.B. Dijksterhuis,    University of Copenhagen 

Dr. A. Heylighen,      Katholieke Universiteit Leuven 

Prof. dr. D.V. Keyson,     Technische Universiteit Delft, reservelid 

 

 

 

 

 

This research was partly supported by MAGW OC grant 400‐03‐131 of the Nether‐lands Organization for Scientific Research (NWO) awarded to H.N.J. Schifferstein. 

 

© ANNA FENKO 

[email protected] 

Page 5: ANNA FENKO - TU Delft

          

CONTENTS  CHAPTER 1. INTRODUCTION ...................................................................... 7 

1. What is multisensory integration ............................................. 8 2. What is product experience.................................................... 20 3. Overview of the thesis ............................................................ 29 

CHAPTER 2. SHIFTS IN SENSORY DOMINANCE BETWEEN VARIOUS STAGES  OF USER‐PRODUCT INTERACTIONS .......................................................... 31 

1. Introduction ............................................................................ 31 2. Method ................................................................................... 34 3. Results .................................................................................... 37 4. Discussion  .............................................................................. 44 4. Conclusions  ............................................................................ 47 

CHAPTER 3. THE ROLE OF SENSORY MODALITIES IN THE DYNAMICS  OF NATURALNESS OF A DEHYDRATED FOOD PRODUCT ........................... 49 

1. Introduction ............................................................................ 49 2. Methods.................................................................................. 57 3. Results..................................................................................... 60 4. Discussion ............................................................................... 72 5. Conclusion............................................................................... 76 

CHAPTER 4. DESCRIBING PRODUCT EXPERIENCE IN DIFFERENT LANGUAGES:  THE ROLE OF SENSORY MODALITIES  ....................................................... 79 

1. Introduction ............................................................................ 79 2. Study 1. Selecting a sensory neutral experience  ................... 89 3. Study 2. Modality importance for sensory descriptions  ........ 94 4. Discussion ............................................................................. 100 5. Conclusion............................................................................. 104 

Page 6: ANNA FENKO - TU Delft

 

CHAPTER 5. LOOKING HOT OR FEELING HOT:  WHAT DETERMINES THE PRODUCT EXPERIENCE OF WARMTH? ........... 107 

1. Introduction .......................................................................... 107 2. Pre‐study .............................................................................. 113 3. The main study ..................................................................... 116 4. Interview study  .................................................................... 121 5. Discussion  ............................................................................ 123 6. Conclusion............................................................................. 125 

CHAPTER 6. WHAT MAKES PRODUCTS FRESH:  THE SMELL OR THE COLOR? .................................................................. 127 

1. Introduction .......................................................................... 127 2. Pre‐study............................................................................... 133 3. Main study ............................................................................ 137 4. Discussion ............................................................................. 142 5. Conclusions ........................................................................... 145 

CHAPTER 7. NOISY PRODUCTS: DOES APPEARANCE MATTER?............ 147 1. Introduction .......................................................................... 147 2. Pre‐study............................................................................... 151 3. Main study  ........................................................................... 156 4. Discussion ............................................................................. 160 

CHAPTER 8. DISCUSSION ....................................................................... 165 

1. General findings  ................................................................... 165 2. Theoretical problems  ........................................................... 169 3. Perspectives for future research .......................................... 176 4. Implications for designers .................................................... 180 

SUMMARY ............................................................................................... 181 

SAMENVATTING ...................................................................................... 187 

REFERENCES............................................................................................ 193 ACKNOWLEDGEMENTS ........................................................................... 219 

ABOUT THE AUTHOR .............................................................................. 221 LIST OF PUBLICATIONS ........................................................................... 223 

 

 

 

Page 7: ANNA FENKO - TU Delft

 

 

 

 

 

 

 

 

 

CHAPTER 1 

INTRODUCTION  

When  I  started writing  this  thesis, my  friends gave me an appropriate present: a 

fountain pen. Of course,  it was never used to write the thesis, but  I enjoy writing 

with it on special occasions. Adjusting my writing habits to this elegant instrument 

was not easy. In the beginning the pen felt too wide and heavy to hold, and when I 

pushed down too hard, I scraped the paper and created little balls of fiber and ink 

on  the end of  the nib. Pulling  them off  the pen made a mess  to my  fingers, and 

when  I  tried  to  refill  the pen,  I  smeared my blouse with  ink. But after  I  stopped 

treating my fountain pen as though it were a ballpoint pen, I started to enjoy writ‐

ing with it. Rather than pressing hard, I simply glide the pen over the paper allow‐

ing the ink to flow freely through the nib to the paper. The result is a more elegant 

and smooth writing stroke that does not tire my hand as quickly as writing with a 

ballpoint pen. I like the way my handwriting looks with the fountain pen; it became 

more elegant and flowing. I enjoy a smooth feel and a soft sound of the gold nib on 

the paper. The wide grip of  the pen  feels  comfortable now and actually  reduces 

hand strain. I started to enjoy handwriting rather than dread it like I used to, when I 

was writing with a ballpoint pen. 

People  perceive  the world  around  them with  their  five  senses:  vision,  audition, 

touch, smell, and taste.  Information  from different modalities  is  integrated  in the 

brain to create a stable and meaningful experience of objects and events. Writing 

with a fountain pen is an example of a multisensory experience. It involves looking 

Page 8: ANNA FENKO - TU Delft

INTRODUCTION 

 

at  the pen and  the  text you are writing, a comfortable  feeling of  the pen  in your 

hand, a soft sound of the nib touching the paper, and a faint but distinctive smell of 

ink. Most  consumer products  evoke  complex multisensory  experiences  that may 

contribute to enjoyment or disappointment of product use.  

In product design, various sensory properties of products (such as color, shape, tex‐

ture,  smell,  sound, etc.)  can be manipulated  to evoke  certain product experiences 

(such as luxury, novelty, grace, and so on). Some sensory modalities may play a more 

important role in product experience than others. Hence, the aim of the present re‐

search  is  to  understand which  sensory modalities  are more  important  for  certain 

product experiences, how  information from different senses  is  integrated, and how 

product experience can be  influenced by various combinations of sensory stimuli  in 

products. We will start with clarifying two general questions: (1) What is multisensory 

integration? and (2) What is product experience?   

 

1. WHAT IS MULTISENSORY INTEGRATION? In  cognitive psychology,  the perception  through different  sensory modalities has been  traditionally  investigated  separately. The main  reason  for  this  separation  is the striking anatomical differences of various perceptual organs and corresponding sensory  receptors, which process different kinds of  information and use different physical principals. As Hermann Helmholtz, one of the early researchers of percep‐tion, pointed out,  

“The distinctions among sensations which belong to different modalities, such as the differences among blue, warm, sweet, and high‐pitched, are so fundamental as to exclude any possible transition from one modality to an‐other and any relationship of greater or less similarity. For example, one cannot ask whether sweet is more like red or more like blue... Comparisons are possible only within each modality; we can cross over from blue through violet and carmine to scarlet, for example, and we can say that yellow is more like orange than like blue” (Helmholtz, 1878). 

Nevertheless, from the very beginning of scientific investigation of sensory percep‐tion, many  findings pointed to the  interrelation of the senses.  In  fact, contrary to Helmholtz’s claim, sweet can be red or blue, at least for a specific group of people called synaesthets. Synaesthesia is a rare familial condition involving a “crossing” of the senses — for example, ordinary activities such as reading or listening to music 

Page 9: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

may be perceived with different colors or tastes (Simner & Ward, 2006).  It ranges in vividness and can occur between different senses. For an adult,  it  is unusual to experience  a  sensation  in one modality  triggered by  a  sensation  in  another. But 

some  researchers  (e.g. Maurer, 1993; Baron‐Cohen, 1996) suggest  that all babies might be synaesthetic at some point in their lives, probably up to about 4 months of age. Early  in  infancy all babies experience sensory  input  in an undifferentiated way. The normal development of perceptual system includes the differentiation of sensory modalities. Adult synaesthesia, according to this view, represents a failure to fully develop the perceptual system. 

More  common everyday phenomena also demonstrate  interactions between dif‐ferent senses. For example, in a cinema we perceive a speaking person every time we hear words and see an actor moving his or her  lips, while  in reality the sound comes from the  loudspeakers behind our back. This phenomenon  is known as the ventriloquism effect (Bertelson, 1999). A ventriloquist synchronizes the movements of a puppet's mouth with his own  speech while avoiding movements of his own head or lips. As a result, the audience perceives a speaking puppet. This effect re‐veals audiovisual integration in the process of speech recognition.   

Multisensory integration is the process by which the brain synthesizes information from different senses to enhance sensitivity to external events (Wallace, 2004). The proper integration of information from the different sensory modalities helps us to perceive the world in an accurate and meaningful way. For example, reaction times are speeded in the presence of congruent multisensory cues when compared with responses to cues presented from a single modality (Hershenson, 1962; Hughes et al., 1994; Forster et al., 2002). Similarly, rapid eye movements are initiated faster in the  presence  of  stimuli  from more  than  a  single  sensory modality  (Frens  et  al., 1995; Harrington & Peck, 1998; Colonius & Arndt, 2001; Amlot et al., 2003).  

Many psychological experiments show the power of multisensory  integration. For ex‐ample, the parchment skin  illusion  illustrates auditory‐tactile  interactions (Jousmäki & Hari, 1998). In this illusion, changing the frequency of a sound coincident with rubbing a textured surface can vary the tactile perception from smooth to rough. Furthermore, the rubber hand illusion (Botvinick & Cohen, 1998) demonstrates a three‐way interac‐tion between vision, touch and proprioception. Watching a rubber hand being stroked synchronously with one’s own unseen hand causes the rubber hand to be attributed to one’s own body, to “feel like it’s my hand.”  

Looking at the crossmodal interactions between hearing and touch, Zampini, Guest, and  Spence  (2003)  asked  their  participants  to  evaluate  the  pleasantness  and 

Page 10: ANNA FENKO - TU Delft

INTRODUCTION 

 

10 

roughness of an electric toothbrush that they had to brush repeatedly across their front teeth. In line with the parchment skin illusion, the toothbrush was judged to be rougher and less pleasant when the overall sound level increased and when the high‐frequency sounds made by the toothbrush were amplified. The same type of auditory manipulation has also been found to affect the level of perceived carbona‐tion in sparkling water drinks (Zampini & Spence, 2005) and the perceived crispness of potato chips (Zampini & Spence, 2004). 

Cross‐modal  interactions  can also  influence affective  information processing. The perception of facial expressions can be modulated by an associated voice (de Geld‐er & Vroomen, 2000; Dolan, Morris, & de Gelder, 2001), and affective impressions of personality  traits  for people can be changed by  fragrance  (Baron, 1981; Fiore, 1992). Furthermore, studies on cross‐modal mere exposure effects between vision and touch (Suzuki & Gyoba, 2008) have demonstrated that previewing objects  in‐creased the preference judged by hand.  

Multisensory  cues  have  also  been  shown  to  play  a  role  in  our  comprehension  of speech. For example, in the McGurk effect (McGurk and MacDonald, 1976), the pair‐ing of the sound of the syllable /ba/ with the sight of a speaker mouthing the syllable /ga/ results  in the perception of neither /ba/ nor /ga/, but rather the syllable /da/, reflecting a synthesis of the information contained in the two sensory channels.  

Recently, it has been shown that an illusory flash can be induced by pairing a single flash with multiple sounds (Shams et al., 2000, 2002). Furthermore, vestibular cues produced by rotatory motion can alter a subject’s visual perception of line orienta‐tion in a darkened room (Graybiel, 1952), and tactile acuity is improved on the fin‐gertips when the subject can view the finger (Taylor‐Clarke et al., 2004). The list of such interactions is ever increasing. Taken together, these phenomena suggest that the human brain is able to synthesize information from the different senses, which allows us to create a multisensory view of the world. 

Behavioral studies have explored  the conditions under which crossmodal  interac‐tions occur. Two  key determinants of  intersensory binding are  synchronicity and spatial  correspondence (Radeau, 1994). Thus, when  two or more  sensory  stimuli occur at the same time and place, they are typically bound into a single percept and detected more  rapidly  than  either  input  alone  (Sekuler  et  al., 1997;  Stein  et  al., 1989). Similar  instances of crossmodal  facilitation have also been shown to affect detection  thresholds.  For  example,  Frassinetti  et  al.  (2002)  found  that  subject’s sensitivity to visual stimuli presented below luminance threshold was increased by a simultaneous accessory sound burst presented at the same spatial location. This 

Page 11: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

11 

effect was eliminated when the two sensory inputs were separated in space or off‐set by more  than 500 ms. Similar crossmodal  influences have also been  reported for auditory and tactile detection thresholds (for reviews see Loveless et al., 1970; Welch & Warren, 1986).  

In addition to the parameters of time and space, the synthesis of multisensory  in‐puts can also be  influenced by their semantic congruence. For example, hearing a dog’s bark emanating from the same approximate location as a visible cat is unlike‐ly to create the impression of a barking cat (Calvert & Thesen, 2004). Other factors that may play a role in mediating crossmodal interactions include task‐related fac‐tor (e.g., whether subjects are required to detect or discriminate a target; Lloyd et al., 2001), as well as the prior sensory bias of the subjects (e.g., whether they are visually or acoustically primed) (Giard & Peronnet, 1999).   

 

1.2. SPECIFICITY OF SENSORY SYSTEMS 

The task of multisensory integration is not easy; it requires the interaction of vari‐ous brain structures, synchronization and  fine‐tuning of  the complex neurological mechanisms. We may  wonder  why  humans  require  all  these  different  sensory channels to acquire information about the world. What determines the unique role of each sensory modality in our experience of the world?  

 

1.2.1. VISION 

One of  the main  characteristics of vision  is  its unique ability  to organize  sensory stimuli into objects. Goodale & Humphrey (1998) suggested that one of the major functions of the visual system  is the creation of an  internal model of the external world – a model that can be used in the recognition of objects and understanding their interrelations:  

“The fundamental task of vision is to construct a representation of the three‐dimensional layout of the world… Vision gives us an experience of the world beyond our immediate body surface, a world full of objects and events that are imbued with meaning and significance… The goal of visual perception is… to reconstruct a detailed and accurate model or replica of the three‐dimensional world on the basis of the two‐dimensional data pre‐sent at the retinas” (Goodale & Humphrey, 1998, p.181). 

Page 12: ANNA FENKO - TU Delft

INTRODUCTION 

 

12 

Behavioral  and neurocognitive  studies have demonstrated  that  visual perception and mental representations share the same mechanisms, and that visual areas of the brain play a functional role in mental imagery (Kosslyn, 1980, 1983). These stu‐dies led to the formulation of the theory that human knowledge is organized in the manner of mental pictures (Block, 1983). 

Visual  representation plays a  crucial  role  in  cognitive activities  related  to  the  se‐mantics  of  objects  and  scenes.  Our  visual  perception  of  the  world  appears  re‐markably rich and detailed, but much of this perceptual representation  is derived from memory  rather  than visual  input  (e.g. McConkie and Currie, 1996; O’Regan, 1992; Rensink et al., 1997). According to Neisser (1994), ‘seeing the world’ requires identification and classification of objects by a recognition system, which constantly relates present sensory input with the stored information about objects. 

The close relationship between vision and cognition has long been a focus of West‐ern  thought  (Weimer,  1977). Modern  cognitive  linguistics  stresses  the  similarity between vision and knowledge by pointing to the abundance of visual metaphors for knowledge and mental processes that exist  in English and other European  lan‐guages (‘as  illustrated by,’  ‘reflect on that,’  ‘see what  I mean,’  ‘a brilliant scholar,’ ‘bright idea,’ ‘a clear explanation’). Sweetser (1990) suggests three reasons for the prominent  relationship between vision and knowledge:  (1) The  focusing ability of the  sense of  vision  enables people  to pick up one  stimulus  from many,  and,  to‐gether with  its  intentionality  and  directionality,  differentiates  it  from  the  other senses.  (2) Vision  is our primary source of objective data about  the world. Visual features are the most marked  in children’s early categorization, and thus form an important  source  for  concept  formation.  Vision  provides  data  from  a  distance, which  through metaphoric  transfer  become  a  characteristic  of  objectivity  in  the intellectual domain.  (3) Vision  is  identical  for different people who  can  take  the same point of view. 

 

1.2.2. HEARING 

Vision and hearing are usually called  ‘distant senses’  (Sekuler & Blake, 1994), be‐cause they do not require direct contact to perceive objects. According to Sweetser (1990), this capacity links these two senses to objectivity and intellect.  

The main function of hearing in humans is communication, which includes intellec‐tual and emotional  influences between people. The sense of hearing  therefore  is connected to internal receptivity and the understanding of ideas. Perhaps, because 

Page 13: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

13 

of  this  readiness  to  respond  to  someone’s  ideas and wishes,  the verb  ‘hear’ also means ‘obey’ in many languages (e.g., in English).  

Auditory information is essential for language acquisition in children (Sloutsky and Napolitano, 2003). Language has been considered a central mechanism of thinking, either in the form of ‘mentalese’, the innate and unconscious ‘language of thought’ (Fodor, 1975), or in the form of ‘inner speech’, central to the socio‐cultural theory of thought (Vygotsky, 1986). Both these theories assume that internal representa‐tions of knowledge are organised  in symbols similar to those  in natural  language, and that thinking is similar to speaking to oneself.  

The question whether thinking is similar to seeing mental pictures or to speaking to 

oneself  is a subject of a  long scientific controversy  (see Block, 1983 for a review). 

According  to pictorialism,  cognition  is  similar  to  vision;  for  the  opposing  theory, 

descriptionalism, understanding  is similar  to hearing words. There  is a third opin‐

ion, called ‘dual coding theory’, according to which cognition involves the activity of 

both systems: a verbal system specialized for dealing with language and an imagery 

system  specialized  for  dealing  with  pictures  (Paivio,  2006).  Paivio  (1971)  and 

Richardson (1977) suggested that individuals can be classified as visualizers versus 

verbalizers according to their specific cognitive style, i.e. the way they acquire and 

process information. According to this view, visualizers rely primarily on visual im‐

ages when they perform cognitive tasks, whereas verbalisers rely primarily on ver‐

bal information and analytical strategies. 

These differences may determine the attention  individuals pay to visual and audi‐

tory  information. Consumer  research has demonstrated  that aesthetic  judgments 

differ among  individuals with different  cognitive  styles  (Childers et al. 1985; Hol‐

brook & Morris, 1986). For example, visualizers respond more sensitively to organic 

patterns  and  feature  interactions, while  verbalizers pay more  attention  to  single 

stimuli  and  isolated details. Analogously,  cognitive  style  could  also  influence  the 

relative importance of vision and hearing in product experience. 

 

1.2.3. TOUCH 

Touch  and  taste  are  often  called  ‘contact  senses,  or  ‘proximity  senses’,  because they require closeness with the perceived object in order to perceive its properties (Sekuler & Blake, 1994).  It may be  suggested  that  these  senses are more  closely 

Page 14: ANNA FENKO - TU Delft

INTRODUCTION 

 

14 

connected  to subjectivity,  intimacy and emotion  than  ‘distant senses’, such as vi‐sion and hearing. 

Touch  is unique among sensory systems.  It  is  the most diverse system, comprising a 

family of submodalities that respond in a variety of ways to mechanical, thermal, and 

chemical energy; it has the largest sensory surface, even if we ignore internal structures 

and consider only the skin. In addition, it is the system that is closely interwoven with 

affect, both positive and negative  (Hollins, 2010).  In everyday  language  the  sense of 

touch has always been related to the field of emotions. Expressions such as “I’m deeply 

touched” or “touching words” are widely used in English. 

Interpersonal touch has both emotional and communicative meaning (see Herten‐stein et al., 2006 for a review). Many social and affiliative behaviors, as well as be‐havior directed towards comfort and a feeling of well‐being, are directed towards pleasant touch (Rolls, 2005). Whether interpersonal touch is perceived as pleasant or unpleasant depends upon the genders and  familiarity  levels of dyad members, and  the  body  site  touched.  Nguyen  and  colleagues  (1975)  found  that  receiving touch from a stranger was  less pleasant and more  invasive than that from a close acquaintance, especially for females receiving touch from males. Touch to the arms and hands were suggested as particularly pleasant. 

The information we receive through touch is often accompanied by affect, which may be either positive (the pleasure of a kiss) or negative (the discomfort of feeling swea‐ty on  a hot day).  Pain  sensations, which  also belong  to  the domain  of  touch,  are strongly connected to negative emotions (Villemure et al., 2003; Bruehl et al., 2006).  But pain is not a necessary component of a negative tactile sensation. Psychological research shows that warm and cold stimuli have affective components (Rolls, 2005). For example, a cool surface applied to the hand often feels unpleasant.  

Touch is more precise than vision in assessing fine differences in textures. Texture, in a broad sense, means any mechanical properties that influence the feel of a surface. A multidimensional scaling study showed that roughness and hardness are the two main dimensions of texture space, with a slipperiness/stickiness dimension playing a secondary role for some individuals (Hollins et al., 2000). Texture perception also has an affective component.   For  instance,  in neurological patients who  lack myelinated mechanoreceptors  and  lost  their  tactile  detection  abilities, Olausson  et  al.  (2002, 2008) found that tactile stimuli elicited vague, pleasant sensations that were difficult to  localize.  The  authors  proposed  that  unmyelinated mechanoreceptive  afferents help trigger the emotional response to interpersonal touch.   

 

Page 15: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

15 

1.2.4. OLFACTION 

Similar to touch, olfaction is strongly connected to affective reactions. It also seems to be the most “subjective” of the senses, because it has a direct non‐reflective link to personal memories.  

Odors can generate pleasant or unpleasant experiences (Schleidt et al., 1988), and the hedonic determination of an odor has been proposed to be the key function of olfaction  (Yeshurun  &  Sobel,  2010).  Pleasant  odors  can  induce  positive moods, whereas  unpleasant  odors  can  induce  negative  moods  (Rétiveau  et  al.,  2004; Schiffman et al., 1995). Studies with newborns suggested that at least some aspects of olfactory pleasantness may be innate (Soussignan et al. 1997, Steiner 1979). 

Odors have the powerful ability to spark off vivid emotional autobiographical memo‐ries (Hinton & Henley, 1993; Chu & Downes, 2000, Willander & Larsson, 2006) and to evoke the associated emotions (Epple and Herz, 1999; Millot and Brand, 2001). Herz and Schooler (2002) found that odor‐evoked memories are more emotion‐laden and make participants feel more "brought back" to the original event as compared with the same cue presented visually or verbally. Stevenson and Boakes (2003) argue that odor perception depends on the implicit memories that an odorant elicits. They sug‐gest  that odor perception  relies more heavily on experience  than do other senses, and that olfactory memory differs  in many ways from other memory systems. Odor identification ability improves over a long period of development, which may extend into  adolescence  and  early  adulthood, while  the major  development of  the  visual system  is nearly complete by 1 year, and audition  (including spoken  language com‐prehension) is fully developed by 5 or 6 years.  

Odors not only elicit moods and emotions (Chu & Downes, 2002; Weber & Heuber‐ger,  2008),  but  odor  perception  also  shares many  common  characteristics with mood and emotion. Pleasantness  is the principal dimension of perception  in both emotion (Fontaine et al., 2007) and olfaction (Khan et al., 2007). When using large numbers  of  verbal  descriptors  to  describe  odorants,  pleasantness  repeatedly emerges as the primary dimension in multidimensional analyses of semantic space (Khan  et  al.,  2007; Moskowitz  et  al.,  1976,  Zarzo,  2008).  Furthermore,  odorant pleasantness was  the  primary  aspect  of odor  spontaneously  used  by  subjects  in olfactory discrimination  tasks  (Schiffman, 1974) and  the primary criterion sponta‐neously used by subjects to combine odorants  into groups  (Berglund et al., 1973; Schiffman et al., 1977).  

Humans are very good at detecting an odorant but are poor at naming  it. People are unable to name by smell at least 50% of the odorous household items they use 

Page 16: ANNA FENKO - TU Delft

INTRODUCTION 

 

16 

daily  (Cain, 1979; de Wijk et  al., 1995;  Lawless & Engen, 1977). The dissociation between knowledge about an odor concurrent with the  inability to name  it  is evi‐dent already at childhood. When smelling dangerous household products, children correctly named 15% of the odors, but correctly rated edibility of 79% of the same odors (De Wijk & Cain 1994).  

Hinton and Henley (1993) argue that the olfactory modality has a nonlexical coding system as opposed to the visual coding system which has lexical nature. There are relatively  few  olfactory  terms  in  English  and most  of  them  refer  to  bad  smells. Metaphorical meanings for smell verbs include the detection of bad characteristics, as in “something about his testimony stinks” (Caplan, 1973; Viberg, 1984; Sweetser, 1990)  and  to  suspect,  to  guess,  to  sense  something  intuitively,  as  in  “she  could smell money, power, victory” (Ibarretxe‐Antuñano, 1999). 

 

1.2.5. TASTE 

While the connotation of ‘smelly’ is often negative, that of ‘tasty’ is positive, which Classen  (1993)  explains  by  the  fact  that  people  are  confronted with  foul  smells more often than with foul tastes: “We can choose our food, but we cannot as read‐ily  close our noses  to bad  smells”  (p. 53). Taste  figuratively means  ‘judgment of what  is beautiful’ and  is, therefore, characterized as a sense of aesthetic discrimi‐nation.  This  relation  between  taste  and  preferences  is  very  common  cross‐linguistically (Buck, 1949). 

Taste is an especially powerful sensory domain for eliciting affective reactions, be‐cause food is a basic biological commodity, and because gustatory stimuli can elicit either  positive  or  negative  affective  valence  (Bartoshuk  &  Beauchamp,  1994; Pfaffmann, 1960; Rozin, 1999; Rozin & Schulkin, 1990). 

Newborn  infants  respond positively  to  sweet  sugars  and  reject bitter  substances (Rozin, 1999). The liking for sweet substances is probably innate in humans and in most omnivores, possibly as an evolutionary selection of a diet rich  in calories, vi‐tamins and minerals, such as ripe fruits and vegetables (Bartoshuk & Beauchamp, 1994).  Innate  rejection of bitter substances may  reflect  the  fact  that many bitter compounds are toxic (Rozin & Schulkin, 1990).  

Although we generally speak of tasting foods and beverages, much of the sensory input involved  is actually olfactory. This  confusion arises because  there are  two  routes by which odorants can reach the odor receptors. One route is taken by the odorant when it is sniffed through the nose. Another route is taken by the odorant when it goes from 

Page 17: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

17 

the mouth up into the nasal cavity during chewing and swallowing. Because of this con‐fusion, the majority of  food studies  investigates olfaction and  taste  together and de‐scribes the integrative food sensation as flavor (Piggott, 1990; Rolls, 2008).  

 

1.3. SENSORY DOMINANCE 

Usually  the  information  from different  senses  is  integrated  into  a  consistent  ex‐perience  of  objects  or  events. However,  if  this  information  is  contradictory,  the brain needs  to  solve  the perceptual problem.  It  can  ignore  the  information  that does not fit the usual picture of the world or it can combine it in some compromise object. For example, in experiments with prismatic lenses that invert the perceived distances  (objects  that  are  close  are  perceived  as  if  they  are  far  away  and  vice versa) a person sees the mask of a face as if it were the inside of the mask (Yellott & Kaiwi, 1979), but perception of human faces remains unchanged. Apparently, the perceptual system allows individuals to see possible or the most likely objects, but ignores impossible or rare variants. 

When the data from the different modalities are taken into account, the contribu‐tion of each modality to the entire experience can be measured experimentally. A lot  of  such  experiments were  performed  by  cognitive  scientists,  and  the  results show the phenomenon of sensory dominance, that is the degree of the prevalence of one modality over the others in conflicting situations. 

One of  the earliest experiments addressing  this  topic was performed  in  the 19th century by  the American psychologist George Stratton. For  several days he wore prismatic lenses that made him see all things inverted, as if the whole field of view had been revolved on the line of sight through an angle of 180 degrees.  

“All images at first appeared to be inverted; the room and all in it seemed upside down. The hands when stretched out from below into the visual field seemed to enter from above. Yet although these images were clear and definite, they did not at first seem to be real things, like the things we see in normal vision, but they seemed to be misplaced, false, or illusory im‐ages between the observer and the objects or things themselves. All movements of the body at this time were awkward, uncertain, and full of surprises... The movement was a series of errors and attempts at correc‐tion... I felt my hand to be in a different position from that in which I saw it, and could not, except by cool deliberation, use its visual image as a sign of impending tactual experience. After a time, however, repeated experience 

Page 18: ANNA FENKO - TU Delft

INTRODUCTION 

 

18 

made this use of the visual image much less strange... The limbs began ac‐tually to feel in the place where the new visual perception reported them to be. The seen images thus became real things just as in normal sight” (Stratton, 1897, p.349).  

One of the conclusions from Stratton’s experience is that different senses may have different degree of reliability. At first, Stratton relied entirely on the felt position of his body, because the visual picture was perceived as false and illusory. But in sev‐eral days the visual picture became more real and “whatever there was of abnor‐mality seemed to  lie  in myself, as  if head and shoulders were  inverted” (Stratton, 1897, p.468).  

Another  important  conclusion  from  Stratton’s  experiment  is  that  the  dominant modality  can  change over  time. At  first  visual  information  felt  completely  inade‐quate. Stratton had to use his memory and tactile  information to assess the  loca‐tion of objects. But after the adaptation he was able to rely on visual information. It was his body localization that felt inadequate now, as if his head was upside down. 

A lot of research that addressed the problem of sensory dominance used distortion of the visual image, but in most cases it was milder than in Stratton’s experiments. When the optical distortion is 10 degrees instead of 180 degrees participants usu‐ally do not even notice it. They rely entirely on vision and completely ignore tactual information (Warren & Cleaves, 1971). For instance, in a classic study by Rock and Victor (1964) participants were presented with an object of which the visual shape, because of optical distortion, differed considerably from its actual shape perceived by  touch. Participants examined a  square object by hands and at  the  same  time saw it through a lens which compressed its visual width to one half its original size. The  conflict between  visual  and  tactual  size was  resolved  completely  in  favor of visual size. Vision was so powerful in comparison to touch that the object actually felt like a rectangle, and most subjects were unaware of a conflict. 

Strong visual dominance over touch has been demonstrated in a variety of percep‐tual  tasks,  involving  the perception of size  (Miller, 1972),  length  (Teghtsoonian & Teghtsoonian,  1970),  curvature  (Easton  & Moran,  1978),  depth  (Singer  &  Day, 1969), and  spatial  location  (Hay, Pick, &  Ikeda, 1965). Some  studies also demon‐strated visual dominance over auditory sensory signals (Bertelson, 1999). 

However, visual dominance does not occur in all instances. For example, in a study in which participants were presented with conflicting information regarding surface texture, vision and touch were weighed equally when subjects performed a texture identification task (Lederman & Abbott, 1981). When subjects were asked to judge 

Page 19: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

19 

surfaces according to their  ‘roughness’, the relative weighing of visual and tactual inputs  appeared  to  be  30%  and  70%,  respectively  (Lederman,  Thorne, &  Jones, 1986). Shapiro et al. (1984) showed that the auditory modality may be superior to vision in alerting the organism to novel events. Furthermore, in a study without the involvement of the visual modality, a significant bias of proprioception on the per‐ceived position of auditory stimuli has been reported (Caclin et al., 2002). 

Developmental studies have highlighted that significant changes take place in mul‐tisensory  processing  as  maturation  progresses  (Lewkowicz,  1996;  Wallace  and Stein, 1997).  For  instance,  Sloutsky  and Napolitano  (2003) demonstrated  that  4‐year‐old children were more  likely to process auditory than visual stimuli. The au‐thors propose that the privileged status of the auditory modality may be function‐ally important for language acquisition:  

“Because visual processing  is parallel and auditory processing  is serial,  in the absence of privileged processing of auditory stimuli the visual modality would completely dominate the auditory modality, thus making the task of language acquisition  very difficult,  if not  impossible”  (Sloutsky & Napoli‐tano, 2003, p.830). 

 

1.3.1. EXPLANATIONS OF SENSORY DOMINANCE 

Several hypotheses were proposed to explain why a particular modality can domi‐nate in particular situations.  

The modality appropriateness hypothesis (Welch, 1999) claims that dominance of a particular modality depends on its appropriateness for a particular task. For exam‐ple, vision is more appropriate for spatial perception, audition is more appropriate for the detection of temporal events (Welch, 1999), while touch  is best suited for tasks that relate to the material properties of objects (Lederman, Thorne, & Jones, 1986).  In  situations of  conflicting  intersensory  stimulation  the modality  that best suits the task will be weighted more heavily. 

Modality dominance can also depend on mechanisms of selective attention. This hy‐pothesis suggests that the modality that receives the observers’ directed attention is the dominant modality (Posner, Nissen, & Klein, 1976). Researchers have shown that people can consciously direct their selective attention to a particular sense (Spence, Nicholls, & Driver, 2001). Attending to one sensory modality can facilitate the percep‐tion of stimuli in that modality relative to when attention is diverted to another sen‐sory modality. Guest & Spence (2003) found that when participants were presented 

Page 20: ANNA FENKO - TU Delft

INTRODUCTION 

 

20 

with both visual and tactual information and asked to ignore the stimuli of one mo‐dality, touch was the more difficult sense to ignore. 

Culture  can play a  role  in determining what  sensory modality  is more  important and to what modality people should pay more attention. Culture has been shown to  influence a  large variety of behaviors across many diverse disciplines,  including consumer behavior (see Hofstede, 2001). McLuhan (1961) believed that the nature of media by which people communicate affects the ratio of their senses. For exam‐ple, the alphabet stresses the sense of sight, which in turn causes people to think in linear, objective terms.  

Language is one of the core components of culture. It is central to communication and closely related to thought. A proposal of  linguistic relativity emphasizes a dis‐tinctive role of  language  in shaping experience and  influencing thought  (Gumperz and Levinson, 1996). In recent years, experimental studies have demonstrated that language has  an  important  influence on  various  aspects of perception,  including perception of space, time, and color (Boroditsky, 2001; Bowerman, 1996; Levinson, 1996). Studies have also shown that language characteristics can influence product experience (Lucy and Gaskins, 2001; Boroditsky et al., 2003).  

Language can also influence the relative importance of sensory modalities through the metaphorical  use  of  sensory  descriptors. Metaphors  often  refer  to  sensory phenomena:  good  ideas  are  described  as  ‘brilliant,’  pleasant  dreams  as  ‘sweet,’ important topics as ‘hot,’ and bright colors as ‘loud.’ These metaphorical meanings of sensory terms may vary between languages, which may affect product categori‐zation and experience.  

 

2. WHAT IS PRODUCT EXPERIENCE?  The theoretical framework of experience has long been at the margins of experimental psychology, being mostly  in  the  focus of phenomenological philosophy  (e.g. Husserl, 1962; Heidegger, 1964; Merleau‐Ponty, 1962) and counseling psychology (Ponterotto, 2005).  Phenomenology  studies  conscious  experiences  as  they  are  lived  and  verbally articulated  from  the  first‐person point of  view,  including  experiences of perception, action, memory, mental imagery, emotion, attention, empathy, dreaming, and so forth. According  to  phenomenology,  all  our  experiences  are  intentional:  they  aim  toward something beyond themselves. When we see, we see something; when we remember, we remember something; when we hope or fear, we hope for or fear something.  

Page 21: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

21 

Any experience is given directly to the person and only to that person. We cannot share  each  other’s  thoughts,  emotions, memories,  pains,  and  so  on.    Such  first‐personal  givenness  is  a  primitive  form  of  self‐consciousness  in  the  sense  that  it does not  require any act of  reflection or  introspection, but occurs spontaneously and involuntarily. In this respect, all experiences are conscious. Although not explic‐itly accessed to attention and reflection, they are accessible  in principle: they can become available to introspection and verbal report (Lutz & Thompson, 2003).  

Although conscious experience is the starting point of phenomenology, experience shades off  into  less overtly  conscious phenomena. As Husserl and other philoso‐phers stressed, we are only vaguely aware of things  in the periphery of attention, and we are only implicitly aware of the wider horizon of things in the world around us. Moreover,  as Heidegger  stressed,  in practical  activities  like walking  along, or hammering a nail, or speaking our native tongue, we are not explicitly conscious of our  habitual  patterns  of  action.  Furthermore,  as  psychoanalysts  have  stressed, much of our  intentional mental  activity  is not  conscious  at  all, but may become conscious in the process of therapy or analysis, as we come to realize how we feel or  think  about  something.  So,  the  domain  of  experience  spreads  out  from  con‐scious experience into semi‐conscious and even unconscious mental activity. 

For  the entire 20th  century, psychology has been dominated by  the positivist  re‐search  paradigm  (Rennie,  2002), which  aims  at  an  explanation  of  psychological processes in order to predict and eventually to control them. Experiential phenom‐ena were considered too vague,  intractable, and therefore, not worth of scientific inquiry. Psychologists have only gradually come to acknowledge the possibility that phenomenology may  be  relevant  for  a  scientific  understanding  of  cognition.  Re‐cently, a growing number of cognitive scientists  recognize  the need  to make sys‐tematic use of  introspective phenomenological reports  in studying the brain basis of consciousness (Jack and Shallice, 2001; Jack and Roepstorff, 2002; Dehaene and Naccache, 2001). A research program of neurophenomenology stresses the impor‐tance of gathering first‐person data as a strategy for describing and quantifying the physiological  processes  relevant  to  consciousness  (Varela,  1996;  Lutz,  2002;  Ru‐drauf  et  al.,  2003). Neurophenomenology  assumes  that  conscious  experience  is grounded in neural activity and embodied action.  

Recently, the concept of experience has also made its way to the field of marketing and consumer behavior. The proclaimed new era of the experience economy and experience marketing (e.g. Pine & Gilmore, 1998; Schmitt, 1999) made experience a focus of consumer behavior research and other areas of applied psychology. Ac‐cording to Schmitt (1999), for example, the ultimate goal of experiential marketing 

Page 22: ANNA FENKO - TU Delft

INTRODUCTION 

 

22 

is  to  create  a  desirable,  coherent,  and  consistent  customer  impression  that  en‐hances  the  brand  image. When  people  repeatedly  encounter  a  particular  brand within the context of a pleasant experience, they are more  likely to develop a fa‐vorable attitude towards this brand. In addition, characteristics of the atmosphere during  the  experience  (e.g., modern,  fresh,  impressive) may  become  associated with the brand.  

Advocates of the “experience economy” emphasize that experiences can be inten‐tionally designed: “An experience occurs when a company  intentionally uses  ser‐vices as the stage, and the goods as props, to engage individual customer in a way that creates a memorable event” (Pine & Gilmore, 1998). They discuss the  impor‐tance of customer’s impressions which should be created by ‘positive cues’ and by the elimination of  ‘negative  cues’:  “The more  senses an experience engages,  the more effective and memorable it can be”. Schmitt (1997) stresses that “in a world in which most consumers have their basic needs satisfied, value  is easily provided by satisfying customers’ experiential needs – their aesthetic needs”.  

Industrial designers  also  realized  that  they do not only make products  to  satisfy consumers' needs, but create new user experiences. In the 1980s, emotional design started in Europe and North America (Norman, 2004; Jordan, 2000; Desmet, 2002). Independently, Kansei engineering spread in Japan from the early 1970s (Nagama‐shi, 1989; Horigushi et al., 1995). Kansei engineering is defined as a product devel‐opment methodology that translates consumers’ and users’  feelings,  impressions, and emotions into concrete design parameters (Nagamashi, 1989). Nowadays, the design for experience approach becomes holistic, considering all sensory properties of  a  product  and  referring  to  the  people’s  global  experience  (Clos &  Bouchard, 2010; Desmet & Schifferstein, 2010). 

Product experience can be defined as “the awareness of the psychological effects elicited by  the  interaction with  a product,  including  the degree  to which  all our senses are stimulated, the meanings and values we attach to the product, and the feelings and emotions that are elicited” (Hekkert & Schifferstein, 2008, p.2). Under‐standing such experiences will allow designers to ‘design for experience’. 

In this thesis we focus on the sensory aspects of product experience. Both in every‐day  life and  in experimental research,  it  is difficult to differentiate the three main components of product experience outlined  in this definition. Sensory perception, especially in the case of visual and auditory modalities, is closely connected to cog‐nitive processes  (Goodale & Humphrey, 1998; Neisser, 1994; Paivio, 2006). Also, some sensory modalities (especially touch, smell, and taste) are strongly associated 

Page 23: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

23 

with  emotions  (Hinton &  Henley,  1993;  Sweetser,  1990).  Furthermore,  emotion and  cognitive processes, which  are usually described  as distinct, often blend  to‐gether  in  everyday  life:  our  attention  is  drawn  to  emotional  stimuli;  emotional events are retained  in memory; and  judgment sometimes  involves weighing emo‐tional consequences (Hollins, 2010).  

In the marketing literature, the utilitarian properties of products are often analyzed separately from their affective and aesthetic properties. Sometimes the separation even reaches  to product groups  in general, some of which are considered strictly utilitarian (‘think products’), while others are completely affective (‘feel products’). ‘Think’ products are bought mainly for utilitarian, cognitive reasons (Claeys, 1995), while purchase of  ‘feel’ products  is driven by the need to satisfy emotional wants (Hirschman  and  Holbrook,  1982),  and  by  value‐expressive  and  affective motives (Park  and Young, 1983),  such  as ego  and  sensory gratification,  social acceptance (Ratchford, 1987; Rossiter and Percy, 1991). 

From the phenomenological perspective, distinctions between ‘thinking’ and ‘feel‐ing’ are not  that clear and straightforward.   Each consumer product can  to some extent be experienced  as utilitarian or  affective depending on  a usage  situation, user personality, and particular product properties. Even practical products such as a dishwashing  liquid can evoke a complex of user experiences, not all of them di‐rectly related to the product’s utilitarian function.  

 

2.1. SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

When people  interact with products, they receive  information through all the dif‐ferent  senses. Because  all  sensory  inputs  can affect  the way  in which users per‐ceive, make  sense  of,  experience,  and  evaluate  a  product,  several  authors  have proposed a multisensory approach  to product design,  in which all perceptual  im‐pressions  elicited when  a potential user  interacts with  the product  are  explicitly created and evaluated  (e.g., Adank and Warell, 2008; MacDonald, 2002; Schiffer‐stein and Desmet, 2008). Nonetheless,  information from some sensory modalities may be more  important  for  certain product  experiences  than others. We define “sensory  importance”  as  the  relative  contribution of  each  sensory modality  to  a particular product experience. The dominant sensory modality is the modality that has the largest effect on the specific experience.  

A general and popular belief seems to be that vision  is the dominant sensory mo‐dality in everyday experience. When people are asked which sensory modality they 

Page 24: ANNA FENKO - TU Delft

INTRODUCTION 

 

24 

would miss most  if  they  lost  it,  the majority  is  likely  to  indicate vision  (Fiore and Kimle, 1997; Schifferstein, 2006).  In addition, when people are asked  to describe objects, they primarily use adjectives that refer to the visual (60%) or tactual (32%) modalities (Stadtlander and Murdoch, 2000).  

Research in the domain of product experience also suggests that visual dominance exists in the experience of consumer products. Schifferstein and Cleiren (2005) per‐formed a study  in which they showed that consumers acquired most of the  infor‐mation on products by vision and  touch:  this  information was most detailed and the  subjects were  surest  of  their  judgments.  The  experiment  also  showed  that products were harder to identify by sound or smell than by vision or touch. Another study on  the  roles of  sensory modalities  in product experience  (Schifferstein and Desmet, 2007) suggested that vision gathers the largest amount of information on a product within the shortest time frame. 

A questionnaire study in which participants reported the importance of the sensory modalities during  the usage of 45 different products  (Schifferstein, 2006) demon‐strated that on average the relative  importance sequence of sensory modalities  is vision, followed by touch, smell, audition and taste. In addition, when people were asked  to  rate how  important  they  found  the different modalities  in  their  lives  in general, most of  them  selected vision as  the most  important modality. However, the  importance  ratings  for  the sensory modalities differed greatly between prod‐ucts. For almost half of the 45 products, the importance of vision was rated lower than one of the other modalities. For example, audition is the most important mo‐dality  for a washing machine and a coffee maker, which can be explained by  the role of the sound in signaling the different stages of the process of washing or mak‐ing coffee. Touch is most important for a computer mouse and a pen, and probably for any other hand tools as well. Smell plays a dominant role for a deodorant and (together with taste) for food products.  

The  relative  importance of  a modality may depend on  various  aspects,  such  as  the availability of sources of sensory stimulation, the degree of variation in sensory stimula‐tion over various products, the usefulness of the sensory information during functional use, the proportion of time a modality is used actively, and the role of the stimulation in enjoying the product. Hence, the roles of the modalities depend not only on whether a certain type of sensory information is present, but also on whether the information is perceived, how it is processed, and how people react to it emotionally.  

 

Page 25: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

25 

2.2. EMPIRICAL CATEGORISATION OF PRODUCT EXPERIENCE 

To  find out how people normally  relate  their product experience  to  sensory  im‐pressions, we  performed  an  exploratory  questionnaire  study. We  asked  partici‐pants  (118 master  students of  the  IDE Faculty of TU Delft)  to describe  the  three most  important experiential aspects of  the products  they had selected previously as  familiar  and  important  for  them  (e.g.,  I  like my mobile  phone  because  it  is thin, elegant and smart, or: I dislike that my car is less powerful and fast than oth‐ers, but  I appreciate that it  is quite silent). For each of these three characteristics separately we asked participants to indicate to what extent they could evaluate this aspect on the basis of (1) the smell, (2) the visual appearance, (3) the sound, (4) the tactual feel, and (5) the taste of the product, using a scale from 1 = not important at all to 5 = very important.  

On  the  basis  of  the  responses, we  created  product  categories  that were  largely homogeneous in type of product (see Table 1.1). The largest categories in our sam‐ple were personal products  (wristwatches, bags,  shoes, etc.), electronics  (mobile phones,  cameras,  computer  accessories),  and  electric  appliances  (such  as  coffee makers and washing machines). 

We also made categories of the various experience aspects that were mentioned (see Table 1.2). The aspects mentioned most frequently by our respondents were related to product usage  (comfortable,  safe, easy  to use, etc.,),  functionality  (durable,  reliable, efficient, etc.), character (professional, smart), and appearance (elegant, simple).  

Then we determined the experiential aspects that were mentioned most often for the different product categories (about 18% or more): 

Appliances:   Usage, functional, character 

Electronics:  Usage, character, functional 

Home:    Character, appearance, functional, usage 

Leisure:    Usage, functional, character 

Personal:  Usage, appearance, character 

Tool:    Functional, usage, emotional 

Vehicles:  Usage, appearance, functional 

 

 

Page 26: ANNA FENKO - TU Delft

INTRODUCTION 

 

26 

TABLE 1.1. PRODUCT GROUPS 

Group  Products  Frequency  % 

Personal  prod‐ucts  

Wristwatch (11), bag (4), backpack (4), shoes (2), perfume bottle (2), razor (2), agenda, Chinese fan, suitcase, tobacco package, wallet, dildo, dressing gown, sponge, toothbrush  

34  30 

Electronics   Mobile phone (7), camera (4), computer mouse (3), laptop (3), videogame  (2),  turntable  (2), earphones, mobile hard disk, mp3 player, speakers, telephone 

26  22 

Electric  appli‐ances 

Coffee maker (6), blender, electric kettle, rice cooker, vac‐uum cleaner, washing machine, fan, sewing machine, elec‐tric drill, electric trimmer 

15  13 

Home  accesso‐ries  

Clock  (2),  lamp  (2), shaker  (2), bed, bench, chaise  longue, candles, light switch, lock, garbage bin, jar, mug 

15  13 

Sport &  leisure equipment 

Flashlight  (2),  skateboard  (2),  camping  stove,  helmet, hockey  guards,  Rubik’s  cube,  snowboard,  tennis  racket, guitar 

12  9 

Tools   Pencil  sharpener  (2), pocket knife  (2), drawing  set, mark‐ers, scissors, utility knife 

8  7 

Vehicles   Car (3), bike (2), airplane, caravan  7  6 

  Total  117  100 

 

These data show that different types of experience aspects are  important  for dif‐ferent product groups. For  instance, appearance  is more  important  for home ac‐cessories, personal products, and vehicles  than  for  the other categories. Further‐more, the emotional aspect of experience (attractive, nice, and friendly) was mainly important for tools. Functional and usage aspects of product experience were the most  important aspects for all product groups. This result suggests that utilitarian product characteristics represent an important component of user experience, and that product  experience  should not be  restricted only  to  affective  and  aesthetic product properties.   

 

 

Page 27: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

27 

TABLE 1.2. ASPECTS OF PRODUCT EXPERIENCE  

Aspect   Examples of aspects mentioned  Frequency  % 

Functionality 

 

reliable (8), efficiency (5), multifunctional (5), durable (4), powerful (3), practical (3) 

64  18 

Usage  comfortable  (24), usage  (12), safety  (7), ergonomics  (6), easy to use (6), feedback (5) 

88  25 

Appearance   

elegant (17), simple (8), appearance (6), design (5), color‐ful (3) 

51  15 

Vision + touch  quality (7), material (3)  20  6 

Touch  light (7) / heavy (2), solid (3), robust (3)  20  6 

Sound  Sound (2), loud (2) / soft, noise (2)  9  2 

Smell and taste  taste of coffee (2), bad smell  6  2 

Character  professional (7), smart (5), original (3), authentic (3)  62  18 

Emotional  im‐pact 

attractive (3), nice (2), friendly (2) 

  

26  8 

  Total  346  100 

 

Subsequently, we calculated the mean importance ratings for all sensory modalities per product category. Our results show that all categories follow the general pat‐tern that was found  in previous studies (Schifferstein, 2006): For most categories, vision is most important on average, followed by touch > audition > smell and taste (see  Table  1.3).  The  only  exception  is  leisure  products  (musical  instruments  and sport equipment), for which touch is more important than vision.  

We also calculated the means for the modality  importance ratings per experience aspect (see Table 1.4). These means per aspect deviate from the rule that modality importance  follow  the  sequence vision >  touch > audition >  taste and  smell. We see, for example, that for emotional and functional aspects of product experience vision and  touch  seem  to be equally  important. As expected,  for  sensory experi‐ences related to sound, touch, and taste/smell, the respective modalities are domi‐nant. For usage and vision/touch aspects, touch seems to be more important than vision. These  findings  indicate  that  the  relative  importance of  sensory modalities 

Page 28: ANNA FENKO - TU Delft

INTRODUCTION 

 

28 

differ  between  the  various  experiential  product  aspects.  Therefore,  it would  be interesting  to  investigate  in more detail  to what extent  sensory modalities affect different aspects of the experience, and to explore why those differences occur. 

 

TABLE 1.3. MEAN MODALITY RATINGS PER PRODUCT GROUP 

Product group  Vision  Audition  Touch  Smell  Taste 

personal  4,17  2,45  3,99  1,85  1,10 

electronics  4,28  3,11  3,83  1,37  1,03 

appliances  4,00  3,53  3,84  1,93  1,47 

home  4,65  2,83  3,81  1,71  1,06 

leisure  4,06  3,28  4,50  1,50  1,12 

tools  4,24  2,60  4,20  1,52  1,04 

vehicles  4,10  3,52  4,10  2,29  1,05 

Total  4,23  2,93  3,98  1,70  1,12 

 

TABLE 1.4. MEAN MODALITY RATINGS PER EXPERIENCE ASPECT 

Experience aspect  Vision  Audition  Touch  Smell  Taste 

functional  4,23  3,30  4,14  1,66  1,08 

usage  3,90  2,88  4,47  1,58  1,02 

appearance  4,90  2,33  3,55  1,75  1,08 

vision/touch  4,38  2,90  4,48  1,38  1,10 

touch  3,75  2,65  4,35  1,40  1,10 

sound  2,75  5,00  2,62  1,25  1,00 

taste/smell  3,17  1,67  1,33  4,83  2,83 

character  4,67  3,00  3,64  1,78  1,19 

emotional  3,89  3,11  3,96  1,89  1,11 

Total  4,24  2,91  3,99  1,70  1,12 

Page 29: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

29 

Some aspects of experience, such as comfortable (24), elegant (17), simple (8), reli‐able  (8),  professional  (7),  light  (7),  safe  (7),  were mentioned  repeatedly  in  our study.  The  analysis of  these  spontaneous descriptions  reveals users’ own under‐standing of product experience and its important aspects. It is interesting to relate this  ‘naive’ understanding of product experience  to  the definition of product ex‐perience proposed by Hekkert and  Schifferstein  (2008), which outlines  the  three main  aspects  of  experience  (sensory,  symbolic,  and  emotional).  Free  elicitation used in our study provided relatively more symbolic aspects (such as elegant, pro‐fessional,  reliable).  Sensory experiences were mentioned  less often and  included mostly  visual  (appearance,  colorful)  and  tactile  properties  (comfortable,  light). Sound and smell/taste characteristics were mentioned only by 2% of participants. Emotional experiences were also mentioned rarely (8%).  

We may wonder to what extent these data reflect the actual importance of affec‐tive or sensory components of product experience, and to what extent they result from differences  in awareness or accessibility of verbal descriptions of various ex‐periential  aspects.  Some  sensory  experiences  (i.e.,  visual)  are  easier  to  verbalize than others (i.e., smell and taste). Also, emotional aspects of experience might be more difficult to verbalize and to report spontaneously than cognitive aspects.  

 

3. OVERVIEW OF THE THESIS The main  focus  of  this  thesis  is multisensory  integration  in  product  experience. Some  of  the  sensory  systems  play  a more  important  role  in  product  experience than others.  It  is often assumed  that vision dominates  the other  senses, but  the previous discussion demonstrates that the importance of a particular sensory mo‐dality may depend on the stage of product usage, on the cultural characteristics of users (such as the language they speak), and on the type of product experience. In the following chapters we will describe the studies that investigated these issues.  

In the first two studies described in Chapters 2 and 3 we investigated dynamic aspects of sensory dominance. When consumers buy a product, they may pay more attention to its visual attributes. But with time other modalities can become more important. No matter how nice new shoes can look, during usage it becomes more important whether they are comfortable or not to wear. Kitchen tools can be too heavy to use; an iron can produce a bad smell while using; new linen may be not as soft as the old, and so on. In Chapter 2 we used a questionnaire approach to investigate the changes in sensory do‐minance over time. We asked participants to describe their experiences with products 

Page 30: ANNA FENKO - TU Delft

INTRODUCTION 

 

30 

in various situations: while buying a product, after the first week, the first month, and the first year of usage. In Chapter 3 we used an experimental approach to investigate how the sensory properties of a specific food product (dehydrated soup) influence the experience  of naturalness of  this product  and  emotional  reactions  to  it  at  different stages of user‐product interaction. 

Verbalization  is an  important component of product experience, because  it allows people to recall their experience and to communicate it to others. People describe their product experiences using adjectives that can be divided in three groups: sen‐sory descriptors  (e.g., hard, red, noisy); symbolic descriptors  (e.g.,  interesting, ex‐pensive, modern); and affective descriptors (e.g., pleasant, beautiful). In the study described  in Chapter 4 we use a questionnaire approach  to quantify  the  relative importance of the five sensory modalities for various descriptors of product experi‐ence. We suggest that modality importance may be influenced by language differ‐ences. To test this hypothesis, we use two groups of participants, native speakers of Dutch and Russian. We also discuss the relationships between literal and meta‐phoric meaning of sensory product properties.  

In  cognitive  science,  the modality  appropriateness  hypothesis  was  proposed  to explain the phenomena of sensory dominance (Welch, 1999). In the area of prod‐uct experience  this hypothesis may be  reformulated as  the experience appropri‐ateness hypotheses. The sensory modality that best suits  the specific product ex‐perience may be more  important  than other  senses. For  instance,  touch may be dominant for the experience of warmth  in products, while audition may be domi‐nant  for  the experience of noisiness. This hypothesis was  tested  in  three experi‐mental  studies discussed  in Chapters 5, 6, and 7. We  investigated  sensory domi‐nance for three product experiences: freshness, warmth, and noisiness. In each of these three experiments, we also  looked at the  importance of sensory modalities for affective product experiences (pleasantness and annoyance).  

In Chapter 8 we discuss the general findings all the studies, outline the possibilities for future research, and formulate the implications of our findings for industrial designers. 

 

 

Chapter 2 was published as:  

Fenko, A., Schifferstein, H.N.J., Hekkert, P. (2010) Shifts  in sensory dominance be‐tween various stages of user‐product  interactions.  Journal of Applied Ergonomics, 41, 34–40.  

Page 31: ANNA FENKO - TU Delft

31 

 

 

 

 

 

 

 

 

 

 

CHAPTER 2 

SHIFTS IN SENSORY DOMINANCE BETWEEN 

VARIOUS STAGES OF USER‐PRODUCT INTERACTIONS 

 

1. INTRODUCTION Consumer  experience with  products  is  always multisensory.  For  example, when making coffee a person sees the coffeemaker, touches its buttons, hears the sound it makes, enjoys the smell of fresh coffee and, eventually, tastes the coffee. All sen‐sory modalities  contribute  to  this  experience  to  some extent. But which  are  the most important? The answer to this question can be of practical interest to indus‐trial designers. If they concentrate their efforts on the most  important modalities, they are more likely to create an agreeable sensory product experience and to pre‐vent user disappointment. Another strategy  is to design new properties for a sen‐sory modality that has until then been considered unimportant,  in order to enrich the overall multisensory product experience.  

 

Page 32: ANNA FENKO - TU Delft

CHAPTER 2 

 

32 

1.1. RESEARCH ON SENSORY DOMINANCE  

In experimental psychological research strong dominance of vision over touch has been demonstrated  in a  variety of perceptual  tasks,  involving  the determination of  shape (Rock  and Victor, 1964),  size  (Miller, 1972),  length  (Teghtsoonian and Teghtsoonian, 1970), curvature (Easton and Moran, 1978), depth (Singer and Day, 1969), and spatial location (Hay et al., 1965).  In addition, some studies have demonstrated visual domi‐nance over auditory sensory signals (Bertelson, 1999). However, vision is not dominant in all situations. When subjects are asked to  judge surfaces according to their rough‐ness,  they exhibit considerably greater emphasis on  the  tactual as opposed  to visual cues (Lederman, Thorne, & Jones, 1986). Furthermore, the auditory modality may be superior to vision in alerting the organism to novel events (Shapiro et al., 1984).  

In one of  the earliest experiments on  sensory dominance, a dominance effect of proprioception  over  vision was  demonstrated.  For  several  days George  Stratton wore prismatic  lenses that made him see objects  inverted, as  if the whole field of view had been  revolved on  the  line of sight  through an angle of 180 degrees. At first Stratton relied entirely on the felt position of his body. The visual picture was perceived as “false” and “illusory”. But  in several days  the visual picture became more “real” and “whatever there was of abnormality seemed to lie in myself, as if head  and  shoulders were  inverted”  (Stratton,  1897).  Stratton’s  experiment  also indicates that the dominant modality can change over time. At first visual informa‐tion felt completely inadequate and Stratton used his memory and touch to assess the location of objects, but after the process of adaptation had taken place, he was able to rely on visual information again.  

Research in the domain of product experience also suggests that visual dominance exists in the experience of consumer products. Schifferstein and Cleiren (2005) per‐formed a study  in which they showed that consumers acquired most of the  infor‐mation on products by vision and  touch:  this  information was most detailed and the  subjects were  surest  of  their  judgments.  The  experiment  also  showed  that products were harder to identify by sound or smell than by vision or touch. Another study on  the  roles of  sensory modalities  in product experience  (Schifferstein and Desmet, 2007) suggested that vision gathers the largest amount of information on a product within the shortest time frame. 

A questionnaire study in which participants reported the importance of the sensory modalities during  the usage of 45 different products  (Schifferstein, 2006) demon‐strated that on average the relative  importance sequence of sensory modalities  is vision, followed by touch, smell, audition and taste. In addition, when people were asked  to  rate how  important  they  found  the different modalities  in  their  lives  in 

Page 33: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

33 

general, most of  them  selected vision as  the most  important modality. However, the  importance  ratings  for  the sensory modalities differed greatly between prod‐ucts. For almost half of the 45 products, the importance of vision was rated lower than one of the other modalities. For example, audition is the most important mo‐dality  for a washing machine and a coffee maker, which can be explained by  the role of  the  sound  in  signalling  the different  stages of  the process of washing or making  coffee.  Touch  is most  important  for  a  computer mouse  and  a  pen,  and probably  for any other hand  tools as well. Smell plays a dominant  role  for a de‐odorant and (together with taste) for food products.  

Culture may play  a  role  in determining what  sensory modality  is more  important. McLuhan  (1961) believed  that  the nature of media by which people  communicate affects the ratio of their senses. For example, the alphabet stresses the sense of sight, which in turn causes people to think in linear, objective terms. He argued that Euro‐peans  and North Americans  live  in  the  visual mode, while  for native Africans  and other non‐literate  societies  the  auditory modality  is dominant. Anthropological  re‐search demonstrates  that  there  is considerable  sensory diversity among non‐literal peoples.  For  example,  Classen  (2005)  describes  three  non‐literate  societies:  the Tzotzil of Mexico, the Ongee of the Little Andaman Island and the Desana of Colum‐bia. The Tzotzil accord primacy to heat  in their cosmology, the Ongee to odor, and the Desana to color. The sensory modality most symbolically elaborated by a culture is not necessarily the modality of most practical importance. All three cultures exam‐ined by Classen (2005) can be classified as oral cultures with regard to their dominant medium of  communication, but  the Tzotzil  symbolically orient  themselves by  tem‐perature, the Ongee by smell, and Desana can be regarded as visualists.  

 

1.2. THE PRESENT STUDY In the present study we address an aspect of sensory dominance that has not been investigated  since  Stratton’s  early  experiments,  that  is  how  sensory  dominance changes over time. We wonder which sensory modalities contribute most to differ‐ent stages of product experiences.  In addition, we  investigate whether any differ‐ences exists in modality importance for pleasant and unpleasant experiences. 

The  importance of the different sensory modalities may vary during different epi‐sodes of product usage. When consumers buy a product they are  likely to pay at‐tention primarily  to  its visual attributes,  simply because other options  to explore most products  in a store are  limited. But with time, other modalities can become 

Page 34: ANNA FENKO - TU Delft

CHAPTER 2 

 

34 

more  important. No matter  how  nice  new  shoes  look,  during  usage  it  becomes more  important whether  they  are  comfortable  or  not;  kitchen  tools  can  be  too heavy to use; an iron can produce a bad smell when used; new linen may not be as soft as the old, and so on. Furthermore, due to the emphasis of Western culture on the visual modality, designers may pay more attention to the visual properties of the products they create. This may result  in a more satisfactory visual product ex‐perience and more disappointing experiences in other modalities.  

We developed a questionnaire with open‐ended questions. We asked respondents to describe how  important they  found  the various sensory modalities  in different stages of the user‐product interaction. Because we wanted to know the context in which  the particular product experience had occurred, we  asked  respondents  to describe  the  situation  and  their experience  in  their own words. To  interpret  the data, we combined qualitative analysis with statistical methods.  

 

2. METHOD 

2.1. PARTICIPANTS 

The questionnaire was distributed among 243 Master students at the Department of Industrial Design, Delft University of Technology. All  respondents were between 22 and 32 years of age; 44% of the participants were women. 

 

2.2. PROCEDURE 

The questionnaire was part of a course assignment. At the beginning of this course each student selected a product that was analyzed repeatedly during the course for various assignments. Students were asked to select a product that they had used themselves and to which they had formed a positive or negative attitude. This pro‐vided a wide variety of different products, which allowed us to generalize our con‐clusions about the influence of sensory modalities on product experience. The par‐ticipants selected 93 different products (see Table 2.1). 

Respondents answered three open‐ended questions. First, they described the most 

pleasant/satisfactory sensory experience with  their product and assessed qualita‐

tively and comparatively  to what extent  the different sensory modalities contrib‐

uted  to  this experience. Second,  they described  the biggest disappointment with 

Page 35: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

35 

their product and assessed to what extent the different senses contributed to this 

experience. Third, they assessed what sensory modality was the most important for 

consumer experience with their product in the following situations: a) choosing the 

product  in the shop; b) during the  first week of usage; c) after the  first month of 

usage; d) after  the  first year of usage. We also asked participants  to explain why 

they thought the particular modality dominated, and, in case any change in impor‐

tance had occurred over time, why the dominant modality had changed. 

 

TABLE 2.1. PRODUCTS CHOSEN BY THE RESPONDENTS  

Categories  Products  

Electronics and electric appli‐ances 

MP3 player (13), camera (12), mobile phone (11), radio alarm clock (8), turn‐table (4), game computer (6), computer mouse (3), musical centre (2), printer (2), tape recorder (2), audio device (2), headphones (2), desk telephone (2), USB‐hub, external hard disc, microphone, pocket PC, laptop, TV 

Tools   Epilator (4), electric toothbrush (3), electric shaver (2), shaver (2), drawing tablet (2), hair dryer (2), flashlight (2), power drill (2), stapler, sewing machine, paintbrush, glue gun, sanding machine, stapler 

Home appli‐ances 

Coffeemaker (24), blender (4), water heater (3), fan (3), tea infuser (2), rice cooker (2), toaster (2), popcorn maker, refrigerator, BBQ, storage containers, saucepan, vacuum cleaner, hot water bottle, ice cream maker, juicer 

Personal acces‐sories 

Shoes (27), wristwatch (8), perfume (3), jacket (2), bag (2), backpack (2), suit‐case, eye glasses, lipstick, wallet, hairbrush, money box, pocket knife, book 

Sport equip‐ment 

Surfboard (2), helmet (2), sport wheelchair, snowboard, hockey stick, power ball, wetsuit, backgammon game, tent, kite, yoga mat 

Musical instru‐ments 

Guitar (4), synthesizer (2), digital piano, metronome 

Vehicles   Bicycle (5), car (2), scooter 

Furniture   Lamp (4), armchair (3) 

Fast moving goods  

Bottle of wine (2), candy, cigarettes 

Non‐consumer products 

Elevator, payment terminal 

* Frequency is given between parentheses if >1. 

Page 36: ANNA FENKO - TU Delft

CHAPTER 2 

 

36 

2.3. DATA ANALYSES 

The results were analyzed both qualitatively and quantitatively by content analysis 

methods usually employed  to analyze  free and  semi‐structured  interviews  (Bran‐

nen, 1992; Krippendorff, 1980). We used semantic categorization of the situations 

in which  pleasant  and  unpleasant  product  experiences  occurred.  The  process  of 

categorization  followed  the procedure of  inductive  category development, which 

aims to reduce the contextual diversity of the material (van Dijk and Kintsch, 1983). 

Tentative  categories were  deduced  step  by  step  from  the  respondents’  descrip‐

tions, and then revised (merging similar categories and renaming the merged cate‐

gories), eventually leading to 10 main categories. 

We also noted the  importance hierarchy of modalities mentioned by respondents 

in reply to the first two questions. We used these to derive a quantitative estimate 

of the importance of the modalities for pleasant and unpleasant experience and for 

various  stages  of  user‐product  interaction. When  only  one  modality  was  men‐

tioned,  it was given  the  rating 1 and all other modalities were given 0.  If several 

modalities were mentioned, they received ratings according to the priority given by 

the respondent, so that the sum of their ratings equaled 1 and the ratios between 

consecutive  ratings were  equal  (2:1).  For  example,  if  two modalities were men‐

tioned, they were given ratings 0.67 and 0.33; for three modalities the ratings were 

0.57, 0.29, 0.14, and so on. By keeping the sum of ratings for each participant equal 

to 1,  the  contribution of each participant  to  the overall  result  is kept equal. The 

ratio between consecutive ratings had to be logically larger than 1, but was chosen 

arbitrarily to be 2:1. By choosing 2:1 the resulting numbers clearly showed that one 

modality was more  important  than  the other one.  In addition,  it made  sure  that 

with 2 or 3 modalities mentioned, all modalities would still have a notable  impact 

on the overall results (all ratings >0.10). Furthermore, analyses performed in which 

only one dominant modality was  recorded per participant, yielded approximately 

similar outcomes.  

The ratings were subjected to repeated measures analysis of variance (see Labovitz 

(1970), who has demonstrated that ordinal variables can be treated as if they con‐

form to interval scales). All post hoc paired comparison tests were performed with 

Bonferroni adjustment. Participants’ explanations of  the change  in dominant mo‐

dality over time were used to illustrate and interpret the data.  

 

Page 37: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

37 

3. RESULTS 

3.1. PLEASANT VERSUS UNPLEASANT PRODUCT EXPERIENCE 

The semantic analysis of the situations in which the most pleasant and unpleasant product experiences occur (Table 2.2) showed that pleasant experiences were most often connected to the everyday usage of a product (making coffee, cooking, print‐ing). We put an experience  into this category when respondents highlighted func‐tional characteristics of a product  (easy  to use, good quality) and expressed  their satisfaction with  the  results of  the  interaction  (fast and efficient  shaving, making delicious rice). The second source of pleasure consisted of pleasant sensory experi‐ences (products look beautiful; have good smell, pleasant sound). About one sixth of the respondents mentioned the first  interaction with their product as the most pleasant and exciting experience.  

 

TABLE 2.2. SITUATIONS ASSOCIATED WITH PLEASANT AND UNPLEASANT 

PRODUCT EXPERIENCES 

Pleasant experience (N=205)  % 

Everyday usage  42.2 

Pleasant sensory experience  32.5 

Enjoying a new product   16.5 

Using on a special occasion   8.8 

Unpleasant experience (N=204)   

Unpleasant sensory experience   31.4 

Accidents causing pain or discomfort  23.0 

Bad design causing problems  22.6 

Difficult to use, special skills needed   11.0 

Damage after long use  8.1 

Disappointment with the first usage  3.9 

 

 

Page 38: ANNA FENKO - TU Delft

CHAPTER 2 

 

38 

The  unpleasant  product  experiences were most  often  connected  to  unpleasant sensations  (bad  smell,  annoying  noise).  The  sudden  dysfunction  of  a  product  or accidents  that happened during usage  (the engine broke down, users burnt  their fingers, or cut themselves with the razor) formed the second source of unpleasant experiences with products. The  third  reason  for disappointment was bad design, such as ‘the screwdriver  is too big for small spaces’, ‘it’s difficult to read the mes‐sage because the screen is too reflective’. 

In a repeated measures ANOVA with the  importance rating as dependent variable and Modality (vision, audition, touch, olfaction, and taste) and Affect (pleasant and unpleasant experiences) as within‐subjects factors, the effect of Modality was sig‐nificant [F (4, 776) = 92.52, p < 0.001, η2 = 0.32], while the Modality*Affect interac‐tion showed a tendency towards significance [F (4, 776) = 3.76, p = 0.054, η2= 0.02]. There was no Affect main effect because the sum of all ratings for both the pleas‐ant and unpleasant interactions equaled 1 for each participant. 

 

0

0,1

0,2

0,3

0,4

0,5

vision audition touch smell taste

impo

rtan

ce r

atin

g

pleasantunpleasant

 FIGURE 2.1. IMPORTANCE RATING (WITH STANDARD ERRORS) FOR SENSORY 

MODALITIES FOR PLEASANT AND UNPLEASANT EXPERIENCES 

 

Because the Modality*Affect interaction tended to be significant, we also investigated the relative importances for the sensory modalities for pleasant and unpleasant experi‐ences separately (Figure 2.1). For pleasant experiences, the effect of Modality was sig‐

Page 39: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

39 

nificant [F (4, 816) = 53.51, p < 0.001, η2 = 0.21]. Vision was mentioned most often as the dominant modality for pleasant experiences, followed by touch, audition, smell and taste.  Post  hoc  paired  comparison  tests  showed  no  significant  differences  between vision and touch (p>0.20). The other differences between modalities were all significant (p < 0.001). For unpleasant experiences the main effect of Modality was also significant [F (4, 812) = 45.63, p < 0.001, η2 = 0.18]. Touch was dominant for unpleasant product experiences, followed by vision, audition, smell and taste. 

 Post  hoc  paired  comparison  tests  significant  differences  between  all  modalities (p<0.001), except  for  the difference between vision and audition  (p>0.20). We also tested for each modality whether it obtained similar importance ratings for pleasant and unpleasant product experiences. These  two‐tailed paired  t‐tests  showed a  sig‐nificant difference for vision (p<0.05) and a tendency towards significance for touch (p=0.060), but no differences for the other modalities (p>0.15). 

 

3.2. CHANGES IN THE DOMINANT MODALITY OVER TIME  

To differentiate the relative modality importance over time, we performed an overall repeated measures ANOVA on  the modality  importance  ratings with Modality and Time as within‐subject  factors. The results showed significant effects of Modality [F (4, 844) = 137.1, p < 0.001,  η2 =0.39] and Time*Modality  [F  (12, 2532) = 69.8, p < 0.001, η2 = 0.25]. There was no Time main effect because the sum of all ratings for each  respondent,  in each period, equaled 1. Separate  repeated measures ANOVAs were performed to test the effect of Time for each modality and to test the differ‐ences between Modalities at each moment of time (see Figure 2.2). 

Analyses for the four different time episodes showed that at the time of buying the effect of modality was significant [F (4, 844) = 525.5, p < 0.001, η2 = 0.71]. The im‐portance of vision was higher than the importance of all the other modalities (p < 0.001). Touch occupied the second position and was significantly different from all the others (p<0.001). Audition, smell and taste played a very small role at the time of buying, and the differences between them were not significant (p>0.20).  

Since at the buying stage the possibility to interact with the products is mostly lim‐ited to the visual modality, we performed an additional analysis  looking at the ef‐fect of  time during only  the  three  in‐use  stages: one week, one month  and one year. This analysis confirmed the overall analysis with significant effects of Modality [F  (4, 844) = 71.3, p < 0.001, η2 = 0.25] and  the Time*Modality  interaction  [F  (8, 1688) = 7.5, p < 0.001, η2 = 0.04].  

Page 40: ANNA FENKO - TU Delft

CHAPTER 2 

 

40 

0

0,2

0,4

0,6

0,8

1

buying 1 week 1 month 1 year

Time

Impo

rtanc

e ra

ting

visionauditiontoucholfactiontaste

 

FIGURE 2.2. CHANGING DOMINANT MODALITY OVER TIME 

 

After  the  first week of usage  the modalities were still significantly different  [F  (4, 844)  =  60.8,  p  <  0.001,  η2  =  0.22].  Vision  and  touch  became  equally  important (p>0.20).  They  showed  the highest  level of  importance  and differed  significantly from  all  the  other modalities  (p<0.001).  Audition  occupied  the  second  position (p<0.001); smell and taste were the  least  important at this stage. The differences between  smell  and  taste were  not  significant  (p<0.05).  After  the  first month  of product usage  the  importance of  the modalities  remained different  [F  (4, 844) = 41.4, p < 0.001,  η2 = 0.16]. The differences between  touch and audition and be‐tween audition and vision were no  longer significant (p>0.20). On the other hand, the  difference  between  touch  and  vision  became  significant  (p<0.05).  Smell  and taste rated significantly less important than the first three modalities (p<0.001), but were not significantly different from each other (p>0.05). After the first year of us‐age  the effect of modality was  still  significant  [F  (4, 844) = 28.5, p < 0.001,  η2 = 0.12], but the differences between vision, audition and touch were no  longer sig‐nificant (p>0.20). Smell rated second, and taste was the least important; both mo‐dalities differed significantly from the rest and from each other (p<0.05). 

Analyses  for each of  the modalities separately showed  that  the  importance of vi‐sion changed significantly over time [F (3, 633) = 149.2, p < 0.001, η2 = 0.41]. It de‐creased from the buying stage to the first week (p<0.001) and from the first week 

Page 41: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

41 

to  the  first month of usage  (p<0.05). The  importance of vision  increased  slightly, but  significantly,  during  the  first  year  (p<0.05).  The  importance  of  touch  also changed significantly [F (3, 633) = 31.8, p < 0.001, η2 = 0.13]. It increased (p<0.001) after  buying,  remained  constant  during  the  first month  (p>0.20),  and  decreased slightly, but significantly, during the first year (p<0.05). The importance of audition also changed significantly over time [F  (3, 633) = 40.8, p < 0.001, η2 = 0.16].  It  in‐creased from the buying stage to the first week (p<0.001) and from the first week to the first month (p<0.001). After that, the importance of audition did not change anymore (p>0.20). The importance of smell also changed significantly [F (3, 633) = 7.73, p < 0.001, η2 = 0.03].  It  increased  (p<0.05) after buying and during the  first week. Then it remained constant (p>0.20). The time variations in the importance of taste were not significant during  the whole period of usage  [F  (3, 633) = 5.7, p > 0.05, η2 = 0.02].  

The qualitative analysis of the students’ answers helps to clarify these results. For the majority of the respondents, vision is the most important modality for choosing a  product  in  a  shop.  Participants  note  that  usually  the  only way  to  explore  the product before buying is simply by looking at it. Some respondents, however, men‐tioned  the  importance of  touch and audition during buying, but commented  that trying a product in a shop is time consuming and sometimes unpractical: “It is em‐barrassing to play all possible instruments inside a shop to hear which one sounds best.” 

After the first week of usage, the importance of vision  is lower than at the buying stage.  Respondents  indicated  that  “It  is  important  that  it  looks  nice  but  how  it works gets more important [sewing machine],” and “Consumers are getting used to or bored by the  looks of the product [alarm clock].” At the same time the  impor‐tance of touch and sound  increase, because they are  important for everyday use: “The touch is most important. When it is uncomfortable to wear, you will not use it [backpack]”.  Sound becomes particularly  important  for products with  an  electric motor: “The sound of the product is really annoying [printer].” 

Many respondents noted that after the product has been acquired, the dominant modality depends on its primary function: “The sound is dominant throughout the life of the product, because the primary function of the guitar is to produce a beau‐tiful sound.” “Touch dominates for a hockey stick because it affects its playing abili‐ties.” “The most important is the smell and taste of the coffee because that is what a coffeemaker is for.” 

Page 42: ANNA FENKO - TU Delft

CHAPTER 2 

 

42 

The  importance  of  the  visual  experience  increases  again  during  the  first  year  of usage, because the product becomes old, dirty, and scratchy: “After one year the iPod looks used, there are scratches visible on the surface.” “The visual aspect be‐comes more important again when the boots get a little damaged.” Another reason is  that after one year  fashion  changes may occur and  the product becomes out‐dated (lipstick, backpack, water heater). The user faces the choice whether to con‐tinue using a product or  to get  rid of  it: “After a  long  time  it becomes more and more important if the device still appeals to you [drawing tablet].” 

 

3.3. SEPARATE PRODUCT CATEGORIES  

A wide  range of products was  included  in our analysis  (93 different products)  to enhance the generalizability of the results. Nevertheless, to check whether our pat‐tern of  results also applies  to various product categories we performed  separate analyses on three homogeneous product groups that occurred relatively frequently in our sample. The first group consisted of hi‐tech devices with a visual display and main auditory function: mobile phones, mp3‐players and radio alarm clocks (N=32). The  second  group  consisted of  shoes  (N=27), while  the  third  group  consisted of electric coffee makers (N=24). The patterns of changes in the importance ratings in modalities over time for these three product groups are shown in Figure 2.3.  

For all three product categories, the pattern of results bears similarities to the ag‐

gregate  results shown  in Figure 2.2. Most  importantly,  for all  three categories vi‐

sion dominates the buying stage. After one week other modalities come into play. 

The exact modalities differ between products.  

For the hi‐tech products, both touch and audition are  important throughout usage. 

The importance of touch increases during the first week, and then remains constant. 

The importance of audition increases during the first week and the first month, when 

it becomes the dominant modality. For shoes, the tactual modality  immediately be‐

comes  important, and after one month  the  importance of olfaction appears  to  in‐

crease. After one year vision and touch become equally  important, and olfaction  is 

the  second  important modality. For  the  coffee maker  touch  is  relatively  important 

after one week, but  audition  and,  to  a  lesser degree, olfaction become  important 

after one month. After one year of usage audition remains dominant.  

 

Page 43: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

43 

Hi-tech products

0

0,2

0,4

0,6

0,8

1

bying 1 week 1 month 1 year

time

mod

ality

impo

rtan

ce

visionauditiontoucholfactiontaste

Shoes

0

0,2

0,4

0,6

0,8

1

bying 1 week 1 month 1 year

time

mod

ality

impo

rtanc

e

visionauditiontoucholfactiontaste

Coffee maker

0

0,2

0,4

0,6

0,8

1

bying 1 week 1 month 1 year

time

mod

ality

impo

rtanc

e

visionauditiontoucholfactiontaste

 

FIGURE 2.3. CHANGING DOMINANT MODALITY OVER TIME FOR HI‐TECH PRODUCTS, SHOES AND COFFEE MAKERS 

Page 44: ANNA FENKO - TU Delft

CHAPTER 2 

 

44 

4. DISCUSSION 

4.1. MAJOR FINDINGS  

Our results show that vision and touch are the modalities that are most often  in‐volved  in positive product experiences. For negative experiences, our participants reported significantly more events involving the tactile modality than the visual or auditory modalities. This might suggest that designers tend to pay more attention to creating beautiful visual properties of products,  rather  than  to  improving  their tactile properties. 

Furthermore, we found that the dominant sensory modality depends on the period of product usage  and  the  type of product. When buying  a product,  vision  is  the most important modality, but after the first week of usage the importance of vision decreases and the other modalities become more important. In general, for the 93 products  analyzed  in  our  study,  vision  is  not  dominant  during  usage.  After  one month of usage  touch becomes dominant,  and  after one  year  vision,  touch  and audition become equally  important. The  analyses of  separate product  categories indicate  that  the  importance  of  the  various  sensory  modalities  is  product‐dependent: after one year of usage audition  is dominant for hi‐tech products and coffee makers, while touch and vision are equally important for shoes.  

The changes  in modality  importance can be explained by changes  in the product‐user  interaction.  In a  shop  the  interaction with  the product  is mostly  visual. But most products are bought for purposes other than visual enjoyment: they are used to cook, print, make coffee, listen to music, and so on. During usage the dominant sensory modality mainly seems to depend on the primary product function: touch for hand  tools,  sound  for an alarm‐clock,  smell and  taste  for  food. However,  the dynamics of sensory dominance can also depend on specific product features that are non‐functional, such as the noise of an electric motor. In addition, the specific characteristics of the user‐product  interaction affect modality  importance. For ex‐ample, wearing shoes for a long time makes them more comfortable for touch but less  pleasant  visually  (as  they  accumulate  dirt  and  scratches)  and  olfactorily  (as they may start to smell bad). 

Our data  suggest  that modality dominance  can depend on  the different  roles 

sensory modalities  play  in  the  everyday  life.  Vision  can  be  important  in  the 

situation  of  buying, when  people  have  to  compare multiple  slightly  different 

products to make an optimal decision, because it gathers the largest amount of 

Page 45: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

45 

information on a product within the shortest time frame (Schifferstein and Clei‐

ren,  2005;  Schifferstein  and  Desmet,  2007).  When  the  user  becomes  ac‐

quainted with  the product,  the need  for other  types of  information  increases, 

such as the motor skills required to operate a product, which increases the im‐

portance  of  touch.  At  the  same  time,  the  emotional  components  of  product 

experience become more  important. Several studies have pointed out the role 

of  audition  and  olfaction  in  emotional  experience  (Hinton  and Henley,  1993; 

Herz, 1998; Schifferstein and Desmet, 2007). For example, after users gain ex‐

pertise with their electric tools, they do not pay attention to their visual attrib‐

utes anymore, but start to notice the sound of the motor, which they often de‐

scribe as  irritating and annoying. Furthermore, even  if a coffee machine makes 

a  loud noise most users will describe  it as pleasant, because  in  their memory 

this sound is associated with the pleasant smell and taste of fresh coffee.  

In  an  ethnographic  study,  Karapanos, Hassenzahl  and Martens  (2008)  suggested 

that  changes  in  user  experience  over  time were  related  to  changes  in  product 

meaning. While early experiences  seemed  to  relate mostly  to hedonic aspects of 

product use, prolonged experiences became  increasingly more tied to aspects re‐

flecting how the product becomes meaningful  in one’s  life. These conclusions are 

quite similar to those obtained  in a study of consumer‐product attachment. Schif‐

ferstein  and  Zwartkruis‐Pelgrim  (2008)  found  that  enjoyment  is one of  the main 

determinants of attachment for recently acquired products (<1 year), while memo‐

ries contributed mainly to the attachment for products owned for a long time (>20 

years). If using a product is pleasurable and evokes positive emotions, the chances 

are high that a user becomes attached to the product (see Mugge et al., 2004). Be‐

cause the enjoyment a product evokes is affected to a large degree by the sensory 

impressions  it creates,  sensory modalities can contribute  to  the  formation of  the 

bond with the product. Furthermore,  it was suggested earlier that touch plays an 

important  role  in experiencing a product as  familiar and  somebody’s own: when 

tactual perception  is blocked, familiar products feel strange and people feel alien‐

ated from their surroundings (Schifferstein and Desmet, 2007). Therefore, we can 

assume  that  tactile experience plays a particularly  important  role  in  the develop‐

ment of product attachment. A good illustration is shoes: some respondents in our 

study admit that they postpone buying new shoes even if their old shoes look worn 

out, because they just love wearing their old shoes.  

 

Page 46: ANNA FENKO - TU Delft

CHAPTER 2 

 

46 

4.2. METHODOLOGICAL LIMITATIONS 

There are some limitations to our data because of the method we used. One of the 

disadvantages of questionnaire research is that respondents describe their experi‐

ences on the basis of their memory. The actual experiences during product usage 

could be different from what they remember. To avoid this downside, longitudinal 

ethnographic  studies  can be performed  to  investigate  the  real‐time dynamics of 

product experience using observations,  in‐depth  interviews, and diaries. Unfortu‐

nately, such studies are restricted to a small amount of products and a small num‐

ber of  respondents  (Karapanos, Hassenzahl and Martens, 2008). The current me‐

thod has  the advantage  that  it allows collecting a wide  range of  information and 

generalizing over products. 

Another problem  in questionnaire studies can arise because experiences of differ‐ent  sensory modalities may  have  different  degrees  of  awareness.  For  example, most respondents found it difficult to describe their olfactory experiences. The fact that  linguistic  categories  of  visual  experiences  are much  richer  and more  elabo‐rated in Western languages could explain why descriptions of other sensory experi‐ences are less common (Hinton and Henley, 1993).  

Furthermore, it can be argued that to use industrial design students as participants in  research  on  product  experience  is  inappropriate,  as  it  is  inappropriate  to use clinical  psychology  students  to  validate  personality  tests. Well‐informed  subjects can  be  inappropriate  indeed  if  experimenters  conceal  the  real  purposes  of  the study. However, our questions were direct and  they did neither  require nor pro‐hibit  any  special  knowledge.  In  fact, we  think  that using design  students was  an advantage, because  they may have better awareness of  their own  consumer ex‐perience than other user groups.  

Also,  the products  in  the sample may have affected  the outcomes of  the current study. As shown  in Figure 3 the dynamics of modality  importance  in user‐product interaction  is  clearly  product‐specific.  The  analyses  of  pleasant  and  unpleasant product experiences suggest that touch, vision, and audition are about equally im‐portant both for pleasant and unpleasant product experiences during usage, while smell and taste play a smaller role. However, this outcome may be due to the small amount of food products in our sample. We plan to gather more data for different product categories and consumer groups to validate the current outcomes. 

Finally, the way we derived  importance rating for the different modalities might be improved in future questionnaire research by using more common rating scales, such 

Page 47: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

47 

as 5‐point  Likert‐type  items  (from  “not  important at all”  to  “very  important”). The advantage of using rating scales would be that we have one explicit rating for each modality. In that case, we would not have to make so many assumptions in the analy‐sis. However, whether individual Likert items can be considered as interval‐level data, or whether they should be considered merely ordered‐categorical data is the subject of disagreement (Meyers et al., 2005). Direct‐rating methods were also criticised for the  lack of nomological  validity  (Harte and Koele, 1995).  In addition,  Jaccard et al. (1986) found low correlations between the importance of attributes measured using the free‐elicitation method and the importance of attributes measured using the di‐rect‐rating method. Because of the lack of convergent validity among different meth‐ods of attribute‐importance measurement  (Van  Ittersum et al., 2007), we are  cur‐rently also exploring methods to measure the degree of sensory dominance through experimental manipulations (Fenko et al., 2009).  

 

4.3. PRACTICAL IMPLICATIONS 

Although it is difficult to formulate practical implications for retailers at this stage of the research, the obvious conclusion is that the presentation of the majority of prod‐ucts in shops is inadequate. “Why decide on vision while the most important modal‐ity  is  touch?”  asked  one  of  our  respondents  describing  the  process  of  choosing  a shaver in a store. To avoid consumers’ disappointment, retailers should think of ways to  demonstrate  the  non‐visual  properties  of  products  at  the  buying  stage  (how  a computer mouse feels, what kind of noise a coffee maker makes, and so on).  

 

5. CONCLUSIONS Information  on  sensory  dominance  is  important  for  product  design.  If  designers know which sensory modality dominates the experience of the particular product, they can concentrate on creating appropriate sensory attributes for it. Our research suggests  that  the dominant  sensory modality  depends  on  the period of  product usage. At  the  time of buying  vision  is  the most  important modality, but  at  later stages touch and audition become equally important, followed by smell and taste. Which modality will dominate at the later stages of product usage depends on the primary function of a product and on the characteristics of the user‐product inter‐action. Our  findings  imply  that  stimuli of any  sensory modality  can provide both 

Page 48: ANNA FENKO - TU Delft

CHAPTER 2 

 

48 

pleasant  and  unpleasant  product  experiences.  To  create  a  rich  and  long‐lasting product experience, it is important to consider user‐product interaction at different stages of product usage, and  to determine which sensory experience  is more  im‐portant for consumers at each stage of usage. 

 

 

 

  

 

 

 

 

 

 

Page 49: ANNA FENKO - TU Delft

49 

 

 

 

 

 

 

 

 

CHAPTER 3 

THE ROLE OF SENSORY MODALITIES IN THE 

DYNAMICS OF NATURALNESS OF A 

DEHYDRATED FOOD PRODUCT 

1. INTRODUCTION During user‐product interaction, different types of experiences can be evoked con‐secutively. At different stages of the user‐product  interaction, various senses may be used and different emotions may be elicited. Product experience  includes  the perception  of  sensory  attributes  (such  as  taste,  smell,  temperature,  etc.),  the meaning attached to products (such as elegant, modern, feminine), and emotional reactions  to products  (such as attraction, surprise, or boredom)  (Hekkert & Schif‐ferstein, 2008).  

In order to provide a more comprehensive  insight  into dynamic changes  in user ex‐perience, we used an interaction with a food product for an experimental study. Food products  are  unique  among  industrial  products,  because  sensory  experience with these products involves all five senses: vision, audition, touch, smell, and taste. That is why  they provide a good example of  the dynamics of sensory dominance during the  various  stages of product usage. We  selected  a dehydrated  food product, be‐cause it  is a packaged product, and the package may play an important role in food 

Page 50: ANNA FENKO - TU Delft

CHAPTER 3 

 

50 

experience. Unlike other packaged products, such as snacks or soft drinks, a dehy‐drated  food product requires preparation before consumption, which makes an  in‐teraction with this product more complex and more interesting to investigate.  

The aim of  this  research  is  to  identify  the  temporal  changes during user‐product interaction with a dehydrated food product. The study aims to highlight the domi‐nant  sensory  dimensions  at  several  stages  of  product  usage:  when  choosing  a product in a supermarket, during package opening, preparation, consumption, and re‐purchase. We are specifically interested in the links between sensory properties of a product, emotional reactions to them, and how they interact to create the ex‐perience of naturalness. 

 

1.1. SENSORY DOMINANCE 

In  the area of product design,  sensory dominance  can be defined as  the  relative importance of different sensory modalities for product experience. Product attrib‐utes may be perceived through various sensory modalities. Nevertheless, it is often assumed that vision dominates the other senses. In a previous study (Fenko, Schif‐ferstein, & Hekkert, 2009), we asked 243 participants to describe their experiences with consumer products in various situations: while buying a product, after the first week, the first month, and the first year of usage. The data suggest that the domi‐nant sensory modality depends on the period of product usage. At the moment of buying, vision is the most important modality, but during the usage the other sen‐sory modalities gain importance. The roles of the different modalities during usage are  product‐dependent. Averaged  over  93  products  analyzed  in  this  study,  after one month of usage touch became more important than vision, and after one year vision,  touch  and  audition  appeared  to  be  equally  important.  To  create  a  long‐lasting positive product experience, designers need to consider user‐product inter‐action at different stages of product usage and to evaluate which roles the sensory modalities play at each stage. 

In the area of food research, the dominant sensation is defined as the most intense 

sensation  (Labbe  et  al.,  2009).  The  evaluation  of  the  temporality  of  the  sensory 

perception in food products is mainly assessed using the time–intensity (TI) meth‐

odology. The  time  intensity method  (TI) was developed  to  take  into account  the 

dynamics  of  perception  (Larson‐  Powers &  Pangborn,  1978).  The  time  intensity 

method consists of recording the evolution of the  intensity of a given sensory at‐

tribute over time using a cursor connected to a recorder. For each evaluation of an 

Page 51: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

51 

attribute,  the panelist  is asked  to move  the cursor along  the sensation scale. The 

panelist moves  the  cursor  up  (or  to  the  right) when  the  perceived  intensity  in‐

creases and down (or to the left) when it drops off.  Time intensity scaling is a well 

known method to record the evolution of one specific attribute over time. This ap‐

proach is useful for studying the temporal aspects of the perception of a given sen‐

sory  attribute  in  a  product.  However,  it  is  still  not  broadly  used  as  a  routine 

method, mainly because it is a time consuming and expensive tool (Le Reverend et 

al., 2008), because one run is needed per attribute (Pineau et al., 2009). 

The  temporal  dominance  of  sensation method  (TDS)  has  been  developed  to 

provide dynamic  information on products  in a quicker way compared to TI and 

to gain additional  information regarding the sequence of sensations (Pineau et 

al., 2003). This descriptive sensory method makes  it possible to collect tempo‐

ral data during one  single evaluation  for up  to 10 attributes on  complex  food 

products. The method consists  in presenting the entire set of attributes to the 

panelist on the computer screen. Along the tasting of one product, the panelist 

has to  indicate what the dominant sensory perception  is, and to score the cor‐

responding attribute. Each  time  the panelist  thinks  the dominant sensory per‐

ception has changed, either  in  intensity or  in quality, he/she has  to  score  the 

new  perception.  For  each  run,  this method  enables  to  collect  a  sequence  of 

sensory attributes (and their respective scoring) quoted at different times along 

the tasting. TDS and TI results exhibit similar patterns, but TDS is found to bet‐

ter discriminate the sequence of the sensations over time (Pineau et al., 2009). 

TDS provides information on the dynamic of perception after product consump‐

tion that is not available using a conventional profiling method and that may be 

critical for the understanding of complex perceptions such as refreshing (Labbe 

et al., 2009). TDS could be used to illustrate product perception as a function of 

time, whereas TI would be better suited if the determination of the dynamics of 

one specific attribute is required (Le Reverend et al., 2008). 

Unlike the TI and TDS methods, in the current study we do not focus on the sepa‐

rate taste sensations during product use. Instead, we ask participants to assess the 

contributions of  the various  sensory modalities  to  the entire product experience. 

Also, we do not register these responses continuously, but we assess them at spe‐

cific stages of the user‐product interaction.  

 

Page 52: ANNA FENKO - TU Delft

CHAPTER 3 

 

52 

1.2. NATURALNESS 

The concept “natural”  is frequently encountered  in modern Western  life, typically in a very positive context. For example,  in five European countries and the United States, free associations to the word “natural” are almost entirely positive. Humans may have an  innate desire (“biophilia”) for the experience of their ancestral envi‐ronment  (Wilson,  1984;  Kellert &Wilson, 1993). Rozin  and  colleagues  (2004) de‐scribed  this phenomenon as “natural preference” and documented  it  for  the do‐main of  foods. These authors  found  that  the great majority of  respondents pref‐ered a natural product to the corresponding commercial product. Natural products are often thought to be healthier, more appealing to the senses, or kinder to the environment  than  products  that  are  not  natural.  In  addition,  the  natural  is  pre‐ferred  just because  it  is  inherently better  ‐ more moral, more aesthetic, or simply “right.” This preference does not shift  to  indifference even when  the natural and commercial products are chemically identical (Rozin et al., 2004).  

Chemical manipulations  ‐  such  as  boiling, minor  subtractions  of  natural  compo‐nents or minor additions of natural or unnatural entities, and genetic engineering ‐ produce major reductions  in naturalness (Rozin, 2005).  In contrast, physical trans‐formations have minor effects. Adding small amounts of unnatural substances re‐duces naturalness substantially (Nemeroff & Rozin, 1994, 2000; Rozin & Nemeroff, 1990). The decrease  in naturalness from adding a small amount of a natural addi‐tive  is  almost  the  same  as  the  decrease  in  naturalness  from  adding  twice  the amount  (the dose‐insensitivity  feature of contagion). The contaminating effect of non‐natural  traces  in  “natural”  substances  is  found not only  for  typical negative contaminants (e.g., a cockroach, a toxin), but even for innocuous, often‐consumed unnatural entities, like minerals. 

The concept of naturalness is widely used by producers and consumers of so called organic food  in contrast to the unnaturalness of products of conventional agricul‐ture.  In qualitative  interviews with  consumers of organic products,  the  following aspects of natural were distinguished (Verhoog et al., 2003):  

The concept of nature has an emotional meaning related to peacefulness, silence, freedom, becoming your own self, and holidays. 

Naturalness  in general  is associated with  simple, pure, non‐artificial, not spoilt, and fair. 

The amount and way of processing food products  influences the natural‐ness. Less processed food, or food processed in a traditional way without additives is associated with more natural food. 

Page 53: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

53 

The more artificial the food production and process, the  less natural  it  is. Genetic engineering is from that point of view very unnatural. 

Production related to season and region, amount of energy input, but also the kind of package material, affects the (perceived) naturalness of food.  

While  ‘natural’  food  is considered healthy by most of  the population  in the West (Rozin, 2005), from a chemical point of view ‘natural’ food additives are not much safer  than  artificial.  Some experts even define  ‘natural’  ingredients  as  ‘the  same ingredients produced by outdated technology’ (Schlosser, 2001).  

A complicating factor in understanding naturalness is the fact that most consumers 

have little knowledge of food production in general. Genetically modified products 

represent the most striking example of the clash between techno‐science and pub‐

lic discourse on  ‘healthy’ and  ‘natural’  food. For  instance, consumers not only re‐

ject genetically engineered  food, but are equally  likely  to  reject  ‘mutated’  foods, 

even if the mutation occurs (as it does in nearly every food product) through cross‐

breeding, hybridization, and other conventional methods. In a survey of consumer 

attitudes  (Hamstra,  1998), only 28% of  respondents  in New  Jersey  thought  they 

had ever eaten a hybrid fruit or vegetable that was the product of traditional cros‐

sbreeding. Moreover, 40% did not approve of making hybrid plants. The consensus 

view held it was best “not to meddle with nature”. 

 

1.3. FOOD AND EMOTIONS 

Our relationship with the world is inherently emotional. Emotions evoked by prod‐ucts enhance  the pleasure of buying, owning, and using  them  (Hirschman & Hol‐brook, 1982).  It has often been argued  that  the emotional quality of products  is becoming more and more important for differential advantage in the marketplace, because products are now often similar with respect to quality and price. In some purchase decisions, especially in food domain, emotional responses may be a deci‐sive factor (e.g., Desmet and Schifferstein, 2008).  

Eating is an activity laden with affect. Almost all potential foods are either liked or disliked; people are  rarely neutral  in  their  judgments about  food products. Some food  likes and dislikes  seem  to be  culture‐wide, while others  vary within  culture (Rozin, 1990). Although cultural forces play a major role  in the acquisition of food likes and dislikes, most people have substantial choice in foods, over and above the “choices” made by  their  culture  (Warde, 1997). As a  result, most people  choose foods  they  like, and hence eating  is,  for  the most part, a positive experience. Al‐

Page 54: ANNA FENKO - TU Delft

CHAPTER 3 

 

54 

though  some  foods  are  consumed primarily out of necessity or  for  instrumental reasons  (gaining  nutrition,  losing weight,  or  being  a member  of  the  group),  the principal basis for food choice is liking for the flavor (Roos and Wandel, 2005). 

It has been suggested that hedonic value is a good predictor of both the amount of food  consumed  and  food  preferences  (McCrory  et  al.,  2006).  Empirical  studies show  that  food  choices directly depend on  sensory  liking  (Eertmans et al., 2001) and that sensory aspects of food products (taste, smell) are the determining factors in food choice and preference (Mela, 2006). According to Birch (1999), food prefer‐ences are established early in infancy based on primary sensory reactions, probably related  to  the  survival of  the  species.  Indeed, babies prefer  some  tastes  such as sweet and reject others such as sour and bitter.  

Food may be both a pleasure and a poison. Negative emotions associated with spe‐cific foods help us to avoid the dangers of eating poisonous food. For  instance, the cultural evolution of disgust  (Rozin et al., 1997)  suggests  that  this emotion derives from the food rejection system of mammals as a response to a bad smell and taste of food (which is usually associated with harmful effects). Food‐related disgust seems to develop somewhere between the ages of 4 and 7 years (Rozin, 1999). 

In  the developed world, where  industrially produced  food has become abundant and excessive, eating is often associated with fear and guilt (Kass, 1994). In the late 20th century, science became obsessed with health promotion, and the dominant scientific discourse focused on the concepts of healthy eating, healthy living, exer‐cising, and environmental pollution  (Turner, 1984).  Information about  the health effects of different patterns of eating and different foods has become widely avail‐able  through  the media  (Kalucy,  1987).  This  has  led  to  frequent  new  concerns about particular dietary  items,  and  resulted  in  the  rise of  confusion  and  anxiety about  food.  For  instance,  in  a  survey of American  college  students, over 10% of women claim that they would be embarrassed to buy a chocolate bar in the store, and about 30% say they would be willing to opt for a nutrient pill, safe, nutritionally complete, and cheap, as a substitute for eating (Rozin, 1999). 

Misinterpretations  of  medical  studies  and  dietician’s  recommendations  lead  to popular beliefs  that certain  food products are either  ‘good’ or  ‘bad’. The  level of intake drops out of the equation. Thus, a lot of people think of fat and salt as tox‐ins: even a trace of each in food is considered unhealthy (Rozin, Ashmore & Mark‐with, 1996). As a  result, consumption of even  small amounts of  fat and  salt may evoke strong emotions of fear and guilt.  

The influence of emotions on food behavior and its health consequences can be illus‐trated by the phenomenon to which European health experts refer as the ‘American paradox’  (Heini & Weinsier, 1997; Astrup, 1998), while  their  colleagues  across  the 

Page 55: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

55 

Atlantic  call  it  the  ‘French paradox’  (Renaud, &  Logeril, 1992).  The American  food market is flooded with fat‐ and sugar‐reduced products, but obesity in the American population is the highest in the world. On the other hand, the French eat a higher fat diet than Americans, have higher  levels of blood cholesterol, and yet have a cardio‐vascular disease rate about one‐third less than Americans. Medical researchers tend to  explain  this  paradox  in  dietary  terms,  such  as  a  protective  effect  of  red wine (Renaud &  Logeril, 1992). An  alternative explanation  is  that,  in  comparison  to  the food‐poison attitude of Americans associated with high stress and anxiety, the French demonstrate a food‐pleasure attitude, which is associated with the lower stress and more positive emotions with respect to food (Rozin, 1999).  

 

1.4. MEASURING EMOTIONS 

The quest  for  instruments  to measure emotions has a  long history. Traditionally, attempts to measure emotions have been done in the field of psychology and soci‐ology. More recently, acknowledging the  important role of emotions  in their field of research, consumer and marketing researchers have developed instruments that measure  consumer  experiences  and  the  emotional  responses  to  advertisement. Both the number of reported instruments and the diversity in approaches to meas‐ure emotions  is abundant. Today’s  instruments range  from simple pen‐and‐paper rating  scales  to dazzling high‐tech  equipment  that measures brain waves or  eye movements. 

Most research on emotions involves some form of self‐report, asking participants in a study to report their feelings in a more or less formal way, from open‐ended ver‐balizations  to  psychometric  scales  and  questionnaires  (Russell,  2003;  Scherer, 2005). One disadvantage of these tools  is their  length, as a  lengthy questionnaire can be less acceptable for the research participants and be too taxing to be admin‐istered  repetitively. When  considering  self‐report  tools  for  the measurement  of emotion, researchers face a trade‐off between the length of the scale and the qual‐ity and richness of the measurement (Laurans & Desmet, 2008). 

Physiological activation  is an  important aspect of neuroscience  research on emo‐tions. A wealth of empirical  research describes  the correlation between different physiological systems and emotion or  related phenomena  (Cacioppo et al, 1993).  Psychophysiological techniques in user experience research have some advantages: they do not demand the users’ attention and avoid some of the biases influencing self‐reports, such as a priming effect of the questions, interference of a self‐report with  the experience, or a  tendency  to give  socially approved answers  (Nisbett & 

Page 56: ANNA FENKO - TU Delft

CHAPTER 3 

 

56 

Wilson, 1977). The disadvantages of these tools  is the need for specific expertise, the  complex  equipment  needed,  and  the  relative  obtrusiveness  of  the  sensors (usually adhesive electrodes attached to the skin with cables running to the ampli‐fication and recording device), reducing  the  freedom of movement of  the partici‐pants (Laurans et al., 2009). Another difficulty of psychophysiological measurement lies  in  its  interpretation. Emotion  is not the only process to affect visceral activity, and  research  failed  to uncover unambiguous  associations between  specific emo‐tions and patterns of bodily activation  (Stemmler, 2003). Therefore,  it  seems ex‐tremely difficult to use psychophysiological recording on  its own to provide an as‐sessment of users’ emotions while interacting with a product.  

Another  class  of  tools  are  the  graphical  self‐report  instruments.  Like  question‐naires, they are based on a person’s own ratings of his or her feelings, but pictures (typically smiling or frowning faces) are used instead of words to anchor the scales. While this technique forgoes the need for translation and avoids a strong tie with a particular language, a graphical questionnaire still needs to be tested with different users as the  interpretation and relevance of a graphical representation of feelings or dimensions of emotional experience might differ between cultures. 

In the area of product experience, the Product Emotion Measurement  instrument (PrEmo) was proposed to assess emotional responses to consumer products (Des‐met,  2003).  The  unique  strength  of  PrEmo  is  that  it  combines  two  qualities:  it measures distinct emotions and it can be used cross‐culturally because it does not ask respondents to verbalize their emotions. In addition, it can be used to measure more  than one emotion experienced  simultaneously,  and  the operation  requires neither  expensive  equipment  nor  technical  expertise.  The  labels  associated with each emotion have been checked with participants in the Netherlands, the US, Ja‐pan and Finland (Desmet, 2002). Considering its flexibility, ease of use and the rich‐ness of  information  it provides, we decided  to use PrEmo as  the most adequate method for measuring emotions at different stages of user‐product interaction.  

 

1.5. THE PRESENT STUDY  

In the current research, we investigate which sensory modalities are important for the experience of a dehydrated food product at different stages of product usage: choosing a product on a supermarket shelf, opening a package, cooking and eating the  food. We are  interested  in  three components of product experience: sensory perception, naturalness, and emotional responses. We developed a questionnaire that is used repeatedly throughout the 5 stages of the experiment. The data were 

Page 57: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

57 

analyzed both qualitatively and quantitatively to determine the changes in product experience at different stages of product usage. 

 

2. METHODS 

2.1. PARTICIPANTS 

The participants were recruited from the consumer panel of the TU Delft that con‐sists of a representative sample of 2000 people (1300 households) from Delft and nearby areas. All participants were occasional consumers of dehydrated food and did  not  suffer  from  any  food  allergies.  47  people  participated  (20 men  and  27 women). Ages ranged from 22 to 61 years, mean age was 39 years. Each consumer evaluated only 1 product. All participants were Dutch‐speaking, and all instructions and questionnaires were presented  in Dutch. Respondents were rewarded with a financial compensation. 

 

2.2. MATERIALS  

Two  equivalent  commercial  brands  of  dehydrated,  vegetable‐based  products  were used. These two products were used in all 5 stages of the experiment. Also present on the shelves were 12 equivalent products, based on the same type of vegetable. 8 prod‐ucts came from the 1st brand, 3 from the 2nd brand, and 1 product from a competing brand. There were also 12 products present on the shelves that were based on differ‐ent types of vegetables. All texts on the packages were in German, and the participants were informed that the products they were going to test were not present yet on the Dutch market.  

 

2.3. PROCEDURE 

The experiment consisted of 5 stages: (1) choosing the product in the supermarket; (2)  opening  the  package;  (3)  preparing  the  food;  (4)  eating  the  food;  (5)  re‐purchasing. At each of the 5 stages we asked the participants to evaluate their ex‐perience with the product in three ways: 

Page 58: ANNA FENKO - TU Delft

CHAPTER 3 

 

58 

1. Sensory  questionnaire.  For  each  of  the  five  sensory modalities,  indicate how important it is in this stage (5‐point scale, from “not important at all” to “very important”). Why are some senses more important than others? 

2. Naturalness  questionnaire.  Indicate  to  what  extent  you  experience  the product as natural  (5‐point scale, from “not natural at all” to “very natu‐ral”). What aspects of the product make it more natural? What aspects of the product make it less natural? 

3. Emotion questionnaire. For each emotion  indicate  if you do not  feel  it,  if you feel it to some degree, or if you feel it strongly (3‐point scale). Which of the emotions do you feel strongest? Why? We used 12 emotions: con‐tempt, admiration, dissatisfaction, satisfaction, unpleasant surprise, pleas‐ant  surprise,  aversion,  attraction,  boredom,  fascination,  sadness,  joy. These emotions are also used in the PrEmo instrument, originally designed to measure  emotional  responses  to  visual  appearances  (Desmet,  2003). These  emotions  have  generally  also  been  shown  to  be  relevant  for  de‐scribing emotional  responses  to  (eating)  food  (Desmet and Schifferstein, 2008). The emotions were presented both  verbally  and non‐verbally  (as cartoon characters) on paper.  

At the first stage participants were standing in front of a set of shelves built at TU Delft, mimicking  supermarket  shelves,  showing  the  target  product  and  several competing products  that  simulated  the  store environment.  For each package 5 samples were present. The experimenter  invited participants to the “store” and explained  the  task: “Please,  imagine  that you are  in a supermarket,  standing  in front of  the  shelves. You would  like  to buy  this dehydrated product  containing the target vegetable. Please select the one you prefer”. 

The participant examined products and chose one product. Then, the experimenter asked 3 open questions: 

1. What did you look for when you tried to find your product? (brand, color, flavor, shape...)  

2. What draws your attention when you stand in front of these shelves?  3. Why did you choose this product?  

Then the experimenter asked the participant to take the test product, being one of the two commercial packages. The participant examined the product and  filled  in the first questionnaire. 

Page 59: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

59 

At the second stage the experimenter invited the participant to a kitchen corner and explained  the  task: “Imagine  that you have already bought the product. You are at home, you feel hungry and you want to eat some food. Please, open the package and put the contents in the pan now. Before you start cooking, please, answers our ques‐tions”. The participant opened the package and filled in the questionnaire. 

At  the  third  stage  the experimenter  showed participants  the  cooking equipment and explained how  to use  it. The  translation of  the German  cooking  instructions from the packages were given to participants. The participant cooked the product. After the food was ready, the participant filled in the questionnaire. 

At the fourth stage, the participants were seated in a dining area and ate the food. After they were finished, they filled in the questionnaire.  

At the fifth stage (directly after the fourth stage), the experimenter invited partici‐pants  back  to  “the  store”  and  asked  them  to  select  a  product  again  from  the shelves. Then three questions were asked: 

1. Why did you choose this product?  2. Why  did  you  choose  another  product  this  time?  (in  case  of  a  different 

choice) 3. Why did you not choose the test product? (in case of a different choice) 

The experimenter filled in the answers and asked participants to fill in the Naturalness and the Emotion questionnaire about the target product for the last time. 

For each participant the experiment took approximately 50 minutes.  

 

2.4. DATA ANALYSIS 

Qualitative data were  first categorized  into groups on  the basis of  their semantic similarity. Three experts agreed on the categories. Then the categories were trans‐lated into English, and the frequencies of each category were computed.  

Repeated measures ANOVAs were performed on the quantitative data to  look for the changes in importance ratings of sensory modalities, ratings of naturalness and emotion ratings for the two target products during five experimental stages. Post‐hoc analyses with Bonferroni adjustment were performed  to  test  the significance of the differences between means. 

 

Page 60: ANNA FENKO - TU Delft

CHAPTER 3 

 

60 

3. RESULTS 

3.1. PURCHASE AND RE‐PURCHASE DECISIONS 

At the buying stage, the two target products were selected most often among the 10 variants of products. Each of the two products was selected approximately by 20 % of participants, while each of the other products was selected by less than 10% of par‐ticipants. The analysis of the reasons participants gave for their first choice indicates that there were two different buying strategies. Most people were looking for famil‐iar brands,  familiar  taste or specific  ingredients  that  they  like  (61%), while approxi‐mately 15% were  looking  for something new and unusual. At the repurchase stage 34% of participants chose the same product they had chosen the first time. A quarter of participants did not like the target product, so they decided to try something dif‐ferent the second time. But even  if participants were satisfied with the target prod‐uct, some of them still wanted to try a new product out of curiosity.   

At the purchase and re‐purchase stages the participants were asked several ques‐tions to clarify their decision‐making process. The stimuli that attracted automatic attention when people were standing  in front of the shelves were the pictures of the  products,  the  color  of  the  vegetables,  and  the  specific  color  scheme  of  the brands.  Some  participants  (13%) mentioned  that  products  on  the  eye  level  at‐tracted most of their attention. Others mentioned specific  ingredients (rice, vege‐tables), variety of colors and shiny packages. When respondents were deliberately trying to choose the product, they paid attention to ingredients (“extra ingredients, something unusual”, “no meat, I am vegetarian") and were trying to guess how the product  would  taste.  One  third  of  respondents  mentioned  package  attributes (“how  the  product  is  presented,  if  the  package  is  attractive”).  Colors  and  brand preferences were mentioned by 12% of participants.  

 

3.2. SENSORY MODALITIES  

Repeated measures ANOVAs were performed on the importance ratings of the sen‐sory modalities with  Stage  and Modality  as within‐subject  factors.  The  effects  of Stage [F (3, 138) = 28.5, p < 0.001], Modality [F (4, 184) = 109.2, p < 0.001] and Stage x Modality interaction [F (12, 552) = 18.1, p < 0.001] were all significant (see Figure 1).  

 

Page 61: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

61 

 

FIGURE 3.1. SENSORY IMPORTANCE AT DIFFERENT STAGES OF PRODUCT USAGE 

 

At  the buying stage, vision was  the most  important modality  (87% of  the partici‐pants rated it as dominant). Participants commented that they looked at the pack‐age very carefully to gather  information on  ingredients (see Table 3.1). They used this  visual  information  to guess about  the  taste of  the products. More  than one third of  the  respondents mentioned  taste as  the most  important modality at  the buying stage. They  imagined how the product would taste, based on the  informa‐tion provided on the package. Touch was  less  important, although 15% of the re‐spondents mentioned that they noticed how the package felt and were able to feel ingredients inside. The anticipation of the smell was also important for 15% of the respondents. The sound was not important, although some people mentioned the sound of the brand name or of the ingredients. 

At the opening stage smell becomes equally important as vision. Participants com‐mented  that  smell  gives  the  first  impression  of  the  taste  and  the  quality  of  the product.  Some  participants  also mentioned  that while  opening  the  package  it  is important to look at the contents (the powder, its color) in order to check what you are going to eat. The  importance of touch  increases slightly at this stage, because people have to open the package and make sure that there is no powder between tear‐strips. 

At the cooking stage smell dominates the product experience, because people find 

it  important  that  the product smells good  (appetizing). They also use olfaction  to 

make sure  the dish does not burn. Vision  is also  important at  this stage, because 

Page 62: ANNA FENKO - TU Delft

CHAPTER 3 

 

62 

the  food  should  look appetizing. Some  respondents commented  that  they notice 

the thickness of the food. People also use visual and auditory cues to know when 

the food begins to boil.  

 

TABLE 3.1. THE MAIN EXPLANATIONS FOR THE MODALITY IMPORTANCE GIVEN BY 

PARTICIPANTS (%*) 

  Buying  Opening  Cooking  Eating 

See  87.2 

Looking at the package to know what to expect (ingredients, taste, etc.) 

61.7 

Looking at the powder (colour) to check what you eat; look how to open the package 

70.2 

If the food looks appe‐tizing; thickness of the food; to see when it boils 

46.8 

The food should look appetizing;  taste and colour on package match 

Taste  34.0 

Imagining the taste 

19.2 

The expectation of the taste 

17.0 

Imagining the taste (by the smell) 

89.4 

The taste is what it is really all about 

Smell  14.9 

Smell of store or imagination of smell of the food 

74.5 

Smell gives first impression of taste and quality 

78.7 

If the food smells good/appetizing and does not burn 

65.9 

Good smell im‐proves the taste 

Touch  14.9 

Feel of the package and the ingredients inside 

27.6 

(How to) open the package, and con‐firm no powder between tear‐strip 

25.5 

Feel the structure /thickness of the food 

36.2 

Thickness food (while stirring) and how it feels in your mouth 

Hear  6.4 

The sound of the name, or of the ingredients 

6.4 

If you hear air escaping the package you know it is fresh 

6.4 

Hearing if the food/ water is boiling 

29.8 

Nothing to hear 

* The percentage of participants who provided an explanation of why a certain modality  is important at a certain stage. The sum of  the answers  is more  than 100%, because partici‐pants were allowed to give several answers. 

Page 63: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

63 

 

At  the  eating  stage  taste  is  dominant,  because,  as  participants  commented, 

“The taste  is what  it's really all about!” Olfactory and visual properties remain 

important at this stage. Participants commented that the food should  look ap‐

petizing, and good smell improves the taste. Some respondents also noted that 

the  color of  the  food on  the package  should match  its actual  color. Although 

touch was not dominant  in any stage, at  the  time of eating  its  importance  in‐

creases, because respondents notice the thickness of food while stirring  it and 

feeling it in the mouth.  

 

3.3. NATURALNESS 

Repeated measures ANOVAs were performed on naturalness ratings with Stage as 

within‐subject factor and Brand as between‐subject factor. The effect of the Stage 

[F (4, 184) = 8.3, p < 0.001] was significant, but the effect of the Brand [F (1, 46) = 

1.2, p > 0.2] was not significant (see Figure 3.2). Post hoc analysis with Bonferroni 

adjustment  showed  that  the  rating of naturalness at  the  cooking  stage  is  signifi‐

cantly higher than at other stages (all p<0.05).   

At the buying stage, the factors that were mentioned to increase the experience 

of naturalness were the pictures of fresh vegetables and fresh spices, the color of 

the  food  in  the picture  and  the  information  about  ingredients  (the  absence of 

unfamiliar  ingredients  and  artificial  taste  enhancers)  (see  Table  3.2).  The most 

important factor that decreased the experience of naturalness at this stage was 

the product type  itself: 40% of respondents mentioned that they did not associ‐

ate dehydrated food powder with naturalness (see Table 3.3). Among other fac‐

tors which decrease the  impression of naturalness was the artificial appearance 

of the package (too many colors, too much unstructured information, the picture 

of the food). Some participants also mentioned unfamiliar ingredients as a factor 

that decreased naturalness. 

At the opening stage, the smell was the most  important factor that  increased the impression of naturalness. The second important factor was the presence of spices or green ingredients in the food powder. The most important factor that decreased the  experience of naturalness was  similar  to  the  first  stage: participants did not 

Page 64: ANNA FENKO - TU Delft

CHAPTER 3 

 

64 

associate dried food powder with naturalness. The second important negative fac‐tor was the color of the powder.  

At  the  cooking  stage,  the majority  of  participants  indicated  the  color  of  the 

food as the main positive driver of naturalness. The second  important natural‐

ness driver was smell. At this stage half of the respondents appeared to change 

their firm beliefs about the absolute  incompatibility between dehydrated food 

and the concept of naturalness, but for 21% the product remained unnatural by 

definition. Other factors that decreased naturalness were the color of the food 

and its smell (too strong).  

 

FIGURE 3.2. CHANGES IN NATURALNESS AT DIFFERENT STAGES OF PRODUCT EXPERIENCE 

 

At the eating stage, the main positive drivers of naturalness were the color and the 

taste of  the  food,  the second  important  factor was  the presence of spices  (green 

pieces) in the food. The negative drivers of naturalness at this stage were the color 

of the food, the absence of vegetable pieces, and the taste (too salty).  

 

Page 65: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

65 

TABLE 3.2. FACTORS THAT INCREASE THE NATURALNESS OF THE FOOD AT DIFFERENT STAGES (% OF ANSWERS) 

What aspects of the product make it more natural?  

purchase  opening  cooking  eating  repurchase 

Picture of fresh vegetables  44.7         

Picture of fresh spices  34.0         

The appearance/color of the food  19.2  8.5  68.1  31.9   

Appearance of the packaging  12.8  6.4      61.7 

Not too many unfamiliar ingredi‐ents 

10.6        10.6 

Nothing  8.5  17.0  6.4  17.0  17.0 

No artificial taste enhancers  6.4         

The spices/green ingredients    27.7  17.0  25.5   

The smell    36.2  34.0  10.6   

The presence of big chunks in the food 

    12.8  6.4   

The taste        27.7   

 

At the fifth stage (re‐purchase), the general appearance of the package was men‐tioned by the majority of respondents as the main positive driver of naturalness. It is important that after participants tasted the food, the number of participants who believed  that  dehydrated  food  could  not  be  natural  by  definition  dropped  from 40% to 28%. Still, almost one third of our respondents did not associate the whole product  category with  naturalness. Artificial  appearance  of  the  package  and  the picture of  the  food were two other negative drivers of naturalness. Some partici‐pants  also mentioned  the negative  experience with  the previous product  as  the factor that reduced naturalness.  

 

 

 

Page 66: ANNA FENKO - TU Delft

CHAPTER 3 

 

66 

TABLE 3.3. FACTORS THAT DECREASE THE NATURALNESS OF THE FOOD AT DIFFERENT STAGES (% OF ANSWERS) 

What aspects of the product make it less natural?  

purchase  opening  cooking  eating  repurchase 

The fact that the food is dried/powder 

40.4  44.7  21.3  8.5  27.7 

Artificial look of the packaging: too much colors, too much infor‐mation, unstructured 

25.5        27.7 

Nothing   17.0  21.3  27.7  12.8   

The picture of the food  10.6        21.3 

Unfamiliar ingredients  6.4        6.4 

The brand color   6.4        6.4 

The appearance/color of the food    27.7  10.6  36.2   

The smell is too strong    12.8  10.6     

No vegetable pieces      12.8  12.8   

The way the food is prepared      8.5     

The foam during preparation      6.7     

Too salty        10.6   

Too much spices        10.6   

Doesn’t taste like vegetable        12.8   

The negative experience with the food 

        19.2 

 

3.4. EMOTIONS 

Repeated measures ANOVAs were performed on emotion  ratings with Stage and Emotion as within‐subject factors and Brand as between‐subject factor. The effects of both Emotion [F (11, 473) = 11.9, p < 0.001] and Stage x Emotion interaction [F (44, 1892) = 3.1, p < 0.001] were significant (see Figure 4). The effect of the brand 

Page 67: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

67 

on the data variance was not significant [F (1, 46) = 0.22, p > 0.20], and the effects of Brand x Emotion [F (11, 473) = 0.72, p > 0.20], and Brand x Stage interactions [F (4, 172) = 1.5, p > 0.20] were also not significant. 

Post‐hoc analyses with Bonferroni adjustment were performed on each emotion separately to test the differences between the 5 stages. Five out of the 12 emo‐tions  showed  significant  changes  over  time:  satisfaction,  pleasant  surprise,  at‐traction,  dissatisfaction,  and  boredom.  These  emotions  were  analyzed  further (Figure 3.3). The overall means  for  the other emotions are: contempt  (1.4); un‐pleasant  surprise  (1.2);  aversion  (1.2);  sad  (1.1);  admiration  (1.3);  fascination (1.4); joy (1.5). 

 

FIGURE 3. EMOTION DYNAMICS AT DIFFERENT STAGES OF PRODUCT EXPERIENCE 

 

At the buying stage the rating of attraction was significantly higher than the rating 

of  dissatisfaction  and  of  pleasant  surprise  (p  <  0.05).  Participants  indicated  that 

they were curious about the new product and about the taste of ‘instant’ product; 

they liked how the food looked and felt like eating it (see participants’ comments in 

Table 3.4). The feeling of satisfaction at this stage was associated with the expecta‐

tion of good food that would taste similar to participants’ taste. The rating of bore‐

dom at this stage was not significantly different from other emotions (all p > 0.20). 

Page 68: ANNA FENKO - TU Delft

CHAPTER 3 

 

68 

Participants indicated that there was nothing special in the product, it was not sur‐

prising, and the food looked boring.  

At  the opening stage  the  rating of satisfaction was significantly higher  than  the 

ratings of dissatisfaction and boredom (p<0.05). At the cooking stage the ratings 

of  three positive emotions  (satisfaction, attraction, and pleasant  surprise) were 

significantly higher than the ratings of the two negative emotions (dissatisfaction 

and  boredom)  (all  p<0.05).  Respondents  indicated  that  the  product met  their 

expectations, it looked and smelled good, they were curious about taste and felt 

like eating the food. 

The analysis of the emotional dynamics shows that the rating of pleasant surprise 

increased significantly between the buying and the cooking stage (p<0.05), and the 

rating of  attraction  increased  significantly between  the opening  and  the  cooking 

stage  (p<0.05). Participants commented  that  the  food  looked and  smelled better 

than they had expected.  

At  the eating  stage  the  feelings of attraction and pleasant  surprise  significantly decreased, and  the  feeling of dissatisfaction  significantly  increased  (all p<0.05). The  ratings of  satisfaction and dissatisfaction were not  significantly different at this stage (p>0.20). Some participants indicated that the food tasted too salty or too sour, other commented that the food tasted good, even though  it was pow‐der food.  

At  the  re‐purchase  stage,  the  rating  of  pleasant  surprise  dropped  significantly 

(p<0.05), and was significantly lower than the ratings of other emotions. The rat‐

ings of satisfaction, attraction, dissatisfaction and boredom did not differ signifi‐

cantly at the final stage (p>0.20).  Some participants liked the taste and smell of 

the food; they noted that the content matched the  information on the package, 

and that for a powder food it was a good food. Others didn’t like the taste, they 

did not find anything special about the product, and did not change their dislike 

for powder foods. 

The comments about the strongest emotions felt by participants are summarized in 

Table 4. The data on the strongest emotions are in agreement with the overall rat‐

ings of the emotions. The strongest positive emotions were attraction, satisfaction, 

and pleasant surprise. Attraction was experienced mostly at the buying and cooking 

stages, satisfaction at the cooking and eating stages, and pleasant surprise at the 

cooking stage. The strongest negative emotions were dissatisfaction (at the eating 

stage) and boredom (at the buying stage). 

Page 69: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

69 

TABLE 3.4. THE STRONGEST EMOTIONS AT DIFFERENT STAGES OF PRODUCT 

USAGE (FREQUENCIES AND EXPLANATIONS) 

The strongest emotion 

purchase  opening  cooking  eating  repurchase 

contempt  5  

negative taste anticipation, dislike for powder food 

bad smell, dislike for powder, de‐ceptive pack‐age 

2  

difficult to mix powder with water 

bad taste  

bad taste, dislike for the product, de‐ceptive pack‐age 

dissatisfied  3 

not much vegetables, not the best food in this store, don’t like dry food 

2  

don’t like dry food, didn’t open the package properly 

color of pack‐age and food don’t match, no good smell 

10 

bad taste, bad smell, too sour, nothing new 

the taste was bad, dislike for powder foods, “not my thing” 

unpleasant surprise 

I wanted other prod‐uct, this one doesn’t look special 

0  0  2 

doesn’t taste like vegeta‐ble, didn’t like the structure 

artificial taste, you don’t get the vegeta‐bles printed on the pack‐age  

aversion  0  0  1  

smells too strong, makes me a bit ab‐horrent 

I don’t taste vegetables 

5  

I didn’t like the food, too thin, too creamy, no chef quality 

bored  9 

nothing spe‐cial, not sur‐prising, food looks boring 

expected more attrac‐tive food, no special smell, no surprises after opening 

5 minutes of stirring is not the best part of cooking 

tastes like normal food 

I didn’t like the taste, nothing spe‐cial 

sad  1  

think of eat‐ing alone with no attention 

0  0  1 

taste is not sufficient, too watery / arti‐

1  

package seemed fine but this 

Page 70: ANNA FENKO - TU Delft

CHAPTER 3 

 

70 

to food  ficial  wasn’t true 

admiration  1  

curious about taste and smell 

0  1  

content meets expec‐tations 

2  

good food, pure taste, smell is simi‐lar to taste, meets expec‐tations 

2  

So many dif‐ferent prod‐ucts, so ful‐filled after just one small package, good food 

satisfied  7  

expectation of good food, taste similar to my taste, not extreme, product is clear 

product meets expec‐tations, looks good, expect good food, look forward to cooking and  eating 

11 

food is ready, it looks and smells good 

12 

food was ok, good for in‐between meals, tastes like vegetable 

5  

for powder food it’s good, food was sufficient and filling 

pleasant sur‐prise 

0  6 

presence of spices, easy to open, good smell 

11 

food looks and smells better than expected 

8  

it tastes good, even though it is powder food 

content matches the information on the pack‐age 

attracted  12 

curious about new product, food looks good, feel like eating food 

feel like eat‐ing food, curious about taste 

12  

curious, the look and smell make me feel like eating the food 

3  

I liked the food, it looks good, like crème food 

I liked the taste and smell of the food 

fascinated  0  3  

How do they make this? 

easy to open, curious about taste 

2  

curious about the taste 

1  

making such a nice food in such a short time 

2  

I hope it will taste differ‐ent this time, curious about other foods  

joy  1  

no comment 

2  

happy with product in‐formation 

I don’t cook often, I want to enjoy the process 

0  0 

Page 71: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

71 

3.5. OVERALL ANALYSIS OF THE DATA 

The importance ratings of sensory modalities, ratings of naturalness and emotions may be related. Therefore, we performed a Principal Component analysis (PCA) on all 18 variables (5 sensory modalities, naturalness and 12 emotions) to look at pos‐sible correlations between the different product experiences. 

 

TABLE 3.5. FACTOR LOADINGS*  

  Factor 1:  positive affect 

Factor 2:  negative  

affect 

Factor 3: eating  

Factor 4: seeing/ boring 

Factor 5:  hearing 

admire  0.76         

satisfied  0.61  ‐0.33      ‐0.30 

pleasant  0.66         

attracted  0.63  ‐0.31       

fascinated  0.69        0.35 

joy  0.75         

contempt  ‐0.34  0.60    0.37   

dissatisfied  ‐0.32  0.72       

unpleasant    0.79       

aversion    0.74       

sad    0.68       

natural  0.38  ‐0.57       

smell      0.78     

touch      0.66     

taste      0.78     

bored        0.72   

see        0.69   

hear          0.81 

* Only loadings bigger than 0.30 are shown. 

Page 72: ANNA FENKO - TU Delft

CHAPTER 3 

 

72 

The results yielded five clear factors that explain 62% of variance (see Table 3.5). 

The first factor (explains 18.2 % of variance) includes six positive emotions: admi‐

ration  (factor  loading  0.76);  satisfaction  (0.61);  pleasant  surprise  (0.66);  joy 

(0.75); attraction (0.63), and fascination (0.67). The second factor  (explains 18.0 

% of variance)  includes  five negative emotions: contempt  (0.60); dissatisfaction 

(0.72); unpleasant surprise (0.79); aversion (0.74), and sadness (0.68). These two 

factors  clearly  represent  positive  and  negative  affective  valence.  Naturalness 

loads on both these factors: positive affect (0.38) and negative affect (‐0.57). The 

third factor (explains 10.3 % of variance) includes three sensory modalities: smell 

(0.78), touch (0.66), and taste (0.78). This factor refers to the senses that are im‐

portant at the active stages of user‐product interaction (cooking and eating). The 

fourth  factor  (explains  8.3 %  of  variance)  includes  two  variables:  one  sensory 

modality  (vision, with factor  loading 0.69) and one negative emotion  (boredom, 

with factor loading 0.72). The last factor (explains 6.9 % of variance) includes only 

one variable, auditory modality (0.81).  

Most of emotional variables  in our data  loaded on positive or negative affective 

factors. The only exception  is boredom, which appeared  to be  related  to visual 

importance. As expected, naturalness was experienced as a positive  food prop‐

erty.  It  correlated  positively with  positive  affect,  and  negatively with  negative 

affect. The data on sensory modalities suggest that the correlations between the 

senses depend  on  the  stage  of  the product usage.  Smell,  taste,  and  touch  are 

important at the eating stage and together they form a separate factor. Vision is 

important at the buying stage, which participants find the most boring part of the 

experiment.  This  may  explain  why  vision  demonstrated  high  correlation  with 

boredom. Hearing was not an important modality at any stage, and it represents 

a separate factor.  

 

4. DISCUSSION This research aimed at understanding how the sensory properties of a food product (dehydrated  food)  influence  the  experience  of  naturalness  of  this  product  and emotional reactions to it at different stages of user‐product interaction. The results demonstrate that all these properties may vary depending on the different stages of product usage.  

 

Page 73: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

73 

4.1. IMPORTANCE OF SENSORY MODALITIES 

At  the buying  stage, people pay most  attention  to  visual properties of  the  food package. This result is in line with our previous research (Fenko et al., 2009), where we demonstrated the dominance of vision at the buying stage  for a variety of  in‐dustrial products. In the present study, participants mentioned several visual prop‐erties  that  attracted  their  immediate  attention  while  standing  in  front  of  the shelves,  including  the  colors  of  the  packages,  the  images  of  specific  ingredients (e.g. vegetables), and the shiny material of the packages. When participants were deliberately trying to choose a product, they paid more attention to the  informa‐tion about the  ingredients and tried to  imagine how the food would taste. That  is why taste was the second important modality at the buying stage. 

At the opening stage participants were able to smell the contents of the package, and at this stage olfaction became as  important as vision. Smell remained  impor‐tant throughout the cooking and eating stages. At the eating stage, taste became dominant, followed by vision and smell. Tactile properties became somewhat more important at the eating stage, because the thickness of the food constitutes one of the components of the eating experience. Hearing was not important at any stage of product usage. 

The most surprising finding in regard to sensory importance dynamics was the rela‐tively high rating of taste at the buying stage, when participants had no actual gus‐tatory  sensations.  This  result  shows  that  participants  interpreted  the  questions about their sensory experience more broadly than we expected, and included their memories  and  expectations  to  imagine  the  sensations  that were not  yet  experi‐enced.  This  suggests  that  in  sensory  dominance  research,  questionnaire  data should be used with caution. The answers may reflect not only actual sensory ex‐periences, but also people’s ideas of what they might or should experience.  

 

4.2. NATURALNESS 

The dynamics of  the naturalness experience demonstrates  the  complex  relation‐ships  between  sensory  experience  and  symbolic  product  properties.  Symbolic meanings people attach to food products depend not only on its perceived sensory properties, but  to a  large extent on people’s attitudes, beliefs and values  (Wood, 1995, Mennel, et al., 1992). 

Page 74: ANNA FENKO - TU Delft

CHAPTER 3 

 

74 

Eating is a social practice. Sharing food is one of the fundamental ways that one can display, establish,  and maintain  interpersonal  intimacy. The dinner  table  is often described as a symbol of ideal family life (Lupton, 1996). Sociological studies dem‐onstrate that traditional meals are perceived as an  important part of producing a home and family (DeVault, 1994). Cooking proper meals is perceived as something women do to care for their family (Murcott, 1983; Charles and Kerr, 1988). Warn‐ings about  the  increase  in  the use of  fast‐food dishes  that can be eaten anytime and anywhere is part of the more general Western critical discourse on the disinte‐gration of family  life and the commercialization of food habits (Warde, 1999). For instance, Falk (1994) argues that one characteristic of the modern consumer soci‐ety  is that food habits disintegrate (e.g. a reduction of ritual dinner meals). Mintz (1985)  ties  these  tendencies  to  the  rapid spread of  fast  food consumption  in  the western world. 

To  address  these  public  concerns,  people  develop  simplifying  heuristics  about appropriate  and  inappropriate  foods  (Holm,  2003).  One  of  such  heuristics  in‐cludes eating “natural” or “pure” foods, i.e., foods that are homemade from raw ingredients and contain few or no additives. For instance, Charles and Kerr (1988) reported  that most of  their  sample of 200 women described  a proper meal  as consisting of  fresh  foods,  and home‐made  food was  considered  as better  than processed food.  

In Western societies, naturalness has become one of the cultural values most peo‐ple wish to achieve in their everyday consumption, especially in food consumption (Rozin, 2005; Verhoog et al., 2003). The eating of “natural” foods is seen as a pro‐tection against known and unknown dangers perceived to be associated with mod‐ern foods (Holm and Kildevang, 1996). The whole group of prefabricated food,  in‐cluding dehydrated foods, is conceptualized as “artificial” and is treated with suspi‐cion. That is why it is so difficult for any member of this product group to stand out as more natural than other foods.  

Our  data  demonstrate  that  at  the  buying  stage  these  attitudes  are  extremely strong. When asked about  the aspects of  the product  that decrease  their experi‐ence of naturalness, 40% of the respondents answered that the very fact that the food consists of dry powder instead of fresh vegetables makes it unnatural. In addi‐tion, 8.5 % said that nothing can be done to improve the impression of naturalness. Nevertheless,  the  actual  experience of  cooking  and  eating  the  food managed  to change this negative attitude to some extent. The color and smell of the food were the main factors that improved the naturalness rating of the product at the cooking stage.  At  this  stage,  the  amount  of  participants who  firmly  believed  that  dehy‐

Page 75: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

75 

drated  food  is artificial by definition, dropped to 21%  (see Table 2). At the eating stage  it dropped even  further,  to 8.5%, but  at  the  repurchase  stage  it  increased again  to 28%.    This dynamics  suggest  that  consumer’s negative  attitudes  to  this product group can be changed by actual positive experience, but it probably needs more than just one positive experience to make this change long‐term. 

 

4.3. EMOTIONS 

The dynamics of emotions confirms our suggestion  that product experience  is  in‐fluenced by both the actual perception of the sensory product properties and pre‐existing attitudes and beliefs about a product. At  the buying stage,  the ratings of satisfaction and dissatisfaction were about equal. But when participants started to actively  interact with the product  (open the package, cook and eat the food), the rating of  satisfaction went up, while  the  rating of dissatisfaction dropped  signifi‐cantly. The most striking dynamics was demonstrated by the emotion of pleasant surprise. Participants commented  that  they did not expect  the  food  to smell and taste so good. Similar to the dynamics of naturalness, the positive impact of actual experience was  not  stable;  at  the  repurchase  stage  the  ratings  of  emotions  re‐turned to the same level as at the first buying stage. 

The analysis of product experience dynamics suggests  that actual sensory experi‐ence has a different  impact on  the product assessment at  the different stages of product usage. At  the buying  stage, when  vision  is  the main and  sometimes  the only source of information about the product, people rely on their pre‐existing atti‐tudes  and beliefs  about  the product more  than on  the  actual perception. When people have more opportunities to use their other senses (touch, smell, and taste), their evaluations may  change. But  these  changes may be only  temporary, unless the pre‐existing beliefs are challenged.  

It may be hard to modify beliefs, based on specific sensory information. The reason for  this may  be  the  different  relationships  of  particular  sensory modalities with cognitive and affective systems. Experimental data suggest that vision and audition are the two sensory modalities that are most closely connected to rational thinking (Goodale & Humphrey, 1998; Neisser, 1994; Paivio, 2006). Touch, smell, and taste are more  strongly  associated with  emotions  (Hinton &  Henley,  1993;  Sweetser, 1990), but these experiences are difficult to verbalize and to recognize consciously (Köster, 2003). In our study of verbal descriptions of product experience, symbolic properties  (such  as modern,  expensive,  and  natural)  demonstrated  strong  visual 

Page 76: ANNA FENKO - TU Delft

CHAPTER 3 

 

76 

dominance, while affective properties  (such as pleasant, exciting, or boring) were equally related to all sensory modalities (Fenko et al., in press).  

We suggest that mere visual exposure to the product may actualize the user’s cog‐nitive attitudes and beliefs about  the product, while other  senses,  such as  smell and  taste, may  trigger mainly  affective  reactions which  are difficult  to  recognize and which often stay at the subliminal level (Berridge & Winkielman, 2003). Even if participants were aware of their emotions and sensory experiences, as they were during our experiment, for some of them sensory and emotional experiences were not enough to change their pre‐existing attitudes about the dehydrated food prod‐ucts.  As soon as they were back to the “supermarket shelf”, their negative cogni‐tive attitudes  towards  this product group were actualized again by  the visual  im‐ages of products.   This may be a  reason why  it  is  so difficult  to  change negative stereotypes about some product groups, even when the actual olfactory and gusta‐tory experiences are positive. 

 

CONCLUSION This  experimental  study was  performed  to  understand  the  temporal  changes  of sensory  importance  during  user‐product  interaction  with  dehydrated  food. We found that at the buying stage, vision is the most important modality, followed by taste. Taste had a  relatively high  rating at  the buying  stage, because participants were trying to imagine how the food would taste. Smell was dominant at the cook‐ing stage, and taste was the most important sensation while eating the food. 

The  dynamics  of  naturalness  and  emotional  experience  suggest  that  different mechanisms are  involved  in the overall product experience at different stages. At the  purchase  and  re‐purchase  stages, where  participants were  able  only  to  use their  vision  and  previous  knowledge,  the  pre‐existing  attitudes  and  stereotypes about the product group seem to play a major role  in participant’s affective reac‐tions  and  naturalness  judgements. When  other modalities  (olfaction,  taste,  and touch) were actively  involved, participants’ emotional  judgements were more di‐rectly linked to their sensory experience. These results confirm the suggestion that tactile, olfactory and gustatory modalities are more closely linked to emotions than vision. Vision, on the other hand, is more closely linked to cognitive processes. That is why visual exposure  to products may actualize existing  cognitive attitudes and stereotypes about the product.  

Page 77: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

77 

In future research, it would be interesting to investigate the relationships of various sensory  experiences  with  cognitive  and  affective  processes  more  closely.  This would be  important both for theoretical understanding of psychological processes involved  in product experience, and for practice of product design, marketing, ad‐vertising, and retail.  

                                 

Page 78: ANNA FENKO - TU Delft

CHAPTER 3 

 

78 

                               Chapter 4 is published as:  

Fenko, A., Otten,  J.J., Schifferstein, H.N.J.  (2010) Describing product experience  in different  languages:  The  role  of  sensory  modalities.  Journal  of  Pragmatics,  42, 3314‐3327.  

Page 79: ANNA FENKO - TU Delft

79 

 

 

 

 

 

 

 

 

 

CHAPTER 4 

DESCRIBING PRODUCT EXPERIENCE  IN DIFFERENT LANGUAGES:  

THE ROLE OF SENSORY MODALITIES 

1. INTRODUCTION When people  interact with products, they receive  information through all the dif‐ferent senses. This sensory information is processed in the brain and contributes to the overall product experience that people are aware of and can describe verbally. In describing  their  experience, people usually use  adjectives  that  reflect  sensory properties (e.g., warm, solid, red, loud), adjectives that refer to symbolic properties (e.g., elegant, expensive, modern), and affective evaluations (e.g., good, bad, beau‐tiful, ugly) of the product (see Hekkert & Schifferstein, 2008).  

All descriptions of product experiences ultimately  rely on  sensory  inputs derived from the product. Nonetheless, information from some sensory modalities may be more important for describing certain product experiences than others. We define 

Page 80: ANNA FENKO - TU Delft

CHAPTER 4 

 

80 

“sensory  importance” as the relative contribution of each sensory modality to the description of a particular product experience. The dominant  sensory modality  is the modality that has the largest effect on the specific description. In this research we wonder what the role of the senses is in the description of product experiences through different types of adjectives. 

 

1.1. FACTORS AFFECTING SENSORY DOMINANCE 

A general and popular belief seems to be that vision  is the dominant sensory mo‐dality in everyday experience. When people are asked which sensory modality they would miss most  if  they  lost  it,  the majority  is  likely  to  indicate vision  (Fiore and Kimle, 1997; Schifferstein, 2006).  In addition, when people are asked  to describe objects, they primarily use adjectives that refer to the visual (60%) or tactual (32%) modalities (Stadtlander and Murdoch, 2000).  

However, vision  is not the dominant modality  in the  interaction with all products. In a study using consumers’ self‐reported  importances, participants reported that they found one of the other sensory modalities more important than vision during product usage for about half of the products (Schifferstein, 2006). For example, for a computer mouse the tactual characteristics were most  important, for a vacuum cleaner the sound it made, for a cleaning product its smell, and for a soft drink its taste. The relative  importance of a modality may depend on various aspects, such as the availability of sources of sensory stimulation, the degree of variation in sen‐sory stimulation over various products, the usefulness of the sensory  information during  functional use,  the proportion of  time a modality  is used actively, and  the role of the stimulation  in enjoying the product. Hence, the roles of the modalities depend not only on whether a certain type of sensory  information  is present, but also on whether the information is perceived, how it is processed, and how people react to it emotionally.  

Culture may play a role  in determining the importance of a sensory modality. Cul‐ture has been shown to influence a large variety of behaviours across many diverse disciplines,  including  consumer  behaviour  (see  Hofstede,  2001).  Schifferstein (2006) suggested  that  the visual system  is  regarded dominant, because  it plays a significant role in many daily activities. The importance of vision in Western socie‐ties may have increased over time due to the products that were created, such as books, television, and computers that require major input from the visual modality. McLuhan (1961) believed that the nature of media by which people communicate 

Page 81: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

81 

affects  the  ratio of  their  senses. For example,  the alphabet  stresses  the  sense of sight, which  in  turn  causes people  to  think  in  linear, objective  terms. He  argued that Europeans and North Americans live in the visual mode, while for native Afri‐cans and other non‐literate societies the auditory modality  is dominant. However, anthropological data suggest that significant diversity exists  in sensory dominance among non‐literal peoples. Classen (2005) describes three non‐literal societies that have three different dominant sensory modalities: temperature is the most impor‐tant  sensory property  for  the Tzotzil of Mexico,  smell  for  the Ongee of  the Little Andaman Island, and vision for the Desana of Columbia. 

 

1.2. CAN LANGUAGE INFLUENCE PRODUCT EXPERIENCE? 

Language is one of the core components of any culture. It is central to communica‐tion and closely  related  to  thought. Much of human cultural heritage  is encapsu‐

lated  in  semantic  concepts  packed  into  words  (Levinson and Jaisson, 2006). Through  language  individuals  have  access  to  the  large  accumulation  of  cultural ideas, practices and technology which constitute a distinct cultural tradition (Gum‐perz and Levinson, 1996). The  learning of natural  languages  is probably  the most complex  cognitive  task  that humans  routinely undertake and  the major pressure for brain evolution in our species (Byrne and Whiten, 1988). 

Languages  differ  fundamentally  in  their  semantic  categories—the  concepts  built into  their  grammars  and  lexicons.  It  is  quite  difficult  to  find  any  exact  cross‐linguistic  matches  between  linguistically‐coded  concepts  (Levinson  and  Meira, 2003). To what extent are our ideas and concepts actually embodied in language? A proposal of  linguistic  relativity emphasizes a distinctive  role of  language  in  inter‐preting  experience  and  influencing  thought  (Gumperz  and  Levinson,  1996).  The idea  that  thought  is  shaped  by  language  is most  commonly  associated with  the writings  of  Benjamin  Lee Whorf. Whorf  (1956)  impressed  by  linguistic  diversity, proposed that the categories and distinctions of each language determine a way of perceiving, analyzing, and acting  in  the world.  In recent years, some studies have claimed evidence that language indeed has an important influence on thinking (e.g. Boroditsky, 2001; Bowerman, 1996; Davidoff et al., 1999; Gentner and Imai, 1997; Levinson, 1996; Lucy, 1992), while others have reported evidence to the contrary (e.g. Heider, 1972; Malt et al., 1999; Li and Gleitrnan, 2002).  

Experimental  evidence has  reopened  the debate  about  the  extent  to which  lan‐guage  influences  cognitive processing of  sensory  information. Extensive  research 

Page 82: ANNA FENKO - TU Delft

CHAPTER 4 

 

82 

has been done in domains such as the experience of space (Levinson, 1996; Li and Gleitman, 2002; Majid et al., 2004), colour (Gilbert et al., 2006; Kay and Kempton, 1984; Robertson, Davies, and Davidoff, 2000), number (Gordon, 2004; Gelman and Gallistel, 2004; Pica et al., 2004), and time (Boroditsky, 2001; Chen, 2007; January and Kako, 2007; Núñez and Sweetser, 2006).  

As regards the time domain, speakers of different  languages might talk about the future as  if  it  lies ahead of us  (in English), behind us  (in Aymara), or below us  (in Mandarin  Chinese).  Behavioural  studies  suggest  that  speakers  of  languages  that use different spatiotemporal metaphors may think differently about time.  Indeed, Mandarin  speakers were  faster  to confirm  that March comes earlier  than April  if they had just seen a vertical array of objects than if they had just seen a horizontal array.  The  reverse was  true  for  English  speakers  (Boroditsky,  2001; Núñez  and Sweetser, 2006). 

Dramatic  cross‐linguistic  differences  have  also  been  reported  for  the  domain  of space  (Levinson, 1996). Whereas most  languages  (e.g. English, Dutch) rely heavily on  relative  spatial  terms  to  describe  the  locations  of  objects  (e.g.  left/right, front/back), Tzeltal (a Mayan language) relies primarily on absolute reference (simi‐lar to the north/south system). In experiments with spatial tasks Dutch and Tzeltal speakers  saw an arrow pointing either  to  the  right  (north) or  to  the  left  (south). They were then rotated 180% and were asked to  identify the arrow “like the one they  saw  before”.  Dutch  speakers  overwhelmingly  chose  the  ‘relative’  solution, while Tzeltal speakers overwhelmingly chose the ‘absolute’ solution. Furthermore, speakers of  languages preferring  absolute  coordinates  show more  accurate  skills when asked  to  indicate  the direction of  familiar  locations  from an unfamiliar site (Levinson 1996), suggesting that the differences may have everyday correlates.  

Studies on occupational accidents  in Finland give another example of everyday con‐sequences  of  language  differences.  Occupational  accident  rates  are  substantially lower  in  Sweden  than  in  Finland,  and  also  among  the  Swedish‐speaking minority within Finland, despite working in the same regions with similar laws and regulations (Salminen  and  Hiltunen,  1995;  Johansson  and  Strømnes,  1995).  Researchers  ex‐plained this difference by structural differences between Swedish and Finnish. Swed‐ish  prepositions  can  be  represented  in  terms  of  vector  geometry  in  a  three‐dimensional space, whereas Finnish cases can be represented  in a two‐dimensional space coupled with a third dimension of time (Strømnes, 1973, 1974). The research‐ers suggest that the Finns organize the workplace in a way that favours the individual worker over the temporal organization of the overall production process. Lack of at‐

Page 83: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

83 

tention to the overall temporal organization  leads to frequent disruptions and acci‐dents (Johansson and Salminen, 1996; Johansson and Strømnes, 1995).  

Studies have also shown that language characteristics (such as grammatical gender) can influence product experience (Lucy and Gaskins, 2001; Boroditsky, 2001; Boro‐ditsky et al., 2003). For example, when German speakers were asked to describe a ‘key’ (a word masculine in German and feminine in Spanish), they were more likely to use words  like  ‘hard, heavy,  jagged, metal, serrated, and useful’, while Spanish speakers were more likely to say ‘golden, intricate, little, lovely, shiny, and tiny’. On the other hand, to describe a ‘bridge’ (a word feminine in German and masculine in Spanish), German  speakers  referred  to  it  as  ‘beautiful,  elegant,  fragile, peaceful, pretty, and slender’, while Spanish speakers more often said ‘big, dangerous, long, strong, sturdy, and towering’.  

Schmitt and colleagues (1994) proposed that structural differences between the Chi‐nese  and  English  language  affect mental  representations which,  in  turn,  influence consumer memory of verbal  information. The authors  showed  that unaided brand recall was differentially affected in Chinese and English when it was spoken compared to when it was written. Chinese brand attitudes were primarily affected by the match between  script  associations  and brand  associations, but brand  attitudes of English names were primarily affected by the match between sound associations and brand associations  (Pan  and  Schmitt,  1996).  Furthermore,  how much  consumers  like  a brand name translation can depend on whether that name depicts phonological or semantic characteristics of the original name (Pan and Schmitt, 1996). 

Schmitt and Zhang  (1998) and Zhang and Schmitt (1998) found that structural as‐pects of a  language can affect one of the most basic aspects of consumer behav‐iour: categorization of products. They demonstrated how classifiers, a widespread lexical and syntactical phenomenon  in the Chinese, Japanese, and Thai  languages, and almost non‐existent  in  Indo‐European  languages such as English, German and Spanish, affect the perceived similarity between objects, attribute accessibility, and concept organization. They also demonstrated the  impact of classifier‐based sche‐mata on inferences about product features and provided evidence of the effect of classifiers on judgment and choice. 

 

1.3. LINGUISTIC DESCRIPTIONS OF SENSORY CHARACTERISTICS 

In the present study, we investigate language differences in descriptions of product experience. We assume that language effects may depend on the type of adjectives 

Page 84: ANNA FENKO - TU Delft

CHAPTER 4 

 

84 

used to describe product experience. Some aspects of product experience seem to be uniquely unimodal: hue can only be experienced by sight, tickle can only be felt by touch, and pitch can only be differentiated by audition. Nevertheless, our per‐ceptual experience of  the world  is  richly multimodal  (Stein and Meredith, 1993). People are able to extract information derived from one sensory modality and use it  in another. People can, for example, know a shape by touch and  identify  it cor‐rectly by sight. Furthermore, people are able  to  integrate  the  impressions gener‐ated by different sensory modalities  into a unified,  rich percept.  In addition, sen‐sory  terms  that describe physical properties of  things  (such as  ‘warm’ and  ‘cold’) also  describe  some  psychological  qualities.  Different  languages  such  as  ancient Greek, Thai, Chinese and Hebrew possess some morphemes that designate physi‐cal‐psychological pairings identical with those found in English (Asch, 1955).  

Human  language  partly  operates  through  metaphors. Metaphors  can  structure 

people’s thoughts, govern their activities, and enable their reasoning from the fa‐

miliar  to  the unfamiliar  (Lakoff and  Johnson, 1980, 1999; Rein and Schön, 1977). 

Metaphors often  refer  to  sensory phenomena. Good  ideas are described as  ‘bril‐

liant,’ pleasant dreams as  ‘sweet,’  important topics as  ‘hot,’ and bright colours as 

‘loud.’ Gallup  and  Cameron  (1992)  argued  that  English  and  other  European  lan‐

guages mainly use visual terms as metaphors to describe or capture mental events 

(‘as  illustrated by,’  ‘reflect on that,’  ‘shed some  light,’  ‘a clear explanation’) by re‐

ferring to human evolutionary heritage and the anatomical structure of the brain. 

Compared with other sensory  input, a disproportionate amount of brain  tissue  is 

devoted to the processing of visual information (van Essen et al., 1990).  

Because people do not rely on each of the five senses equally and use some sen‐

sory modalities more often  than others,  it makes  sense  that certain  sensory mo‐

dalities have greater frequency in linguistic representation. Based on data from 53 

languages from all parts of the world, Viberg has shown that the  large amount of 

polysemy with respect to the sense modalities for verbs is constrained by a hierar‐

chy which  gives  the most  prominent  place  to  vision  and  the  least prominent  to 

smell and taste (Viberg, 1984). The hierarchy can be presented as follows: sight > 

hearing >  touch > smell,  taste. According  to  this hierarchy,  ‘a verb having a basic 

meaning belonging to a sense modality higher (to the left) in the hierarchy can get 

an extended meaning that covers some (or all) of the sense modalities lower in the 

hierarchy’ (p. 136–137). A sense that holds the top position has the highest number 

of polysemic patterns and also the highest number of  lexemes at the  level below 

the basic one. Viberg concludes that the verbs of perception, especially ‘see’, have 

Page 85: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

85 

an extensive range of cognitive meanings, and that ‘know’ and ‘see’ are covered by 

the same word in a number of languages. For example, in English it is acceptable to 

say any of the following: (1) “I listened to the voice to see if I could recognize it.” (2) 

“I felt his hand to see how hot it was.” (3) “I tasted the soup to see whether it was 

ready.”  (4)  “I  smelled  the  air  to  see how  fresh  it was.” Miller  and  Johnson‐Laird 

(1976) also acknowledge the startling complexity of the verb ‘see’ and link it to the 

conceptual complexity of visual processes. However, none of  these  linguists  con‐

siders the extension of the meaning of sensory verbs to be metaphorical. 

English verbs of perception and their complex polysemous structure and etymology have also been studied by Sweetser (1990). Unlike Viberg, Sweetser argues that the connection between  the domain of external, physical  sensation on  the one hand and abstract, cognitive states, on the other hand, is metaphorical in nature. Accord‐ing to the cognitive theory of metaphor, verbal metaphors, including conventional expressions based on metaphor, reflect underlying conceptual metaphors in which people conceptualize vague, abstract domains of knowledge (such as time, causa‐tion, ideas, and emotions) in terms of more specific, familiar, and concrete knowl‐edge (Gibbs, 1994). Human thinking  is organized from simple to complex, and the basic  sensory  experience  (source  domain)  is  used  to  understand more  abstract phenomena  (target domain).  Sweetser  suggests  three  reasons  for  the prominent relationship between vision  (as a source domain) and knowledge  (as a  target do‐main): (1) The focusing ability of the sense of vision enables people to pick up one stimulus from many, and, together with its intentionality and directionality, differ‐entiates it from the other senses. (2) Vision is our primary source of objective data about the world. Visual features are the most marked in children’s early categoriza‐tion,  thus  form an  important  source  for  concept  formation. Vision provides data from  a  distance, which  through metaphoric  transfer  become  a  characteristic  of objectivity in the intellectual domain. (3) Vision is identical for different people who can take the same point of view.  

Some authors argue  that visual metaphors  for mental processes are of  relatively recent date. The salient position vision occupies  in conceptualization of  the  intel‐lect  is  not  shared  by  all  cultures  or  even  present  in  older  stages  in  the  Indo‐European  culture  (Ibarretxe Antuñano, 2008). Classen  (1993)  suggests  that  sight‐based words such as bright, brilliant, and lucid have been used to mean intelligent only since the Enlightenment period, “perhaps  in consequence of the general rise of visualism at that time” (p. 58). Many more terms of thought are tactile or kinaes‐thetic, such as apprehend, comprehend, conceive, grasp, ruminate and understand. 

Page 86: ANNA FENKO - TU Delft

CHAPTER 4 

 

86 

“Thought is, or was, experienced primarily in terms of touch […]. Knowing was less like seeing than like holding” (Classen, 1993: 58).  

Auditory  verbs  rarely  serve  as metaphors  for  thought  or  intelligence  in  English, perhaps because hearing  is conceived of as a passive sense, receiving  information but not probing  it. The main  function of hearing  is  communication, and  it  is  the major path of interpersonal influence. Therefore, hearing is associated with obedi‐ence and emotional  receptivity  (‘being deaf  to someone’s plea’)  rather  than with intelligence (Sweetser, 1990). However, languages differ in the metaphorical use of auditory verbs. For example, Evans and Wilkins (2000) showed that Australian lan‐guages recruit verbs of cognition like ‘think’ and ‘know’ from ‘hear’, not from ‘see’. Suya Indians of Brazil use the same verb for ‘listen’, ‘understand’, and ‘know’ (See‐ger, 1975), and in Russian the word понимать (‘understand’) has the same root as внимать (‘listen carefully’).  

Describing the semantics of three Russian perception verbs, воспринимать (‘per‐ceive’),  ощущать  (‘sense’),  and  чувствовать  (‘feel’),  Iordanskaja  (1979)  notes that Russian semantics reflects the 'naive' conception of smelling, feeling, and taste as opposed to sight and hearing. The verb воспринимать (‘perceive’) can stand for any mental ability, including reason, sight, and hearing, while the verbs ощущать (‘sense’) and чувствовать (‘feel’) stand for olfaction, touch, taste, kinesthetic and other internal senses. In addition, чувствовать (‘feel’) can be metaphorically used for  intuition,  instinct  and  the  unconscious mind.  Optical  or  acoustic  perception cannot be called ощущать or чувствовать in Russian, which coincides with psy‐chological  theories  that  consider  sight  and hearing  to be  the most  sophisticated forms of perception, in contrast to other three senses. 

A lot of perceptual (or sensory) adjectives, such as sharp, dull (touch), sweet, sour (taste), loud, and quiet (sound), often extend their meaning from one basic, or pro‐totypical,  sense modality  to  one  or more  secondary modalities.  From Ullmann’s (1957) work based on an investigation of 19th century poetry the following two hi‐erarchies be derived: touch > taste > scent and touch > sound, sight. Out of a total of 2009 transfers, he found 1665 upward transfers (from left to right) and only 344 downward transfers. The  largest number of transactions occurred between touch and sound. 

Ullmann's study mainly deals with literary metaphor, and many of the examples he provides are certainly not part of established usage. Williams (1976) looked at his‐torical changes of English adjectives referring to sensory experience in English dic‐tionaries. He  found  that such adjectives  transferred according  to a hierarchy  that 

Page 87: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

87 

accords well with the one presented by Ullmann. Williams noticed that metaphori‐cal  transfers, with  relatively  few exceptions, go  in one direction. That  is, a  touch word may transfer to taste or directly to sound (‘hot’ or ‘sharp’) or sight (‘soft’). A taste word may  transfer  to  smell or  to  sound  (‘sour  sound’,  ‘sweet music’),  and sight words may transfer to sound and vice versa (‘loud colour’, ‘clear sound’). Wil‐liams  assumes  the  hierarchy  to be biologically based.  The  transfers  go  from  the physiologically least differentiating, evolutionary and ontogenetically primitive sen‐sory modalities to the most differentiating, most advanced. 

It is puzzling that the hierarchies of adjectives and verbs differ substantially. While the verb “to see” can have many non‐visual meanings, the touch‐related adjectives can transfer their meanings to other sensory domains. The reason behind this dis‐crepancy may be that transfer of meaning in verbs and adjectives refer to different areas of human experience. Lehrer (1978) suggests that the transfer of meaning in adjectives  is mainly based on the general experiential dimensions of  intensity and evaluation. In the early 1950s, Osgood asked participants to evaluate different con‐cepts and objects on a number of bipolar scales based on semantic opposites, such as  good‐bad,  soft‐hard,  fast‐slow,  clean‐dirty,  valuable‐worthless,  fair‐unfair,  and so on (Osgood, 1962; Osgood et al., 1957). Osgood called them “semantic differen‐tial”  scales, because  they differentiated personal attitudes based on a  subjective understanding  of  the  connotative meanings  of words.  Factor  analysis  of  the  re‐sponses generally yielded three dimensions: Evaluation or valence (e.g. good‐bad, pleasant‐unpleasant), Activity (e.g. fast‐slow, stimulating‐relaxing, lively‐quiet), and Potency  (e.g.  weak‐strong,  gentle‐tough,  persistent‐accommodating).  Evaluation (valence) and Activity  (arousal) are often seen as the two primary dimensions  for emotional experiences (see, e.g., Feldman Barrett and Russell, 1998). According to Lehrer (1978), they may also be the most important dimensions for the transfer of meaning in sensory adjectives.  

The sense of  touch has always been related  to the  field of emotions. Expressions such as “I’m deeply  touched” or “touching words” are widely used  in English. Al‐ready  in 1921, Hans Kurath  classified  sense perception with  respect  to emotions and stated how “the kinaesthetic, the visceral, and the tactual perceptions have a relatively  stronger  tone  than  those of  hearing  and  especially  of  sight,  the  taste‐smell perceptions taking a middle ground” (p.39). Kurath explained this transfer of meaning from sense perception to emotion on the basis of the similarity of feelings that both domains share. This connection can also be explained by the etymology. Buck (1949) points out that the general word in West Germanic languages for ‘feel’ 

Page 88: ANNA FENKO - TU Delft

CHAPTER 4 

 

88 

refers not only to tactile perception but also to emotions, even in the earliest peri‐ods of the languages.  

Viberg  (1984)  argues  that,  unlike  the  hierarchy  of  sensory  adjectives,  which  is probably based on the Evaluation and Activity dimensions, the hierarchy of sensory verbs  is  based  on  another  experiential  dimension,  related  to  the  degree  of  cer‐tainty. For example, the statement “I saw that the building was huge” implies more certainty than “I heard that the building was huge”, which depends on the reliabil‐ity of the source of information. The statement “I feel that Mary has doubts about her marriage”  is the  least certain. The dimension of certainty  is most relevant for verbs connected to sight, hearing, and touch.  

For the verbs connected to taste and smell the evaluative component seems domi‐nant  (‘to  taste  freedom’,  ‘to  smell  treason’).  There  are  relatively  few  olfactory terms in English and most of them refer to bad smells. Metaphorical meanings for smell verbs include the detection of bad characteristics, as in “Something about his testimony stinks” (Caplan, 1973; Viberg, 1984; Sweetser, 1990) and to suspect, to guess, to sense something intuitively, as in “She could smell money, power, victory” (Ibarretxe‐Antuñano,  1999). While  the  connotation  of  ‘smelly’  is  often  negative, that of  ‘tasty’  is positive, which Classen explains by  the  fact  that people are con‐fronted with  foul  smells more  often  than with  foul  tastes:  “We  can  choose  our food, but we cannot as readily close our noses to bad smells” (p. 53). Taste figura‐tively means  ‘judgment of what  is beautiful’ and  is,  therefore, characterized as a sense of aesthetic discrimination. This  relation between  taste and preferences  is very common cross‐linguistically (Buck, 1949). 

 

1.4. PRESENT RESEARCH 

The present research aims to  find out which sensory modalities are dominant  for different descriptors of product experiences. We assume that for sensory descrip‐tors of product experience the corresponding sensory modalities will be dominant, such as taste for ‘sweet’, touch for ‘hard’, and audition for ‘loud’ If an adjective has both a literal and a figurative meaning, we expect the link with the modality corre‐sponding to the literal meaning to be stronger, because this is the original meaning. As an original source is likely to have a stronger effect than a derivative, we expect the original modality to have a stronger connection to the descriptor than any of the modalities that are involved in the figurative meanings. Symbolic and affective descriptors of product experience are likely to be multisensory, but possibly one or 

Page 89: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

89 

more sensory modalities play a more dominant role  in assessing how  interesting, modern, exciting or funny people find a particular product.  

In Study 1 we try to find experience descriptors that are related to all sensory mo‐dalities to the same degree and are not biased towards any modality. According to our assumption, we expect to find these among the symbolic and affective descrip‐tors.  In Study 2 we  look  for adjectives  that are mainly related  to a single sensory modality. We expect to find these mostly among the sensory descriptors. In Study 2 we also  test whether a difference exists between  the sensory dependence of  the three different types of product experience descriptors (sensory, symbolic and af‐fective) for participants with different native languages. 

 

2. STUDY 1. SELECTING A SENSORY NEUTRAL EXPERIENCE  The aim of this study was to find product experience descriptions that are not bi‐ased  towards any of  the modalities. We assumed  that concrete, sensory descrip‐tors are likely to refer to a single sensory modality. An adjective that can be used in a  literal sense  for one modality and  in a metaphorical sense  for other modalities (e.g., rough, heavy) is likely to be dominated by the modality for which the descrip‐tor has  a  literal meaning  (i.e.,  touch).  Therefore, we were  looking  for  adjectives that were abstract and had a meaning that could be applied to the same degree to all sensory modalities. 

To develop a new  index for evaluating aesthetic  impressions in close relation to hu‐man modalities, Suzuki and Gyoba  (2001)  introduced  the  sensory  relevance  coeffi‐cient. This coefficient indicates the extent to which pairs of adjectives are related to sensory modalities. For each adjective pair, the coefficient reflects the proportion of participants that indicated that a particular modality was related most to this adjec‐tive. Examples of adjective pairs  that were  judged  to be primarily  related  to visual perception were beautiful‐ugly  (0.95), gay‐sober  (0.99), and clear‐cloudy  (0.86). Pri‐marily tactual were wet‐dry  (0.91), soft‐hard  (0.96), cold‐warm  (0.91), and smooth‐rough (0.78). The only adjective pair that was mainly auditory was living‐quiet (0.83). No adjective pairs in their study were primarily olfactory or gustatory. 

In a more recent study Gyoba, Suzuki, Kawabata, Yamaguchi, and Komatsu (2005) instructed  participants  to  indicate  the  degree  of  sensory  relevance  of  adjective pairs on 6‐point rating scales for 10 sensory modalities, including the perception of pain, warmness, coldness, equilibrium, and kinaesthesia. In general, they obtained 

Page 90: ANNA FENKO - TU Delft

CHAPTER 4 

 

90 

high ratings for vision for most pairs of adjectives. They found that adjective pairs 

that usually  span  the Evaluation  factor  in  Semantic Differential  studies (Osgood, 1962; Osgood et al., 1957), such as pleasant‐unpleasant and good‐bad  tended  to get high ratings for multiple modalities. Adjective pairs that span the Activity factor, such as quiet‐noisy, static‐dynamic, and calm‐turbulent received primarily high rat‐ings for the auditory and kinaesthetic modality. Adjective pairs that are related to the  Potency  factor,  such  as  smooth‐rough,  soft‐hard,  small‐big,  and  feminine‐masculine were rated high for the tactile sense. These outcomes have been partly supported by brain activity measurements while  the participants  rated stimuli on semantic differential scales (Suzuki et al., 2005). Activation patterns of the tempo‐ral and parietal regions of the brain were significantly related to the semantic po‐larities of Activity and Potency, but no changes in brain activity were related to the ratings on the Evaluation scale. 

To determine the extent to which adjective pairs are related to the different sen‐sory modalities, we need to assess the extent to which all modalities are related to the adjectives. In line with the studies performed by Gyoba, Suzuki, and colleagues, we asked participants to rate the relevance of four sensory modalities for each se‐mantic scale on a 5‐point scale. Because the research was performed in the context of the evaluation of durable consumer products (see Schifferstein et al., 2010), the taste modality was not included in this study.  

 

2.1. PARTICIPANTS 

A  convenience  sample  consisting  of  colleagues,  friends  and  acquaintances  was used, consisting of 57 participants  (42 male and 15  female). All participants were native Dutch speakers. One female participant was removed from the sample, be‐cause  she was  unable  to  smell  anything.  Ages  varied  between  20  and  60  years (mean 35.4). 

 

2.2. PROCEDURE 

Participants  received an email  that  invited  them  to go  to a  specific website. The 

introduction  to  the  study  pointed  out  that  the way  in which  people  experience 

products depends on the information perceived through the senses. Judging differ‐

ent aspects of the product experience, however, would not necessarily rely on each 

Page 91: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

91 

sensory modality  to  the  same  degree.  Participants  then  answered  the  question 

‘Suppose that you would have to rate a product on the following semantic scale, to 

what extent would the sensory modalities play a role?’ on a 5‐point scale, ranging 

from ‘not at all’ on the left side to ‘very large extent’ on the right side. A set of 39 

possible descriptors was derived from previous studies that used the Semantic Dif‐

ferential Method. In addition, 20 scales were derived from a set of unipolar product 

personality descriptors  (Govers, 2004) by creating  items  that  ranged  from  ‘not at 

all’ to ‘very’. All original items and their translations can be found in Table 4.1.  

 

TABLE 4.1. MEAN SENSORY RELEVANCE RATINGS FOR BIPOLAR AND UNIPOLAR 

DESCRIPTIVE ATTRIBUTES ON A 5‐POINT SCALE 

Modality English  Dutch  Ver‐sion 

Vision  Audition  Touch  Smell 

Evaluation 

Beautiful – ugly  Mooi – lelijk  A  5.0  3.7  2.9  2.2 

    B  4.7  3.4  2.7  2.6 

Pleasant – unpleas‐ant 

Aangenaam ‐ onaan‐genaam 

A  4.5  4.3  4.5  4.5 

Agreeable ‐ disagree‐able 

Prettig ‐ onprettig  B  3.8  3.9  3.9  3.9 

Good – bad  Goed ‐ slecht  B  3.6  3.4  3.2  2.8 

Friendly – unfriendly  Vriendelijk ‐ onvriendelijk  B  3.9  3.7  2.9  2.1 

Attractive – unattrac‐tive 

Aantrekkelijk ‐ onaantrek‐kelijk 

A  4.7  3.7  3.9  4.0 

Seductive ‐ Repulsive  Verleidelijk – afstotelijk  A  4.7  3.7  3.9  4.4 

Sympathetic – un‐sympathetic 

Sympathiek ‐ onsympa‐thiek 

B  3.8  3.8  2.5  2.4 

Inviting – rejecting  Uitnodigend – afwijzend  A  4.3  3.9  3.2  3.1 

Activity 

Active – relaxed  Actief ‐ ontspannen  A  4.3  3.7  3.0  2.1 

    B  3.9  3.4  3.5  2.3 

Stimulating – relaxing  Stimulerend ‐ ontspan‐ B  3.4  3.8  3.5  2.9 

Page 92: ANNA FENKO - TU Delft

CHAPTER 4 

 

92 

nend 

Excited – calm  Opgewonden ‐ kalm  B  4.1  4.3  3.4  2.6 

Fast – slow  Snel ‐ langzaam  B  4.2  3.6  2.8  1.5 

Fussy ‐ lethargic  Druk ‐ sloom  A  4.5  3.8  2.1  1.4 

Tense ‐ fusty  Gespannen ‐ duf  A  4.1  3.1  3.1  2.0 

Flashy ‐ discrete  Flitsend ‐ discreet  A  4.6  3.5  2.3  1.9 

Conspicuous ‐ incon‐spicuous 

Opzichtig ‐ onopvallend  B  4.6  3.4  2.0  2.7 

Lively – quiet  Levendig ‐ rustig  A  4.5  4.5  3.1  2.5 

Potency 

Gentle – tough  Schattig ‐ stoer  A  4.8  3.7  3.2  2.2 

    B  4.4  3.4  3.3  2.3 

Masculine – feminine   Mannelijk ‐ vrouwelijk  B  4.5  3.7  3.8  3.5 

Strong – weak  Sterk ‐ zwak  B  3.7  3.0  3.9  2.3 

Careful – brave  Voorzichtig ‐ dapper  A  4.3  3.0  2.6  1.9 

Momentous – humble  Gewichtig ‐ nederig  A  4.3  3.3  2.6  1.7 

Impressive ‐ meaning‐less 

Indrukwekkend ‐ nietszeggend 

A  4.5  3.9  2.8  2.4 

Persistent ‐ accommo‐dating 

Vasthoudend ‐ meegaand 

A  3.6  3.2  2.9  1.7 

Other 

Rough – soft  Ruw ‐ zacht  A  4.2  3.0  5.0  2.0 

    B  3.8  2.5  4.7  1.6 

Fresh – musty  Fris ‐ muf  A  3.4  1.9  2.6  5.0 

    B  3.0  2.0  2.6  4.7 

Loud – quiet  Luid ‐stil  A  2.3  4.9  2.2  1.3 

    B  1.9  4.9  2.2  1.4 

Mature – youthful  Volwassen ‐ jeugdig  B  4.1  3.4  2.5  2.1 

Young – old   Jong ‐ oud  B  4.3  3.1  2.8  3.3 

Interesting – boring  Interessant‐ saai  B  4.3  3.9  3.0  2.5 

Funny‐ serious  Grappig‐ serieus  B  4.0  4.0  2.2  1.6 

Safe – dangerous  Veilig ‐ gevaarlijk  B  4.3  3.9  3.5  3.2 

Page 93: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

93 

Severe – mild  Streng ‐ mild  B  3.4  3.5  2.7  2.0 

Expensive – cheap  Duur ‐ goedkoop  A  4.4  2.7  3.7  2.6 

Valuable – worthless  Waardevol – waardeloos 

A  4.6  3.6  3.8  2.6 

Comprehensible – in‐comprehensible 

Begrijpelijk ‐ onbegri‐jpelijk 

A  4.2  3.1  2.3  1.3 

Predictable –unpredictable 

Voorspelbaar ‐ onvoor‐spelbaar 

A  4.0  3.5  3.0  2.6 

Modern – traditional  Modern ‐ traditioneel  A  4.7  3.4  2.9  2.8 

Personality 

Dominant  Dominant   B  4.0  4.1  2.4  2.4 

Cute  Schattig  B  4.3  3.2  3.1  2.1 

Provocative  Uitdagend   B  4.3  2.9  2.7  2.2 

Cheerful  Vrolijk  B  4.3  4.2  2.3  1.7 

Pretty  Leuk  B  4.1  3.7  3.0  2.2 

Childish  Kinderachtig   B  4.0  3.6  2.6  2.0 

Interesting  Interessant   B  4.2  3.9  3.1  2.5 

Silly  Dom   B  3.5  3.7  2.0  1.4 

Relaxed  Relaxed   B  3.4  3.9  3.4  2.2 

Idiosyncratic  Eigenzinnig   B  3.7  3.8  2.6  2.3 

Lively  Pittig   A  2.9  1.8  4.1  2.2 

Open   Open   A  4.4  3.1  2.9  1.6 

Aloof  Afstandelijk   A  4.2  3.1  3.0  2.1 

Honest   Eerlijk   A  3.8  3.6  2.7  2.1 

Untidy  Slordig   A  4.7  2.8  2.8  1.8 

Boring   Saai  A  4.5  3.9  2.7  2.3 

Modest  Bescheiden   A  4.0  4.1  2.5  2.1 

Easy‐going  Vlot   A  4.5  3.8  2.3  2.0 

Obtrusive  Opdringerig   A  4.3  4.3  3.0  3.2 

Serious  Serieus   A  4.2  4.0  2.7  2.0 

 

Page 94: ANNA FENKO - TU Delft

CHAPTER 4 

 

94 

To  limit  the  number  of  questions  the  participants  had  to  answer,  two  separate questionnaires were used,  each  containing  about half of  the  semantic  items.  To evaluate between‐sample  reliability,  six  items were  identical  in  the  two versions. The order  in which the modalities were presented was different for both surveys. For each semantic item, the sensory relevance was rated for olfaction, touch, audi‐tion, and vision (version A) or audition, olfaction, vision, and touch (version B), con‐secutively. Thirty‐three (version A) or 32 (version B) items were rated. Twenty‐nine participants filled out version A, 28 filled out version B. The time needed to fill out the questionnaire was approximately 15 min. 

 

2.3. RESULTS 

For each  item, responses on the 5‐point scales were subjected to repeated meas‐ures  ANOVA with Modality  as within‐subjects  factor. Most  items  showed  biases towards a sensory modality (see Table 4.1).  In contrast to Gyoba et al. (2005), we found that most  items that referred to the Activity and Potency dimensions were most  strongly  related  to  the  visual  modality.  In  accordance  with  Gyoba  et  al. (2005), most of the  items spanning the Evaluation factor were related to multiple modalities. Only four semantic scales did not show a significant difference between sensory modalities: agreeable‐disagreeable (p=0.83), pleasant‐unpleasant (p=0.55), stimulating‐relaxing  (p=0.05), and good‐bad  (p=0.05). Therefore,  the Pleasantness of the product seems to be a sensory neutral product experience. This experience can be assessed by averaging the responses on three items that appear to be sen‐sory neutral (agreeable‐disagreeable, pleasant‐unpleasant, and good‐bad).  

To  test  if age or gender differences between  respondents  influenced our results, we performed  repeated measures ANOVA on  the A and B data  sets, with Modality and Experience as within‐subject factors and Age and Gender as between‐subject factors. All effects that included the two demographic factors were not significant (all p > 0.20).  

 

3. STUDY 2. MODALITY IMPORTANCE FOR SENSORY 

PRODUCT DESCRIPTIONS Almost all product experiences have associations with more  than one sensory 

modality.  The  color  of  a  dress  can  be  ‘loud’;  the  voice  of  a  singer  can  be 

Page 95: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

95 

‘sweet’,  and  so  on.  The  primary  aim  of  the  second  study was  to  determine 

which sensory descriptors of product experience rely mainly on a single modal‐

ity  and  which  are  primarily multisensory. We  included  several  symbolic  and 

affective experiences  in this study as control variables. Based on the results of 

Study 1, we assumed that symbolic and affective experiences would be mostly 

multisensory.  

Another  aim of  Study 2 was  to  look  at  language differences  in modality domi‐

nance  for  three  types  of  product  descriptors:  sensory,  symbolic,  and  affective. 

When the meaning of a product is expressed in words, it may be interpreted dif‐

ferently  in  different  languages. Most  adjectives  that  describe  product  experi‐

ences have  several meanings, and usually not all  these meanings  can be  trans‐

lated adequately to another  language. For example, the English word ‘fresh’ has 

16 different meanings (Simpson and Weiner, 1989), which can be roughly divided 

into two groups: 1) new, recent, newly made, recently arrived, retaining its origi‐

nal qualities, not deteriorated or changed by lapse of time; 2) pure, invigorating, 

refreshing  (said  especially  of  air  and water),  not  stale, musty,  or  vapid.  In  the 

Dutch  language two different words are used to  indicate these two meanings  in 

the  case  of  food  products  (vers  for  the  first meaning  and  fris  for  the  second 

meaning). When a  text  is  translated  from English  into Dutch,  the  translation of 

the word  ‘fresh’  is  likely  to have a more  restricted meaning and  fewer associa‐

tions in Dutch than in English. We assumed that some descriptors of product ex‐

perience  are  culturally  specific  and bear  linguistic  associations.  To  test  this  as‐

sumption, we conducted the study with two groups of respondents: native Dutch 

speakers and native Russian speakers. 

 

3.1. PARTICIPANTS 

The sample consisted of 57 Dutch‐speaking students of TU Delft  (30 men and 27 

women) and 55 Russian‐speaking students (27 men and 28 women) of the Higher 

School of Economics (Moscow, Russia). For the Dutch sample, ages ranged from 21 

to 34 years, mean age was 23.5 years. For the Russian sample, ages ranged from 17 

to 36, mean age was 20.2 years. 

 

 

Page 96: ANNA FENKO - TU Delft

CHAPTER 4 

 

96 

3.2. PROCEDURE 

Respondents  received  paper  questionnaire  forms  at  a  lecture  and  filled  them  in during the break. The questionnaire asked respondents to indicate to what extent different  sensory modalities  contributed  to  the evaluation of 34 product proper‐ties. The list included sensory descriptors: tactile (such as warm, sharp, and rough.); auditory (such as noisy, quiet, and loud); visual (such as colourful, shiny, and clear); olfactory (such as fresh and stale), and gustatory (such as bitter and sweet); affec‐tive descriptors (such as exciting, funny, and cute); and symbolic descriptors (such as complex, modern, and luxurious). Respondents were asked to think of any prod‐uct  (a coffee maker, shoes, cheese, a  tooth brush, shampoo, a camera, a chair, a soft drink, a bag, etc.) and to answer the question: “To what extent do the follow‐ing senses contribute to your evaluation of a product as…?” They assessed the im‐portance of 5 sensory modalities on 5‐point scales from ‘not important’ (1) to ‘very important’  (5).  Because  the  second  study was  performed  in  the  context  of  the evaluation of a set of products that  included a soft drink (see Fenko et al., 2009), taste was added to the list of modalities in this study.  

There were two types of questionnaires which differed with respect to the order of the sensory modalities. The A‐form used the following sequence: audition, olfaction, touch, taste, and vision. The B‐form used the sequence: touch, taste, vision, olfaction, and audition. In the Dutch sample, 26 filled out the A‐form, and 31 filled out the B‐form. In the Russian sample, 27 respondents filled out the A‐form, and 28 filled out the B‐form. It took participants 10 to 15 minutes to fill out the questionnaire. 

3.3. RESULTS 

We performed repeated measures ANOVA with Descriptor and Modality as within‐subjects  factors  and  Language  as  between‐subjects  factor.  The main  effects  for Descriptor, Modality and Language were significant: F (33, 3234) = 39.9; p < 0.001 for Descriptor; F (4, 392) = 332.8; p < 0.001 for Modality; and F (1, 98) = 4693.5; p < 

0.001 for Language. The Modality × Language interaction was not significant: F (4, 392) = 1.5; p > 0.2. The Modality × Descriptor interaction [F (132, 12936) = 133.5; p < 0.001] and  the Descriptor × Modality × Language  interaction  [F  (132, 12936) = 13.8, p < 0.001] were both significant.  

To  test  if age or gender differences between  respondents  influenced  results, we also performed repeated measures ANOVA with Descriptor and Modality as within‐subjects  factors  and  Language, Age  and Gender  as between‐subject  factors.  The 

Page 97: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

97 

effects of both demographic factors and their interactions with other factors were not significant (all p > 0.20).  

Overall mean ratings of the 34 experiences showed that vision had the highest rat‐ing  for 14 descriptors,  touch  for 10,  taste  for 4, and audition and olfaction  for 3 descriptors  each.  Symbolic  descriptors  tended  to  rate  high  on  all modalities,  al‐though all of them demonstrated visual dominance. As predicted, sensory descrip‐tors  demonstrated  the  highest  ratings  for  the  corresponding  sensory modalities (e.g., touch for ‘warm’, taste for ‘bitter’, audition for ‘loud’, olfaction for ‘fresh’).  

Because the three‐way interaction was highly significant, responses for each adjec‐tive were subjected to repeated measures ANOVA with Modality as within‐subjects factor and Language as between‐subjects factor. The effect of Modality was signifi‐cant  for all descriptors  (p < 0.01). For 24 out of 34 descriptors,  the effect of  the 

Modality × Language interaction was significant (see Table 4.2). 

Some of the cultural differences were big enough to affect the importance hierar‐chy  of modalities  for  particular  descriptors.  For  example,  ‘spicy’  appeared  to  be mainly a gustatory experience for Dutch respondents (kruidig) and olfactory for the Russian sample (ароматный). The dominant modality for ‘mild’ in the Dutch group (mild) was also taste, but  in the Russian group (мягкий) touch was dominant, fol‐lowed by vision. Olfaction was dominant  for  ‘stale’  in  the Dutch group  (muf), but for Russians ‘stale’ (старый) was a visual experience. For Dutch respondents ‘pure’ (puur) was a gustatory experience, while for Russians (строгий) it was visual.  

These  differences  encouraged  us  to  perform  further  analysis  on  the  items. We looked at  the response distribution patterns of all  the  items  to evaluate whether the responses were distributed normally within each language group. In the Dutch sample 13 bi‐modal distributions  (7.6 %) were observed by visual  inspection: one for olfaction  (for  colourful), 4  for  audition  (hard, pure, modern,  and  cute), 4  for touch (luxurious, pure, modern, and quick), and 4 for taste (sharp, luxurious, clean, and conspicuous). In the Russian sample there were 31 bi‐modal distributions (18.2 %): 3 for vision (sharp, sweet, and sour); 8 for audition (exciting,  luxurious, rough, complex, pure, modern, cute, and clear), 9  for  touch  (exciting, sharp, quiet, com‐plex, pure, funny, modern, beautiful, and conspicuous), 5 for olfaction (bitter, luxu‐rious, modern, cute, and beautiful), and 6 for taste (warm, hard,  luxurious, rough, complex, and clean). The bi‐modal distributions suggest that there are sub‐groups of participants within each sample that differ  in their opinion on modality  impor‐tance for a particular descriptor.  

 

Page 98: ANNA FENKO - TU Delft

CHAPTER 4 

 

98 

TABLE 4.2. MEAN RELEVANCE RATINGS OF MODALITIES FOR 34 DESCRIPTORS OF PRODUCT EXPERIENCES 

Language  Modality 

English  Dutch/Russian  Vision  Audition  Touch  Olfaction  Taste 

Visual descriptors 

fleurig  4.7  1.7  1.9  3.7  2.1 **Colourful 

красочный  4.9  1.4  1.6  1.4  1.4 

schoon  4.6  1.4  3.7  4.3  2.7 **Clean 

чистый  4.8  2.0  3.9  3.5  2.6 

glanzend  4.9  1.5  3.6  1.1  1.1 **Shiny 

яркий  5.0  1.6  1.7  1.6  1.8 

helder  4.7  3.0  2.1  2.1  2.3 Clear 

ясный  4.5  2.8  2.2  2.1  1.8 

opvallend  4.7  3.6  3.5  3.4  3.2 *Conspicuous 

заметный  4.7  3.7  2.8  3.0  2.6 

mooi  4.9  3.2  3.4  2.5  2.1 Beautiful 

красивый  4.8  2.9  2.9  2.3  2.0 

breekbaar  4.4  2.9  4.3  1.1  1.2 **Breakable 

хрупкий  4.2  2.1  4.4  1.3  1.5 

snel  4.7  3.9  2.4  1.5  1.3 *Quick 

быстрый  4.8  3.3  2.4  1.6  1.4 

Tactile descriptors 

warm  3.3  1.9  4.6  2.1  2.6 **Warm 

теплый   2.3  1.3  4.6  2.0  3.5 

scherp  4.0  2.3  4.4  2.4  3.2 **Sharp 

острый  2.9  1.6  3.4  2.5  4.3 

hard  3.8  3.0  4.7  1.3  1.9 **Hard 

твердый  3.3  1.6  4.8  1.3  2.4 

**Rough  ruw  4.1  2.2  4.8  1.4  1.8 

Page 99: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

99 

грубый  4.1  3.1  4.3  1.9  2.5 

sterk  4.2  2.5  4.4  2.3  2.5 *Strong 

прочный  3.9  2.0  4.5  1.3  2.1 

zwaar  4.1  2.5  4.6  1.4  1.2 *Heavy 

тяжелый  4.0  2.6  4.3  1.4  1.8 

flexibel  4.0  2.2  4.6  1.2  1.2 Flexible 

гибкий  4.1  2.0  4.7  1.3  1.8 

vochtig  3.9  1.7  4.7  2.5  2.8 Moist 

влажный  3.8  1.8  4.7  2.9  3.4 

Auditory descriptors 

luid  2.7  4.9  1.6  1.2  1.2 *Loud 

громкий  2.1  4.8  1.6  1.4  1.3 

stil  2.6  4.9  2.1  1.2  1.2 **Quiet 

спокойный  4.1  4.3  3.1  2.4  1.8 

lawaaierig  3.1  4.9  1.8  1.3  1.2 Noisy 

шумный  3.1  4.9  1.6  1.3  1.2 

Gustatory descriptors 

bitter  2.2  1.2  1.5  3.2  4.8 **Bitter 

горький   1.8  1.3  1.4  2.4  4.9 

zoet  3.1  1.4  1.8  4.3  4.9 *Sweet 

сладкий  3.2  1.4  2.1  3.7  4.9 

zuur  2.4  1.1  1.3  3.8  4.9 Sour 

кислый  2.8  1.2  1.6  3.5  4.9 

kruidig  2.6  1.1  1.6  4.2  4.7 **Spicy 

ароматный  2.2  1.5  1.6  4.6  3.1 

mild  2.6  1.8  2.4  3.5  4.4 **Mild 

мягкий  3.4  2.1  4.8  1.8  2.6 

puur  3.9  2.4  2.6  3.6  4.2 **Pure 

строгий  3.9  2.8  2.4  1.7  1.8 

Page 100: ANNA FENKO - TU Delft

CHAPTER 4 

 

100 

Olfactory descriptors 

fris  3.9  1.8  2.7  4.4  4.1 Fresh 

свежий  4.1  1.9  3.3  4.6  4.5 

muf  3.0  1.6  2.2  4.7  3.4 **Stale 

старый  4.5  2.2  3.7  3.0  2.5 

Symbolic descriptors 

luxueus  4.7  2.8  3.7  3.1  2.9 Luxurious 

роскошный  4.9  2.9  3.3  3.0  2.9 

complex  4.9  2.9  3.4  1.9  2.1 Complex 

сложный  4.6  3.1  3.2  2.1  2.6 

modern  4.8  2.6  3.2  2.0  1.8 **Modern 

современный  4.5  3.5  2.6  2.5  2.2 

interessant  4.7  3.5  3.9  3.1  3.0 *Interesting 

интересный  4.6  4.0  3.6  3.2  3.2 

Affective descriptors 

opwindend  4.3  2.9  3.8  3.3  3.2 **Exciting 

волнующий  3.8  3.1  3.1  3.4  2.6 

grappig  4.6  3.9  3.2  2.2  2.1 Funny 

забавный  4.4  3.6  2.8  2.1  2.0 

schattig  4.7  3.0  3.6  2.1  1.6 **Cute 

изящный  4.8  2.8  3.5  2.3  2.7 

** The effect of Modality X Language interaction is significant at 0.01 level.  

* The effect of Modality X Language interaction is significant at 0.05 level. 

 

4. DISCUSSION 

4.1. SENSORY RELEVANCE OF ADJECTIVES IN TWO LANGUAGES 

We set out to determine the  importance of various sensory modalities for different types of descriptors  of product  experience. As predicted,  the  affective descriptors 

Page 101: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

101 

(pleasant‐unpleasant,  stimulating‐relaxing,  and good‐bad)  relied equally on all  sen‐sory modalities. The  symbolic descriptors  (complex,  luxurious, modern,  interesting) were  also multisensory, but  relied mainly on  the  visual modality.  The  latter  result corresponds to findings of visual dominance both  in the field of product experience (Schifferstein, 2006) and  in  the area of  sensory  semantics  (Viberg, 1984; Sweetser, 1990). As we predicted, for many sensory descriptors the corresponding sensory mo‐dality was dominant (e.g., audition for noisy, taste for bitter, touch for hard, vision for shiny). Nevertheless, we also found several sensory descriptors for which more than one modality was important. For example, fresh is mostly an olfactory adjective, but it has also high importance ratings for taste, vision, and touch in both the Dutch and the Russian sample. Most tactile adjectives (rough, heavy, moist, warm, flexible) also have high importance ratings for the visual modality. This agrees to Williams’ (1976) suggestion that touch is the main source domain and vision is the main target domain for the metaphorical transfer of meaning in sensory adjectives.  

Significant  language differences were found for various sensory descriptions (such as colorful, rough, quiet, spicy, and stale). The possible explanation for this result is that semantic connotations for sensory adjectives differ considerably between the languages. For  instance, scherp  in Dutch and острый  in Russian are both equiva‐lents  to  the English word  ‘sharp’, of which  the  literal meaning  is  tactile  (having a keen edge or point). In English and Dutch it is also possible to characterize a high‐pitch sound or pungent food as sharp. In Russian another word is used for charac‐terizing a  sharp  sound  (резкий),  so audition has a  low  importance  rating  for  the experience of sharp (1.6). On the other hand, gustatory associations with sharpness are even stronger for the Russian sample than tactile associations (importance rat‐ing 4.3 for taste vs. 3.4 for touch). This can be due to the fact that for the Russians associations were stronger with food than with tools like knives or scissors.  

Vice versa, for Dutch respondents gustatory associations with ‘pure’ (puur) were as strong as the visual associations (importance rating 4.2 for taste vs. 3.9 for vision). Although the visual rating of ‘pure’ (чистый) was identical for the Russian sample (3.9), taste had  little  importance for pure (1.8). The same was true for  ‘mild’. The Dutch word mild showed the highest importance rating for taste (4.4), but the Rus‐sian equivalent мягкий was mostly tactile (4.8). Apparantly, mild in Dutch is similar to  ‘mild’  in English when applied to food: not sharp, strong, or hot  in flavour, not pungent, while мягкий  in Russian  is more synonymous to  ‘soft’ and refers mostly to tactile properties. 

When participants assess particular sensory modalities as  important  for  judging a specific product attribute, they probably imagine a product that has this attribute. 

Page 102: ANNA FENKO - TU Delft

CHAPTER 4 

 

102 

These products may be considered as ‘prototypical’ in a sense that they first come to mind  in  association with  a  certain  attribute  (Rosch,  1978; Mervis  and  Rosch, 1981).  The  fact  that  the  assessment  of  sensory modalities  varies  between  lan‐guages may  suggest  that prototypical products may  also  vary  cross‐linguistically. For  example,  knives  and  scissors  may  be  prototypical  ‘sharp’  products  for  the Dutch, while spicy foods may be prototypical for the Russians. Therefore, the  lan‐guage differences found in the present study may be related to the categorization processes which were  shown  to  differ  between  speakers  of  different  languages (Schmitt and Zhang, 1998; Zhang and Schmitt, 1998). Future research can evaluate whether prototypical products for specific attributes differ between languages. 

The  language differences we  found  in  the present study might have an effect on how  people  experience  products.  According  to  the  linguistic  relativity  proposal (Whorf,  1956),  speakers  of  different  languages  perceive  and  conceive  the world differently. Language may act as a  filter  through which people view reality  in  the process of perception, categorization and  the  interpretation of  information  (Hunt and Agnoli, 1991). Differences in the linguistic coding have been shown to correlate with differences  in  the non‐linguistic conceptual coding  (Levinson, 2003; Majid et al., 2004; Levinson and Wilkins, 2006), suggesting that  linguistic distinctions affect how we  think  (see also Lucy, 1992; Boroditsky, 2001). Lexical differences can  im‐pact  perception  (Kay  and  Kempton,  1984;  Davidoff  et  al.,  1999)  and  cognition (Gordon,  2004).  Further  research  is  needed  to  evaluate  whether  differences  in product  experiences  exist  between  people  speaking  different  languages,  and whether any such differences can be  traced back  to differences  in sensory adjec‐tives and sensory metaphors in different languages.  

Some  authors  argue  that  each  culture  has  its  own mental  frames  (Hong  et  al., 2000),  which  are  learned  and  used  in  conjunction  with  that  culture’s  language (Foucault, 1972). As a result, words in two different languages that may seem to be exact translations of each other are  likely to have different sets of culture‐specific connotations (Kroll and De Groot, 1997), reflecting the differences in cultural frame content.  

Because the interpretations of words vary among cultures and individuals, it is dif‐ficult to standardize verbal communication with research participants from differ‐ent cultures.  Incompatibility of  testing methods creates methodological problems for  interpreting cross‐cultural differences.  It  is difficult to reach  item equivalence, to make sure that the  instruments used  in the research are similar, even with the most accurate translation and back translation (Brislin, 1980). 

Page 103: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

103 

In  cross‐cultural  studies,  it  is  extremely  difficult  to  separate  a  cultural  effect 

from a  language effect, because people  from different  cultures  tend  to  speak 

different  languages.  This  is why people who  speak  two  languages  (bilinguals) 

and people who internalized the values, beliefs, and norms of two cultures (bi‐

culturals) have become the focus of research in psycholinguistics, anthropology 

and consumer studies. Biculturals often report feeling “like a different person” 

when  they  speak  a  different  language  (LaFromboise,  Coleman,  and  Gerton, 

1993). This suggests that biculturals may possess two different culture‐specific 

mental  frames  for  a  single  word  in  two  different  languages  (translation‐

equivalent words)  (e.g.,  Hong  et  al.,  2000).  Using  bicultural  bilingual  partici‐

pants can be a useful option in the future studies of cross‐cultural and language 

differences in product experience.  

 

4.2. COGNITIVE AND SENSORY MODES OF INFORMATION PROCESSING 

In  this  study we  found more  language  differences  for  sensory  descriptors  of 

product experiences  than  for  symbolic descriptors. This  result may be  related 

to  the differences between cognitive and  sensory modes of  information proc‐

essing (Hirschman, 1984; Pearson, 1970). These different modes  initiate differ‐

ent mental and physical activities. Cognitive modes of  information processing 

refer  to  logical,  rational,  sequential  thought  processes  and  verbal modes  of 

information processing  in  contrast  to holistic, gestalt, and visual modes of  in‐

formation processing  (Childers et al., 1985). Cognitive experiences are  related 

to the need for being adaptively oriented to the environment and for achieving 

a sense of meaning, while sensory and affective experiences are related to the 

need  to  achieve  a  feeling  of  satisfaction  and  to  attain  emotional  goals 

(McGuire, 1976). We can assume that  logical thinking and rationality  is univer‐

sal across cultures, while sensory experiences and emotional goals will vary not 

only between social and cultural groups, but also between  individuals. The  lat‐

ter  suggestion was  supported  by  our  finding  that modality  ratings  for  some 

sensory experiences demonstrated bi‐normal distribution patterns, even within 

a  relatively  uniform  group  of  students  of  the  same  age  and  education  level. 

Additional research is needed into the associations people have with particular 

descriptions  in  various  situations,  in  order  to  specify  the  exact  meaning  of 

these adjectives for different products and different user groups.  

Page 104: ANNA FENKO - TU Delft

CHAPTER 4 

 

104 

5. CONCLUSIONS This research  investigated the  importance of sensory modalities for various descrip‐tors of product experience. The results showed that only for the affective descriptors of  product  experience  (pleasant‐unpleasant,  good‐bad)  all modalities  were  about equally  important.    Symbolic descriptors  (such as modern, expensive, or  feminine) generally  convey  the  social  or  personal  meaning  of  products.  These  descriptors tended to be multisensory, but most of them demonstrated visual dominance. Sen‐sory descriptors (such as colorful,  loud, or soft) reflect the perception of sensory  in‐formation.  In  addition,  sensory meaning may  transfer  to other domains of  experi‐ence. We  found  that  sensory descriptors of product experience  showed  significant language differences. The latter result can be explained by the fact that metaphorical meanings of sensory descriptors differ between the languages. 

Further research with a more broad variety of  languages  is needed to  investigate the differences  in  sensory metaphors.  In  cognitive  linguistics,  sensory metaphors (such as KNOWLEDGE  IS VISION or UNDERSTANDING  IS GRASPING) are viewed as conceptual metaphors which  are  fundamental  to  language,  thought,  and  experi‐ence (Lakoff & Johnson, 1980, 1999; Sweetser, 1990). Whether such metaphors are universal or vary between languages and cultures is a question of significant impor‐tance  for current cognitive  linguistic  theory. Most cognitive  linguists assume  that conceptual metaphors are explicit abstract representations of embodied cognitive structures. According  to Lakoff &  Johnson  (1999), conceptual metaphors  (such as POSITIVE IS UP, NEGATIVE IS DOWN) could be established as people implicitly learn associations between physical experiences and emotional states  that  typically co‐occur. Kovecses (2005) distinguishes between universal metaphors that are rooted in  the  universal  bodily  experience  and  conventional metaphors  that  vary  across cultural, social, regional, style, developmental and other dimensions. Some scholars suggest that conceptual metaphors are not merely representative of universal body experience, but are tied to specific socio‐cultural cognition  (Kimmel, 2006; Zlatev, 2006). Metaphors  could be  learned  from patterns  in  language  and  culture  (Ibar‐retxe‐Antuñano, 2008);  they  are  not  necessarily  “embodied”  (Boroditsky,  2000; Gentner et al., 2001). Even if direct bodily experience is necessary on the timescale of biological or cultural/linguistic evolution,  it may not be necessary on  the  time‐scale of the conceptual development of an  individual  (Tomasello, 2003; Vygotsky, 1986). Further  research  into  the  cultural differences  in  sensory metaphors might help to resolve this theoretical argument. 

Page 105: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

105 

As concerns practical  implications,  it  should be noted  that  sensory adjectives are frequently used in advertising to describe sensory and symbolic properties of prod‐ucts  (Ruiz,  2006).  Since  most  sensory  adjectives  have  additional  metaphorical meanings which can differ between  the  languages,  it  is  important  for advertisers who want to use the same slogans and product descriptions  in countries with dif‐ferent  languages,  to  be  aware  of  the  polysemy  of  sensory  adjectives  and  cross‐cultural differences in their metaphorical meanings. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Page 106: ANNA FENKO - TU Delft

CHAPTER 4 

 

106 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Chapter 5 is published as: 

Fenko, A., Schifferstein, H.N.J., Hekkert, P. (2010) Looking hot or feeling hot: What determines  the  product  experience  of  warmth? Materials  &  Design,  31,  1325–1331. 

Page 107: ANNA FENKO - TU Delft

107 

 

 

 

 

 

 

 

 

CHAPTER 5 

LOOKING HOT OR FEELING HOT:  

WHAT DETERMINES THE PRODUCT 

EXPERIENCE OF WARMTH?  

 

1. INTRODUCTION Our perceptual  experience of  the world  is  richly multimodal.  People  are  able  to extract information derived from one sensory modality and use it in another. Peo‐ple can, for example, know a shape by touch and identify it correctly by sight (Cal‐vert et al., 2004). Furthermore, people are able to integrate the impressions gener‐ated by different sensory modalities into a unified, rich percept (Stein & Meredith, 1993). When users interact with products, more than one sensory system is usually involved in this interaction. People can see the product, touch it, hear the sound it makes and smell  it. All the senses are  involved simultaneously, but their contribu‐tion to the overall experience is not necessarily equivalent. It is interesting to know which sensory modality plays a  leading role  in a particular experience, so that de‐signers may concentrate on the creation of the most relevant product properties.  

Page 108: ANNA FENKO - TU Delft

CHAPTER 5 

 

108 

1.1. EXPERIENCING WARMTH 

The aim of our research is to investigate sensory dominance in the product experi‐ence of warmth. In a literal sense, terms such as “warm” and “cold” describe ther‐mal properties of products and thus refer to tactual perception. However, warmth is a multisensory product experience that may also include visual, olfactory and, in some cases, also gustatory and auditory components. The  factors  that contribute to the subjective experience of warmth will be discussed below. 

Objects made  of  different materials  feel  thermally  different.  For  instance, wood generally  feels warmer than metal, even though both materials are at room tem‐perature.  This  effect  is  caused by differences  in  the  thermal properties of  these materials. An object feels cold if it extracts warmth from the skin. To produce this sensation, the object does not only need to have a temperature below body tem‐perature,  it also has  to extract warmth at a  fast rate:  the material needs  to have low temperature resistance. Examples of such materials are glass and metal (Ashby &  Johnson,  2002). Materials with  high  temperature  resistance,  such  as wood  or plastics,  generally  feel  ‘warm’ even  if  their  temperature  is below body  tempera‐ture. When the material is hotter than the observer’s hand, the subjective coldness is reversed. For example, a copper sample that felt colder than a wood sample at room  temperature, will  feel warmer  than  the  same wood  sample when both are felt at a high temperature.  

The perceived warmth depends on  the  thermal conductivity and heat capacity of the material and on the object’s geometry. High thermal conductivity allows heat extracted from the finger to spread quickly to other parts of the object, thus ena‐bling the object to extract heat from the finger faster. A high heat capacity means that the object does not warm up very much from outside heat, which enables it to continue extracting heat from the finger. The geometry of the object also plays an important role: a thick bar can conduct heat away from the finger more easily than a thin foil (Bergmann Tiest & Kappers, 2008).  

But  the  product  experience  of warmth  goes  beyond  the  perception  of  thermal properties and  is,  therefore, not necessarily  restricted  to material properties. For instance,  certain  colors  are  experienced  as warm  and  others  as  cool  (Arnheim, 1959; Itten, 1961; Ross, 1938). Although there are some inconsistencies and ambi‐guities, most  psychological  research  has  supported  the  conclusion  that  red  and yellow are perceived as warmer than green and blue. For example, Lewinski (1938) and  Ross  (1938)  found  that  projected  red  colors were  rated warmer  than  blue ones, and Tinker  (1938) and Newhall  (1941)  found  that surface reds were consis‐

Page 109: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

109 

tently  described  as  warmer  than  surface  blues. Wright  (1962)  and Wright  and Rainwater (1962) demonstrated that there is a well‐defined effect of hue on judg‐ments of the apparent warmth of colored squares, independent of brightness and saturation. They also demonstrated that darker and more saturated colors tended to be viewed as warmer.  

It  is still not clear  if  these associations are  innate or  if  they are acquired  through experience. Morgan  et  al.  (1975)  explored  age differences  in  the  associations of temperatures with  specific  colors.  Their  results  demonstrated  the  relatively  late development (between the ages of 6 and 12) and the fragile nature of the conven‐tional associations between temperature and color. The authors argue that these associations are based on cultural norms, which are  learned only gradually during later  childhood  and  adolescence  and  are  far  from  universal,  even  in  college‐age subjects. But more research is needed to test this hypothesis. 

Another  interesting question  is whether temperature associations with color exist only for abstract color samples (paper chips) or can also be found in real products. In a color‐meaning study that used 5 objects painted in 6 different colors (Osgood et al., 1957),  consistent  color effects were  found  for warmth, e.g. all objects ap‐pearing  in red were consistently rated as warmer than those  in other colors. Taft (1997) compared semantic ratings of color chips with those of the same colors ap‐plied to a variety of familiar objects. He reported that colors were rated nearly the same on the warm‐cold scale despite context differences, although the color chips were rated as slightly warmer than the objects of the same color. 

There  is also evidence that color can  induce thermal sensations. Michael and Rol‐hion  (2008) asked  subjects  to  sniff a bottle  containing distilled water mixed with odorless yellow, green or red coloring and one colorless solution. Participants re‐ported  feeling a  cool  sensation  in  their nasal  cavity when  they  sniffed  the green solution and a warm sensation when they sniffed the red solution.  

According to people’s self‐reports, they find touch more important than vision for judging warmth.  In a questionnaire study (Fenko, 2008) respondents assessed the relative importance of five sensory modalities for 34 product experiences. The do‐minant modality  for  the warm experience was  touch  (mean  rating 4.6 out of 5), followed by vision (3.3) and taste (2.6).  

In reporting their sensory experience people can use the terms “warm” and “cold” either  in a  literal or  in a  figurative  sense. Only  the  tactile  component of warmth refers to its literal meaning. The experience of warm colors may be induced by as‐sociations with hot and cool elements (such as fire and water). Warmth in a figura‐

Page 110: ANNA FENKO - TU Delft

CHAPTER 5 

 

110 

tive sense may also be associated with affection and tenderness, comfort and cozi‐ness, sexuality, anger, and so on. Sensory terms that describe physical properties of things also describe psychological qualities in different languages (Asch, 1955). For example,  the morpheme  for  ‘hot’ stands  for  rage or wrath  (Hebrew), enthusiasm (Chinese, Malayalam), sexual arousal (Thai), worry (Thai), energy (Hausa, a member of the Sudanese family), or nervousness (Shilha, a Berber language). In spite of the differences between languages, all psychological meanings of “hot” refer to height‐ened activity, increased energy and emotional arousal. In contrast, the morpheme for  ‘cold’  stands  for  self‐possession  (Hebrew),  indifference  or  hostility  (Chinese), loneliness (Thai), laziness or apathy (Hausa) (Asch, 1955, p.33). 

Asch (1958) believed that when we describe psychological events in the same way as  the  physical  aspects  of  things, we  are  referring  to  functional  properties  they share.  Contemporary  cognitive  linguists  also  believe  that  people  conceptualize their psychological experiences by analogy  to  the physical world  (Johnson, 1987). According  to  the  theory of embodied cognition  (Lakoff &  Johnson, 1999), mental metaphors, such as positive is up, negative is down, could be established as people implicitly  learn  associations  between  physical  experiences  and  emotional  states that typically co‐occur (people spontaneously assume upright body postures when they  are  feeling  good).  Embodiment  theorists  point  out  that  objects  and  events that produce  the  same quality  of  affective  response  are  categorized  together  in memory  (Niedenthal  et  al.,  1999).  Therefore,  the  feelings  of warmth when  one drinks a hot coffee or takes a warm bath might activate memories of other feelings associated with warmth (trust and comfort). 

 

1.2. SENSORY DOMINANCE 

Previous studies have shown that people tend to consider vision as more important for product experience  than other  sensory modalities.  For example,  Schifferstein and Cleiren (2005) demonstrated that consumers acquired most of the information on products by vision and touch: this  information was most detailed and the sub‐jects were  surest of  their  judgments. The experiment also  showed  that products were harder to identify by sound or smell than by vision or touch. A questionnaire study in which participants reported the importance of the sensory modalities dur‐ing the usage of 45 different products  (Schifferstein, 2006) demonstrated that on average the relative importance sequence of sensory modalities is vision, followed by  touch, smell, audition and  taste.  In addition, when people were asked  to  rate 

Page 111: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

111 

how important they found the different modalities in their lives in general, most of them selected vision as the most important modality.  

However,  the  importance  ratings  for  the  sensory modalities  differed  greatly  be‐

tween products. For about half of  the 45 products,  the  importance of vision was 

lower than for other modalities. For example, audition  is the most  important mo‐

dality  for a washing machine and a coffee maker, which can be explained by  the 

role of  the  sound  in  signalling  the different  stages of  the process of washing or 

making  coffee.  Touch  is most  important  for  a  computer mouse  and  a  pen,  and 

probably  for any other hand  tools as well. Smell plays a dominant  role  for a de‐

odorant  and  (together with  taste)  for  food  products.  Furthermore,  it  has  been 

demonstrated  that  the dominant modality depends on  the period of product us‐

age. At the moment of buying, vision  is the most  important modality, but at  later 

stages other modalities become more important (Fenko et al., 2010).  

Several hypotheses have been proposed to explain why specific sensory modalities 

dominate  in different situations. The modality appropriateness hypothesis  (Welch 

& Warren, 1986)  claims  that dominance of  a particular modality depends on  its 

appropriateness  for a particular  task. For example, vision  is more appropriate  for 

spatial  perception,  audition  is  more  appropriate  for  the  detection  of  temporal 

events (Welch, 1999), while touch is best suited for tasks that relate to the material 

properties of objects (Lederman et al., 1986).  

Another  explanation  for modality  dominance  points  to mechanisms  of  selective 

attention. This hypothesis suggests that the modality that receives the observers’ 

directed  attention  becomes  dominant  (Posner  et  al.,  1976).  Researchers  have 

shown  that people  can  consciously direct  their  selective attention  to a particular 

sense  (Spence et al., 2001). Attending  to one  sensory modality  can  facilitate  the 

perception of stimuli in that modality relative to when attention is diverted to an‐

other  sensory modality.  Spence, Nicholls  and Driver  (2001)  found  that  the  costs 

associated with shifting attention toward touch from either vision or audition were 

larger  than  the  costs associated with  shifting attention  in  the opposite direction. 

These results suggest that touch is, in a sense, ‘sticky’: once our attention has been 

directed  to what we  are  feeling, we  find  it harder  to  shift our  attention  toward 

what we are seeing or hearing  instead. Spence et al.  (2001) argued that this may 

occur because  it  takes people  longer  to  shift  attention between  internal  (touch) 

and external (vision, audition) perspectives. 

 

Page 112: ANNA FENKO - TU Delft

CHAPTER 5 

 

112 

1.3. PRESENT STUDY 

People’s perception of  the attributes of a product  in a given  sensory modality  is frequently affected by  the sensations  that are simultaneously being perceived by another modality. Consumers are typically completely unaware of the occurrence (or nature) of these crossmodal effects (Schifferstein & Spence, 2008). Thus, intro‐spective  consumer  reports often  fail  to provide accurate  insight  into  the  relative importance of sensory modalities. Therefore, experimental  research on  individual cases of multisensory perception  is necessary  in order  to determine  the  sensory dominance in product usage situations. 

In the present study we used an experimental approach to determine the degree of sensory  dominance  in  product  experience.  We  manipulated  product’s  sensory properties of  two modalities  (vision and  touch)  to determine what was more  im‐portant  for  the  experience of  a product  as warm:  its  tactual properties  (such  as material, fabric and texture) or its visual properties (such as color). In the pre‐study we  created a basis  for multisensory  comparison by asking  respondents  to assess the warmth of various sensory stimuli on a 10‐point scale. We assumed that a spe‐cific ‘warmth’ rating of a visual stimulus is equivalent to the same tactual ‘warmth’ rating of a material measured with the same scale.  

In the main study we created products using warm and cold stimuli of both modali‐ties  in  four different  combinations: 1)  cold  color  +  cold material, 2)  cold  color  + warm material, 3) warm color + cold material and 4) warm color + warm material. Participants assessed the overall ‘warmth’ and ‘pleasantness’ of these products on a 10‐point scale. We expected the products with the combination of warm material and warm  color  to  have  the  highest warmth  ratings,  and  the  products with  the combination of cold material and cold color to have the lowest warmth ratings. We expected the  inconsistent combinations to have ratings  in between these two ex‐tremes. The dominance of touch or vision was determined by investigating the dif‐ferences between these ratings.  

In the main study were did not specify which meaning of warmth we wanted par‐ticipants  to evaluate:  the  literal one or  the  figurative one. The participants were free  to make  their own decision about  the  importance of each aspect of warmth for  the experimental products. To determine  to what extent  literal and  figurative meaning contributes to the overall experience of warmth we performed the follow‐up  interview study. We asked participants about the products, activities and envi‐ronments they associated with the experience of warmth. 

 

Page 113: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

113 

2. PRE‐STUDY 

2.1. PARTICIPANTS 

The participants were  recruited  from  students and  staff of TU Delft. Twenty one participants took part  in the experiment, 30% of them were women. Ages ranged from 18 to 32, the mean age was 24.  

2.2. MATERIALS 

Two different products were chosen for which ‘warmth’ is an important charac‐

teristic: a scarf and a breakfast tray. The function of a scarf is to keep the body 

warm, and  thus  it  refers  to  the  literal meaning of warmth. The breakfast  tray, 

as a part of the home  interior and as a breakfast accessory, refers more to the 

figurative meaning of warmth.  It  contributes  to  the  cosy and  intimate atmos‐

phere of a private home environment. Both products can be easily manipulated 

experimentally. 

For each product two sets of stimuli were prepared. Ten different materials were 

collected for a scarf: silk, viscose, nylon, denim, yarn, rumpled cotton, thin cotton, 

wool,  thick wool, and  fleece. Also,  ten different materials were  collected  for  the 

breakfast tray: aluminum, marble, tile, wood, pressed wood, ripple wood, plastic, 

polystyrene, and rubber. The material samples were about the same size, 300*300 

mm, big enough to touch with a whole hand. We used the same 10 visual stimuli 

for both products: purple,  red, magenta, orange, yellow,  light green, dark green, 

green blue, cyan, and dark blue. Each color was presented to the participants as a 

90*90 mm paper card.  

 

2.3. PROCEDURE 

The  experiment  took place  at  the  Faculty of  Industrial Design of  TU Delft under 

natural  lighting conditions. Tactile stimuli were presented to the respondents one 

by one  in a closed box, so  that  they could explore  them by  touch without seeing 

them. Visual stimuli were presented one by one on a  light grey  table. Half of  the 

respondents assessed visual stimuli  for both products  first, and  the other half as‐

sessed tactile stimuli for both products first. The sequence of the stimuli varied for 

each  respondent. Participants were asked  the question:  “How warm do you  find 

Page 114: ANNA FENKO - TU Delft

CHAPTER 5 

 

114 

this material  (color)  for a scarf  (a  tray)”? They  indicated  their answers on a scale 

from 1  (‘very cold’) to 10  (‘very warm’). Spontaneous comments made by partici‐

pants during the experiment were recorded.  

 

2.3. STATISTICAL ANALYSIS 

For  each  product,  repeated  measures  ANOVAs  on  warmth  ratings  were  per‐

formed with material or  color as within‐subjects  factor. Post‐hoc analyses with 

Bonferroni adjustment were performed to test the significance of the differences 

between means.  

 

2.4. RESULTS 

The  analysis  showed  significant main  effects of both material  (F  (9, 171)  = 16.8, 

p<0.001)  and  color  (F  (9,  171)  =  12.8,  p<0.001)  on  the warmth  ratings  for  the 

scarves and significant main effects of material (F (9, 171) = 18.9, p<0.001) and col‐

or (F (9, 171) =6.9, p<0.001) for the trays. 

Fleece and wool were  found  to be  the warmest and viscose and  silk  the  coldest 

fabrics for a scarf (see Table 5.1). The thick woolen scarf with a lot of holes in it had 

a second warmth rating, while the three coldest fabrics had a dense structure. All 

samples of wood and plastic, both rough and smooth, were found warmer than the 

plates of metal for a tray. At the same time, the smooth plate of steel was found 

colder than the rough plate of aluminum.  

The perception of warm colors  in scarves was dominated by two factors: hue and 

brightness. The warmest colors for a scarf were both reddish and dark. The colors 

that were perceived as warm for a breakfast table tended to be reddish and yellow. 

Blue  colors were experienced  as  the  least warm  for  a breakfast  table  and green 

colors as medium warm. 

 

 

Page 115: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

115 

TABLE 5.1. MEAN WARMTH RATINGS (+SE) OF MATERIALS AND COLORS FOR 

TWO PRODUCTS 

Material  Mean (+SE)  Color  Mean (+SE) 

Scarf 

Silk  1.80 (.35)  Cyan*  2.65 (.47) 

Viscose*  2.00 (.51)  Yellow  2.75 (.50) 

Nylon  3.30 (.50)  Light green  4.10 (.55) 

Denim  4.20 (.55)  Dark blue  4.50 (.53) 

Yarn  4.50 (.34)  Green blue  5.90 (.40) 

Rumpled cotton  4.70 (.30)  Dark green  6.35 (.48) 

Thin cotton  5.55 (.29)  Magenta  6.40 (.58) 

Wool*  6.75 (.44)  Orange  6.65 (.48) 

Thick wool   6.80 (.46)  Red*  7.30 (.57) 

Fleece  7.90 (.46)  Purple  8.40 (.53) 

Tray 

Steel  1.90 (.34)  Dark blue   3.40 (.67) 

Aluminium*  3.80 (.56)  Cyan*  3.90 (.60) 

Marble  4.60 (.47)  Green blue   4.45 (.48) 

Tile  4.70 (.41)  Dark green   4.60 (.60) 

Wood   4.85 (.49)  Light green  4.80 (.58) 

Plastic  5.05 (.67)  Magenta   5.50 (.64) 

Pressed wood   5.50 (.43)  Yellow  6.25 (.46) 

Ripple wood*   7.55 (.50)  Purple  6.65 (.53) 

Polystyrene   7.90 (.34)  Red*  7.50 (.55) 

Rubber  9.15 (.33)  Orange   7.95 (.53) 

* Stimuli chosen for the main experiment.  

 

Page 116: ANNA FENKO - TU Delft

CHAPTER 5 

 

116 

Participants’  comments given  in  the pre‐study  suggested  that  the product ex‐

perience of warmth for a scarf had strong affective components related to the 

perceived pleasantness and comfort. For example, denim was often assessed as 

the  least warm because of  its stiffness, and fleece was  judged as warmer than 

wool because  it was very soft and did not  irritate the skin. The comments sug‐

gested that the comfort of wearing a scarf affected participants’  judgments of 

perceived warmth. Respondents noted  that  they did not  like  rubber as a  sur‐

face  for a breakfast  tray, because  it  felt  sticky and unpleasant  to  touch. They 

also commented that they did not like a polystyrene surface for a tray because 

it felt dusty and unclean. But in case of the tray, pleasantness did not affect the 

warmth ratings of materials.  

 

3. THE MAIN STUDY 

3.1. PARTICIPANTS 

The participants of the main study were 14 male and 8 female students of TU Delft. Ages ranged from 19 to 27, mean age was 22.  

 

3.2. MATERIALS 

Based on the assessments of stimuli made  in the pre‐study, we created 4 vari‐

ants of scarves and 4 variants of trays, combining two warm and two cold stim‐

uli of the visual and tactile modalities, according to a full factorial (2x2) design. 

We have chosen cyan and  red as cold and warm colors,  respectively,  for both 

products. The materials were chosen to match the warmth ratings of colors for 

the product in question as closely as possible: viscose and wool for a scarf, alu‐

minum and wood for a tray. The mean ratings of the stimuli for the main study 

are shown  in Table 1. All differences between cold and warm stimuli were sig‐

nificant at the 0.01 level.  

For scarves we used two white 100% viscose scarves  (H&M department store) 

measuring  0.7*1.7 m  and  two  white  80%  woolen  scarves  (V&D  department 

store) measuring 0.3*1.9 m. We dyed them  in red and blue colors using  ‘scar‐

let’ and ‘kingfisher’ colors of the Dylon brand. Because of the differences in the 

Page 117: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

117 

fabric’s  structure  and material  the  resulting  colors were  slightly  different  for 

viscose and wool: the color of the red viscose scarf was similar to Munsell code 

5R 5/12,  the  red woolen scarf 5R 4/12;  the blue viscose scarf 5B 7/6, and  the 

blue woolen scarf 7.5B 5/10. 

The trays were constructed using textured wood and aluminum sheets. They were 

identical in size (500*310*10 mm) and painted in ‘red’ and ‘fresh blue’ colors of the 

Gamma aerosol paint. The  resulting colors were  identical  for wooden and alumi‐

num trays: red trays had Munsell code 7.5 R4/12, blue trays 7.5B 4/8. The resulting 

products are shown in Figure 1. 

 

          

                               A                                                                  B 

FIGURE 5.1. EXPERIMENTAL PRODUCTS: A) TRAYS; B) SCARVES 

 

3.3. PROCEDURE 

The experiment took place at the Faculty of Industrial Design of TU Delft, in a room with natural lighting. Participants were given products one by one in a random or‐der (scarves and trays separately) and were asked to explore them. For each prod‐uct participants filled in a separate questionnaire assessing how warm and pleasant they found the product on a 10‐point scale (from “not at all” to “very”). Additional questions were asked about the comfort of wearing scarves and about the weight of trays to conceal the real aim of the experiment. Participants were instructed to take their time to observe the products before filling in the questionnaire. All par‐ticipants wanted to put on the scarves and did not only hold them in their hands. It 

Page 118: ANNA FENKO - TU Delft

CHAPTER 5 

 

118 

took about ten to fifteen minutes for each participant to do the test. Participants’ spontaneous comments during the experiment were recorded.  

 

3.4. STATISTICAL ANALYSIS 

Repeated measures ANOVA was performed on warmth and pleasantness  ratings with Color and Material as within‐subjects factors, for scarves and trays separately.  

 

3. 5. RESULTS 

Repeated measures ANOVA showed significant main effects of both Color (F (1, 20) = 22.6, p<0.001) and Material (F (1, 20) = 16.3, p<0.001) on the warmth ratings for a scarf. The Color*Material interaction was not significant (F (1, 20) = 0.93, p>0.20). The results for a tray also showed significant main effects of Color (F (1, 20) = 27.6, p<0.001) and Material  (F  (1, 20) = 19.2, p<0.001), but no significant effect of  the Color*Material interaction (F = (1, 20) = 0.93, p>0.20) on the warmth ratings. 

As  predicted,  woolen  scarves  were  evaluated  as  warmer  than  viscose  scarves, wooden trays were warmer than aluminum trays, and red products were warmer than blue products (Figure 2). The mean differences between the two inconsistent products (red viscose/aluminum and blue wool/wood) were not significant for both scarves and trays (p>0.20 in a separate T‐test). This result suggests that touch and vision are equally important for the experience of warmth in these two products.  

We also analyzed the pleasantness ratings of both products to see if the interaction pattern of pleasantness matched the pattern of warmth (Figure 3). The main effect of Material on the pleasantness ratings for a scarf was significant (F (1, 20) = 67.4, p<0.001), but the effect of Color (F (1, 20) = 1.42, p>0.20) and the Color*Material Interaction were not significant  (F  (1, 20) = 0.006, p>0.20). This suggests  that  the pleasantness  of  scarves was mainly  affected  by  the material: woolen  scarves  of both  colors were  significantly  less  pleasant  than  viscose  scarves.  There were  no significant effects of either Color (F (1, 20) = 3.8, p>0.05), Material (F (1, 20) = 0.09, p>0.2) or Color*Material  interaction ((F (1, 20) = 0.9, p>0.05) on the pleasantness ratings for a tray. 

 

 

Page 119: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

119 

 

Scarf

1

2

3

4

5

6

7

8

9

10

viscose wool

War

mth blue

red

 

Tray

1

2

3

4

5

6

7

8

9

10

aluminium wood

War

mth blue

red

 

FIGURE 2. WARMTH RATINGS FOR SCARVES AND TRAYS 

 

 

 

 

Page 120: ANNA FENKO - TU Delft

CHAPTER 5 

 

120 

 

Scarf

1

2

3

4

5

6

7

8

9

10

viscose wool

Plea

sant

ness

bluered

 

Tray

1

2

3

4

5

6

7

8

9

10

aluminium wood

Plea

sant

ness

bluered

 

FIGURE 3. PLEASANTNESS RATINGS FOR SCARVES AND TRAYS 

 

 

Page 121: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

121 

4. INTERVIEW STUDY The  experimental procedure we used  in  the main  study  revealed  the  composite assessment of warmth. But  in their evaluation of product warmth participants re‐lied both on the literal and on the figurative meaning of warmth. To determine to what extent  literal and metaphorical meanings contributed to the product experi‐ence  of warmth, we  did  an  additional qualitative  study. We  interviewed  partici‐pants  about  the  products,  activities  and  environments  they  associated with  the experience of warmth.  

 

4.1. PARTICIPANTS 

The participants were 20 native Dutch speakers. Ages ranged from 19 to 62, mean age was 37.  

4.2. PROCEDURE 

The  interviews were  performed  at  participants’  homes,  as  one‐to‐one  conversa‐tions.  The  interviewer  explained  that  designers were  interested  in  studying  the experience of warmth  in products and asked participants to name three products for which warm was important.  

Then the interviewers asked what sensory properties (color, shape, material, smell, sound,  etc.)  of  the  products  mentioned  were  relevant  for  the  experience  of warmth.   Next,  the  interviewer asked about  the environments and activities  that participants associated with the experience of warmth, and what feelings they had in such environments and during such activities. Each interview took approximately 20 minutes.  

 

4.3. ANALYSIS  

Free categorization was used to categorize the data. The process of categorization followed the procedure of  inductive category development, which aims to reduce the  contextual diversity of  the material  [32].   Products,  situations, activities, and emotional associations mentioned by participants were  categorized  in groups on the basis of their similarity. Rare answers that did not fit in any category (such as ‘a stone’ for a product) were excluded from further analysis.  

Page 122: ANNA FENKO - TU Delft

CHAPTER 5 

 

122 

4.4. RESULTS 

Two product groups were mentioned most frequently as categories for which per‐

ceived warmth  is an  important product characteristic: 1) clothes and textile (23%) 

and 2) home  interior products  (21%). Other product groups  that were also men‐

tioned frequently were 3) food (16%), 4) heat radiating products (16%), and 5) holi‐

day related items (9%). These results demonstrate that our choice of experimental 

products was adequate: a scarf and a breakfast tray are members of the two most 

frequently mentioned product groups.  

As regards the environments associated with warm experiences, home was men‐

tioned most  frequently  as  a warm  place  (36%  of  the  answers).  Holidays  on  a 

beach were also mentioned  frequently as a warm environment  (23%). Activities 

which  were most  frequently  associated  with  a  warm  experience  included:  1) 

spending  time with  friends  and  family  (having  a  family  dinner,  drinking  coffee 

with  friends, celebrating a birthday, walking a dog; 35% of the answers), and 2) 

physical  relaxation  (sleeping,  taking a warm bath, having a massage, enjoying a 

work break; 30 %).  

The emotional experiences most often associated with “warm” can be divided in 

three groups  (Figure 4). The  first group had to do with physical comfort and  in‐

cluded associations with  “comfortable”  (13%),  “relaxed”  (7%),  “cosy”  (7%), and 

“familiar”  (3%).  The  second  group was  associated with  intimacy  and  included 

“loving”  (10%),  “being  together”  (11%),  “atmosphere”  (10%),  and  “memories” 

(4%). The  third group was associated with energy and  included  “active”  (10%), 

“energized”  (8%),  “excited”  (8%),  “creative”  (3%),  “proud”  (3%),  and  “healthy” 

(3%). 

Only 30% of all respondents associated warmth with physical warmth and com‐

fort. The rest associated warmth either with activity (35% of respondents) or with 

intimacy,  i.e.  loving  atmosphere  and being  together  (35%).    These  results may 

help to understand why we found color to be equally important as the product’s 

material for the experience of warmth in the main study: warm materials refer to 

the  literal meaning of warmth  (physical comfort), while warm colors are associ‐

ated with the metaphorical meaning of warmth (activity, energy and excitement). 

Apparently, both  literal and  figurative meanings of warmth contribute  to  its ex‐

perience.  

 

Page 123: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

123 

5. DISCUSSION 

5.1. PRE‐STUDY 

The  type  of  product was  an  important  factor  that  influenced  the  experience  of warmth. There were inconsistencies in the warm color ratings for the two products. According  to color  theory  (Arnheim, 1959;  Itten, 1961; Ross, 1938), dark green  is not a warm color. Nevertheless,  for a scarf  it was considered as equally warm as magenta, probably because of its darkness. On the other hand, yellow is generally considered  a warm  color, but  respondents  found  it  too bright  for  a warm  scarf. Cyan was  the  least warm  color  for  a  scarf, probably because  it  is both  cold  and light. The colors  that were perceived as warm  for a breakfast  table  tended  to be reddish and yellow. In contrast to the results for a scarf, they were more bright and “sunny”. The  inconsistencies between color  ratings  suggest  that  the warmth of a color  depends  on  the  product  context.  Similar  product  dependencies  were  re‐ported  for  color  pleasantness  judgments  by  Holmes  and  Buchanan  (1984), who showed that although people’s favorite color was blue, this was not the case for a sofa, walls, a carpet, or a chair.  

 

5.2. MAIN STUDY  

The results of the main study demonstrate that color (vision) and material (touch) make an equal contribution to the experience of warmth  in scarves and breakfast trays.  These results deviate from people’s self‐reports, in which touch was judged to be the dominant sensory modality for warmth (Fenko, 2008). This may suggest that  people  tend  to  overestimate  the  impact  of  the  literal meaning  of warmth (touch)  and underestimate  the  impact  of  its  figurative meaning  (vision)  on  their product  experience.  Alternatively,  the  discrepancy  between  the  results  of  ques‐tionnaire and experimental studies may be explained by the fact that  in the ques‐tionnaire  study no  specific products were mentioned. Participants were asked  to think of any relevant product, and possibly they thought of products for which tac‐tile characteristics were more important than for a scarf and a tray.   

We did not find correspondences between warm and pleasant ratings for the final products,  although  for  the materials of  scarves  such  connection was  revealed  in participants’  comments  during  the  pre‐study.  However,  in  the  main  study  the pleasantness of  the  scarves depended mainly on  the material properties  (viscose 

Page 124: ANNA FENKO - TU Delft

CHAPTER 5 

 

124 

scarves were  significantly more  pleasant  than woolen  scarves),  and  not  on  the color. These results  indicate that the pleasantness of complex products cannot be predicted  from  the pleasantness of  their  separate  sensory properties.  Instead,  it may depend on other variables. For instance, spontaneous comments suggest that participants  in  the present  study assessed dark  scarves as warmer, because  they considered dark colors more suitable for a winter outfit than  light colors. Surpris‐ingly, wool had a  lower warmth rating than  fleece;  in this case respondents com‐mented that they did not like wool because it irritated their skin. 

 

5.3. INTERVIEW STUDY  

The semantic analysis of the associations with warmth demonstrated that the figu‐rative meaning of warmth (social activity, love and intimacy) was mentioned twice more often than its literal meaning (physical warmth and comfort).  

The  importance of the social aspect of warmth  is not surprising.  In social psychol‐ogy,  the warm‐cold dimension  is considered  the basic criterion  for social percep‐tion. The first research on this topic was published in 1946 by Solomon Asch. In his famous study undergraduates formed impressions of another person based on lists of trait adjectives (e.g. determined, practical, industrious, intelligent, skilful), which also  included  either  ‘warm’  or  ‘cold’  depending  on  the  experimental  condition. Warm and cold were ‘central traits’ that dramatically altered impressions of people (in a positive way for warm, and in a negative for cold). 

In the past few years, research has clearly established that perceived warmth is one of the  two universal dimensions of human social cognition  (the second  is competence). The basic dimensions of warmth and competence account for 82% of the variance  in perceptions of everyday social behaviors (Wojciszke et al., 1998). The warmth dimen‐sion captures traits that are related to perceived intent, including friendliness, helpful‐ness, sincerity, trustworthiness and morality (Fiske et al., 2006). Warmth is judged be‐fore competence, and warmth judgments carry more weight in affective and behavioral reactions. When judging faces after an exposure time of 100 ms, social perceivers judge trustworthiness most reliably,  followed by competence  (Willis & Todorov, 2006). The importance of these fast visual impressions of social warmth may explain why vision is also important for the judgment of warmth in products.  

Recently, Williams and Bargh (2008) showed that people who had just briefly held a hot  cup  of  coffee,  perceived  a  target  person  as  being  significantly warmer  than those who had briefly held a  cup of  iced  coffee.  In addition, participants primed 

Page 125: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

125 

with physical coldness were more likely to choose a gift for themselves rather than a gift for a friend as a reward for participating in the study, whereas those primed with physical warmth were more  likely  to choose a gift  for a  friend. The authors argue that because of frequent early  life experiences with the  loving family mem‐bers  (Harlow,  1958;  Bowlby,  1969)  a  close mental  association may  develop  be‐tween the concepts of physical warmth and psychological warmth.  

The  results of Williams and Bargh’s  (2008) experiment  can be useful  for product designers. Designers often  try  to create a warm atmosphere  in home  interiors or spaces  used  for  relaxation  and  informal meetings  (bars,  hotel  lobbies,  cafeteria, etc.). Williams  and Bargh  (2008) demonstrate  that  such warm environments  can facilitate pro‐social behavior. Our data are helpful  in creating warm surroundings because they suggest that material and color play an equally important role in cre‐ating such warm experience. 

 

CONCLUSION Our  research  aimed  to  determine  to  what  extent  the  product  experience  of warmth depended on tactile and visual stimuli. The results demonstrated that both color and material contribute equally to the judgments of warmth in two products, scarves and breakfast tables.  However, the pleasantness of a product could not be predicted on the basis of the pleasantness attributed to its color and material. In a qualitative study we asked about the meaning people attribute to warmth in their product experience. The results revealed that the experience of warmth  is a com‐bination of literal and figurative meaning. The literal meaning is related to physical warmth and comfort, while the figurative meaning is associated with social activity, intimacy  and  friendly  atmosphere.  The  figurative meaning was mentioned more often in association with products than the literal meaning. 

 

 

 

 

 

 

Page 126: ANNA FENKO - TU Delft

CHAPTER 5 

 

126 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Chapter 6 was published as:  

Fenko, A., Schifferstein, Huang, T.‐C., & Hekkert, P.  (2009) What makes products fresh: The smell or the colour? Food Quality and Preference, 20, 372–379. 

Page 127: ANNA FENKO - TU Delft

127 

 

 

 

 

 

 

 

 

 

CHAPTER 6 

WHAT MAKES PRODUCTS FRESH:  

THE SMELL OR THE COLOR?  

1. INTRODUCTION When people  interact with products, more than one sensory system  is usually  in‐volved  in this  interaction. People can see the product, touch  it, hear the sound  it makes and often smell it. All the senses are involved simultaneously, but their con‐tribution  to  the  overall  product  experience  is  not  necessarily  equivalent.  Since product experience is multisensory, it is interesting to know which sensory modal‐ity plays a leading role in a particular experience, so that designers can concentrate their efforts on the creation of the most relevant product properties.  

 

1.1. SENSORY DOMINANCE 

Most designers believe  that vision  is  the most  important modality  in product ex‐perience. For instance, Crilly and colleagues (2004) argue that “judgments of prod‐ucts  are  based  largely  on  visual  information”  (p.  547).  In  addition,  Bloch  (1995) 

Page 128: ANNA FENKO - TU Delft

CHAPTER 6 

 

128

claims that “the visual appearance of products is a critical determinant of consumer response and product success”  (p.16). The  importance of vision has been empha‐sized for such product properties as elegance (Coates, 2003), functionality (Monö, 1997),  and  social  significance  (Dittmar,  1992).  Although  visual  information  fre‐quently dominates our culture and environment  (Postrel, 2003; Schroeder, 2002), the full range of human senses influence responses to design (Macdonald, 2000). It is  important that a product’s appearance  is congruent with other sensory aspects of design (Smets & Overbeeke, 1995), because the visible product form creates an expectation of what will be perceived by the other senses (Monö, 1997). 

In  cognitive psychology  the  topic  of  sensory  dominance has  been  addressed  ex‐perimentally. For example,  in their classic study Rock and Victor (1964) presented participants with an object of which the visual shape, because of optical distortion, differed  considerably  from  its  actual  shape  perceived  by  touch.  The  conflict  be‐tween visual and tactual size was resolved completely in favor of vision, and most participants were unaware of any conflict. Strong visual dominance over touch has been demonstrated in a variety of perceptual tasks, involving the determination of size (Miller, 1972), length (Teghtsoonian & Teghtsoonian, 1970), curvature (Easton & Moran, 1978), depth (Singer & Day, 1969), and spatial location (Hay et al., 1965). Studies also demonstrated visual dominance over auditory signals, such as  in  the well known “ventriloquism effect” (Bertelson, 1999).  

Empirical studies on product experience have shown that consumers also tend to 

consider  vision  as more  important  than  other  sensory modalities.  For  example, 

Schifferstein  and  Cleiren  (2005)  demonstrated  that  consumers  acquired most  of 

the  information on products by  vision  and  touch:  this  information was most de‐

tailed  and  the  subjects  were  surest  of  their  judgments.  The  experiment  also 

showed that products were harder to  identify by sound or smell than by vision or 

touch. Blocking vision  increases  task difficulty and  task duration, up  to  the point 

where simple tasks can no longer be completed without help from others (Schiffer‐

stein and Desmet, 2007). A questionnaire study in which participants reported the 

importance of the sensory modalities for the usage of 45 different products (Schif‐

ferstein, 2006) demonstrated that on average the relative importance sequence of 

sensory modalities was vision, followed by touch, smell, audition and taste. In addi‐

tion, when people were asked to rate how important they found the different mo‐

dalities for their  lives  in general, most of them selected vision as the most  impor‐

tant modality.  

Page 129: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

129

However,  the  importance  ratings  for  the  sensory modalities  differed  greatly  be‐tween products. For about half of  the 45 products,  the  importance of vision was lower than for one of the other modalities. For example, audition  is the most  im‐portant modality  for  a  vacuum  cleaner  and  a  coffee maker,  and  for many other products with electric motors. Touch is most important for a computer mouse and a pen, and probably for many other hand tools as well. Smell plays a dominant role for a deodorant and  (together with  taste)  for  food products  (Schifferstein, 2006). Evaluations are  likely  to  reflect  the  importance of  sensory modalities among  the existing  products  in  the  described  category.  New  product  introductions  can  in‐crease or decrease the relative  importances. For  instance, the  introduction of cell phones that are carried all day long has probably made haptic properties more im‐portant for telephones.  

Furthermore,  it has been demonstrated  that  the dominant modality depends on the period of product usage. At the moment of buying vision is the most important modality, but at  later stages the other modalities become more  important (Fenko et al., 2008). Modality  importance  in product usage also depends on  the product characteristic a person  is  judging. When respondents rated the  importance of the various sensory modalities for the evaluation of three product aspects (safety, ease of use, and enjoyment), the  impact of sensory  information was highest for enjoy‐ment, lower for ease of use, and lowest for safety (Schifferstein, 2006).  

We suspect that the importance of various sensory modalities for the judgment of different product experiences  (such as  freshness, warmth, and pleasantness)  can also vary. Therefore,  in the present study we will focus on the roles of the senses for a specific experience (freshness) in a number of products.  

 

1.2. EXPERIENCING FRESHNESS 

For many personal care products, cleaning products, beverages and food products it is important to evoke freshness. Freshness plays an important role in many eve‐ryday experiences, such as shaving, brushing teeth, taking a shower, and drinking soda. It is also possible to talk about “the fresh look” of a new fashion collection or “the fresh color” of a product.  

Freshness is a multisensory product experience that includes visual, olfactory, tac‐tile, and, in some cases, also gustatory and auditory components. The Oxford Eng‐lish Dictionary (1989) gives 16 different meanings of “fresh”, several of which can 

Page 130: ANNA FENKO - TU Delft

CHAPTER 6 

 

130

be applied to products. These meanings can be roughly divided into two groups: 1) new, recent, newly made, recently arrived, retaining its original qualities, not dete‐riorated or changed by  lapse of  time; 2) pure,  invigorating,  refreshing  (said espe‐cially of air and water), not stale, musty, or vapid.  

In this paper we will focus on the second definition of freshness. As the study was 

conducted  in  the Netherlands,  these  two meanings can be separated empirically, 

because in the case of food products different words are used to indicate these two 

meanings (“vers” for the first meaning and “fris” for the second meaning). For the 

other  two  product  categories we  used  (dishwashing  liquid  and  scented  candles) 

only the second meaning is applicable. Therefore, our results on perceived “fresh‐

ness” are equivalent to and can be directly compared to results obtained  in other 

studies on  the perception of  “refreshing”. Previously,  Labbe et al.  (2009, p. 100) 

have used similar arguments to relate the perception of “refreshing” in food prod‐

ucts to Westerink and Kozlov’s (2004) study on “freshness” in oral care. In studies 

on  beverages,  one  of  the  characteristics  of  “refreshing”  is  the  ability  to  quench 

thirst (see Labbe et al., submitted). Following the line of reasoning set out by Zell‐

ner and Durlach  (2003, p. 635), we  consider  “thirst‐quenching”  as  an  aspect of 

“refreshing” relevant to food and beverages.  

According to several studies, the most important characteristics of refreshing foods 

or  beverages  are  temperature‐related  tactile  attributes  (cool,  cold).  Zellner  and 

Durlach (2002) found that “cold”, “cool” and “icy” were the sensory characteristics 

that were most commonly expected by participants for refreshing food and bever‐

ages. In the area of oral care, coldness was also mentioned as a tactile characteris‐

tic related to oral freshness (Westerink & Kozlov, 2004).  

The second important attribute of refreshing in food and beverage products is fla‐

vor. According to Zellner and Durlach  (2002), orange and strawberry  flavors were 

judged as the most refreshing for food and beverages. Labbe et al. (2009) demon‐

strated  that mint odorants were  scored  as  the most  refreshing.  The  flavor most 

commonly listed as not refreshing is chocolate (Zellner & Durlach, 2002). The gusta‐

tory  sensation most often associated with a  refreshing experience  is high acidity 

(Labbe et al., 2009). According to McEwan and Colwill (1996), a carbonated lemon 

drink was more thirst‐quenching for consumers than orange juice and some other 

drinks  (orange  squash,  cola,  isotonic,  sparkling  water,  diet  cola  and  strawberry 

milk), mainly due to its high acidity.  

Page 131: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

131

Color also affects the perceived freshness of beverages. Experimental data suggest that  judgements  on  color  freshness  depend  on  the  associations with  particular products. For example, Clydesdale et al. (1992) found that consumers expect clear and brown non‐alcoholic beverages  to satisfy  their  thirst more  than other colors, because of their association with water and colas, respectively. Of the other colors, red and orange beverages were perceived as more thirst‐quenching than green or purple ones. Zellner and Durlach (2002) found that the color most frequently asso‐ciated with  refreshing  foods/beverages was clear,  followed by  red.  In contrast  to Clydesdale et al. (1992), these authors found that the least likely color for a refresh‐ing food/beverage was black, followed by brown. 

Studies also suggest  that auditory cues can modulate  the perception and evalua‐tion of  food  freshness, despite  the  fact that consumers are often unaware of  the influence of such auditory cues. For example, potato chips were perceived as being both crisper and fresher when the sound of biting was amplified, or when the high frequency sounds were selectively amplified  (Zampini & Spence, 2004). However, the  freshness of potato  chips  is probably quite different  from  the  freshness of a soft drink. The former  is probably closer to the first meaning given by the Oxford Dictionary (newly made, retaining its original qualities, not deteriorated). A similar meaning of freshness has been investigated in several food studies. For example, in the study of freshness of apples (Peneau et al., 2006) the perception of freshness was best described by  taste,  crispness  and  juiciness.  In  the  study of  fresh bread positive drivers of freshness were appearance  (“porous”), odor (“floury”, “malty”, and “toasted”) and flavor (“sweet”, “buttery” and “oily”) (Heenan et al., 2008). 

 

1.3. PRESENT STUDY 

In the present study we  investigate what  is more  important for the experience of 

product  freshness:  the smell or  the color. According  to  the  results of a question‐

naire study in which respondents assessed the relative importance of different sen‐

sory modalities  for various product experiences  (Fenko, 2008),  the dominant mo‐

dality for the experience of freshness was olfaction (mean rating 3.4 out of 4), fol‐

lowed by taste  (3.1) and vision  (2.9). The difference between the ratings of olfac‐

tion and vision was significant (paired samples two‐tailed T‐test, p<0.05). Because 

we included non‐food products in our study, we will not discuss taste further.  

Page 132: ANNA FENKO - TU Delft

CHAPTER 6 

 

132

The importance ratings of the modalities reported in Schifferstein (2006) and Fenko 

(2008) were obtained  through self‐report. However,  the questionnaire methodol‐

ogy  is often  criticized  for  the  lack of ecological  validity, especially  in  the  area of 

product experience, because people are often unaware of why they make certain 

judgments (e.g., Slovic & Lichtenstein, 1971). People’s evaluation of a product may 

not be based on  rational decision making, but  rather on  emotional  and  sensory 

processes they are not aware of. This is especially true for the olfactory properties 

of  the  products  (Köster,  2003).  People  seem  to  have  difficulties  communicating 

about smell experiences through words (Engen, 1982), which limits the usefulness 

of verbal communication in research on olfaction. 

Therefore,  in  the present study we used an experimental approach  to determine 

the degree of sensory dominance in product experience. In the pre‐study we asked 

respondents to assess the freshness and pleasantness of sensory stimuli (colors and 

smells)  on  a  9‐point  scale.  Several  studies  have demonstrated  that  pleasantness 

was highly  correlated with oral  freshness  (Westerink & Kozlov, 2004),  refreshing 

sensation (Lee & O’Mahony, 2005; Labbe et al., 2009), or thirst‐quenching qualities 

of drinks (Clydesdale et al., 1992; McEwan & Colwill, 1996). To control for the effect 

of pleasantness on freshness, we tried to select stimuli with equal  levels of pleas‐

antness. In the main study we created products using fresh and non‐fresh stimuli of 

both modalities in four different combinations: 1) fresh color + fresh smell (CFSF), 2) 

fresh color + non‐fresh smell (CFSNF), 3) non‐fresh color + fresh smell (CNFSF) and 4) 

non‐fresh  color  +  non‐fresh  smell  (CNFSNF). We  asked  respondents  to  assess  the 

freshness and pleasantness of each product on a 9‐point scale.  

We expect the products with the combination of fresh smell and fresh color (CFSF) 

to have  the highest  freshness  ratings, and  the products with  the  combination of 

non‐fresh  smell and non‐fresh  color  (CFSNF)  to have  the  lowest  freshness  ratings. 

We expect  the  inconsistent  combinations  (CNFSF and CFSNF)  to have  ratings  in be‐

tween  these  two extremes.  If both olfaction and vision contribute equally  to  the 

overall experience of freshness, the ratings of the inconsistent combinations should 

be equal.  If vision  is more  important  than olfaction,  then  the  ratings of  the CFSNF 

products will be higher  than  the  ratings of  the CNFSF products.  If olfaction domi‐

nates the freshness experience, then we expect CNFSF products to score higher than 

CFSNF products.  

 

Page 133: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

133

2. PRE‐STUDY 

2.1. PARTICIPANTS 

The participants were 40 students and staff members of  the Faculty of  Industrial Design Engineering, Delft University of Technology. Their ages ranged from 18 to 52 years; mean age was 26.0 years.  

 

2.2. MATERIALS AND METHODS 

Two products were selected for which ‘freshness’ is likely to be an important char‐

acteristic: a soft drink and a dishwashing liquid. The importance of freshness for a 

soft drink is associated with its refreshing and thirst‐quenching qualities, while the 

concept of freshness for a dishwashing liquid refers to its cleaning power. We used 

a scented candle as the control product for which freshness is an optional, but not 

a necessary  characteristic. A  scented  candle  is mainly used  to  create  an  interior 

atmosphere, which is not necessarily fresh, but can also be relaxing, warm, or cozy. 

Because sensory stimuli are experienced differently when presented separately or in 

real product form (Holmes & Buchanan, 1984; Clydesdale, 1993), we created realisti‐

cally  looking product  samples.  For  each product 10 olfactory  and 11  visual  stimuli 

were prepared (odors and colors). As olfactory stimuli for soft drinks and dishwashing 

liquids, we used water‐based perfume and beverage odors produced by Givaudan 

Nederland B.V. (patchouli, almond, rose, cinnamon, pineapple, apple, cherry, grape‐

fruit, bergamot lemon, and peppermint). As olfactory stimuli for candles, we used oil‐

based  odorants  provided  by  Senta Multisensory  Concepting  (vanilla,  sandalwood, 

pine,  lavender,  peony,  lilac,  freesia,  jasmine, water  lily,  and  bergamot).  Fragrance 

bottles were stored at about 4  to 80C  to prevent chemical degradation. Fragrances 

were  dissolved  in Neutral  unscented  soap  basis  (Sara  Lee)  for  dishwashing  liquid, 

were dissolved in pure water for soft drink, or were applied to tea lights for candles. 

All olfactory stimuli were presented to the participants  in closable, transparent PVC 

containers  (50 mm high and 70 mm  in diameter). The dishwashing  liquid and  soft 

drink samples contained 20 ml of scented fluid. The candle samples contained a sin‐

gle  tea  light. Equal amounts of  fragrances  (2 drops) were used  for each  sample  in 

Page 134: ANNA FENKO - TU Delft

CHAPTER 6 

 

134

order to keep the intensity of the odors equal. Olfactory stimuli were prepared every 

day at least one hour before the experimental session.  

To manipulate colors for dishwashing liquid, we dissolved Brouwland food pigments in Neutral soap basis (Sara Lee). The solutions were presented to the participants in half‐full 500 ml  fully  transparent dishwashing  containers. For  soft drink  colors,  the same food pigments were dissolved in carbonated water (C1000 food retailer private brand). The soft drinks were presented to the participants in 500 ml completely filled clear plastic bottles of non‐distinctive shape. The color set  for soft drinks and dish‐washing liquid consisted of dark brown (Munsell code 5YR 2.5/1), purple (7.5RP 4/6), red (5R 5/14), orange (5YR 7/12), clear (N 9.5), green (5G 5/8), pink (10RP 6/10), blue (2.5PB 4/10), yellow (7.5Y 9/6), light blue (5B 7/8) and light green (5GY 7/8).  

To manipulate colors in candles we used colored candles (from ‘Xenos’ store chain) of 

equal size and shape: 70 mm high and 68 mm in diameter. The colors of the candles 

included  brown  (Munsell  code  10YR  5/4), marble  green  (2.5GY  5/2), white  (2.5Y 

8.5/2), cyan (5B 5/8), light green (7.5GY 8/8), orange (2.5YR 6/14), red (5R 5/14), yel‐

low (2.5Y 8.5/10), magenta (10RP 5/12), light purple (5P 7/8), and pink (7.5RP 8/6). 

All  stimuli were  presented  to  the  participants  under  natural  lighting  conditions. 

Munsell coding occurred under similar conditions.  

 

2.3. PROCEDURE 

The smells and the colors for each product were evaluated  in separate parts of the 

session.  For  the  evaluation of  the  visual  stimuli, participants  looked  at  the  stimuli 

presented on a white background. For the olfactory stimuli, participants opened con‐

tainers one by one, and first smelled them from a distance. If they didn’t smell any‐

thing, they were instructed to bring the container closer to their nose. There was a 15 

sec break between odor  trials. Participants were  instructed  to clean  their noses by 

sniffing their lower arm between the trials. All stimuli were identified by a three‐digit 

code and were  randomized between  respondents. Participants  filled out one ques‐

tionnaire on a separate sheet  for each stimulus. Each sample was assessed on a 9‐

point  scale  for  fresh  (fris),  colorful  (fleurig), natural  (natuurlijk)  and pleasant  (aan‐

genaam) from “not at all” to “very”. Each sample was evaluated by 20 participants. 

 

Page 135: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

135

2.4. STATISTICAL ANALYSIS 

For each product, repeated measures ANOVAs on freshness and pleasantness rat‐ings were performed with smell or color as within‐subjects factor. Post‐hoc analy‐ses with Bonferroni adjustment were performed to test the significance of the dif‐ferences between means.  

 

2.5. RESULTS 

Repeated measures ANOVA showed significant main effects of both smell and color on the freshness ratings for all three products (all p<0.01). The main effects of color and smell on the pleasantness ratings were also significant for soft drinks and dis‐hwashing  liquids  (p<0.02), but not  for candles  (p>0.10). The detailed statistics  for each product and each factor are shown in Table 1.  

The smells associated with  fresh soft drinks were peppermint  (mean rating 7.40), 

followed by bergamot lemon (7.05). The least fresh smell was patchouli (2.00), fol‐

lowed  by  almond  (3.9)  and  rose  (4.05).  Bergamot  lemon  (6.40)  and  peppermint 

(6.25) were also considered as the freshest smells for a dishwashing  liquid. Again, 

patchouli  (2.10) was  judged  as  the  least  fresh  smell,  followed by  almond  (3.35). 

Bergamot  pure  reggio  (6.40), water  lily  (5.80),  jasmine  (5.70),  and  freesia  (5.10) 

were perceived as fresh smells for a candle, and vanilla (3.65), a very sweet smell, 

was evaluated as the least fresh smell. 

The color most  frequently associated with  fresh soft drinks was  ‘clear'  (6.90),  fol‐

lowed by light green (6.75). The most unlikely color for a fresh soft drink was dark 

brown (2.60), followed by dark green (3.60). For the dishwashing liquid, light green 

was perceived as the freshest color (6.95). Light blue (6.90), yellow (6.80), and blue 

(6.40) were  also  highly  associated with  freshness.  Purple  (2.95)  and  dark  brown 

(1.1) were considered as non‐fresh  for a dishwashing  liquid. Pink  (7.25), magenta 

(6.90)  and  yellow  (6.75) were  highly  associated with  a  fresh  color  for  a  candle. 

Brown (2.35) and marble green (3.40) were rated as the least fresh colors.  

The purpose of the pre‐study was to select fresh and non‐fresh stimuli for the main study. We selected stimuli with mean ratings from 2.90 to 4.05 as non‐fresh, and from 6.40 to 7.40 as fresh (Table 2). All differences in freshness between fresh and non‐fresh stimuli for a product were significant at 0.01 levels.  

Page 136: ANNA FENKO - TU Delft

CHAPTER 6 

 

136

TABLE 6.1. F VALUES AND SIGNIFICANCE LEVELS FROM ANOVA FOR THE FRESHNESS AND PLEASANTNESS RATINGS FOR SMELL AND COLOR STIMULI 

Product  Factor  F  df  p 

Fresh smell   20.6  9  <0.001 

Fresh color  11.6  10  <0.001 

Pleasant smell  13.5  9  <0.001 

Soft drink 

Pleasant color  7.1  10  <0.001 

Fresh smell   9.4  9  <0.001 

Fresh color  29.0  10  <0.001 

Pleasant smell  4.8  9  <0.001 

Dishwashing liquid 

Pleasant color  16.1  10  <0.001 

Fresh smell   3.8  9  <0.001 

Fresh color  15.7  10  <0.001 

Pleasant smell  1.5  9  0.139 

Candle 

Pleasant color  1.0  10  0.423 

 

It was not possible  to  choose  stimuli with  the  same pleasantness  ratings, be‐

cause pleasantness was highly correlated with  freshness. The Pearson correla‐

tions between  freshness and pleasantness of colors were 0.78  for a dishwash‐

ing liquid, 0.61 for a soft drink, and 0.42 for a candle. The correlations between 

freshness and pleasantness of  smells were 0.72  for a dishwashing  liquid, 0.75 

for a  soft drink, and 0.59  for a  candle. All  correlations were  significant at  the 

0.01  level  (two‐tailed). For  the colors  the sizes of  the  three correlation coeffi‐

cients differed  significantly  (p<0.001),  as  shown  in paired  comparisons of  the 

corresponding regression coefficients (see Snedecor & Cochran, 1980). For the 

smells the correlation  for the candles was significantly  lower than the correla‐

tions  for  dishwashing  liquid  and  soft  drink  (p<0.05), while  the  difference  be‐

tween the latter two did not reach significance (p>0.20). 

 

Page 137: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

137

TABLE 6.2. FRESH AND NON‐FRESH STIMULI FOR THE MAIN EXPERIMENT 

Product     Smell  Mean fresh (SE) 

Mean pleas‐ant (SE) 

Color  Mun‐sell code 

Mean fresh (SE) 

Mean pleas‐ant (SE)  

Non‐fresh  

rose  4.05 (.48) 

2.90 (.43) 

dark green 

5G 5/8  3.60 (.37) 

3.45 (.49) 

Soft drink 

Fresh  pepper‐mint 

7.40 (.44) 

4.50 (.43) 

light green 

5GY 7/8 

6.75 (.40) 

5.20 (.50) 

Non‐fresh 

almond  3.35 (.50) 

3.30 (.44) 

purple  10RP 4/6 

2.95 (.32) 

3.75 (.49) 

Dishwash‐ing liquid 

Fresh   bergamot lemon 

6.40 (.54) 

5.90 (.54) 

blue  2.5PB 4/10 

6.40 (.29) 

5.90 (.46) 

Non‐fresh 

vanilla  3.65 (.41) 

5.20 (.59) 

brown  10YR 5/4 

2.90 (.34) 

5.05 (.44) 

Candle 

Fresh  bergamot  6.40 (.56) 

5.20 (.52) 

pink  7.5RP 8/6 

7.00 (.37) 

5.40 (.50) 

 

3. MAIN STUDY 

3.1. PARTICIPANTS 

Participants were  recruited among  students and  staff of  the Faculty of  Industrial 

Design Engineering, Delft University of Technology. Two groups of 20 respondents 

participated in the main study. In the first group ages ranged from 18 to 28 years, 

mean  age was 22,  and 62% were women.  In  the  second  group  the  ages  ranged 

from 18 to 32 years, mean age was 23, and 60% were women. All participants were 

naive with  regard  to  the purpose of  the  study  and none of  the participants was 

color blind, as indicated by self‐report. 

 

 

Page 138: ANNA FENKO - TU Delft

CHAPTER 6 

 

138

3.2. MATERIALS 

For each product  (soft drink, dishwashing  liquid, and candle) 4 variants were cre‐

ated:  1)  fresh  smell  +  fresh  color;  2)  fresh  smell  +  non‐fresh  color,  3)  non‐fresh 

smell + fresh color; 4) non‐fresh smell + non‐fresh color. In total 12 products were 

created for the main experiment. We used the combinations of the color pigments 

from Brouwland and fragrances from Givaudan Nederland B.V. dissolved in carbon‐

ated water (C1000) or unscented soap (Neutral from Sara Lee) to create soft drinks 

and  dishwashing  liquids,  respectively.  Soft  drinks were  presented  to  the  partici‐

pants  in 500 ml clear plastic bottles of non‐distinctive shape. The dishwashing  liq‐

uids were presented  in the half‐full 500 ml clear dishwashing containers. The col‐

ored candles were bought at Xenos and impregnated with selected odorants from 

Senta Multisensory Concepting (one drop of odorant was applied to the top of the 

candle and then dried with a hair drier). The smells and colors of all the products 

were identical to the selected stimuli used in the pre‐study (Table 2). 

 

3.3. EXPERIMENTAL DESIGN 

To  prevent  participants  from  guessing  the  purpose  of  the  experiment,  the main 

experiment was performed in two separate groups with different stimuli. The first 

group evaluated 6 consistent products. By consistent products we mean that their 

properties in both modalities conveyed the same degree of freshness: fresh color + 

fresh smell (CFSF) and non‐fresh color + non‐fresh smell (CNFSNF) combinations. The 

second group evaluated 6  inconsistent products,  implying that their sensory prop‐

erties contained both fresh and non‐fresh stimuli in different modalities: non‐fresh 

color + fresh smell and fresh color + non‐fresh smell (CNFSF and CFSNF) combinations. 

The sequences  in which products were presented were randomized between par‐

ticipants  in  such a way  that one  set of  soft drink, dishwashing  liquid, and candle 

was  followed by another set with  the same sequence, so  that products  from  the 

same category never followed each other. 

 

 

 

Page 139: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

139

3.4. PROCEDURE 

At each  session participants evaluated 6 products  (2  soft drinks, 2 dishwashing 

liquids, and 2 candles) with respect to their freshness, pleasantness, colorfulness, 

and naturalness on 9‐point scales (from “not at all” to “very”). The samples were 

presented one by one on an A4 white paper. Participants were told to take their 

time  to  explore  the product.  They were  instructed  to pick up  the product  and 

open  the bottles of  the  soft drinks  and  the dishwashing  liquids  to  smell  them. 

Participants were not allowed to taste the soft drinks. The candles were stored in 

closed plastic containers  to avoid  release of  its  smell. Before each  trial,  the ex‐

perimenter opened  the container,  took out  the candle, and presented  it  to  the 

participant. Participants had to fill  in a separate questionnaire for each product. 

There was a small break for 10 to 15 s between the trials. Participants were asked 

to smell their lower arm after each trial to get rid of the previous smell. Each ses‐

sion lasted about 5 minutes. 

 

3.5. DATA ANALYSIS 

Of the four variants of each product, two were evaluated by one group of partici‐

pants, while  the other  two were evaluated by another group. As a consequence, 

some of these effects could be tested using within‐participants comparison, while 

the others  should be  tested using between‐participants  comparisons. Due  to  the 

particular design of the study, ANOVA F‐tests were hard to interpret. Therefore, we 

decided  to use T‐tests only. For each product we started out by performing  two‐

tailed T‐tests for paired samples on the pairs of consistent and  inconsistent prod‐

ucts  separately. The data were analyzed  further  to  check  for  the differences be‐

tween the products with fixed visual or olfactory stimuli using two‐tailed t‐tests for 

independent samples. 

 

3.6. RESULTS 

Within‐participants  comparisons  showed  that  the mean  freshness  ratings  for 

the 4 soft drinks and the 4 dishwashing  liquids were significantly different be‐

tween the two consistent products (CFSF versus CNFSNF) and between the two 

Page 140: ANNA FENKO - TU Delft

CHAPTER 6 

 

140

inconsistent products  (CNFSF versus CFSNF)  (p<0.01). As predicted, CFSF prod‐

ucts  rated higher  than CNFSNF products  for both  soft drinks and dishwashing 

liquids. Furthermore,  the  ratings of CNFSF  soft drinks and dishwashing  liquids 

were significantly higher than the ratings of CFSNF products. This indicates that 

olfaction  is  more  important  for  the  experience  of  freshness  than  vision  for 

these two products.  

Between‐participants  comparisons  indicated  that  freshness  ratings  for  CFSF 

samples were significantly higher than for CFSNF samples (p<0.001)  in both soft 

drink  and  dishwashing  liquid.  Similarly,  CNFSF  rated  significantly  higher  than 

CNFSNF  samples  (p<0.001). The other differences were not  significant  (p>0.20). 

These differences  indicate that smell was  important  for the overall experience 

of freshness  in both soft drinks and dishwashing  liquids and that color did not 

seem to contribute to freshness for these two products (Figure 1). 

Mean  freshness  ratings  for  the  4  candles  showed  a  different  pattern.  The 

difference between the two consistent products (CFSF and CNFSNF) was signifi‐

cant  (p<0.01),  but  there  was  no  difference  between  the  two  inconsistent 

products, CNFSF and CFSNF  (p>0.20). This  indicates  that both vision and olfac‐

tion contribute equally  to  the experience of  freshness  for  this product. The 

between‐participants  comparisons  only  showed  a  significant  difference  be‐

tween CNFSNF on the one hand, and the two inconsistent stimuli on the other 

hand (p<0.05).  

We analyzed the pleasantness ratings of all three products to see if the inter‐

action pattern of pleasantness matched the pattern of freshness. However, in 

the within‐participants  comparisons  CFSF  and  CNFSNF  did  not  differ  in  pleas‐

antness  for  all  3  products  (p>0.10).  These  results  do  not  allow  us  to  deter‐

mine  the degree of  sensory dominance  for  the pleasantness  ratings.  In  fact, 

for candles no significant differences in pleasantness ratings were found at all 

(all p>0.05). For both soft drinks and dishwashing  liquids, the combination of 

fresh  color and non‐fresh  smell was  considered as  the  least pleasant by our 

respondents (Figure 6.1).  

Page 141: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

141

Soft drink

1

2

3

4

5

6

7

8

9

CFSF CNFSFl CFSNF CNFSNF

Fres

hnes

s

Soft drink

1

2

3

4

5

6

7

8

9

CFSF CNFSF CFSNF CNFSNF

Ple

asan

tnes

s

Dishwashing liquid

0

1

2

3

4

5

6

7

8

9

CFSF CNFSFl CFSNF CNFSNF

Fres

hnes

s

Dishwashing liquid

1

2

3

4

5

6

7

8

9

CFSF CNFSF CFSNF CNFSNF

Ple

asan

tnes

s

Candle

1

2

3

4

5

6

7

8

9

CFSF CNFSFl CFSNF CNFSNF

Fres

hnes

s

Candle

1

2

3

4

5

6

7

8

9

CFSF CNFSF CFSNF CNFSNF

Ple

asan

tnes

s

 

** Differences between means are significant at .01 level, * at .05 level (2‐tailed T‐test). 

FIGURE 6.1. MEAN FRESHNESS AND PLEASANTNESS RATINGS FOR THREE PRODUCTS 

(WITH SE) 

** 

** 

** 

** 

** 

** 

** 

** 

**

Page 142: ANNA FENKO - TU Delft

CHAPTER 6 

 

142

4. DISCUSSION 

4.1. FRESHNESS 

Our  results  indicate  that when  smell was  fixed,  the degree of  freshness  for  soft drinks and dishwashing  liquids was not changed by different colors. On the other hand, when the color was fixed, people perceived significantly different degrees of freshness because of the change in smell. Therefore, olfaction is the dominant mo‐dality when assessing freshness for a soft drink and a dishwashing liquid. This find‐ing was  supported  by  participants’  spontaneous  remarks  and  confirms  the  out‐comes of our previous questionnaire study (Fenko, 2008). A different result, how‐ever, was  found  for candles. Here, smell and color both  influenced  the degree of freshness  to  approximately  the  same degree.  These  results  could  imply  that  the dominance of smell for the experience of freshness depends on the role of fresh‐ness  in the product experience.  In candles, for which freshness  is not a necessary property, vision and smell both contribute to the degree of freshness. On the other hand, for products such as soft drinks and dishwashing  liquids, where freshness  is related to their main function, the experience of freshness mainly depends on the olfactory properties of the product.  

Most of the results of our pre‐study were in agreement with previous studies. For instance, peppermint scored as a  fresh smell both  for a soft drink and  for a dish‐washing liquid in our study. This is in line with findings from Westerink and Kozlov (2004), who found that mint was strongly associated with oral freshness, and Labbe et al. (2009), who found that mint was one of the drivers of the refreshing sensa‐tion. Furthermore, smells associated with sour taste, such as lemon and grapefruit, scored higher on freshness for both a dishwashing liquid and a soft drink. This is in line with McEwan and Colwill (1996)’s suggestion that the thirst‐quenching proper‐ties of a drink may increase with an increase in perceived acidity (p. 108). Our par‐ticipants,  like  those  of  Clydesdale  et  al.  (1992)  and  Zellner  and  Durlach  (2003), found clear to be the most refreshing/thirst‐quenching color for a soft drink. How‐ever,  Clydesdale  and  colleagues  found  brown  to  be  a  thirst‐quenching  color, whereas we  found  it  to be non‐fresh. This difference may be  caused by  cultural differences: Clydesdale et al. (1992) did the experiment  in the USA and suggested that brown was experienced as a thirst‐quenching color because of the associations with colas. In our study, dark brown was seen as non‐fresh, even though some re‐spondents  identified  it with  the color of cola during  the experiment, as could be derived  from  their spontaneous comments. Zellner and Durlach  (2003) suggested 

Page 143: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

143

that  brown  colas  and  vanilla  drinks  can  be  refreshing  because  people  consider brown  the  appropriate  color  for  these  drinks,  whereas  brown  lemon  and mint drinks are inappropriate and therefore not refreshing.  

In  our  study  “freshness”  has  the  similar meaning  to  “freshness”  in  the  study  of Westerink and Kozlov  (2004) and “refreshing”  in  the studies of Zellner & Durlach (2002; 2003) and Labbe and colleagues (2009). Furthermore, “freshness” may have another  meaning:  “new,  retaining  its  original  qualities;  not  deteriorated  or changed”  (Zampini & Spence, 2005; Peneau et al., 2006; Heenan et al., 2008).  In addition, a person can  feel “refreshed” after a good night sleep or after  taking a holiday. Refreshing in this sense means energizing or invigorating. The fresh quali‐ties of products can be associated with particular environments  (such as a forest) or activities  (such as sailing). Therefore, additional research  in different situations may reveal other associations people have with “freshness” that may help to spec‐ify the meaning of freshness for other product groups.  

The current results add to the body of knowledge on different types of color‐odor interactions.  For  instance,  Zellner  and  Kautz  (1990)  reported  that  the  perceived intensity of food odors was increased when the odors were presented in colored as compared  to  colorless  liquids, but  the appropriateness of  the  color  for  the  smell had little or no effect on the perceived intensity of the odor: having any color at all seemed to be the most important characteristic for increasing the subjective inten‐sity of  smell. On  the other hand,  several  studies  claimed  that appropriate  colors significantly  facilitate  odor  identification  and  increase  odor  liking  (Davis,  1981; Bone & Jantrania, 1992). Zellner and Durlach (2003) demonstrated that the effect of color on perceived refreshment might be  flavor‐specific. The present study ex‐pands this type of research by demonstrating that the way in which color and smell contribute  to  freshness  is  product‐dependent.  Nevertheless,  more  research  is needed to be able to generalize not only beyond the products used in the present study, but also beyond the sensory stimuli (smells and colors) that we used. 

The present study has demonstrated that evaluations of freshness can be predicted as an additive combination of  the evaluations of  two component stimuli  (a smell and a color in this case). This opens up the possibility to further test additive mod‐els of information integration (e.g. Anderson, 1981) to determine how various sen‐sory inputs contribute to the experience of freshness. In the present study, we did not  investigate  the  influence  of  touch,  taste  and  audition  on  the  experience  of freshness. Other  studies  indicate  that  tactile  sensations  (especially  the  feeling of temperature) can play an important part in creating a refreshing sensation (Zellner 

Page 144: ANNA FENKO - TU Delft

CHAPTER 6 

 

144

& Durlach, 2002; Westerink & Kozlov, 2004; Labbe et al., 2009). For food products, gustatory  properties  are  also  important  determinants  of  refreshing/thirst‐quenching (McEwan and Colwill, 1996; Labbe et al., 2009). Further experiments are needed  to  investigate how  these sensory modalities contribute  to  the experience of freshness for different product groups. In addition, the current approach can be used  to  study  how  olfaction  contributes  to  other  aspects  of  human  experience, such as the perceived healthiness, naturalness, or cleanness of products. 

 

4.2. PLEASANTNESS  

Previous research has found that pleasantness is strongly related to the refreshing 

sensation (Lee & O’ Mahony, 2005; Labbe et al., 2009) and thirst‐quenching quali‐

ties of drinks  (Clydesdale et al., 1992; McEwan & Colwill, 1996). Analogously, our 

pre‐study  found  that  pleasantness  ratings  correlated  significantly with  freshness 

ratings of single modality stimuli. These correlations were higher  for dishwashing 

liquid and soft drink than  for candles, suggesting that  freshness and pleasantness 

are more closely connected for products for which freshness  is a necessary, func‐

tional property. Despite  these correlations  found  for single, component stimuli, a 

comparison of  the product means  shows  that  there was no  correlation between 

freshness and pleasantness for the final products. When specific smells were com‐

bined with specific colors, the freshness of the combinations could be predicted on 

the basis of the freshness of the separate stimuli (Figure 1), but the pleasantness of 

the combinations could not  (Figure 2). The most pleasant dishwashing  liquid had 

both a non‐fresh smell and a non‐fresh color. For candles, in the main study all the 

final products were considered equally pleasant. These outcomes are  in  line with 

those  obtained  by  Schifferstein  and  colleagues  (submitted),  who  also  failed  to 

demonstrate  the  link between  the pleasantness  ratings of  single  stimuli  and  the 

overall pleasantness of the products. In their study, the pleasantness of visual, tac‐

tual, auditory, and olfactory unisensory stimuli was determined for two test prod‐

ucts (a portable air purifier and a table lamp).  

Pleasantness has been suggested as an odor's most salient attribute (Engen, 1982) 

and a lot of data indicate the strong link between the sense of smell and affective 

reactions  (Herz, 1998a; 1998b; Herz and Schooler, 2002). Based on  the extensive 

literature on the affective character of olfactory perception, we would expect smell 

to have more influence on the pleasantness rating than color, but our results only 

Page 145: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

145

partly confirm this assumption. The most unpleasant samples of both soft drink and 

dishwashing  liquid were  those with  fresh  color  and  non‐fresh  smell.  These  out‐

comes  indicate  that negative affective reactions might be due  to the  inconsistent 

combinations of olfactory and visual stimuli rather than by the unpleasant smell as 

such. Probably, evaluations of pleasantness are more dependent on the combina‐

tions of stimuli used and their degree of (in)congruence than evaluations of fresh‐

ness (e.g., Schifferstein & Verlegh, 1996).  

 

5. CONCLUSIONS Olfaction was found to be more important than vision in determining the degree of 

freshness for those products for which freshness is relevant for the product’s main 

function (such as a soft drink and dishwashing liquid). On the other hand, both ol‐

faction  and  vision  seem  to  contribute equally  to  the experience of  freshness  for 

those products  for which  freshness  is optional, but not a necessary characteristic 

(such as a scented candle). These results can help to understand the experience of 

freshness in different product contexts and can be used in product development.  

 

  

 

 

 

Page 146: ANNA FENKO - TU Delft
Page 147: ANNA FENKO - TU Delft

147 

 

 

 

CHAPTER 7 

NOISY PRODUCTS:  

DOES APPEARANCE MATTER?  

 

1. INTRODUCTION 

1.1. MULTISENSORY INTEGRATION IN PRODUCT EXPERIENCE 

Many everyday products make noise while operating. For some products noise is a consequence of their functioning, like the rotating noise of a washing machine, the mechanical noise of a hammer, or  the blowing noise of a hair dryer. Other prod‐ucts, such as kitchen timers or fire sirens, are designed to make noise  in order to attract people’s attention. Özcan and Van Egmond (2008) refer to these two types of sounds as consequential and  intentional,  respectively. Although  there  is an  in‐creasing interest in the sound design of domestic appliances, most sound designers are concentrating on  the  reduction of  industrial and  traffic noise and diminishing the loudness of domestic appliances (Lyon, 2000). 

Page 148: ANNA FENKO - TU Delft

CHAPTER 7 

 

148

Sound  is not the only sensory product property that can be associated with noisi‐ness. Our perceptual experience of the world is richly multimodal (Stein and Mere‐dith, 1993). Different sensory modalities make different contribution to the overall product experience. For instance, color and material have been found to contribute about  equally  to  the  experience  of  ‘warmth’  in  scarves  and  trays  (Fenko  et  al., 2010), while  the  experience of  ‘freshness’  in  soft drinks  and dishwashing  liquids depends more on smell than on color (Fenko et al., 2009). Ludden and Schifferstein (2007)  have  also  found  that  the  expression  of  a  product’s  sound  influenced  the overall expression of dust busters and juicers. 

In this study we wonder to what extent vision  influences the product experience of noisiness. According to people’s self‐reports they find audition more important than vision for judging noisiness. In a questionnaire study, in which respondents assessed the relative importance of five sensory modalities for 34 product experiences (Fenko et al., submitted), the dominant sensory modality for ‘noisiness’ was audition (mean rating 4.9 out of 5), while the second most important modality was vision (3.1).  

The results of questionnaire studies on sensory  integration should be  interpreted with caution, because people are usually unaware of perceptual mechanisms un‐derlying their experience. A good example is the McGurk effect (McGurk and Mac‐Donald,  1976)  that  demonstrates  an  interaction  between  audition  and  vision  in speech perception. When a recording of the sound ‘ba’ is combined with a video of lip movements saying  ‘ga’, most people think they are hearing  ‘da’. When people close their eyes, they hear  ‘ba’ correctly. The brain combines the two types of  in‐formation  automatically;  knowledge  about  the  effect  has  little  effect  on  one's perception of  it. Another example of audio‐visual  interaction  is  the ventriloquism effect  (e.g.,  Bertelson,  1999).  A  ventriloquist  synchronizes  the movements  of  a puppet's mouth with his own speech while avoiding movements of his own head or lips. As a result, the audience perceives a speaking puppet. This effect  is regularly experienced when watching television and movies, where voices seem to emanate from  the actors'  lips  rather  than  from  the  loudspeakers. The ventriloquism effect does  not  depend  on  the  direction  of  deliberate  or  automatic  visual  attention (Bertelson et al., 2000; Vroomen et al., 2001).  

Both these examples demonstrate visual dominance over audition  in multisensory speech perception. Most people are unaware of these phenomena and believe that speech  is perceived only by hearing. People  can be equally mistaken  in  their as‐sessment of the relative contribution of different senses to a product experience, such as noisiness. That  is why experimental manipulation  is necessary to establish which sensory modality is more important for a specific product experience.  

Page 149: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

149

1.2. EXPERIENCING NOISINESS 

1.2.1. AUDITORY NOISINESS 

From the physical point of view, any complex sound may be described as a combi‐nation of pure tones of various amplitudes, frequencies and phases. Musical tones typically  contain  only  a  few  components,  for which  the  frequencies  and  phases have simple relationships. More complex sounds contain many components, which may not be systematically related in frequency and phase. The ultimate complexity is “white noise” that contains all frequencies. The sound from a water‐spray or a jet of air is an approximation of white noise (Northwood, 1963). The subjective rating of “noisiness” increases as 1) the pitch of a sound is raised (Hellman, 1984); 2) the complexity of  the  spectrum  is  increased  (Hellman,  1985);  and  3)  the duration  is increased beyond 200 milliseconds  (Kryter, 1966). The  loudness of a  sound does not usually contribute to its subjective noisiness. Annoyance from a noise increases with the increase of the duration of the event: longer exposures to low sound lev‐els are equally annoying as shorter exposures to higher sound levels (Hiramatsu et al., 1983; Fujii et al., 2001).  

Although  in  the  psychoacoustic  literature  the  terms  “noisy”  and  “annoying”  are often used as synonyms (Hellman, 1982; Takeshima et al., 1991; Bowsher & Robin‐son, 1962; Schultz, 1978; Berlung et al., 1976), in this study we treat noisiness and annoyance as separate experiences. Noisiness refers to particular characteristics of the sound that may or may not cause annoyance. On the other hand, not all annoy‐ing  sounds  are necessarily noise:  loud music or  conversation  can  also be experi‐enced as annoying under certain circumstances, although these sounds do not have a noisy character (Northwood, 1963). Any sound may become annoying if it is able to distract  listeners  from  their  activities.  Speech  sounds  are particularly  trouble‐some  if  they  are  intelligible  or  nearly  so.  Sudden  impacts,  startling  or  alarming sounds, and sounds with marked pitch or rhythm (for example, a dripping faucet) are particularly distracting and  therefore annoying  (Job, 1988). Similarly, a  single identifiable source of noise is more troublesome than the same loudness level pro‐duced by a  random assortment of many noises  from many  sources  (Northwood, 1963). Random and meaningless sounds sometimes mask other sounds that would be distracting. Paradoxically, the white noise from ventilators and air‐conditioners can  successfully mask  a particular  intruding  sound.  The widespread use of back‐ground music  in  restaurants  is  another  example  of  the  use  of  sound  to  protect people from the noise made by their neighbors. Unfortunately, it can also be very annoying. 

Page 150: ANNA FENKO - TU Delft

CHAPTER 7 

 

150

Annoyance from noise is usually described as the unpleasantness of noise. The ca‐pacity  of  noise  to  induce  annoyance  depends  on  many  acoustical  and  non‐acoustical  factors. Few sounds are  intrinsically unpleasant. The most common ex‐amples include the sound of a metal saucepan being scraped with a knife, a pneu‐matic drill at close  range, and a powerful motor cycle accelerating  (Bowsher and Robinson, 1962). Loudness was found to account for a relatively small percentage of  the variability  in self‐reported unpleasantness, usually 10%  to 25%  (Griffiths & Langdon, 1968; McKennel, 1963).  

 

1.2.2. VISUAL NOISINESS 

The English word  ‘noisy’ can be used without  implication of audible sound.  In the latter  sense  it means  “showy,  ostentatious;  conspicuous;  overwhelming;  gaudy, loud, or bright  in  color”  (Simpson  and Weiner,  1989).  This meaning of noisiness refers to the use of brighter colours and to visual complexity.  

The perceived visual complexity can increase with the quantity and range of objects and the variety of colors, materials and surface styles, while the number of objects and surfaces remain constant (Heaps and Handel, 1999). The perceived complexity of an  image also depends on  the degree of perceptual grouping, a  characteristic independent of  the quantity of parts. Regularities,  such as  symmetry,  repetition, and  similarity,  simplify  a  visual  pattern,  making  it  less  noisy  (Feldman,  1997; Palmer, 1999; van der Helm, 2000). Similarly, textures with repetitive and uniformly oriented patterns are  judged as  less complex than disorganized and cluttered pat‐terns (Heaps & Handel, 1999; Oliva & Torralba, 2001).  

Visual complexity has been reported to  influence aesthetic  judgment  (e.g., Berlyne, 1963; Eisenman, 1967; Eysenck & Hawker, 1994). Berlyne  (1971) believed  that  the pleasure or hedonic value of an object depends on  its capacity for arousal. He sug‐gested that the same characteristics determine pleasure and arousal: intensity, color, pitch, novelty, and complexity. As the arousal potential of an object increases, one's pleasure at first rises but eventually peaks and then falls down again. Thus, pleasure is a function of an arousal change. Represented graphically, the curve is bell‐shaped. The bell‐shaped curve goes back to Wundt's theory of sensory affect, depicting pleas‐antness as a function of stimulus intensity (Wundt, 1904).  

In a number of experiments Berlyne and his co‐workers have shown that medium levels of complexity for visual or auditory stimuli are liked most, while low and high 

Page 151: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

151

levels  are  liked  the  least  (Berlyne, 1971).   Possibly,  low  levels of  visual noisiness may be experienced as boring and thus not particularly pleasant, while high levels of visual noisiness may cause high levels of arousal and thus be experienced as an‐noying and unpleasant. We expect  the medium  level of visual noisiness  to be ex‐perienced as the most pleasant. This was tested in the pre‐study. 

 

1.3. PRESENT STUDY 

In  the  present  study we manipulated  the  auditory  and  visual  properties  of  two products  in  order  to  find  out  to what  extent  the  overall  experience  of  product noisiness depended on  the  specific  sounds  these products made and  their visual appearance. In the pre‐study we asked respondents to assess the noisiness, pleas‐antness and annoyance of various product sounds or product images on a 10‐point scale. In the main study we created products combining noisy and quiet stimuli of both sensory modalities in four different combinations: 1) noisy sound + noisy vis‐ual pattern, 2) noisy sound + quiet visual pattern, 3) quiet sound + noisy visual pat‐tern and 4) quiet  sound + quiet  visual pattern. Participants  indicated how noisy, pleasant and annoying they found these products on a 10‐point scale. 

 

2. PRE‐STUDY 

2.1. METHOD 

2.1.1. PARTICIPANTS  

Two  groups  of  Dutch‐speaking  participants  took  part  in  the  pre‐study.  The  first group (10 men and 12 women, age ranged from 17 to 54, mean age 29) assessed the  sounds  of  alarm  clocks  and  kettles,  and  the  second  group  (14 men  and  12 women,  age  ranged  from  22  to  45, mean  age  32)  assessed  pictures  of  the  two products with various types of decorative patterns. 

 

2.1.2. MATERIALS 

Two  products  were  used  in  the  experiment:  a  kettle  with  removable  whistle (HEMA)  and  an electronic  radio  alarm  clock  (AKAI). We  selected  these products, 

Page 152: ANNA FENKO - TU Delft

CHAPTER 7 

 

152

because  for  both  of  them  noisiness  is  a  functional  property:  the  sounds  of  the products are designed intentionally to attract people’s attention.  

Ten  alarm  sounds were  created using digital  sound  samples  and  recordings. Ten different whistling sounds were created by manipulating the whistles of the kettles (e.g., enlarging holes, drilling additional holes, and filling whistles with various ma‐terials). All sounds were recorded with a semi‐professional digital recorder with a range of 20.000 Hz at 1.5 m distance from the object. Each sound lasted for 10 s. 

Decorative visual patterns were created in which the amount of elements and col‐ours,  the  size of  the elements, and  regularity and  symmetry of  the pattern were manipulated.  These patterns were  applied  to  the photographs of  the  kettle  and alarm clock using Adobe Photoshop. The resulting images (12 for each product) are shown in Figures 6.1 and 6.2.  

 

FIGURE 7.1. IMAGES OF ALARM‐CLOCKS USED IN THE PRE‐STUDY (WITH MEAN 

NOISINESS RATINGS) 

3.30  3.50  3.55 

4.25  4.25  5.20 

5.65  5.75  5.85 

5.90  6.45  6.80 

Page 153: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

153

 

FIGURE 7.2. IMAGES OF WHISTLE KETTLES USED IN THE PRE‐STUDY (WITH MEAN 

NOISINESS RATINGS) 

 

2.1.3. PROCEDURE 

Since  the  study was  performed  in  the Netherlands with  Dutch‐speaking  partici‐pants, all the questionnaires and instructions were provided in Dutch. When prod‐uct experience is expressed in words, it may be interpreted differently in different languages. Most adjectives that describe product experiences have several mean‐ings, and usually not all these meanings can be translated adequately into another language (Brislin, 1980).  

3.05  3.11  3.58 

4.11  4.42 

4.58  5.53  5.74 

5.89  6.89  7.17 

4.37 

Page 154: ANNA FENKO - TU Delft

CHAPTER 7 

 

154

In English, the word “noisy” carries both auditory and visual meanings. The Ox‐

ford English Dictionary defines noisy as 1)  full of or characterized by noise; 2) 

showy, ostentatious; conspicuous; overwhelming; gaudy,  loud, or bright  in col‐

our  (Simpson & Weiner,  1989).  In  the  Dutch  language  several  adjectives  are 

used  to  capture  auditory  and  visual  aspects of noisiness  (Van Dale,  2008):  1) 

lawaaierig – noisy,  full of or  characterized by  loud and nonmusical  sounds; 2) 

luidruchtig – loud, with high volume; noisy, full of or characterized by loud and 

nonmusical sounds; 3) druk – restless, overwhelming (of artwork etc.); with too 

many  flowers,  lines,  figures;  loud  (of  colors). That  is why  in  the pre‐study we 

used several Dutch adjectives to assess auditory and visual noisiness. We ana‐

lysed the correlations between various noisiness ratings to establish their sen‐

sory connotations, and finally selected the word  lawaaierig as the most appro‐

priate multisensory measure of product noisiness. 

The sounds of kettles and alarm clocks were presented to the participants  in two sets.  Half  of  the  group  evaluated  kettle  sounds  first,  and  half  evaluated  alarm sounds first. Within each set all the sounds were randomized between participants. Participants filled out the questionnaire evaluating how loud (hard), pleasant (pret‐tig),  noisy  (lawaaierig  and  luidruchtig)  and  annoying  (irritant)  they  found  each sound on a 10‐point scale (from “not at all” to “extremely”). 

The images of products were presented to the participants in two series. Half of the group evaluated 12 whistle kettles  first, and half evaluated 12 alarm  clocks  first. The  images  of  the  products  were  presented  in  random  order  on  a  17"  laptop screen. Participants filled out the questionnaire evaluating how pleasant (prettig), noisy (lawaaierig), loud (druk), colorful (kleurrijk) and annoying (irritant) they found each image on a 10‐point scale (from “not at all” to “extremely”).  

 

2.1.4. STATISTICAL ANALYSIS 

For each product, repeated measures ANOVAs on the ratings were performed with sound or visual pattern as within‐subjects factor. Post‐hoc analyses with Bonferroni adjustment were  performed  to  test  the  significance  of  the  differences  between means. Pearson correlations between all evaluative ratings were computed.  

 

Page 155: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

155

2.2. RESULTS 

The  statistical analysis  showed  significant main effects of  sound on  the noisiness (lawaaierigheid) for both alarm clocks (F (9, 180) = 11.0, p < 0.001) and kettles (F (9, 180)  =  36.6,  p  <  0.001).  The main  effect  of  visual  pattern  on  the  noisiness  (la‐waaierig) was also  significant both  for both alarm  clocks  (F  (11, 209) = 13.7, p < 0.001) and kettles  (F  (11, 198) = 11.9, p < 0.001). The main effects of sound and visual pattern on the ratings of pleasantness (prettig) and annoyance (irritant) were also significant for both products (all p < 0.001).  

Mean noisiness  (lawaaierig)  ratings  for  sounds  varied  between  1.76  and  9.48 

for whistle kettles and between 4.76 and 8.24  for alarm clocks. The acoustical 

analysis of sounds, performed with Adobe Audition 3.0, showed  that high  fre‐

quency  sounds were  assessed  as  noisier  than  low  frequency  sounds.  Sounds 

with multiple and fluctuating frequencies were assessed as noisier than sounds 

with relatively constant frequency.  

Noisiness  (lawaaierig) was positively  correlated with  annoyance  (irritant; 0.62 

for alarm clocks and 0.77  for kettles) and negatively correlated with pleasant‐

ness (prettig; ‐ 0.40 and ‐ 0.61 for both products, respectively). All correlations 

were significant at the 0.01 level (2‐tailed). Noisiness was also positively corre‐

lated with loudness (hard; 0.76 and 0.90) and with the other noisiness measure 

(luidruchtig; 0.85 and 0.90). 

Mean noisiness (lawaaierig) ratings for the visual patterns varied between 3.05 and 

7.17  for kettles and between 3.30 and 6.80  for alarm clocks  (see Figures 7.1 and 

7.2). In both products the ratings of noisiness increased as the amount of elements 

and  colors  increased. Regular patterns  (checked,  striped) were perceived  as  less 

noisy  than  irregular patterns. For  flowery patterns,  the degree of clutter was  the 

most important factor that increased noisiness. 

The correlations between noisiness (lawaaierig) and annoyance (irritant) were posi‐

tive for both alarm clocks (0.74) and kettles (0.74). Pleasantness (prettig) was nega‐

tively correlated with noisiness (lawaaierig) both for kettles (‐0.43) and alarm clocks 

(‐  0.16). All  correlations were  significant  at  the  0.05  level  (2‐tailed).  In  addition, 

noisiness  ratings were positively  correlated with  colorfulness  (kleurrijk;  0.37  and 

0.24) and the other noisiness measure (druk, 0.64 and 0.69).  

 

Page 156: ANNA FENKO - TU Delft

CHAPTER 7 

 

156

3. MAIN STUDY 

3.1. METHOD 

3.1.1. PARTICIPANTS 

The participants were 40 Dutch‐speaking students and staff of TU Delft (17 women and 23 men). Ages ranged from 18 to 51 years, mean age was 24.  

 

3.1.2. MATERIALS 

The  stimuli  for  the main  study were  selected  in  such  a way  that differences be‐tween auditory and visual stimuli were similar for each product. For the kettles we selected sounds with mean ratings 5.57 and 3.57 (difference 2.0) and visual stimuli with mean ratings 5.89 and 3.58 (difference 2.31). For the alarm clocks we selected sounds with mean  ratings 8.05 and 4.86  (difference 3.19) and  images with mean ratings  6.80  and  3.30  (difference  3.50).  All  differences  between  noisy  and  quiet stimuli for each product were significant at the 0.01 level, but mean differences for alarm clocks were bigger than for kettles. This was the result of technical difficulties in creating different kettle sounds. 

For each product, we made a computer video clip in Adobe Flash CS4. In the clips, the products were presented in a naturalistic environment: alarm clocks in the bed‐room, kettles in the kitchen (see Figure 7.3). To increase the degree of realism, the steam of boiling kettles was added  to  the  latter  clips. For each product, a visual image was  combined with  a  sound  according  to  a  full  factorial  (2x2)  design:  1) sound noisy +  image noisy  (SNIN), 2) sound noisy +  image quiet  (SNIQ), 3) sound quiet + image noisy (SQIN) and 4) sound quiet + image quiet (SQIQ).  

 

3.1.3. PROCEDURE 

To prevent participants from guessing the purpose of the study, the experiment was performed with two groups of participants. One group assessed products that com‐bined noisy sound with noisy image (SNIN) and quiet sound with quiet image (SQIQ). Another group assessed products  that combined noisy and quiet stimuli  (SNIQ and SQIN). Each group evaluated 4 products (two kettles and two alarm clocks). The se‐

Page 157: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

157

quences in which products were presented were randomized between participants in such a way that products from the same category never followed each other. 

 

      

                

FIGURE 7.3. IMAGES OF PRODUCTS USED IN THE MAIN STUDY 

 

The experiment took place in a quiet room with natural lightning. One laptop com‐puter with  fixed  visual and  sound  settings was used  throughout  the experiment. The experimenter  first explained  to  the participants  that  they were about  to  see four different products and asked them to answer 5 evaluative questions for each product.  After  participants  filled  in  their  personal  information,  they  clicked  the “start” button, and the first product appeared on a 17" diagonal computer screen. Participants saw each product and heard the product sound for 10 seconds. After a couple of seconds the first question appeared below the image. The next question appeared automatically after the previous question was answered. For each prod‐uct, participants assessed how noisy (lawaaierig, druk, and schreeuwerig), annoying (irritant) and pleasant  (prettig)  they  found  the product on a 10‐point scale  (from “not at all”  to “extremely”).  In addition, spontaneous comments given by partici‐pants during the experiment were recorded. 

 

Page 158: ANNA FENKO - TU Delft

CHAPTER 7 

 

158

3.1.4. DATA ANALYSIS 

Of the four variants of each product, two were evaluated by one group of participants, while  the other  two were evaluated by  another  group. As  a  consequence,  some of these effects could be tested using within‐participants comparisons, while the others should be tested using between‐participants comparisons. Due to the particular design of the study, ANOVA F‐tests were hard to interpret. Therefore, we used T‐tests only.  

 

3.2. RESULTS 

On  the noisiness  rating  (lawaaierig),  the  two alarm clocks with noisy sound were assessed as noisier than the two alarm clocks with quiet sound (T  (SNIQ – SQIQ) = 7.0, p = 0.001; T (SNIN – SQIN) = 4.2; p = 0.002). For kettles, only the noisiness of the  product with  noisy  sound  and  quiet  image was  significantly  higher  than  the noisiness of the product with quiet sound and noisy image (T (SNIQ – SQIN) = 2.3; p = 0.03). For kettles, the differences between the two products with the same image and different  sounds were not  significant. Overall,  these  results  suggest  that  the overall noisiness of both products (lawaaierigheid) depends on sound;  it  is not  in‐fluenced by the visual noisiness.  

The pattern of annoyance  ratings  roughly matched  the noisiness  ratings  for both products  (see  Figure  7.4).  The  pleasantness  ratings  for  alarm  clocks  showed  the reversed pattern, although somewhat less pronounced: two alarm clocks with quiet sound were assessed as more pleasant  than  two alarm clocks with noisy sounds. However,  for  kettles  the  only  significant  difference  in  pleasantness  ratings was found between the unpleasant combination of noisy stimuli (SNIN) and the combi‐nation of noisy sound and quiet image (SNIQ). 

The  ratings  of  product  noisiness  (lawaaierig) were  positively  correlated with  the ratings of annoyance (irritant; 0.77 for alarm clocks and 0.69 for kettles) and nega‐tively correlated with the ratings of pleasantness (prettig; ‐ 0.56 and ‐ 0.27 for both products,  respectively).  All  correlations  were  significant  at  the  0.05  level  (two‐tailed). Noisiness  (lawaaierig) was positively correlated with one of  the  two addi‐tional noisiness measures (schreeuwerig; 0.37 and 0.35). The correlation between lawaaierig  and  the  second  noisiness measure  (druk) was  positive  only  for  alarm clocks (0.40), but was not significant for kettles.  

Page 159: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

159

Alarm clock

0

1

2

3

4

5

6

7

8

9

10

Sound NoisyImage Noisy

Sound NoisyImage Calm

Sound CalmImage Noisy

Sound CalmImage Calm

Noi

sy (L

awaa

ierig

)

Kettle

0

1

2

3

4

5

6

7

8

9

10

Sound NoisyImage Noisy

Sound NoisyImage Calm

Sound CalmImage Noisy

Sound CalmImage Calm

Noi

sy (L

awaa

ierig

)

 

Alarm clock

0

1

2

3

4

5

6

7

8

9

10

Sound NoisyImage Noisy

Sound NoisyImage Calm

Sound CalmImage Noisy

Sound CalmImage Calm

Ann

oyin

g (ir

ritan

t)

Kettle

0

1

2

3

4

5

6

7

8

9

10

Sound NoisyImage Noisy

Sound NoisyImage Calm

Sound CalmImage Noisy

Sound CalmImage Calm

Ann

oyin

g (Ir

ritan

t)

Alarm clock

0

1

2

3

4

5

6

7

8

9

10

Sound NoisyImage Noisy

Sound NoisyImage Calm

Sound CalmImage Noisy

Sound CalmImage Calm

Ple

asan

t (pr

ettig

)

Kettle

0

1

2

3

4

5

6

7

8

9

10

Sound NoisyImage Noisy

Sound NoisyImage Calm

Sound CalmImage Noisy

Sound CalmImage Calm

plea

sant

(pre

ttig)

 

** Differences between means are significant at .01 level, * at .05 level (2‐tailed T‐test). 

FIGURE 7.4. MEAN NOISINESS, PLEASANTNESS AND ANNOYANCE RATINGS FOR KETTLES AND ALARM CLOCKS (WITH SE). 

**

**

**

**

**

**

**

**

** 

*

**

**** 

** 

Page 160: ANNA FENKO - TU Delft

CHAPTER 7 

 

160

4. DISCUSSION 

4.1. PRE‐STUDY 

In agreement with previous studies, our sound pre‐study found a high correlation between  noisiness  and  annoyance  (Bowsher  and  Robinson,  1962;  Griffiths  and Langdon, 1968; Berglund et al., 1976). In addition, our pre‐study demonstrated that noisiness of a product sound depends on the sound frequency  (high‐pitch sounds were experienced as noisier than low‐pitch sounds) and the complexity of the fre‐quency spectrum (sounds with multiple frequencies were assessed as noisier than sounds  with  one  dominant  frequency).  These  results  are  in  line  with  psycho‐acoustical data  (Kryter, 1966; Hellman, 1982; Takeshima et al., 1991).  In our pre‐study, rhythmical sounds were assessed as noisier than constant sounds. This cor‐responds with  other  findings  that  sounds with marked  rhythm  are  perceived  as more annoying than random sounds (Northwood, 1963).  

We also found some differences in sound experience between products. For whis‐tle  kettles  the mean  noisiness  ratings  varied  between  1.76  and  9.48, while  for alarm  clocks  they  varied between 4.76 and 8.24.  In other words, even  the most quiet alarm clock sound was assessed as quite noisy.  It  is possible  to explain  this result by a difference in meanings and associations attached to these two products. The sound of a whistling kettle may be associated with the pleasant experience of making tea, while the sound of an alarm clock  is typically associated with the un‐pleasant  experience  of waking  up  early  in  the morning.  These  associations may influence the experience of noisiness and annoyance of product sounds.  

The results of the visual pre‐study demonstrate that noisiness of an image depends on the amount of elements and colors, regularity, symmetry and clutter in the vis‐ual  pattern.  These  findings  correspond  to  the  existing  data  on  perceived  visual complexity (Feldman, 1997; Palmer, 1999; van der Helm, 2000; Heaps and Handel, 1999; Oliva & Torrabla, 2001).  

For  both  products we  found  strong  positive  correlations  between  noisiness  and annoyance.  The  correlation  between  noisiness  and  pleasantness  was  negative. Studies  in experimental aesthetics generally  tend  to  show  that medium  levels of visual complexity are perceived as pleasant, while low and high levels are perceived as unpleasant (Berlyne, 1971).  In our pre‐study the mean noisiness for visual pat‐terns varied  from 3.05  to 7.17  for kettles and  from 3.30  to 6.80  for alarm clocks. 

Page 161: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

161

The negative correlation between noisiness and pleasantness may suggest that our visual stimuli varied in the range of medium to high complexity.  

In our pre‐study  the  link between  visual noisiness  and  visual  complexity was  as‐sumed, but not tested experimentally. Further research is needed to systematically investigate this relationship. For this research we suggest to select a broader range of stimuli, so that the ratings are not limited to the noisy/complex pole of the scale. This will give researchers the possibility to find out whether the relation between noisiness and pleasantness follows the bell‐shaped curve pattern similar to the re‐lation between complexity and pleasantness (Berlyne, 1971; Messinger, 1998).  

 

4.2. MAIN STUDY 

The results of the main study demonstrated that the overall experience of noisiness (lawaaierig) was influenced mainly by the sound for both alarm clocks and kettles. The annoyance  rating  for both products was also dominated by sound, while  the contribution of  the visual pattern was not significant. The ratings of pleasantness for the alarm clock were also dominated by sound. For the kettle, however, there were not enough significant differences in mean pleasantness ratings. 

The differences between  the  two products  in our  study  can be explained by  the choice of stimuli for the main experiment. The differences between the noisy and quiet  stimuli  chosen  for  kettles  (2.0  for  sounds  and  2.31  for  images) were  less prominent than the noisiness differences between the alarm clock stimuli (3.19 for sounds and 3.50 for  images). This can explain why  in the main study the effect of sound was more pronounced for alarm clocks than for kettles.  

Spontaneous  comments made  by  participants  during  the  experiment  indicated  that both alarm clock  sounds were quite  irritating. Several participants asked  to  stop  the sound; others commented that they always found alarm clocks annoying. On the other hand, the cheerful visual patterns that had been chosen for kettles were perceived as pleasant by most participants. Several participants commented: “What a nice kettle!”  

 

4.3. GENERAL DISCUSSION 

Product experience can be defined as “the awareness of the psychological effects elicited by  the  interaction with  a product,  including  the degree  to which  all our 

Page 162: ANNA FENKO - TU Delft

CHAPTER 7 

 

162

senses are stimulated, the meanings and values we attach to the product, and the feelings and emotions that are elicited” (Hekkert & Schifferstein, 2008, p.2). In this study,  noisiness was  analyzed  as  the multisensory  product  experience  that  inte‐grated auditory and visual sensory perceptions and evoked affective reactions (an‐noyance and unpleasantness).  

Many perceptual  (or  sensory) adjectives,  such as  sharp, dull  (touch),  sweet,  sour (taste), loud, and quiet (sound), extend their meaning from one basic or prototypi‐cal sense modality to one or more secondary modalities (Ullmann, 1957; Williams, 1976). For instance, Sweetser (1990) argues that the connection between different meanings of sensory words is metaphorical in nature.  

When we analyze multisensory experiences such as “noisy”,  it  is difficult to deter‐mine which meaning of the word is literal and which is metaphorical, because both meanings may  refer  to  basic  perceptual  processes.  For  example,  Rakova  (2003) argues  that  synaesthetic  and  double‐function  adjectives  (such  as  ‘bright’  and ‘sharp’) are literal rather than metaphorical because all meanings of such adjectives refer to the same psychologically primitive concept. For example, both tactile and gustatory meanings of the term ‘hot’ are literal. It does not seem right to say that thermal heat is more primary than food spiciness, because virtually the same neu‐ral mechanisms underlie the perception of thermal heat and spicy heat. Other au‐thors  (e.g. Nogales, 2007) argue that a sharp sound  is metaphorically related to a sharp knife because of a particular shared property. Specifically,  the sharp sound plays  the  same  role with  silence  and  human  ears  as  the  sharp  knife  plays with bread and cheese.  

In the case of “noisy”, complexity can be a property that is shared by auditory and visual noisiness. Complexity in the auditory modality refers to the presence of dif‐ferent frequencies and to phases that are not systematically related to each other. The  ultimate  auditory  complexity  is  “white  noise”  that  contains  all  frequencies. Complexity in the visual modality refers to the large amount of different elements, colors,  and  textures  that  are put  together without  any  regularity and  symmetry. Complexity may not be the only property common to visual and auditory noisiness. A high frequency sound and a pure bright color are not complex, but they can be perceived as quite noisy. And vice versa, a painting or a piece of music can be per‐ceived as complex but not noisy.  

Lehrer (1978) suggests that the transfer of meaning in adjectives is mainly based on the general experiential dimensions of  intensity and evaluation.  It  is possible that 

Page 163: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

163

noisiness  refers  to certain  sensory properties  (such as pitch, hue,  intensity,  loud‐ness,  brightness,  etc.)  that  are  perceived  as  unpleasant,  because  they  attract automatic attention and distract people from their normal activity. It is also possi‐ble  that  they are experienced as unpleasant because  they evoke arousal  that ex‐ceeds the optimum arousal level (Berlin, 1971).  

The correlation between auditory noisiness and annoyance  is widely documented (Bowsher and Robinson, 1962; Griffiths and Langdon, 1968; Berglund et al., 1976; Schultz, 1978) and was  supported by our  findings. We also  found correlation be‐tween visual noisiness and annoyance. Nevertheless, noisiness  is not necessarily a negative experience. If we look at the language differences, we can find some posi‐tive connotations with noise. In English there is no special verb to indicate an activ‐ity of making noise, but such verbs exist in German (rauschen), French (bruire), and Russian  (шуметь).   They are used mostly  to  indicate  the  sounds of nature  (e.g., trees, water), of happy crowds of people (e.g., an outdoor festival, a ball, a party), of playing children, etc. All these situations refer to the positive meaning of noise‐making. This suggests that the correlations between noisiness, annoyance and un‐pleasantness may not be universal.  

Since sensory connotations of English and Dutch terms for noisiness are different, product experiences of noisiness might also differ between participants who speak different  languages.  In recent years, many studies have claimed experimental evi‐dence that language has an important influence on cognitive processing of sensory information  (Levinson,  2003;  Li  and  Gleitman,  2002;  Robertson  et  al.,  2000; Gordon, 2004; Núñez and Sweetser, 2006). Although  in our  study we  found high positive  correlations  between  Dutch  synonyms  of  noisiness  (such  as  lawaaierig, druk,  schreeuwerig, and  luidruchtig), our  results might have been different  if we had  used  any  of  these  synonyms,  because  their  exact meanings  differ.  Further cross‐cultural  and  cross‐language  research  is  needed  into  the  relationships  be‐tween the various noisiness synonyms, between noisiness and complexity, and into the  relationships  between  noisiness,  annoyance  and  unpleasantness  in  different product‐usage  situations.  Such  research  can  clarify  the  experiential  basis  and metaphorical meaning  of  noisiness  and  help  designers  to  create  products  with more pleasurable sensory properties. 

 

 

Page 164: ANNA FENKO - TU Delft
Page 165: ANNA FENKO - TU Delft

165 

 

 

 

 

 

 

 

 

 

 

CHAPTER 8 

DISCUSSION  

The studies reported in this thesis aimed at understanding perceptual mechanisms of sensory  integration  in product experience.  In the following chapter we will dis‐cuss our  general  findings,  theoretical problems of  the  research,  perspectives  for future research, and the implications of our findings for industrial design and prod‐uct development. 

 

1. GENERAL FINDINGS In the previous chapters we described several studies that addressed the problem of sensory dominance from various angles. We have demonstrated that the impor‐tance of particular sensory product experiences depend on the type of product and experience, the stage of user‐product interaction, and the language that is used to conceptualize and communicate  the experience. These  findings demonstrate  that the concept of sensory dominance  is multilayered and can be addressed at differ‐ent levels. 

 

Page 166: ANNA FENKO - TU Delft

DISCUSSION 

 

166

1.1. SENSORY DOMINANCE  

In  the  Introduction we mentioned  that  sensory dominance  can manifest  itself  in two variants. When a perceptual conflict takes place between two or more sensory modalities, different outcomes can follow: (1) a person may ignore the information of one modality and rely entirely on the others,  in which case one modality really dominates the other ones; (2) the data from different modalities can be taken into account  to  some extent,  resulting  in  some kind of  intermediate percept. The ex‐perimental studies described in Chapters 5, 6, and 7 illustrate both variants. For the experiences of freshness and noisiness, the data demonstrate the total dominance of one modality (olfaction in case of freshness, audition in case of noisiness), while visual  information did not  influence  the experience  ratings.  In  the experiment on warmth  the  two modalities  (vision  and  touch)  contributed  about  equally  to  the experience ratings. 

Various psychological mechanisms may determine which modality dominates and to what extent. Our  findings suggest that  for products for which a certain experi‐ence  is  not  important  (e.g.,  freshness  for  candles),  the  effect  of modality  domi‐nance is not as pronounced as for products for which the same experience is very important  (e.g.,  freshness  for  soft  drinks  and  dishwashing  liquids).  These  results may  be  explained  by mechanisms  of  selective  attention, which  allow  people  to concentrate on particular events of  interest and  importance  in their environment. Research has shown that people can direct their selective attention to a particular sense (e.g., Spence et al, 2001), which can facilitate the perception of stimuli in that modality if it is important for a particular product property. When a certain sensory property  is not  important  (e.g.,  smell  for  candles),  attention may be diverted  to another sensory modality or divided simultaneously between several senses. This may explain the differences in sensory integration effects between products.  

Furthermore, modality  importance reported  in questionnaire studies may deviate from modality  importance  found  in  experimental  research,  in which participants are unaware of experimental manipulations. For instance, in case of warmth, ques‐tionnaire data indicated the dominance of touch, while experimental data demon‐strated equal  impact of vision and  touch on  the experience of warmth. This  sug‐gests  that  people may  be  unaware  of  the  relative  influence  of  various  sensory stimuli on their overall product experience.  

 

Page 167: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

167

1.2. DYNAMICS OF SENSORY DOMINANCE 

In the studies described in Chapters 2 and 3 we investigated the temporal changes in sensory modality  importance  in product experience. The  first study used ques‐tionnaire methodology and a wide range of products. The results of this study indi‐cate  that  vision  is dominant at  the buying  stage, while other modalities become important at  later stages. Which modality will dominate at  later stages of product usage  probably  depends  on  the main  product  functions,  characteristics  of  user‐product  interactions, and  the  frequency with which a certain  sensory modality  is involved in product use.  

The  second  study, which  employed  an  experimental methodology,  confirmed  the leading role of vision at the buying stage. However, when people were asked about the importance of various modalities in the buying situation, they also mentioned the importance of sensations which they could not actually experience at that moment (e.g., taste). This suggests that certain sensory properties (such as the taste of food) may be taken  into account on the basis of memory and  imagination; people do not necessarily need  to experience  these  sensations  in order  to be able  to  judge  their importance. On  the other hand,  this  research also suggests  that actual  sensory ex‐periences with the product (e.g., the liking for a flavor of food) do not automatically change negative attitudes toward certain product groups (such as dehydrated food).  

This  research  indicates  that  relationships between  sensory experiences,  affective reactions and cognitive attitudes toward products may differ between sensory mo‐dalities. We suggest that visual and auditory experiences may be more closely con‐nected  to  cognitive  processes,  while  touch,  olfaction  and  taste  may  be  more strongly associated with affective processes. These differences may have an impact on the overall product experience. For  instance, vision may be dominant  in situa‐tions that  involve cognitive tasks and rational decisions, while touch and olfaction may be dominant in situations that involve affective judgments. 

 

1.3. THE ROLE OF LANGUAGE 

In the study described in Chapter 4 we investigated the importance of sensory mo‐dalities for various descriptors of product experience. The results showed that for the affective descriptors of product experience (pleasant‐unpleasant, good‐bad) all modalities were equally important.  Symbolic descriptors (such as modern, expen‐sive, or feminine) tended to be multisensory, but most of them demonstrated vis‐

Page 168: ANNA FENKO - TU Delft

DISCUSSION 

 

168

ual dominance. Sensory descriptors (such as colorful, loud, or soft) typically showed dominance  for  the corresponding primary modality.  In addition,  they showed sig‐nificant language differences. The possible reason for these differences is that sen‐sory adjectives usually have several meanings. They can transfer their meaning to other sensory modalities  (e.g.,  ‘sweet sound’,  ‘warm color’) and  to other psycho‐logical  domains  (e.g.,  ‘hard  decision’,  ‘clear  explanation’).  Language  differences between  sensory  descriptions  can  be  explained  by  the  fact  that  metaphorical meanings of sensory descriptors differ between the languages. 

These findings make us wonder to what extent our experience is actually embodied in  language. Can  the words we use  to describe our experience actually  influence what we  feel or  think, or what we see or smell? The phenomenological  tradition emphasizes  the  exceptionally  important  role  of  language  in  human  experience. Dilthey (1985) argues that language can be seen as a linguistic map that names the possibilities  of  human  experiences.  In  phenomenological  philosophy,  human  ex‐perience is seen as essentially linguistic, and language is understood not as a thing people  create and possess but as  the ongoing origination of meaning  (Gadamer, 1975). Merleau‐Ponty  (1973)  argues  that  speech  is  not  the  clothing  of  thought, rather  it  is  its  body;  speech  does  not  simply  transmit  thought,  rather  it  accom‐plishes, or completes, it. "Understood in this way, human being would be bespoken by  language"  (Heidegger, 1971, p. 192). The socio‐cultural  theory of  thought  (Vy‐gotsky, 1986) shares the view of language as the central mechanism of thinking. In the Vygotskyan theory, the first speech of the child  is a form of social  interaction, which later acquires a second function; social speech turns into inner speech, which becomes a means for cognitive self‐regulation, a vehicle for thinking. 

A proposal of linguistic relativity emphasizes a distinctive role of language in inter‐preting experience and  influencing thought (Gumperz and Levinson, 1996). Whorf (1956) proposed that the categories of each language determine a way of perceiv‐ing,  analyzing,  and  acting  in  the world.  In  recent  years,  extensive  research  has demonstrated  language  influence  on  the  experience  of  space  (Levinson,  1996), color  (Kay  and Kempton, 1984; Robertson et  al., 2000), number  (Gordon, 2004), and time (Boroditsky, 2001; Núñez and Sweetser, 2006).  

However,  the emphasis on  language  limits  the area of experience  to phenomena that can be consciously processed and verbally  reported. However, experience  is not  easily  captured  in  language;  it  is  always more  immediate, more  enigmatic, more complex, more ambiguous than any description can do justice to. Some com‐ponents of experience may be more difficult to verbalize than others. For instance, Izard  (2009) argues  that  feeling an emotion does not guarantee  that  it will be  la‐

Page 169: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

169

beled, articulated, or sensed in reflective consciousness or at a high level of aware‐ness. The  level of awareness of an emotion depends  in part on  its  intensity and expression, and in part on labeling, articulating, and acknowledging the emotional experience.  Infants  experience  and  express  basic  emotions  long  before  they  ac‐quire language (Izard et al., 1995). Moreover, acquiring language does not guaran‐tee that affective experiences can always be identified and communicated verbally. Even adults have great difficulty articulating a precise description of their feelings (cf. Langer, 1967/1982).  

Olfactory  sensations  represent  another example of experience  that  is difficult  to verbalize. Humans are very good at detecting an odorant but are poor at naming it (Yeshurun  and  Sobel,  2010).  For  instance,  the  odorant  ethylmercaptan, which  is often added to propane as a warning agent, can be detected at concentrations be‐low 1 part per billion  (ppb), which  is equivalent  to approximately  three drops of odorant within an Olympic‐size swimming pool (Whisman et al., 1978). At the same time, people are unable to name by smell at  least 50% of the odorous household items  they use daily  (Cain, 1979; de Wijk et al., 1995; Lawless & Engen, 1977).  In our  interview  research,  visual  and  tactile  properties  of  products were  reported more  often  than  auditory  and  olfactory/gustatory  properties.  In  Schifferstein’s (2006) questionnaire study of sensory modalities  importance, effects of smell and audition on product experience were also rated low. These results may reflect the differences in availability of specific vocabulary for various sensory modalities. 

To understand product experience, we need to use language. Although verbal self‐reports are the main source of data in research on experience (Barrett et al., 2007; Russell, 2003),  the relevance of  these data may be different  for different compo‐nents of experience. For  instance, affective components of experience, as well as some of sensory components (such as olfaction) are especially difficult to put  into words.  To  investigate  these  experiences,  we  need  to  supplement  verbal  self‐reports with non‐verbal measurements. 

 

2. THEORETICAL PROBLEMS  As a starting point of our research we used the definition of product experience as “the awareness of the psychological effects elicited by the interaction with a prod‐uct, including the degree to which all our senses are stimulated, the meanings and values we attach to the product, and the  feelings and emotions that are elicited” (Hekkert  &  Schifferstein,  2008,  p.2).  This  definition  emphasizes  the  intrinsically 

Page 170: ANNA FENKO - TU Delft

DISCUSSION 

 

170

conscious character of experience and distinguishes between  three main compo‐nents: sensory, symbolic and affective experiences.  

Throughout our research, we repeatedly encountered phenomena that were diffi‐cult to describe using this theoretical model. Below we will discuss how the results of our studies  illustrate two main theoretical problems of product experience: (1) the degree of awareness in product experience, and (2) the relationships between sensory, symbolic, and affective components of product experience. 

 

2.1. THE PROBLEM OF AWARENESS IN PRODUCT EXPERIENCE 

Experiential phenomena are  conscious by definition, but experiences of different sensory modalities may have different degrees of awareness. The degree of aware‐ness of a specific sensory property may depend on selective attention that  is paid to this property, which in turn may depend on the importance of this property for a particular product. For example, people may pay more attention to olfactory prop‐erties of food and cleaning products than to the smell of home decoration products (such as candles), because they find freshness more important for food and clean‐ing products than for other product groups.  

The  awareness  of  product  experience may  be  represented  as  the  flow  of  con‐sciousness  as  it  was  first  described  by  the  pioneers  of  scientific  psychology, Wilhelm Wundt and William  James. According to Wundt, consciousness  is a  func‐tion of the scope of attention, which may either actively focus on a perceived rep‐resentation or passively find certain representations suddenly thrusting themselves into the center of attention (Wundt, 1874, p. 267; 562). Attention may be under‐stood  in  terms  of  the  differing  degrees  to which  representations  are  present  in consciousness. Wundt appeals to an analogy with the field of vision: 

If we say that the representations present at a particular moment are in con‐sciousness's field of vision, then that part of the field upon which our attention is turned may be called the inner focal point of vision. The entry of a represen‐tation into the field of inner vision we call “perception,” and its entry into the focal point of vision we call “apperception” (Wundt, 1874, p. 267). 

For James, our capacity for attention to one thing rather than another is the sign of an “active element in consciousness” (James, 1890, p. 285). Modern psycholo‐gists often link selective attention to the affective processes, which are necessary 

Page 171: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

171

for  the  simplest  forms  of  exploration  and  learning  as well  as  for  higher‐order cognition and organized behavior (i.e., Izard, 2009). Affective mechanisms play a crucial role in selecting relevant sensory information from the outside world and combining  it  into meaningful experience  (Frijda, 2006; Russell, 2003). They  can enhance sensory processing, so that a person can efficiently assess the relevance or value of  the stimulus,  thereby  influencing which contents are experienced  in the moment and which are more  likely to be stored  in  long‐term memory (Edel‐man & Tononi, 2000).  

Another explanation for different degrees of awareness is the availability of verbal descriptions. For example, most respondents found it difficult to describe their ol‐factory  experiences.  The  fact  that  linguistic  categories  of  visual  experiences  are much richer and more elaborated in Western languages could explain why descrip‐tions of other  sensory  experiences  are  less  common  (Hinton  and Henley,  1993). Nonetheless, Izard argues that the lack of linguistic accessibility does not automati‐cally make an experience unconscious. He gives an example of “a feeling that you know  you  are  experiencing  but  cannot  specifically  identify  or  describe”  (Izard, 2009, p.16). Inability to put the feeling into words bars it from linguistic accessibil‐ity,  but  not  from  phenomenal  consciousness  and  various  levels  of  awareness. A feeling  in phenomenal  and other nonlinguistic  levels of  consciousness  retains  its properties, including its power to motivate and regulate cognition and action. Such nonlinguistic levels of consciousness may explain well‐known examples of influence of smell on product  liking and quality  judgments, of which people seem to be un‐aware (Larid, 1932; Demattè et al., 2006). 

A vast amount of the processes in the brain and the rest of the body (blood circula‐tion, digestion) usually do occur without any awareness. Self‐reports are often not useful  to  explain why  people  experience what  they  do,  because  people  are  not aware of  the  causal processes  that  influence  their experience  (Nisbett & Wilson, 1977; Wilson & Dunn, 2004). For  instance, people are usually unaware of percep‐tual mechanisms underlying  their multisensory experience.  Illusions  that demon‐strate visual dominance over audition  in multisensory speech perception, such as the McGurk  effect  (McGurk  and MacDonald,  1976)  and  the  ventriloquism  effect (e.g., Bertelson, 1999), are experienced automatically and depend neither on  the direction of selective attention nor on the knowledge of  the  illusion  (Vroomen et al., 2001).  

People can be mistaken in their assessment of the relative contribution of different senses to a product experience, as  illustrated by our experiment on the multisen‐

Page 172: ANNA FENKO - TU Delft

DISCUSSION 

 

172

sory  integration  in  the  experience  of  warmth:  participants  believed  that  their judgments  of warmth were  dominated  by  tactile  properties, while  experimental findings  suggest  that  vision plays  an  equally  important  role  in  the  experience of warmth in products.  

Although there  is considerable agreement on the psychological processes that are accessible to consciousness, there  is no consensus on the contents and processes of the unconscious  (Bargh & Morsella, 2008).  Izard  (2009) argues that dividing all mental processes  into two domains —conscious and unconscious — might be the greatest oversimplification in current psychological science. 

In  the context of product experience,  it might be useful  to differentiate between three types of unconscious phenomena: (1) pre‐conscious experiences that are not deliberately attended to and therefore represent a periphery of consciousness, but can be accessed in principle; (2) experiences that are difficult to report, because of the lack of accessible verbal descriptions; and (3) automatic mental processes (such as mechanisms of multisensory integration) that are not, under any circumstances, accessible  to  introspection. Clearly,  these three  types of unconscious phenomena require different methods of investigation. In the first case, attention variables can be manipulated experimentally (i.e., Spence & Dryver, 2004). In the second case, it is possible  to use non‐verbal measurements of experience,  such as  facial expres‐sions (Ekman & Friesen, 1978) or physiological activation (Cacioppo et al., 1993). In the  third  case,  traditional  experimental methods  can  be  supplemented  by  new neurobiological methods  (Damasio,  1999;  leDoux,  2000),  including  neuroimaging (fMRI and PET) research (Murphy et al., 2003; Phan et al., 2002).  

 

2.2. INTERACTIONS BETWEEN DIFFERENT COMPONENTS OF PRODUCT 

EXPERIENCE 

This thesis is focused on sensory properties of products, which represent the most ‘objective’ component of product experience. Indeed, whether we like it or not, we smell the food while cooking it, hear the sound of a vacuum cleaner while cleaning a room, and feel the bumps of the road while driving. Sensory experiences tie us to the material world and to the objects that make up our environment. Nevertheless, our findings suggest that sensory experiences are not immune to the subjectivity of our emotions and beliefs.  

 

Page 173: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

173

2.2.1. INTERACTIONS BETWEEN SENSORY AND AFFECTIVE EXPERIENCES 

The results of our experimental research on warmth, freshness, and noisiness sug‐gest  that  sensory  product  experiences  have  affective  components.  We  found strong  correlations  between  freshness  and  pleasantness  of  visual  and  olfactory stimuli and between noisiness and annoyance of auditory and visual  stimuli. The suitability of a certain sensory aspect for a product may influence this relationship. For example,  fruity smells were chosen more often as  ‘fresh’  for a softdrink, and the  floral  smell was  judged as  the most  ‘fresh’  for a dishwashing  liquid. Respon‐dents commented that fruit smells seemed more ‘nice’ and suitable for a softdrink, while  the  floral  smell was  ‘strong’  and  ‘pleasant’  and  suitable  for  a dishwashing liquid.  In  some  cases,  respondents may have  confused  sensory  and  affective  ex‐periences. For instance, fleece scored higher than wool on the scale of ‘warmth’ for a scarf, because ‘it does not irritate your skin’.  

Affective properties of  sensations were already noticed by early experimental psy‐chologists. According to Wundt, elementary (or pure) sensations display three com‐ponents: quality,  intensity,  and  affective  tone  (Wundt, 1874, p. 282‐283). Modern psychologists also suggest that an affective tone can be experienced as a property of the external world  rather  than as  the person’s  reaction  to  it  (Barrett et al., 2007). Indeed, we experience some people as nice and others as mean, some foods as deli‐cious and others as distasteful, some pictures as beautiful and others as ugly. 

The impact of sensory stimuli on affective experience was also reported in empiri‐cal studies on consumer behavior. The classic research in this domain is Zajonc and colleagues’ investigation of mere exposure, where simply being exposed to a stimu‐lus leads to more positive affective reactions (Kunst‐Wilson & Zajonc, 1980; Zajonc, 1968) due  to  familiarity  (Whittlesea, 1993). Consumer studies also  find a positive relationship between  typicality of an object  (or  similarity  to other members of a category) and affective evaluation of  this object.  In general, more  typical objects are  liked better  (Carpenter & Nakamoto, 1996; Simonin & Ruth, 1998; Veryzer & Hutchinson, 1998; Zhang & Sood; 2002). 

Berlyne  (1970) considered whether  stimulus  complexity mediates  favorable  judg‐ments  of  novel  or  familiar  objects. Other  structural  sensory  variables  that were reported  to  influence affective  judgments  include  symmetry  (Frith & Nias, 1974) and other principles of good gestalts (Arnheim, 1954).  

In our  studies on multisensory  integration, positive correlations between  sensory (warm, noisy) and affective (pleasant, annoying) ratings were found mainly for sin‐

Page 174: ANNA FENKO - TU Delft

DISCUSSION 

 

174

gle modality  stimuli  (colors,  smells,  sounds, etc.). When  these  stimuli were  com‐bined  in  real products, correlations with affective  judgments were  found only  for noisiness, but not  for warmth  and  freshness.  These  findings  suggest  that  simple one‐to‐one correspondences between sensory and affective experiences may exist only  for stimuli created  in a  laboratory. When sensory stimuli are experienced as properties of real, complex products, situational variables and symbolic meaning of products may  influence  affective  reactions.  For  instance,  in  Chapter  6  the  least fresh  combination of  color  (purple) and  smell  (vanilla) was assessed as  the most pleasant  for  the dishwashing  liquid, probably because  the dishwashing  liquid  is a boring product, and users appreciate some newness and surprise in it. 

 

2.2.2. INTERACTIONS BETWEEN SENSORY AND SYMBOLIC EXPERIENCES  

Our findings demonstrate that sensory product experiences can be influenced by the symbolic meanings people apply to products. For instance, red and yellow are usually perceived as warm colors, while blue and green are perceived as cold colors (Wright, 1962). However, in our study green scarves were rated higher on warmth than yellow scarves. Participants commented  that yellow  is  ‘too bright and more  suitable  for a spring scarf’, while green is ‘darker and more suitable for a warm winter scarf’.  

Furthermore,  in  the  study  on  noisiness  of  alarm  clocks  and whistle  kettles, we found  differences  in  sound  experience between  products. While mean  noisiness ratings for whistle kettles varied depending on sound frequency and complexity, all alarm clock sounds were assessed as quite noisy. This result may be explained by a difference in meanings and associations attached to these two products. The sound of a whistling kettle may be associated with the pleasant experience of making tea, while  the  sound of an alarm  clock  is  typically associated with  the unpleasant ex‐perience of waking up early  in  the morning. These memories might  influence the experience of noisiness of product sounds.  

The  results  of  our  study  on  naturalness  dynamics  in  a dehydrated  food  product suggest that mere visual exposure to the product may activate users’ cognitive atti‐tudes and beliefs about  the product, while other senses, such as smell and  taste, may trigger mainly affective reactions. In spite of the positive sensory experiences and  emotions  during  cooking  and  eating  the  product, most  participants  did  not change  their  pre‐existing  attitudes  about  the  dehydrated  food.  In  a  re‐purchase situation  their negative cognitive attitudes  towards  this product group were acti‐

Page 175: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

175

vated again by  the visual  images of products, while previous positive sensory ex‐periences were ignored. 

Consumer studies also demonstrate  that cognitive processes  (concepts, attitudes, ideas) can influence affective and sensory product experiences. In one of the classic studies (Allison & Uhl, 1964), consumers who drank beer with visible brands judged those beers as highly variable  in taste, whereas consumers who drank unbranded beers tended to rate them all as tasting similar.  In addition, consumers tended to prefer the beer with their favorite label. In a recent investigation, Lee et al. (2006) asked participants to drink a glass of a commercially available beer and a glass of ‘MIT brew,’ which consisted of the same beer with the addition of one  ingredient — balsamic  vinegar. Before people  learned  that  the MIT brew had  vinegar,  they liked the beer just fine. But when they knew about the vinegar, the beer was per‐ceived as repulsive.  

The impact of cognitive beliefs and expectations on preferences is powerful enough 

to be observed in brain imaging studies. McClure et al. (2004), for example, asked 

participants who preferred Coke to Pepsi to drink Coke and Pepsi; participants pre‐

ferred Coke, but only when  they  knew  it was Coke. The analysis of  fMRI  images 

revealed  that  these preferences were  reflected by  recruitment of brain areas  in‐

volved  in  the  processing  of  reward.  In  a  similar  investigation,  Plassmann  et  al. 

(2008) asked participants  to  taste one wine several  times but  told  them  that  the 

wine was  either  cheap  or  expensive.  Consumption  of  ‘higher‐priced’ wines was 

related to greater recruitment of reward circuitry in the brain. In a field study, Shiv 

et al. (2005) showed that people who had caught colds rated their cold remedies as 

more effective  if  they had paid  full price  for  them. Another  investigation demon‐

strated similar placebo effects  for analgesics pills: Participants who were  told  the 

pill had been discounted were unable  to  tolerate as much physical pain as  those 

who were told the pill was not discounted (Waber et al., 2008).  

Taken  together,  these studies suggest  that preconceptions and  ideas about prod‐

ucts may modify  affective  and  sensory  components  of  product  experience.  The 

taste of beer, wine and Coke, as well as the feeling of physical pain, can be affected 

by  price  and  brand  information.  Sensory  experiences  (such  as  noisiness)  can  be 

affected by pleasant and unpleasant memories associated with products  (whistle 

kettles and alarm clocks). Furthermore, negative cognitive attitudes  toward some 

Page 176: ANNA FENKO - TU Delft

DISCUSSION 

 

176

product groups (such as a dehydrated food) may prevent people from appreciating 

their sensory properties (such as good taste and smell).  

All these examples demonstrate the complex  interactions between sensory, affec‐tive and cognitive components of product experience, which cannot be easily sepa‐rated and  investigated  independently.  In a  continuously  changing  stream of  con‐sciousness, different  sensory properties of products  such as  color,  shape,  sound, smell, as well as affective and cognitive contents  like beliefs, memories and emo‐tions, bind together  into a single experience  (Barrett et al., 2006). Therefore,  it  is important  for a theoretical model of product experience to take  into account the integrative processes  that bind  together  information  from different  sensory mo‐dalities with affective reactions and higher order cognitive processes.  

 

3. PERSPECTIVES FOR FUTURE RESEARCH 

3.1. HOW ARE SENSORY AND AFFECTIVE PRODUCT EXPERIENCES 

RELATED? 

In product design, it is important to promote positive, pleasurable product experi‐ences and  to prevent  frustration, pain, and  stress associated with product usage (Hancock  et  al.,  2005).  Traditionally,  designers  have  been  concentrating  on  pre‐venting bad  things  from happening during product usage, but  recently  they have started paying more and more attention to creating pleasurable products (Jordan, 2000), exciting product experiences (Norman, 2004; Desmet, 2002), and enjoyable product‐user  interactions  (Han  &  Hong,  2003;  Helander  &  Tham,  2003;  Khalid, 2004; Krippendorff, 2004; Liu, 2003).  

An  interesting  avenue  of  investigation  is  to  look  at  the  relative  contribution  of stimuli of various sensory modalities to the affective product experience. Multiple data point to the differences between sensory modalities in respect to their ability to produce strong affective reactions. Researchers argue that touch (Hollins, 2010) and olfaction  (Yeshurun &  Sobel, 2010) are more  closely  connected  to emotions than vision and audition. It might be  interesting to compare the relative  influence of sensory properties of various modalities on affective product experience using experimental methodology with systematic variable manipulations. 

Page 177: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

177

The idea of one‐to‐one correspondences between sensory attributes of products and users’  affective  reactions  seems  very  tempting. Various  researchers  (e.g., Norman, 2004; Hanckock et al., 2005) have created  lists of attributes that could make users' experiences of products and services enjoyable, pleasurable, and fun. Some examples of attributes  that  induce positive affect are  smiling  faces; warmth;  comfort;  sweet tastes;  pleasant  smells;  bright  and  highly  saturated  hues;  harmonious music  and sounds; rhythmic beats; symmetrical, round, smooth shapes. Attributes that  induce negative affect include darkness; harsh, abrupt sounds; startling noises; bitter tastes; sharp objects; rotting smells; decaying foods; sudden, bright lights; extreme tempera‐tures. However, some researchers point out that empirical studies are needed to link these attributes to experience (Murphy, Stanney, & Hancock, 2003).  

The  studies  in Chapters 5, 6  and 7 demonstrate  that,  although  correspondences between sensory stimuli and affective reactions can be  found within one sensory modality, they can not predict the affective reactions to the combinations of stimuli in real products (see also Schifferstein et al., 2010). Probably, affective experiences are more  dependent  on  the  combinations  of  stimuli  used  and  their  degree  of (in)congruence than sensory experiences. Future research  is needed  into sensory‐affective correspondences, which may concentrate on complex multisensory stim‐uli  and  look  at  such  variables  as  consistency/inconsistency  or  congru‐ence/incongruence of  sensory product properties  (Schifferstein & Verlegh, 1996; Ludden et al., 2007).  

 

3.2. HOW ARE SENSORY AND SYMBOLIC PRODUCT EXPERIENCES 

RELATED? 

No matter how pleasant  and  enjoyable  sensory properties of  a product  are,  the decision to buy and use the product might be influenced by people’s attitudes, be‐liefs and values (Wood, 1995, Mennel, et al., 1992). For  instance, the dynamics of the naturalness experience  in dehydrated food demonstrates that people may  ig‐nore positive gustatory and olfactory experiences  if  they have strong pre‐existing ideas about the naturalness of the product. Another striking example of an  ‘ideo‐logical’ influence on sensory experience is reported by Annemarie Mol:  

“Let me tell you, I don’t like Granny Smiths. In the late 1970s and early 1980s we (my political friends and myself) invested a lot in disliking Granny Smiths. At the time they were always imported from Chile, and thus stained with the 

Page 178: ANNA FENKO - TU Delft

DISCUSSION 

 

178

blood spilled by Pinochet and his men. Once Pinochet had gone, it turned out to be difficult to re‐educate my taste. It should be possible, but so far I have not  succeeded.  Yes,  I  can  eat a Granny  Smith apple: bite,  chew,  swallow, gone. But it does not give me pleasure” (Mol, 2008, p.29).  

In future research  it might be  interesting to  investigate the complex relationships between  sensory  experience  and  symbolic  product  properties. What  sensory  in‐formation  is used  to  form specific product attitudes? How can  these attitudes be changed? Our  findings suggest  that  it may be difficult  to modify beliefs based on sensory experience of olfaction and  taste, probably because  these modalities are more strongly associated with emotions than with cognitive processes. Experimen‐tal data  suggest  that vision and audition are  the  two  sensory modalities  that are most closely connected to rational thinking (Goodale & Humphrey, 1998; Neisser, 1994; Paivio, 2006).  In addition,  in our study of verbal descriptions of product ex‐perience,  symbolic  properties  (such  as modern,  expensive,  and  natural)  demon‐strated strong visual dominance, while affective properties (such as pleasant, excit‐ing, or boring) were equally related to all sensory modalities. Both  for theoretical understanding of product experience and for the practice of product development, it is important to investigate the influence of the information from sensory modali‐ties on the formation and change of attitudes towards different product groups.  

 

3.3. SENSORY METAPHORS IN PRODUCT EXPERIENCE 

In empirical research, it is often difficult to make a distinction between sensory and symbolic components of product experience, because almost any sensory property of a product also has a figurative meaning. The color of a dress can be  ‘loud’; the sound of music  can be  ‘sweet’,  and  so on.  Sensory  terms  that describe physical properties of  things  (such as warm, cold, heavy,  sweet, bitter, etc.) also describe psychological qualities. In future research, we could investigate the interaction be‐tween  sensory  and  symbolic  product  properties  further,  and we may determine how they contribute to the overall product experience.  

We suggest that future research may concentrate on  language differences  in sen‐sory metaphors and employ a broad variety of  languages.  In cognitive  linguistics, sensory metaphors (such as “knowledge is vision”, “understanding is grasping”) are viewed as  fundamental  to  language,  thought, and experience  (Lakoff &  Johnson, 1980,  1999;  Sweetser,  1990).  Most  cognitive  linguists  assume  that  such  basic metaphors are explicit representations of embodied cognitive structures. However, some scholars suggest  that metaphors are not merely representative of universal 

Page 179: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

179

body  experience, but  are  tied  to  specific  socio‐cultural  cognition  (Kimmel,  2006; Zlatev,  2006).  Further  research  into  cultural  differences  in  sensory  metaphors might help  to  resolve  this  theoretical argument.  It might also help  to understand cultural differences  in product  experience  and  to  create products  that  are more closely tailored to various cultural groups of users. 

 

3.4. INDIVIDUAL DIFFERENCES IN SENSORY EXPERIENCE 

Recently, designers have been trying more and more to take into account individual differences between product users by employing technologies that allow customiza‐tion and individuation of products (Hancock, 2003; MacDonald, 2003). In this respect, it might be  interesting and useful to  investigate  individual differences  in sensory  in‐formation processing. According to dual coding theory, individuals can be classified as visualizers  versus  verbalizers. Visualizers  rely primarily on  visual  information when attempting to perform cognitive tasks, whereas verbalizers rely primarily on verbal‐analytical strategies. Here is how Temple Grandin describes visual thinking: 

“Thinking in language and words is alien to me. I think totally in pictures. The process is like trying to figure out what the picture on a jig saw puzzle is, when only some of the pieces are put together. A piece is put on one corner and then another corner and after about one fourth of the pieces are in place, a person can tell that the puzzle has a picture of a house on it. When I think about abstract concepts, such as relationships with people, I use visual images, such as a sliding glass door. Relationships must be ap‐proached gently because barging forward too quickly may shatter the door” (Grandin, 1995). 

These differences may determine  the attention  individuals pay  to visual  informa‐tion  in  comparison  to  information  from  other  sensory modalities.  Consumer  re‐search has demonstrated  that aesthetic  judgments differ among  individuals with different cognitive styles (Childers et al. 1985; Holbrook & Morris, 1986). For exam‐ple,  visualizers  respond more  sensitively  to organic patterns and  feature  interac‐tions, while  verbalizers pay more  attention  to  single  stimuli  and  isolated details. These differences may also affect product experience. In future research,  it would be  interesting  to  investigate  how  people with  different  cognitive  styles  process information from various sensory modalities, and how this information affects their overall product experience. The results of this research may help designers to cus‐tomize products and services for people with different cognitive styles. 

Page 180: ANNA FENKO - TU Delft

DISCUSSION 

 

180

4. IMPLICATIONS FOR DESIGNERS When designers create new products, they manipulate physical properties of these products (material, shape, color, smell, etc.)  in order to provide more pleasurable qualities for consumers.  

In our research we have demonstrated that although simple one‐to‐one correspon‐dences may  be  found  between  single modality  stimuli  and  affective  experiences (pleasure and annoyance), it is much more difficult to predict the affective reactions to combinations of stimuli in real products. Therefore, in design practice it is impor‐tant to bear in mind that the pleasantness of complex products may depend on other variables, such as the degree of congruence between the various sensory properties of products, and not on the pleasantness of the individual sensory property per se.  

Furthermore,  our  findings  suggest  that  in  order  to  create  a  long‐lasting  positive product experience, designers need to consider user‐product  interaction at differ‐ent stages of product usage. 

When trying to create a pleasurable sensory experience,  it  is also  important to take into account the symbolic meaning of products. As one of our studies demonstrated, positive sensory experiences can be ignored by consumers if these experiences con‐tradict  pre‐existing  negative  attitudes  towards  the  product.  This  suggests  that  for creating a  successful product  it might not be enough  to select  sensory pleasurable properties  and  to  integrate  them  into  a  congruent  combination.  The product  also needs to communicate symbolic meaning that fits consumers’ beliefs and values. 

Interrelations between sensory experiences and symbolic product properties need to be  investigated  further.  But  our  findings  suggest  that  sensory  product  properties might be linked to symbolic properties through the metaphorical meaning of sensory adjectives. Communicating product experience  through multiple sensory modalities may evoke different associations  in different cultures. For designers  involved  in de‐veloping products for various cultural groups,  it may be useful to consider the sym‐bolic associations of sensory product properties in different languages. 

Every year,  industries  spend huge amounts of  time and money on product evalua‐tion. For designers and product developers, it is vital to understand how products are experienced by users. This  thesis provides  insights  into multisensory  integration  in product experience. Nevertheless, our findings are limited to specific experiences and product  groups,  and more  research  is needed  to understand  the  relationships be‐tween the sensory, affective, and symbolic components of product experience. 

Page 181: ANNA FENKO - TU Delft

181 

 

 

  

 

 

 

 

 

 

 

SUMMARY  

People perceive the material world around them with their five senses. Information from different sensory modalities  is  integrated  in the brain to create a stable and meaningful experience of objects, including industrial products that accompany us in our everyday  life. Some of  the  sensory  systems play a more  important  role  in product experience than others. The aim of this research was to understand which sensory modalities are more  important for specific product experiences, and how product experience can be influenced by various combinations of sensory stimuli in products. We investigated the relative importance of various sensory modalities at different stages of product usage, for specific product experiences (freshness, natu‐ralness, warmth, and noisiness), and for different groups of users.  

In the first two studies described in Chapters 2 and 3 we investigated dynamic as‐pects  of  sensory  dominance.  In  the  first  study  the  questionnaire  approach was used. We asked participants to describe their experiences with products in various situations: while buying a product, after  the  first week,  the  first month, and  the first year of usage. The data demonstrate that the dominant sensory modality de‐pends on the period of product usage. At the moment of buying, vision is the most important modality,  but  during  usage  the  other  sensory modalities  gain  impor‐tance. The  roles of  the different modalities during usage are product‐dependent. Averaged over 93 products analyzed in this study, after one month of usage touch becomes more important than vision, and after one year vision, touch and audition appear to be equally important. We conclude that to create a long‐lasting positive 

Page 182: ANNA FENKO - TU Delft

SUMMARY 

 

182

product experience, designers need to consider user‐product  interaction at differ‐ent stages of product usage and  to determine which sensory modality dominates product experience at each stage. 

In the second study described  in Chapter 3 we used an experimental approach to investigate  how  the  sensory  properties  of  a  dehydrated  vegetable  food  product influence the experience of naturalness and emotional reactions at different stages of user‐product interaction. The results demonstrate that both emotional and sym‐bolic experience  (naturalness) change during product usage. At  the buying  stage, people pay most attention to visual properties of the food package. At the opening and cooking stages olfaction becomes as  important as vision. At  the eating stage taste  becomes  dominant.  The  importance  of  tactile  properties  increased  at  the eating  stage,  because  the  thickness  of  the  food  product  constitutes  one  of  the components of eating experience.  

The research demonstrates that different sensory experiences may have a different impact on the overall product assessment. At the buying stage, when vision  is the main  source  of  information  about  the  product,  people  rely mainly  on  their  pre‐existing attitudes and beliefs about the product. When they have more opportuni‐ties to use other senses (touch, smell, and taste), their assessment of the product relies more on  the  actual  sensory experience. But  the  influence of  these experi‐ences on conscious attitudes towards the product may be limited, because tactile, olfactory and gustatory experiences are more difficult to verbalize and to recognize consciously than visual experiences.  

Verbalization  is an  important component of product experience, because  it allows people to recall their experience and to communicate it to others. People describe their product experiences using adjectives that can be divided in three groups: sen‐sory descriptors  (e.g., hard, red, noisy); symbolic descriptors  (e.g.,  interesting, ex‐pensive, modern); and affective descriptors (e.g., pleasant, beautiful). In the study described  in  Chapter  4 we  developed  a  questionnaire  approach  to  quantify  the relative importance of the five sensory modalities for various descriptors of product experience. We  suggested  that modality  importance may  be  influenced  by  lan‐guage differences. To test this hypothesis we used two groups of participants, na‐tive speakers of Dutch and Russian. The results demonstrate that the importance of a sensory modality differs for various groups of descriptors. Most symbolic descrip‐tors demonstrate strong visual dominance. Affective descriptors are equally related to all sensory modalities. Sensory descriptors show significant cross‐cultural differ‐

Page 183: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

183

ences. The latter result can be explained by the different associations between lit‐eral and metaphorical meanings of sensory adjectives in different languages.  

The  validity of  the questionnaire methodology  in  sensory dominance  research  is limited, because when people report their experience they may not be aware of all the sensory  information  they  take  into account. Some of  the  information may be processed  subliminally, but  still  influence  the  assessment of  the product. That  is why in the three studies described in Chapters 5, 6, and 7 we used an experimental approach  to  investigate  sensory dominance  for  three product experiences:  fresh‐ness, warmth, and noisiness. The design of each study contains the following steps. In  a  pre‐study,  participants  rate  unisensory  visual,  tactual,  auditory,  or  olfactory stimuli on  a  specific  target experience dimension  (e.g., warmth,  freshness, noisi‐ness). Subsequently,  for each sensory dimension we select one stimulus pair  that differs  approximately  to  the  same  extent  on  this  target  dimension.  In  the main study,  these  stimuli  are  combined  in  all  possible  combinations,  yielding  a  set  of multisensory products, which are then evaluated by a new group of participants. 

Warmth  is an  important  characteristic  for  clothes, home  interior and  some  leisure related products. The experience of warmth has both a literal aspect, associated with the thermal characteristics of products, and a figurative aspect, associated with the metaphorical meaning of warmth  (i.e.,  intimacy, coziness). Chapter 5 describes  the experimental study aimed to determine the relative importance of material and color for the product experience of warmth. We designed products (scarves and breakfast tables) using warm and cold stimuli (colors and materials) in four different combina‐tions and asked respondents to evaluate the warmth and pleasantness of each prod‐uct. The results demonstrated that both color and material contribute equally to the judgments of warmth  in both products. We  also performed  a  follow‐up  interview study  to  clarify  the  literal and  figurative meanings people attach  to warmth  in  the context of product usage. We  found  that  the  literal meaning  is  related  to physical warmth and comfort, while the figurative meaning  is associated with social  interac‐tion, intimacy and friendly atmosphere. The figurative meaning was mentioned more often in association with products than the literal meaning. 

Freshness  is  important  for  food products, soft drinks, personal care products and cleaning products. Chapter 6 describes  the experimental  study on  sensory domi‐nance  in  the product experience of  freshness. We  created products  (soft drinks, dishwashing liquids, and scented candles) using fresh and non‐fresh stimuli (colors and smells)  in four different combinations and asked respondents to evaluate the freshness and pleasantness of each product. The results demonstrated that smell 

Page 184: ANNA FENKO - TU Delft

SUMMARY 

 

184

dominated  the  judgments  of  freshness  for  soft  drinks  and  dishwashing  liquids. However, for scented candles smell and color were equally important in determin‐ing  freshness. This  suggests  that  the dominant  sensory modality  for  the product experience of freshness depends on the characteristics of the particular product. 

Product noisiness is a multisensory product experience that integrates auditory and visual components.  In  the study described  in Chapter 7 we manipulated auditory and visual properties of alarm clocks and whistle kettles to find out to what extent the overall product noisiness depended on sounds these products made, or could be influenced by the visual appearance of the products. We created products com‐bining noisy and quiet stimuli of both sensory modalities according to a full factorial design  and  asked  participants  to  assess  how  noisy,  pleasant  and  annoying  they found  these products. The results demonstrated  that noisiness and annoyance of the products were generally dominated by the product sound. The noisiness of the sound had a negative influence on the overall pleasantness of the products. 

Sensory pleasure  is  an  important  component of product  experience.  In  all  three experiments  the  ratings of  the  target experiences  (fresh, warm, and noisy) were compared with  the  ratings of affective experiences  (pleasant and annoying). The correlations between noisiness and annoyance were  found both  for single stimuli and for the final products. However, the correlations between freshness and pleas‐antness and between warmth and pleasantness were found only for single stimuli. There were no such correlations for the final products. These findings suggest that the pleasantness of a product  could not always be predicted on  the basis of  the pleasantness attributed to its sensory components. Apparently, the combination of two pleasant stimuli does not guarantee the pleasantness of the final product. 

In the final chapter of the thesis we discuss our general findings, theoretical chal‐lenges emphasized in the research, perspectives for future research, and the impli‐cations of our findings for industrial design and product development. Based on our own results and the analysis of the  literature, we formulate two theoretical prob‐lems that need to be addressed in future research: (1) the problem of awareness in product  experience,  and  (2)  the  relationships  between  various  components  of product experience. We also suggest several topics for empirical  investigation,  in‐cluding  interactions  between  sensory,  affective,  and  symbolic  components  of product experience, sensory metaphors  in product experience, and  individual dif‐ferences in sensory information processing.  

Page 185: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

185

Designing pleasurable products can enrich user satisfaction and contribute  to  the well‐being of people and society. While designing products,  it  is  important for de‐signers to be aware of the complex relationships between various sensory product properties.  In  this  thesis we  have demonstrated  that  the  importance of  sensory modalities may depend on  the  stage of user‐product  interaction, on  the  specific experience that designers aim to induce, and on the language differences between users. Other  factors  (such  as  individual  differences  or  situational  variables) may also play a role in sensory dominance. Therefore, research on this topic should con‐tinue in the future. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Page 186: ANNA FENKO - TU Delft
Page 187: ANNA FENKO - TU Delft

187 

 

 

 

 

 

 

 

 

 

SAMENVATTING  

Mensen ervaren de materiële wereld om zich heen met hun vijf zintuigen. Informa‐tie van de verschillende sensorische modaliteiten wordt geïntegreerd  in de herse‐nen om een stabiele en betekenisvolle ervaring van objecten te creëren, met inbe‐grip van  industriële producten die ons omringen  in ons dagelijks  leven. Sommige zintuiglijke systemen spelen een belangrijkere rol in de productervaring dan ande‐ren. Het doel van dit onderzoek was om te begrijpen welke sensorische modalitei‐ten belangrijker zijn voor specifieke productervaringen, en hoe de productervaring kan worden beïnvloed door  verschillende  combinaties  van  zintuiglijke prikkels  in producten. We onderzochten het relatieve belang van de verschillende zintuiglijke modaliteiten  in  verschillende  stadia  van het productgebruik,  voor  specifieke pro‐ductervaringen (frisheid, natuurlijkheid, warmte en lawaaierigheid en voor verschil‐lende groepen gebruikers.  

In de eerste twee studies beschreven in hoofdstukken 2 en 3 hebben we de dyna‐mische  aspecten  van  de  zintuiglijke  dominantie  onderzocht.  In  de  eerste  studie vroegen we deelnemers aan de hand van vragenlijsten om hun ervaringen  te be‐schrijven met producten  in verschillende situaties: bij het kopen van een product, na de eerste week, de eerste maand, en het eerste  jaar van gebruik. De gegevens tonen aan dat de dominante sensorische modaliteit afhankelijk  is van de periode van gebruik van het product. Op het moment van aankoop is de visuele modaliteit het belangrijkst, maar  tijdens het gebruik van het product winnen de andere zin‐tuiglijke modaliteiten aan belang. De rollen van de verschillende modaliteiten zijn tijdens het gebruik productafhankelijk. Gemiddeld over 93 producten, vonden we 

Page 188: ANNA FENKO - TU Delft

SAMENVATTING 

 

188

dat na een maand tactiliteit belangrijker wordt dan visuele waarneming en dat na een  jaar zien, voelen en horen even belangrijk  lijken te zijn. We concluderen dat, om een  langdurige positieve productervaring te creëren, ontwerpers genoodzaakt zijn de mens‐product  interactie tijdens verschillende stadia van productgebruik te analyseren, om te bepalen welke sensorische modaliteit domineert in elke fase van het gebruik.  

In de  tweede  studie beschreven  in hoofdstuk 3 hebben we gebruik gemaakt van een experimentele benadering om te onderzoeken hoe de sensorische eigenschap‐pen van een voedingsmiddel  (een gedroogd groenteproduct) de ervaring van na‐tuurlijkheid en emotionele reacties beïnvloeden tijdens de verschillende fasen van de mens‐product  interactie. De resultaten tonen aan dat zowel de emotionele als de  symbolische  ervaring  (natuurlijkheid)  veranderen  tijdens  het  productgebruik. Tijdens het aankoopstadium besteedt men de meeste aandacht aan de visuele ei‐genschappen van de verpakking. Tijdens het openen en bereiden van het product wordt de geur even belangrijk als de visuele eigenschappen. Tijdens het eten wordt smaak dominant. Het belang van de tactiele eigenschappen neemt toe tijdens het eten: de dikte van het product is een onderdeel van de gehele eetervaring.  

Het onderzoek toont aan dat verschillende zintuiglijke ervaringen een verschillende uitwerking kunnen hebben op de totale productervaring. Tijdens de aankoopfase, wanneer het zicht de belangrijkste  informatie geeft over het product, vertrouwen mensen  vooral  op  hun  reeds  bestaande  opvattingen  en  overtuigingen  over  het product. Wanneer er mogelijkheden  zijn om andere  zintuigen  (tast,  reuk,  smaak) mee te nemen bij de beoordeling van een product, baseert men zich meer op de werkelijke zintuiglijke ervaring. Maar de invloed van deze ervaringen op de bewus‐te houding ten opzichte van het product zijn beperkt, omdat tactiele, olfactorische en smaak ervaringen moeilijker onder woorden  te brengen en bewust  te herken‐nen zijn dan visuele ervaringen.  

Verbaliseren  is een belangrijk onderdeel van de productervaring, omdat het men‐sen in staat stelt om hun ervaring te herinneren en te communiceren naar anderen. Mensen  beschrijven  hun  productervaringen met  bijvoeglijke  naamwoorden,  die kunnen  worden  onderverdeeld  in  drie  groepen:  zintuiglijke  descriptoren  (zoals: hard,  rood,  lawaaierig),  symbolische  descriptoren  (zoals:  interessant,  duur, mo‐dern) en affectieve descriptoren (zoals: aangenaam, mooi). In de studie beschreven in hoofdstuk 4 hebben we een vragenlijst ontwikkeld om het relatieve belang van de  vijf  zintuiglijke modaliteiten  te  kwantificeren  voor  verschillende  descriptoren tijdens  de  algehele  productervaring. We  suggereerden  al  dat  het  belang  van  de 

Page 189: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

189

verschillende modaliteiten kan worden beïnvloed door verschillen in taal. Om deze hypothese  te  testen hebben we gebruik gemaakt van  twee groepen deelnemers: mensen met als moedertaal Nederlands of Russisch. De resultaten  tonen aan dat het  belang  van  een  sensorische modaliteit  verschilt  voor  verschillende  groepen descriptoren.  De meeste  symbolische  descriptoren  vertonen  een  sterke  visuele dominantie. Affectieve descriptoren hebben in gelijke mate betrekking op alle zin‐tuiglijke  modaliteiten.  Zintuiglijke  descriptoren  laten  significante  crossculturele verschillen  zien. Dit  laatste  resultaat  kan worden  verklaard door  verschillende  in associaties bij de letterlijke en metaforische betekenissen van sensorische adjectie‐ven in verschillende talen.  

De  validiteit  van  de  vragenlijst‐methodologie  in  het  onderzoek  naar  zintuiglijke dominantie  is beperkt, omdat mensen zich mogelijk niet bewust zijn van alle zin‐tuiglijke informatie die van belang is, wanneer zij verslag van hun ervaringen doen. Sommige delen van de informatie kunnen subliminaal worden verwerkt, maar nog steeds van invloed zijn op de beoordeling van het product. Dat is de reden waarom we  in de drie studies beschreven  in de hoofdstukken 5, 6 en 7 een experimentele benadering hebben  gebruikt  om  de  zintuiglijke  dominantie  te  onderzoeken  voor drie soorten productervaringen: versheid, warmte en  lawaaierigheid. Het ontwerp van  elk  onderzoek  bevat  de  volgende  stappen.  In  een  voorstudie  rangschikken proefpersonen unisensorische  visuele,  tactiele,  auditieve of olfactorische prikkels voor  een  specifieke doel‐ervaring  (bijv. warmte,  frisheid,  lawaaierigheid). Vervol‐gens selecteerden we voor elke sensorische dimensie twee stimuli die ongeveer in dezelfde mate verschilden met betrekking tot deze doel‐dimensie. In het hoofdon‐derzoek zijn deze stimuli vervolgens samengevoegd  in alle mogelijke combinaties tot een reeks van multi‐sensorische producten, die vervolgens werden beoordeeld door een nieuwe groep proefpersonen.  

Warmte  is een belangrijk kenmerk voor kleding,  interieur en een aantal verwante vrijetijdsproducten. De ervaring van warmte heeft zowel een letterlijk aspect, gere‐lateerd aan de thermische eigenschappen van producten, als een figuratief aspect, gerelateerd  aan  de metaforische  betekenis  van warmte  (intimiteit,  gezelligheid). Hoofdstuk  5 beschrijft  een  experimentele  studie  gericht op het bepalen  van het relatieve  belang  van materiaal  en  kleur  in  de  productervaring  van warmte. We ontwierpen producten (sjaals en dienbladen) op basis van warme en koude stimuli (kleuren  en materialen)  in  vier  verschillende  combinaties  en  vroegen de  respon‐denten om de warmte en aangenaamheid van elk product te evalueren. De resulta‐ten toonden aan dat kleur en materiaal in gelijke mate bijdragen aan de beoorde‐

Page 190: ANNA FENKO - TU Delft

SAMENVATTING 

 

190

lingen van warmte in beide producten. Ook deden we een follow‐up interviewstu‐die, om de letterlijke en figuurlijke betekenissen te verhelderen, die mensen hech‐ten  aan warmte  in de  context  van productgebruik. We  vonden dat de  letterlijke betekenis is gerelateerd aan fysieke warmte en comfort, terwijl de figuurlijke bete‐kenis wordt geassocieerd met sociale interactie, intimiteit en een vriendelijke sfeer. De  figuurlijke betekenis werd vaker genoemd  in associatie met producten dan de letterlijke betekenis.   

Frisheid is belangrijk voor voedingsmiddelen, frisdranken, producten voor persoon‐lijke verzorging en schoonmaakartikelen. Hoofdstuk 6 beschrijft een experimentele studie naar  sensorische dominantie  voor de  frisheid  in  een productervaring. We creëerden producten  (frisdranken, afwasmiddel, en geurkaarsen) met behulp van frisse en niet‐frisse stimuli (kleuren en geuren) in vier verschillende combinaties en we vroegen respondenten om de frisheid en de aangenaamheid van elk product te beoordelen. De  resultaten  toonden  aan dat  geur domineerde bij de beoordeling van de frisheid van frisdranken en afwasmiddel. Bij de geurkaarsen waren geur en kleur even belangrijk bij het bepalen van de  frisheid. Dit suggereert dat de domi‐nante sensorische modaliteit voor de productervaring van frisheid afhankelijk is van de eigenschappen van het product.  

De lawaaierigheid van een product integreert auditieve en visuele componenten in een multi‐zintuiglijke ervaring.  In de studie beschreven  in hoofdstuk 7 hebben we de auditieve en visuele eigenschappen van wekkers en  fluitketels gemanipuleerd, om te bepalen in welke mate de lawaaierigheid van producten bepaald werd door het geluid dat producten maakten, of door het uiterlijk van deze producten. We creëerden producten met lawaaiige en rustige stimuli van beide zintuiglijke modali‐teiten  volgens  een  volledige  factoriëel  design  en  vroegen  proefpersonen  om  te beoordelen hoe  lawaaierig, aangenaam en  irritant zij deze producten vonden. De resultaten toonden aan dat lawaaierigheid en irritantheid van de producten in het algemeen werden gedomineerd door het geluid van de producten. De  lawaaierig‐heid van het geluid had een negatieve  invloed op de aangenaamheid van de pro‐ducten. Zintuiglijk genot is een belangrijk onderdeel van de productervaring. In alle drie de experimenten zijn de beoordelingen voor de doel‐ervaringen (fris, warm en lawaai‐erig)  vergeleken met  de  oordelen  van  affectieve  ervaringen  (aangenaam  en  irri‐tant). Tussen  lawaaierigheid en  irritatie werden  correlaties gevonden  voor  zowel enkelvoudige stimuli als voor de eindproducten. Maar de correlaties tussen frisheid en aangenaamheid en  tussen warmte en aangenaamheid vonden we alleen voor 

Page 191: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

191

enkelvoudige stimuli. Deze waren afwezig bij de eindproducten. Deze bevindingen suggereren dat de aangenaamheid  van een product niet altijd kan worden  voor‐speld op basis van de aangenaamheid toegeschreven aan zijn sensorische compo‐nenten. Blijkbaar leidt combinatie van twee aangename prikkels niet altijd tot een aangenaam eindproduct. 

In het  laatste hoofdstuk van het proefschrift bespreken we onze algemene bevin‐dingen, theoretische uitdagingen benadrukt  in het onderzoek, perspectieven voor toekomstig onderzoek en de implicaties van onze bevindingen voor industrieel de‐sign  en productontwikkeling. Gebaseerd op onze  eigen  resultaten  en de  analyse van de literatuur, formuleren we twee theoretische problemen die dienen te wor‐den meegenomen in toekomstige onderzoeken: (1) het probleem van het bewust‐zijn in productervaring en (2) de relaties tussen de diverse onderdelen van de pro‐ductervaring. We suggereren ook een aantal onderwerpen voor empirisch onder‐zoek, waaronder interacties tussen zintuiglijke, affectieve en symbolische onderde‐len van de productervaring, zintuiglijke metaforen in productervaring, en individue‐le verschillen in zintuiglijke informatieverwerking. 

Het ontwerpen van aangename producten kan de  tevredenheid van de gebruiker verhogen en kan bijdragen aan een hoger welzijn van mensen en samenleving. Tij‐dens het ontwerpen van producten  is het belangrijk voor ontwerpers om zich be‐wust te zijn van de complexe relaties tussen de verschillende sensorische product‐eigenschappen.  In dit proefschrift hebben we aangetoond dat het belang van de sensorische modaliteiten kan afhangen van de fase van de mens‐product interactie, de specifieke ervaring die de ontwerpers trachten op te roepen, en van de taalver‐schillen tussen gebruikers. Andere factoren, zoals individuele verschillen of situati‐onele variabelen, kunnen ook een rol spelen  in zintuiglijke dominantie. Daarom  is het noodzakelijk dat onderzoek over dit onderwerp ook in de toekomst plaats blijft vinden.  

Page 192: ANNA FENKO - TU Delft
Page 193: ANNA FENKO - TU Delft

193 

 

 

 

 

 

 

 

 

 

 

REFERENCES  Adank, R., & Warell, A. (2008). "Five senses testing" ‐ Assessing and predicting sensory experi‐

ence of product design. In P. M. A. Desmet, J. van Erp & M. Karlsson (Eds.), Design & emotion moves. Newcastle upon Tyne, UK: Cambridge Scholars Publishing, pp. 35‐58. 

Allison, R.I., & Uhl, K.P.  (1964).  Influence of beer brand  identification on  taste perception. Journal of  Marketing Research,  1, 36–39. 

Amlot, R., Walker, R., Driver, J., & Spence, C. (2003). Multimodal visualsomatosensory  inte‐gration in saccade generation. Neuropsychologia, 41, 1–15. 

Anderson, N.H.  (1981).  Foundations of  information  integration  theory. New  York: Academic Press.  

Arnheim, R.  (1959). Art and  visual perception: A psychology of  the  creative eye. Berkeley: University of California Press. 

Asch, S.E. (1946). Forming impressions of personality. Journal of Abnormal Social Psychology, 42, 258–290. 

Asch, S.E. (1955). On the use of metaphor in the description of persons. In: H. Werner (Ed.). On Expressive Language. Worcester: Clark University Press, p. 86‐94. 

Asch, S.E.  (1958). The metaphor: A psychological  inquiry.  In: R. Tagiuri & L. Petrullo  (Eds.). Person  Perception  and  Interpersonal  Behavior.  Stanford,  CA:  Stanford University Press, p. 86‐94. 

Ashby, M., & Johnson, K. (2002) Materials and design: the art and science of material selec‐tion in product design. Oxford, UK: Butterworth – Heinemann.  

Astrup, A. (1998). The American paradox: The Role of Energy‐Dense Fat‐Reduced Food in the Increasing Prevalence of Obesity. Current Opinion  in Clinical Nutrition and Meta‐bolic Care, 1, 573–577. 

Page 194: ANNA FENKO - TU Delft

REFERENCES 

 

194

Bargh, J.A., & Morsella, E. (2008). The unconscious mind. Perspectives of  Psychological Science, 3, 73–79. 

Baron, R. A. (1981). The role of olfaction in human social behavior: Effects of a pleasant scent on  attraction  and  social  perception.  Personality  and  Social  Psychology  Bulletin,  7, 611‐617. 

Baron‐Cohen, S. (1996). Is there a normal phase of synaesthesia in development? Psyche, 2, 27.  Barrett,  L.F., Mesquita, B., Ochsner, K.N., & Gross,  J.J.  (2007). The experience of emotion. 

Annual Review of Psychology, 58, 373–403. Bartoshuk, L.M., & Beauchamp, G.K.  (1994). Chemical Senses. Annual Review of   Psychology, 

45, 419–449. Berglund, B., Berglund, U., and Lindvall, T. (1976). Scaling  loudness, noisiness, and annoyance 

of community noises. Journal of the Acoustical Society of America, 60 (5), 1119–1125. Berglund, B., Berglund, U., Engen, T., & Ekman, G.  (1973). Multidimensional analysis of 21 

odors. Scandinavian Journal of Psychology, 14, 131–137. Bergmann Tiest, W.M., & Kappers, A.M.L. (2008) Thermosensory reversal effect quantified. 

Acta Psychologica, 127, 46–50. Berlyne, D. E. (1970). Novelty, complexity and hedonic value. Perception and Psychophysics, 

8, 279–286. Berlyne, D.E.  (1963). Complexity  and  incongruity  variables  as determinants of exploratory 

choice and evaluative ratings. Canadian Journal of Psychology, 17, 274‐290. Berlyne, D.E. (1971). Aesthetics and psychobiology. New York: Appleton‐Century‐Crofts. Berridge, K. & Winkielman, P. (2003). What is an unconscious emotion? (The case for uncon‐

scious “liking”). Cognition & Emotion, 17 (2), 181‐211. Bertelson, P. (1999). Ventriloquism: A case of cross‐modal perceptual grouping. In: G. Aschersle‐

ben, T. Bachmann & J. Müsseler (Eds.). Cognitive contributions to the perception of spa‐tial and temporal events. Amsterdam: Elsevier, pp. 347‐362. 

Bertelson, P., de Gelder, B., and Driver, J.  (2000). The ventriloquist effect does not depend on  the direction of deliberate visual attention, Perception and Psychophysics, 62 (2), 321–332. 

Birch, L.L. (1999). Development of food preferences. Annual Review of Nutrition, 19, 41–62. Bloch, P.H.  (1995). Seeking the  ideal form: Product design and consumer response. Journal 

of Marketing, 59, 16–29. Block, N. (1983). Mental Pictures and Cognitive Science. Philosophical Review, 93, 499‐542.  Bone, P. F., & Jantrania, S. (1992). Olfaction as a cue for product quality. Marketing Letters, 

3, 289‐296. Boroditsky,  L.  (2000). Metaphoric  structuring:  Understanding  time  through  spatial meta‐

phors. Cognition 75, 1–28. Boroditsky,  L.  (2001). Does  language  shape  thought? Mandarin and English  speakers'  con‐

ceptions of time. Cognitive Psychology, 43, 1‐22. 

Page 195: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

195

Boroditsky, L., Schmidt, L., & Webb, P. (2003). Sex, syntax, and semantics. In: Gentner, D., & Goldin‐Meadow, S. (Eds.), Language  in Mind: Advances  in the Study of Language and Cognition. MIT Press, Cambridge, MA, pp. 61‐80. 

Botvinick, M., & Cohen, J. (1998). Rubber hands "feel" touch that eyes see. Nature, 391, 756. Bowerman, M.  (1996). The origins of  children's  spatial  semantic  categories:  cognitive versus 

linguistic determinants.  In: Gumperz,  J., &  Levinson, S.  (Eds.), Rethinking  Linguistic Relativity. Cambridge University Press, New York, pp. 145‐176. 

Bowlby, J. (1969). Attachment and Loss. Hogarth Press, London. Bowsher, J.M. and Robinson, D.W. (1962). On scaling the unpleasantness of sounds. British 

Journal of Applied Physics, 13, 179‐181. Brannen, J. (1992). Combining qualitative and quantitative approaches: an overview. In: Mix‐

ing methods: qualitative and quantitative research. Avebury, Aldershot, pp. 3‐37. Brislin, R. (1980). Translation and content analysis of oral and written material. In: Triandis, 

H. & Berry, J. (Eds.) Handbook of Cross‐Cultural Psychology, Vol.2, Allyn and Bacon, Boston, MA, pp. 389‐444. 

Bruehl, S., Chung, O.Y., & Burns, J.W. (2006). Anger expression and pain: an overview of find‐ings and possible mechanisms. Journal of  Behavioral Medicine, 29, 593–606. 

Buck,  C.D.  (1949). A Dictionary  of  Selected  Synonyms  in  the  Principal  Indo‐European  Lan‐guages. Chicago University Press, Chicago. 

Byrne, R., & Whiten, A.  (Eds.)  (1988). Machiavellian  Intelligence:  Social  Expertise  and  the Evolution of Intellect in Monkeys, Apes, and Humans. Oxford University Press, New York. 

Cacioppo, J. T., Klein, D. J., Berntson, G. G., & Hatfield, E. (1993). The Psychophysiology of Emo‐tion. In M. Lewis & J. M. Haviland (Eds.) Handbook of Emotions. New York: The Guil‐ford Press, pp. 119‐142. 

Caclin, A., Soto‐Faraco, S., Kingstone, A., & Spence, C. (2002). Tactile "capture" of audition. Perception & Psychophysics, 64, 616‐630. 

Cain, W.S. (1979). To know with the nose: keys to odor identification. Science, 203, 467–470. Calvert, G.A., & Thesen, T. (2004). Multisensory integration: methodological approaches and 

emerging principles in the human brain. Journal of Physiology, 98, 191–205. Calvert,  G.A.,  Spence,  C.,  &  Stein,  B.E.  (2004).  The  handbook  of multisensory  processes.  

Cambridge, MA: MIT Press. Caplan, D. (1973). A note on the abstract readings of verbs of perception. Cognition, 2, 269–

277.  Carpenter, G.S., & Nakamoto, K. (1996). Impact of consumer preference formation on mar‐

keting objectives and  second mover  strategies.  Journal of Consumer Psychology, 5(4), 325–358. 

Charles, N., &  Kerr, M.  (1988). Women,  Food  and  Families. Manchester University  Press, Manchester/New York. 

Chen, J.‐Y. (2007). Do Chinese and English speakers think about time differently? Failure of replicating Boroditsky (2001). Cognition 104, 427–436. 

Page 196: ANNA FENKO - TU Delft

REFERENCES 

 

196

Childers, T. L., Houston, M.J, & Heckler, S.E. (1985). Measurement of individual differences in visual  versus  verbal  information  processing.  Journal  of  Consumer  Research,  12, 125‐134. 

Chu, S., & Downes, J.J. (2000). Odour‐evoked autobiographical memories: Psychological inves‐tigations of proustian phenomena. Chemical Senses 25, 111–116. 

Claeys, C., Swinnen, A., & Vanden A.P. (1995). Consumers' means‐end chains for "think" and "feel" products. International Journal of Research in Marketing, 12, 193‐208. 

Clark, A.  (1997). Being  There:  Putting Brain, Body,  and World  Together Again. MIT  Press, Cambridge, MA. 

Classen, C. (1993). Worlds of Sense. Exploring the Senses in History and Across Cultures. Lon‐don, Routledge.  

Classen, C. (2005). McLuhan in the rainforest. In: Howes, D. (Ed.), Empire of the Senses. Berg, Oxford, pp. 147‐163. 

Clos, F., & Carole Bouchard, C. (2010). Building relevant contexts to design experiences. Pa‐per presented at the International Conference on Kansei Engineering and Emotion Research, Paris, March 2‐4 2010, pp. 1161‐1171. 

Clydesdale, F.M. (1993). Color as a factor in food choice. Critical Reviews in Food Science and Nutrition, 33, 83‐101. 

Clydesdale, F.M., Gover, R., Philipsen, D.H., & Fugardi, C. (1992). The effect of color on thirst quenching,  sweetness,  acceptability  and  flavor  intensity  in  fruit  punch  flavored beverages. Journal of Food Quality, 15, 19–38. 

Coates, D.  (2003). Watches tell more than time: Product design,  information and the quest for elegance. London: McGraw‐Hill. 

Colonius, H., & Arndt, P.  (2001). A  two‐stage model  for visual–auditory  interaction  in  sac‐cadic latencies. Perception & Psychophysics, 63, 126–147. 

Crilly, N., Moultrie, J., & Clarkson, P.J. (2004). Seeing things: Consumer response to the visual domain in product design. Design Studies, 25, 547–577. 

Damasio, A.R.  (1999).  The  Feeling  of What Happens: Body and  Emotion  in  the Making  of Consciousness. New York: Harcourt Brace. 

Davidoff, J., Davies, I., & Robertson, D. (1999). Colour categories of a stone‐age tribe. Nature 398, 203‐204. 

Davis, R. G. (1981). The role of nonolfactory context cues in odor identification. Perception & Psychophysics, 30, 83‐89. 

de Gelder, B., & Vroomen, J. (2000). Perceiving emotions by ear and by eye. Cognition and Emotion, 14, 289‐311.  

de Wijk, R.A., & Cain, W.S.  (1994). Odor  identification by name and by edibility:  life‐span development and safety. Human Factors, 36, 182–187. 

de Wijk, R.A., Schab, F.R., & Cain, W.S. (1995). Odor identification. In Schab, F.R. (Ed.) Mem‐ory for Odors. Mahwah, NJ: Erlbaum, pp. 21–37. 

Dehaene, S. & Naccache, L. (2001). Towards a cognitive neuroscience of consciousness: basic evidence and a workspace framework. Cognition, 79, 1–37. 

Page 197: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

197

Demattè, M.  L.,  Sanabria, D.,  Sugarman, R., &  Spence, C.  (2006). Crossmodal  interactions between olfaction and touch. Chemical Senses, 31, 291‐300. 

Desmet, P. M. A. (2003). Measuring emotion: development and application of an instrument to measure emotional responses to products. In M. A. Blythe, A. F. Monk, K. Over‐beeke & P. C. Wright  (Eds.), Funology: from usability to enjoyment (pp. 111‐123). Dordrecht, the Netherlands: Kluwer. 

Desmet, P.M.A. (2002). Designing Emotions. Phd Thesis, Delft University of Technology. Desmet,  P.M.A., &  Schifferstein,  H.N.J.  (2008)  Positive  and  negative  emotions  associated 

with food experience. Appetite, 50, 290‐301. Desmet, P.M.A., & Schifferstein, H.N.J. (Eds) (2010). From floating wheelchairs to mobile car 

parks: A collection of 35 experience‐driven design projects. Lemma, Den Haag.  DeVault, M.  L.  (1991).  Feeding  the  Family. The  Social Organization of Caring as Gendered 

Work. University of Chicago Press, Chicago. Dilthey, W. (1985). Poetry and experience. Princeton, NJ: Princeton University Press. Dittmar, H. (1992). The social psychology of material possessions: To have is to be. New York: 

St. Martin’s Press. Dolan, R. J., Morris, J. S., & de Gelder, B. (2001). Crossmodal binding of fear in voice and face. 

Proceedings of the National Academy of Sciences of the United States of America, 98, 10006‐10010. 

Easton, R., & Moran, P.W. (1978). A quantitative confirmation of visual capture of curvature. Journal of General Psychology, 98, 105‐112. 

Edelman, G.M., & Tononi, G. (2000). A Universe of Consciousness. New York: Basic Books. Eertmans, A., Baeyens, F., & van Den Bergh, O. (2001). Food likes and their relatives impor‐

tance  in human eating behaviour: Review and preliminary suggestions  for health promotion, Health Education Research 16, 443–456.  

Eisenman, R. (1967). Complexity–simplicity: Preference for symmetry and rejection of com‐plexity. Psychonomic Science, 8, 169‐170. 

Ekman, P., & Friesen, W.V.  (1978). Facial Action Coding System: A  technique  for  the meas‐urement of facial movement. Palo Alto, CA: Consulting Psychologists Press. 

Engen, T. (1982). The Perception of Odors. New York: Academic Press. Epple, G., & Herz, R.S.  (1999). Ambient odors associated  to  failure  influence cognitive per‐

formance in children. Developmental Psychobiology, 35, 103–107. Evans, N. & Wilkins, D.  (2000).  In  the mind’s  ear:  The  semantic  extensions  of perception 

verbs in Australian languages. Language 76, 546‐592.  Eysenck, H. J., and Hawker, G. W. (1994). The taxonomy of visual aesthetic preferences: An 

empirical study. Empirical Studies of the Arts, 12, 95‐101. Falk, P. (1994). The Consuming Body. London: Sage. Feldman  Barrett,  L., &  Russell,  J.  (1998).  Independence  and  bipolarity  in  the  structure  of 

current affect. Journal of Personality and Social Psychology 74, 967‐984. Feldman,  J.  (1997).  Regularity‐based  perceptual  grouping.  Computational  Intelligence,  13(4), 

582‐623. 

Page 198: ANNA FENKO - TU Delft

REFERENCES 

 

198

Fenko, A, Schifferstein, H.N.J., & Hekkert, P.  (2010). Shifts  in  sensory dominance between various stages of user‐product interactions. Applied Ergonomics, 41, 34–40. 

Fenko, A. (2008). Sensory dominance  in product experience: The paradox of theory and prac‐tice. Paper presented at the 7th Nordcode Seminar, Lund, Sweden, 28‐30 May 2008. 

Fenko, A., Otten, J.J., & Schifferstein, H.N.J. (2010). Describing product experience in different languages: The role of sensory modalities. Journal of Pragmatics, 42, 3314‐3327. 

Fenko, A., Schifferstein, H.N.J., & Hekkert, P.  (2009). Shifts  in sensory dominance between various stages of user‐product interactions. Applied Ergonomics, 41, 34‐40. 

Fenko, A., Schifferstein, H.N.J., & Hekkert, P. (2008). Which senses dominate  in product ex‐perience. Paper presented  at  the Design Research  Society Conference,  Sheffield, UK, 16‐19 July. 

Fenko, A., Schifferstein, H.N.J., & Hekkert, P. (2010). Looking hot or feeling hot: What deter‐mines the product experience of warmth? Materials and Design, 31, 1325–1331. 

Fenko, A., Schifferstein, H.N.J., Huang, T.‐C., & Hekkert, P. (2009). What makes products fresh: The smell or the colour? Food Quality & Preference, 20, 372–379.  

Fiore, A. M.  (1992). Effect of  composition of olfactory  cues on  impressions of personality. Social Behavior and Personality, 20, 149‐161. 

Fiore, A. M., & Kimle, P. A. (1997). Understanding aesthetics for the merchandising and de‐sign professional. New York: Fairchild. 

Fiske, S.T., Cuddy, A.J.C., & Glick, P. (2006). Universal dimensions of social cognition: warmth and competence. Trends in Cognitive Science, 11, 77‐83. 

Fodor, J.A. (1975). The language of thought. New York: Harvard University Press. Fontaine, J.R., Scherer, K.R., Roesch, E.B., Ellsworth, P.C. (2007). The world of emotions is not 

two‐dimensional. Psychological Science, 18, 1050–1057. Forster, B., Cavina‐Pratesi, C., Aglioti, S.M., & Berlucchi, G.  (2002). Redundant target effect 

and  intersensory  facilitation  from  visual–tactile  interactions  in  simple  reaction time. Experimental Brain Research, 143, 480–487. 

Foucault, M. (1972). The Archaeology of Knowledge, trans. A. M. Sheridan Smith. Pantheon, New York. 

Frassinetti, F. Pavani, F., & Ladavas, E. (2002). Acoustical vision of neglected stimuli: interac‐tion among spatially converging audiovisual  inputs  in neglect patients,  Journal of Cognitive Neuroscience, 14 (1), 62–69. 

French, S. J., Read, N. W., Booth, D. A. & Arkley, S. (1993). Satisfaction of hunger and thirst by foods and drinks. British Food Journal, 95, 19‐26. 

Frens, M.A., van Opstal, A.J., & van der Willigen, R.F.  (1995). Spatial and  temporal  factors determine auditory–visual interactions in human saccadic eye movements. Percep‐tion & Psychophysics, 57, 802–816. 

Frijda, N.H. (2006). The Laws of Emotion. Mahwah, N.J.: Erlbaum. Frith, C. D., & Nias, D. K. B. (1974). What determines aesthetic preferences? Journal of Gen‐

eral Psychology, 91, 163–173. 

Page 199: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

199

Fujii, K., Soeta, Y., and Ando, Y. (2001). Acoustical properties of aircraft noise measured by tempo‐ral and spatial factors. Journal of Sound and Vibration, 241, 69‐78. 

Gadamer, H.‐G. (1975). Truth and method. New York: Seabury Press. Gallup, G.  Jr., & Cameron, P.  (1992). Modality  specific metaphors:  Is our mental machinery 

"colored" by a visual bias? Metaphor and Symbolic Activity 7, 93‐98. Gatignon, H.,  Eliashberg,  J., & Robertson,  T.  (1989). Modeling multinational diffusion pat‐

terns: an efficient methodology. Marketing Science 8, 231‐47.  Gelman, R., & Gallistel, C.  (2004).  Language and  the origin of numerical  concepts. Science 

306, 441–443. Gentner, D., &  Imai, M.  (1997). A  cross‐linguistic  study  of  early word meaning: Universal 

ontology and linguistic influence. Cognition, 62, 169‐200. Gentner, D., & Goldin‐Meadow, S. (Eds.), 2003. Language in Mind: Advances in the Study of 

Language and Cognition. MIT Press, Cambridge, MA. Gentner, D., Bowdle, B., Wolff, P., & Boronat, C. (2001). Metaphor is like analogy. In Dedre

G., K.J. Holyoak, & B.N. Kokinov (Eds.), The Analogical Mind: Perspectives from Cognitive Science. Cambridge, MA: MIT Press, pp. 199–253. 

Giard, M.H., &  Peronnet,  F.  (1999).  Auditory–visual  integration  during multimodal  object recognition  in  humans:  a  behavioral  and  electrophysiological  study.  Journal  of Cognitive Neuroscience, 11, 473–490.  

Gibbs, R.W., Jr. (1994). The Poetics of Mind: Figurative Thought, Language, and Understand‐ing. Cambridge and NewYork: Cambridge University Press. 

Gilbert, A., Regier, T., Kay, P., &  Ivry, R.  (2006). Whorf hypothesis  is supported  in the right visual field but not the left. Proceedings of the National Academy of Sciences USA 103, 489–494. 

Glenberg, A. (1997). What memory is for. Behavioural Brain Science 20, 1–55. Goodale, M.A., & Humphrey, G.K.  (1998). The objects of action and perception. Cognition, 

67, 181‐207. Gordon, P.  (2004). Numerical  cognition without words:  Evidence  from Amazonia.  Science, 

306, 496. Govers,  P.  (2004).  Product  Personality.  Unpublished  PhD  dissertation,  Delft  University  of 

Technology, Delft, the Netherlands. Grandin, T. (1995). Thinking in pictures. New York: Doubleday. Graybiel, A. (1952). Oculogravic illusion. AMA Archives of Ophthalmology, 48, 605–615. Griffiths,  I. D. &  Langdon, F.J.  (1968). Subjective  response  to  road  traffic noise.  Journal of 

Sound and Vibration, 8, 16‐32. Guest, S., & Spence, C. (2003). Tactile dominance in speeded discrimination of textures. Ex‐

perimental Brain Research, 150, 201–7. Gumperz, J., & Levinson, S. (Eds.) (1996). Rethinking Linguistic Relativity. Cambridge Univer‐

sity Press, New York. Gyoba,  J., Suzuki, M., Kawabata, Y., & Hiroshi, K.  (2005). Analyses of  sensory‐relevance of 

adjective  pairs  frequently  used  in  the  semantic  differential method.  Paper  pre‐

Page 200: ANNA FENKO - TU Delft

REFERENCES 

 

200

sented at the 6th Annual Meeting of the  International Multisensory Research Fo‐rum, Rovereto, Italy, 5‐8 June 2005. 

Hamstra, I.A. (1998). Public Opinion about Biotechnology: A Survey of Surveys. Delft: Euro‐pean Federation of Biotechnology, Task Group on Public Perceptions of Biotech‐nology, ISBN 90‐76110, http://www.kluyver.stm.tudelft.nl/efb/tgppb/pdf/survey.pdf. 

Han, S. H., & Hong, S. W. (2003). A systematic approach for coupling user satisfaction with product design. Ergonomics, 46, 1441‐1461. 

Hancock, P.A., Pepe, A.A., & Murphy,  L.L.  (2005). Hedonomics: The Power of Positive and Pleasurable Ergoriornics. Ergonomics in Design, Winter, 8‐14. 

Harlow, H.F. (1958). The nature of love. American Psychologist, 13, 673‐685. Harrington,  L.K., & Peck, C.K.  (1998). Spatial disparity affects visualauditory  interactions  in 

human sensorimotor processing. Experimental Brain Research, 122, 247–252. Harte, J.M., & Koele, P. (1995). A comparison of different methods for the elicitation of attrib‐

ute weights:  Structural modeling,  process  tracing,  and  self‐reports. Organizational Behavior and Human Decision Processes, 64, 75‐117. 

Hay,  J.C., Pick, H.L., &  Ikeda, K.  (1965). Visual  capture produced by prism  spectacles. Psy‐chonomic Science (2), 215‐216. 

Heaps, C., and Handel, C.H. (1999). Similarity and features of natural textures. Journal of Experi‐mental Psychology: Human Perception and Performance, 25, 299‐320. 

Heenan, S.P., Dufour,  J.‐P., Hamid, N., Harvey, W., & Delahunty, C.M.  (2008). The  sensory quality of fresh bread: Descriptive attributes and consumer perceptions. Food Re‐search International, 4, 989–997. 

Heidegger, M. (1962). Being and Time. Translated by John Macquarrie and Edward Robinson. New York: Harper. 

Heidegger, M. (1971). On the way to language. New York: Harper & Row. Heider, E. (1972) Universals  in color naming and memory. Journal of Experimental Psychol‐

ogy 93, 10‐20. Heini, A. F. and R. L. Weinsier, (1997). Divergent Trends  in Obesity and Fat Intake Patterns: 

The American paradox. American Journal of Medicine, 102(3), 259–264. Hekkert, P., & Schifferstein, H.N.J. (2008).  Introducing product experience.  In: Schifferstein, 

H.N.J., & Hekkert, P. (Eds.). Product experience. Amsterdam: Elsevier, pp.1‐8. Helander, M. G., & Tham,M. P.  (2003) Hedonomics: Affective human  factors design. Ergo‐

nomics, 46, 1269‐1 272. Hellman,  R.P.  (1982).  Loudness,  annoyance,  and  noisiness  produced  by  single‐tone–noise 

complexes. Journal of the Acoustical Society of America; 72, 62–73. Hellman, R.P.  (1984). Growth  rate of  loudness, annoyance, and noisiness as a  function of 

tone location within the noise spectrum. Journal of the Acoustical Society of Amer‐ica, 75(1): 209–218. 

Hellman,  R.P.  (1985).  Perceived magnitude  of  two‐tone–noise  complexes:  loudness,  annoy‐ance, and nosiness. Journal of the Acoustical Society of America; 77 (4): 1497–1504. 

Page 201: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

201

Helmholtz, H. (1878). The Facts of Perception from Selected Writings of Hermann Helmholtz: Wesleyan University Press. 

Hershenson M.  (1962). Reaction  time as a measure of  intersensory  facilitation.  Journal of Experimental Psychology, 63, 289–293. 

Hertenstein, M.J., Verkamp, J.M., Kerestes, A.M., & Holmes, R.M. (2006). The communicative functions of touch in humans, nonhuman primates, and rats: A review and synthesis of the empirical research. Genetic, Social, and General Psychology Monographs, 132, 5–94. 

Herz, R. S. (1998a). Are odors the best cues to memory? A cross‐modal comparison of associa‐tive memory stimuli. Annals of the New York Academy of Sciences, 855, 670‐674. 

Herz, R. S. (1998b). An examination of objective and subjective measures of experience asso‐ciated to odors, music, and paintings. Empirical Studies of the Arts, 16, 137‐152. 

Herz,  R.  S.,  &  Schooler,  J. W.  (2002).  A  naturalistic  study  of  autobiographical memories evoked by olfactory and visual  cues: Testing  the Proustian hypothesis. American Journal of Psychology, 115, 21‐32. 

Herz, R.S., & von Clef, J. (2001). The influence of verbal labeling on the perception of odors: Evidence for olfactory illusions? Perception, 30, 381‐391. 

Hinton, P.B., & Henley, T.B. (1993). Cognitive and affective components of stimuli presented in three modes. Bulletin of the Psychonomic Society, 31, 595‐598. 

Hiramatsu, K., Takagi, K., and Yamamoto, T. (1983). Experimental investigation on the effect of some temporal factors of nonsteady noise on annoyance, Journal of the Acousti‐cal Society of America, 74, 1782‐1793. 

Hirschman, E. (1984). Experience seeking: A subjectivistic perspective for consumption. Jour‐nal of Business Research 12, 115‐136. 

Hirschman, E.C., & Holbrook, M.B. (1982). Hedonic consumption: Emerging concepts, meth‐ods and propositions. Journal of Marketing, 46, 92‐101. 

Hofstede, G. (2001). Culture's Consequences: Comparing Values, Behaviours,  Institutions, and Organizations across Nations, 2nd ed. Sage, Thousand Oaks, CA.  

Holbrook, M. B. (1986). Aims, Concepts, and Methods for the Representation of Meaningful Heterogeneity  in Evaluative  Judgments of Esthetic Objects With Known Features. The Journal of Consumer Research, 13, 337‐347. 

Hollins, M. (2010). Somesthetic Senses. Annual Review of Psychology, 61, 243–271. Hollins, M., BensmaЁıa, S.J., Karlof, K., & Young, F. (2000).  Individual differences  in percep‐

tual space for tactile textures: evidence from multidimensional scaling. Perception & Psychophysics, 62, 1534–1544. 

Holm, L. (2003). Food health policies and ethics: Lay perspectives on functional foods. Jour‐nal of Agricultural and Environmental Ethics, 16, 531–544. 

Holm, L. and Kildevang, H.  (1996). Consumers’ Views on Food Quality. A Qualitative  Inter‐view Study. Appetite 27, 1–14. 

Holmes,  C.B., &  Buchanan,  J.A.  (1984).  Color  preference  as  a  function  of  the  object  de‐scribed. Bulletin of the Psychonomic Society, 22, 423‐425. 

Page 202: ANNA FENKO - TU Delft

REFERENCES 

 

202

Hong, Y.‐Y., Morris, M., Chiu, C.‐Y., &   Benet‐Martínez, V.  (2000). Multicultural minds: A dy‐namic  constructivist  approach  to  culture  and  cognition. American Psychologist 55, 709–720. 

Horigushi, A. et al. (1995). A Kansei approach to a driver / vehicle system. International Jour‐nal of Industrial Ergonomics, 5, 25‐37. 

Hughes, H.C., Reuter‐Lorenz, P.A., Nozawa, G., & Fendrich, R. (1994). Visual‐auditory interac‐tions  in  sensorimotor  processing:  saccades  versus manual  responses.  Journal  of Experimental Psychology: Human Perception and Performance, 20, 131–153. 

Hunt, E., & Agnoli, F. (1991). The Whorfian hypothesis: a cognitive psychology perspective. Psychological Review 98, 377–389. 

Husserl, E. (1964). Cartesian Meditations. Translated by Dorion Cairns. The Hague: Nijhoff. Ibarretxe‐Antuñano, I. (1999). Metaphorical mappings in the sense of smell. In: Gibbs, R., &

Steen, G.  (Eds.) Metaphor  in Cognitive Linguistics,  John Benjamins, Amsterdam, pp. 29–45. 

Ibarretxe‐Antuñano,  I.  (2008).  Vision  metaphors  for  the  intellect:  Are  they  really  cross‐linguistic? Atlantis. Journal of the Spanish Association of Anglo‐American Studies 30, 15–33. 

Iordanskaja,  L.  (1979). The  semantics of  three Russian  verbs of perception.  Linguistics 17, 825‐842. 

Itten, J. (1961). Art of Color. New York: Van Nostrand Reinhold. Izard,  C.E.  (2009).  Emotion  theory  and  research:  Highlights,  unanswered  questions,  and 

emerging issues. Annual Review of Psychology, 60, 1–25. Izard, C.E., Fantauzzo, C.A., Castle, J.M., Haynes, O.M., Rayias, M.F., Putnam, P.H. (1995). The 

ontogeny and significance of infants’ facial expressions in the first 9 months of life. Developmental Psychology, 31, 997–1013. 

Jaccard, J., Brinberg, D., & Ackerman, L.J. (1986). Assessing attribute importance: A compari‐son of six methods. Journal of Consumer  Research, 12, 463‐468. 

Jack, A.I., & Roepstorff, A. (2002). Introspection and cognitive brain mapping: from stimulusre‐sponse to script‐report. Trends in Cognitive Sciences, 6, 333–339. 

Jack, A.I., & Shallice, T.  (2001).  Introspective physicalism as an approach  to  the  science of consciousness. Cognition, 79, 161–96. 

Jacob,  T.J.C.,  Fraser,  C.S., Wang,  L., Walker,  V.E.,  &  O'Connor,  S.  (2003).  Psychophysical evaluation of responses to pleasant and mal‐odour stimulation in human subjects; adaptation,  dose  response  and  gender  differences.  International  Journal  of  Psy‐chophysiology, 48, 67‐80. 

James, W.  (1890). The Principles of Psychology, Cambridge, MA: Harvard University Press, 1981. Originally published in 1890. 

January, D., & Kako, E. (2007). Re‐evaluating evidence for linguistic relativity: Reply to Boro‐ditsky (2001). Cognition 104, 417–426. 

Page 203: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

203

Job, R.F.S. (1988). Community response to noise: a review of factors influencing the relation‐ship  between  noise  exposure  and  reaction,  Journal  of  the  Acoustical  Society  of America, 83, 991–1001. 

Johansson A., & Strømnes, F.  (1995). Cultural differences  in occupational accidents. Part  I: Theoretical background. Presented at  the 44th Nordic Meeting on Work Environ‐ment, Nantali, Finland. 

Johansson,  A., &  Salminen,  S.  (1996).  Different  languages:  different  information  processing systems? Part  I: Why can we expect differences  in occupational accidents between language groups. Presented at the International Symposium on Work in the Informa‐tion Society, 20‐22 May 1996, Helsinki, Finland.  

Johnson, M.  (1987). The Body  in  the Mind: The Bodily Basis of Meaning,  Imagination, and Reasoning. University of Chicago Press, Chicago. 

Johnson, M. (1991). Knowing through the body. Philosophical Psychology 4, 3–20. Jordan, P. (2000). Designing pleasurable products. London: Taylor and Francis. Jousmäki, V., & Hari, R. (1998). Parchment‐skin illusion: sound biased touch. Current Biology, 

8, 190. Kalucy, R.S. (1987). The “new” nutrition. The Medical Journal of Australia, 147, 529—530.  Karapanos, E., Hassenzahl, M., & Martens,  J.‐B.  (2008). User experience over  time. CHI  '08 

extended abstracts on Human factors in computing systems. ACM, Florence, Italy. Kass, L. (1994). The Hungry Soul. New York: The Free Press. 

Kay, P., & Kempton, W. (1984). What is the Sapir‐Whorf hypothesis? American Anthropolo‐gist 86, 65–79. 

Kellert, S.R., & Wilson, E.O.  (Eds.)  (1993). The biophilia hypothesis. Washington, DC:  Island Press. 

Khalid, H.M.  (2004).  Conceptualizing  affective  human  factors  design.  Theoretical  lssues  in Ergonomics Science, 5(1), 1‐3. 

Khan, R., Luk, C., Flinker, A., Aggarwal, A., Lapid, H., et al. (2007). Predicting odor pleasant‐ness  from odorant  structure: pleasantness  as  a  reflection of  the physical world. Journal of Neuroscience, 27, 10015–10023. 

Kimmel, M. (2006). Culture regained: situated and compound image schemas. In: Hampe, B. (Ed.), From Perception to Meaning:  Image Schemas  in Cognitive Linguistics. Mou‐ton de Gruyter, Berlin, pp. 285–312. 

Kosslyn, S. (1983). Ghosts in the Mind's Machine. New York: W.W. Norton & Co.  Kosslyn, S.M. (1980). Image and Mind. Cambridge, Mass.: Harvard University Press.  Köster, E.P. (2003). The psychology of food choice: Some often encountered fallacies. Food 

Quality and Preference, 14, 359–373. Kövecses, Z.  (2005). Metaphor  in  culture: Universality and variation. Cambridge University 

Press, New York. Krippendorff, K. (1980). Content analysis: an introduction to its methodology. Sage, London. Krippendorff, K. (2004).  Intrinsic motivation and human‐centered design. Theoretical  Issues 

in Ergonomics Science, 5(I), 43‐ 71. 

Page 204: ANNA FENKO - TU Delft

REFERENCES 

 

204

Kroll,  J., & de Groot, A.  (1997).  Lexical  and  conceptual memory  in  the bilingual: mapping form to meaning in two languages. In: de Groot, A., & Kroll, J. (Eds.), Tutorials in Bi‐lingualism: Psycholinguistic Perspectives. Erlbaum, Mahwah, NJ, 169–99. 

Kryter, K.D. (1966). Psychological reactions to aircraft noise. Science, 151, 1346‐1355. Kunst‐Wilson, W.R., & Zajonc, R.B. (1980). Affective discrimination of stimuli that cannot be 

recognized. Science, 207, 557–558. Kurath, H.  (1921). The Semantic Sources of the Words  for the Emotions  in Sanskrit, Greek, 

Latin and the Germanic Languages. George Banta Publishing, Menasha, WI. Labbe, D., Almiron‐Roig, E., Hudry, J., Leathwood, P., Schifferstein, H.N.J., & Martin, N. (2009). 

Sensory basis of refreshing perception: Role of psychophysiological factors and food experience. Physiology & Behavior, 98, 1‐9. 

Labbe, D., Gilbert, F., Antille, N., & Martin, N.  (2009). Sensory determinants of  refreshing. Food Quality and Preference, 20, 100‐109. 

Labbe, D., Schlich, P., Pineau, N., Gilbert, F., & N. Martin, N. (2009). Temporal dominance of sensations  and  sensory profiling: A  comparative  study.  Food Quality and  Prefer‐ence, 20, 216–221. 

Labovitz, S.  (1970). The assignment of numbers  to  rank order  categories. American Socio‐logical Review, 35, 5151‐5524. 

LaFromboise,  T.,  Coleman,  H., & Gerton,  J.  (1993).  Psychological  impact  of  biculturalism: Evidence and theory. Psychological Bulletin 114, 395–412. 

Laird, D.A.  (1932). How  the  consumer  estimates  quality  by  subconscious  sensory  impres‐sions. Journal of Applied Psychology, 16, 241‐246. 

Lakoff, G., &  Johnson, M.  (1980). Metaphors We Live By. University of Chicago Press, Chi‐cago. 

Lakoff, G., &  Johnson, M.  (1999). Philosophy  in  the  flesh: The embodied mind and  its chal‐lenge to Western thought. Chicago: University of Chicago Press. 

Langer, S.K. (1967/1982). Mind: An Essay on Human Feeling. Baltimore, MD: Johns Hopkins University Press. 

Larson‐Powers, M., & Pangborn, R. M. (1978). Paired comparison and time‐intensity measure‐ments of the sensory properties of beverages and gelatins containing sucrose or syn‐thetic sweeteners. Journal of Food Science, 43, 41–46. 

Laurans, G., & Desmet, P. (2008). Speaking in tongues – Assessing user experience in a global economy. Paper presented at Design & Emotions Conference, Hong Kong, 6‐9 Oc‐tober, 2008. 

Laurans, G., Desmet, P., & Hekkert, P. (2009). Assessing emotion in interaction: Some prob‐lems and a new approach. Paper presented at International Conference on Design‐ing Pleasurable Products and Interfaces, 13‐ 16 October 2009, Compiegne, France. 

Lawless, H., & Engen, T. (1977). Associations to odors: interference, mnemonics, and verbal label‐ing. Journal of Experimental Psychology: Human Learning & Memory, 3, 52–59.  

Page 205: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

205

Le Reverend, F.M., Hidrio, C., Fernandes, A., & Aubry, V. (2008). Comparison between tem‐poral dominance of sensations and time intensity results. Food Quality and Prefer‐ence, 19, 174–178. 

Lederman, S.J., & Abbott, S.G. (1981). Texture perception: studies of  intersensory organiza‐tion using a discrepancy paradigm, and visual versus tactile psychophysics. Journal of Experimental Psychology: Human Perception and Performance, 4, 902‐915. 

Lederman, S.J., Thorne, G., &  Jones, B.  (1986). Perception of  texture by vision and  touch: multidimensionality and intersensory integration. Journal of Experimental Psychol‐ogy: Human Perception and Performance, 12, 169‐180. 

LeDoux, J.E. (2000). Emotion circuits in the brain. Annual Review of Neuroscience, 23, 155–184. Lee, H.S., & O’ Mahony, M.  (2005). Sensory evaluation and marketing: Measurement of a 

consumer concept. Food Quality and Preference, 16, 227–235. Lee,  L., Frederick, S., & Ariely, D.  (2006). Try  it, you’ll  like  it:  the  influence of expectation, 

consumption,  and  revelation on preferences  for beer. Psychological  Science,  17, 1054–1058. 

Lee, M., & Lou, Y. –C.  (1996). Consumer  reliance on  intrinsic and extrinsic cues  in product evaluations: a conjoint approach. Journal of Applied Business Research 12, 21‐29.  

Lehrer, A. (1978). Structures of the lexicon and transfer of meaning. Lingua 45, 95‐123. Leibovitz, M., London, P., Cooper, L., & Hart, J. (1972). Dominance in mental imagery. Educa‐

tional and Psychological Measurement 32, 679‐703. Levinson, S. (1996). Language and space. Annual Review of Anthropology, 25, 353‐382. Levinson,  S.  (2003).  Space  in  Language  and  Cognition:  Explorations  in  Cognitive Diversity. 

Cambridge University Press, New York. Levinson, S., &  Jaisson, P.  (Eds), 2006. Evolution and Culture: A Fyssen Foundation Sympo‐

sium. The MIT Press, Cambridge, MA. Levinson, S., & Meira, S. (2003). 'Natural concepts' in the spatial topological domain ‐ adposi‐

tional meanings  in  cross‐linguistic perspective: An exercise  in  semantic  typology. Language 79, 485‐516. 

Levinson, S., & Wilkins, D. (Eds.), 2006. Grammars of Space. Cambridge University Press, New York. 

Lewinski,  R.H.  (1938).  An  investigation  of  individual  responses  to  chromatic  illumination. Journal of Psychology, 6, 155‐160. 

Lewkowicz, D.J. (1996). Perception of auditory‐visual temporal synchrony in human infants. Jour‐nal of Experimental Psychology: Human Perception and Performance, 22, 1094 –1106. 

Li, P., & Gleitman, L.  (2002). Turning  the  tables: Language and spatial reasoning. Cognition 83, 265–294. 

Liu, Y.  (2003). The aesthetic and  the ethic dimensions of human  factors and design. Ergo‐nomics, 46, 1293‐1305. 

Lloyd, D.M., Calvert, G.A., Hansen, P.C., & McGlone, X.L.F. (2001). Visuo‐tactile Integrations Sites are Modulated by Task and Attention. Society for Neuroscience, San Diego, CA. 

Lord,  F.  (1980). Applications of  Item Response Theory  to Practical Testing Problems.  Law‐rence Erlbaum Associates, Hillsdale, NJ. 

Page 206: ANNA FENKO - TU Delft

REFERENCES 

 

206

Loveless, N.E.,  Brebner,  J., & Hamilton,  P.  (1970).  Bisensory  presentation  of  information. Psychological Bulletin, 73, 161–199. 

Lucy, J. (1992). Language Diversity and Thought: A Reformulation of the Linguistic Relativity Hypothesis. Cambridge University Press, New York. 

Lucy, J., & Gaskins, S. (2001). Grammatical categories and the development of classification preferences:  A  comparative  approach.  In:  Bowerman, M., &  Levinson,  S.  (Eds.). Language Acquisition  and  Conceptual Development.  Cambridge University  Press, New York, pp. 257‐283. 

Ludden, G. D.  S., &  Schifferstein, H. N.  J.  (2007). Effects of  visual–auditory  incongruity on product expression and surprise. International Journal of Design, 1 (3), 29‐39.  

Lupton, D. (1996). Food, the Body and the Self. London: Sage. Lutz,  A. &  Thompson,  E.  (2003). Neurophenomenology:  integrating  subjective  experience 

and brain dynamics in the neuroscience of consciousness. Journal of Consciousness Studies, 10, 31‐52. 

Lutz, A. (2002). Toward a neurophenomenology as an account of generative passages: a first empirical case study. Phenomenology and the Cognitive Sciences, 1, 133–67. 

Lyon, R. H. (2000). Designing for product sound quality. New York: Marcel Dekker, Inc. MacDonald,  A.  (2002).  The  scenario  of  sensory  encounter:  Cultural  factors  in  sensory‐

aesthetic experience. In W.S. Green & P.W. Jordan (Eds.), Pleasure with products: Beyond usability. London: Taylor & Francis, pp. 113‐123. 

Macdonald, A. S.  (2000). Aesthetic  intelligence: Optimizing user centred design.  Journal of Engineering Design, 12, 37–45. 

MacDonald, A.S. (2003). Affective technology and affective human factors design in an era of rapid technological change. In Proceedings of the XVth Triennial Congress of the In‐ternational Ergonomics Association. lEA, Seoul, Korea. 

Majid, A., & Levinson, S. (2008). Language does provide support for basic tastes. Behavioral and Brain Sciences 31, 86‐87. 

Majid, A., Bowerman, M., Kita, S., Haun, D, & Levinson, S. (2004). Can language restructure cogni‐tion? The case for space. Trends in Cognitive Sciences 8, 108–114. 

Malt, B., Sloman, S., Gennari, S,, & Wong, Y. (1999). Knowing versus naming: similarity and the linguistic categorization of artefacts. Journal of Memory and Language 40, 230‐262. 

Martindale, C., & Moore, K.  (1988). Priming, prototypicality, and preference.  Journal of Experi‐mental Psychology: Human Perception and Performance, 14, 661–670. 

Maurer,  D.  (1993).  Neonatal  synesthesia:  implications  for  the  processing  of  speech  and faces. In de Boysson‐Bardies, B., de Schonen, S., Jusczyk, P., McNeilage, P., & Mor‐ton, J. (Eds) Developmental Neurocognition: Speech and face processing in the first year of life. Kluwer Academic Publishers, Dordrecht. 

McClure, S.M., Li, J., Tomlin, D., Cypert, K.S., Montague, L.M., & Montague, P.R. (2004). Neu‐ral  correlates of behavioral preference  for  culturally  familiar drinks. Neuron, 44, 379–387. 

Page 207: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

207

McConkie, G.W., & Currie, C.B.,  (1996). Visual stability across saccades while viewing com‐plex pictures. Journal of Experimental Psychology: Human Perception and Perform‐ance 22, 563–581. 

McCrory, M., Saltzma, E., Rolls, B.J., & Roberts, S.B.  (2006). A  twin  study of  the effects of energy density and liking on energy intake of individual foods. Physiology & Behav‐ior 87, 451–459.  

McEwan, J. A., & Colwill, J. S. (1996). The sensory assessment of the thirstquenching charac‐teristics of drinks. Food Quality and Preference, 7, 101‐111. 

McGuire, W. (1976). Some internal psychological factors influencing consumer choice. Jour‐nal of Consumer Research, 2, 302‐319. 

McGurk, H., & MacDonald, J. (1976). Hearing lips and seeing voices. Nature, 264, 746–748. McKennel, A.C. (1963). Aircraft noise annoyance around London (Heathrow) airport. London: 

Central Office of Information. McLuhan, M. (1961). Inside the five sense sensorium. The Canadian Architect, 6 (June), 49‐54. Mela, D.J. (2006). Eating for pleasure or just wanting to eat? Reconsidering sensory hedonic 

responses as a driver of obesity. Appetite, 47, 10–17.  

Mennel, S., Murcott, A. & van Otterloo, A.H. (1992). The Sociology of Food: Eating, Diet and Culture. Sage Publications, London. 

Merleau‐Ponty, M. (1962) Phenomenology of Perception, trans. Colin Smith, New York: Hu‐manities Press. 

Merleau‐Ponty,  M.  (1973).  Consciousness  and  the  Acquisition  of  Language,  trans.  Hugh Silverman, Evanston: Northwestern Univ. Press.  

Mervis, C., & Rosch, E. (1981) . Categorization of natural objects. Annual Review of Psychol‐ogy 32, 89‐115. 

Messinger, S.M.  (1998). Pleasure and  complexity: Berlyne  revisited.  Journal of Psychology, 132, 558‐564.  

Meyers,  L.S., Guarino, A., & Gamst, G.  (2005). Applied Multivariate Research: Design  and Interpretation. Sage Publications, p. 20. 

Michael, G.A., &  Roihion,  P.  (2008).  Cool  colors:  Color‐induced  nasal  thermal  sensations. Neuroscience Letters, 436, 141‐144. 

Miller, E.A.  (1972).  Interaction of vision and  touch  in conflict and nonconflict  form percep‐tion tasks. Journal of Experimental Psychology, 96, 114‐123. 

Miller, G., & Johnson‐Laird, P. (1976). Language and Perception. Cambridge University Press, New York. 

Millot, J., & Brand, G.  (2001). Effects of pleasant and unpleasant ambient odors on human voice pitch, Neuroscience Letter, 297, 61–63. 

Mintz, S.W. (1985). Sweetness and Power: The Place of Sugar in Modern History. New York: Viking. 

Mol, A. (2008). I eat an apple. On theorizing subjectivities. Subjectivity, 22, 28–37. Monö, R. (1997). Design for product understanding. Stockholm: Liber. 

Page 208: ANNA FENKO - TU Delft

REFERENCES 

 

208

Morgan, G.A., Goodson. F.E., & Jones, T. (1975) Age differences in the associations between felt temperatures and color choices. American Journal of Psychology, 88, 125‐130. 

Moskowitz, H.R., Dravnieks, A.,  Klarman,  L.A.  (1976). Odor  intensity  and  pleasantness  for  a diverse set of odorants. Perception & Psychophysics, 19, 122–128. 

Mugge, R., Schifferstein, H.N.J., & Schoormans, J.P.L.  (2004). Personalizing product appear‐ance: the effect on product attachment. In: A. Kurtgözü (Ed.) 4th International Con‐ference on Design and Emotion, Ankara, Turkey. 

Murcott, A.  (1982). On  the Social Significance of  the Cooked Dinner  in South Wales. Social Science Information, 25, 677–96. 

Murphy F.C., Nimmo‐Smith,  I., & Lawrence, A.D.  (2003). Functional neuroanatomy of emo‐tions: a meta‐analysis. Cognitive, Affective, & Behavioral Neuroscience, 3, 207–233. 

Murphy,  L.L.,  Stanney  K.,  &  Hancock,  P.  A.  (2003).  The  effect  of  affect:  The  hedonomic evaluation of human‐computer  interaction.  In Proceedings of  the Human Factors and Ergonomics Society 47th Annual Meeting. Santa Monica, CA: Human Factors and Ergonomics Society, pp. 764‐767. 

Nagamashi, M. (1989). Kansei Engineering. Kaibundo, Tokyo. Neisser, U. (1994). Multiple systems: a new approach to cognitive theory. European Journal 

of Cognitive Psychology 6, 225–241. Nemeroff,  C., &  Rozin,  P.  (1994).  The  contagion  concept  in  adult  thinking  in  the  United 

States: Transmission of germs and interpersonal influence. Ethos, 22, 158–186. Nemeroff, C., & Rozin, P.  (2000). The makings of the magical mind.  In K.S. Rosengren, C.N. 

Johnson, & P.L. Harris (Eds.), Imagining the impossible: Magical, scientific, and reli‐gious thinking in children (pp. 1–34). New York: Cambridge University Press. 

Newhall, S.M. (1941). Warmth and coolness of colors. Psychological Record, 4, 198‐212. Newton, N. (1996). Foundations of Understanding, Benjamins, Amsterdam. Nguyen, T.D., Heslin, R., and Nguyen, M.L.  (1975). The meanings of touch: Sex differences. 

Journal of Communication 25, 92–103. Niedenthal, P.M., Halberstadt, J.B., & Innes‐Ker, Å.H. (1999). Emotional response categoriza‐

tion. Psychological Review, 106, 337‐361. Nisbett, R., and Wilson, T. (1977). Telling more than we can know: Verbal reports on mental 

processes. Psychological Review, 84, 231‐259. Nogales, P. (2007). The extent of the literal by Marina Rakova. Mind, 116, 467‐471. Norman, D. (2004). Emotional Design: Why We Love (or Hate) Everyday Things. Basic Books, 

New York. Northwood, T.D. (1963). Sound and People. Canadian Building Digest, Vol. 41, 973‐978. Núñez, R., & Sweetser, E. (2006). Looking ahead to the past: Convergent evidence from Ay‐

mara  language and gesture  in the crosslinguistic comparison of spatial construals of time. Cognitive Science 30, 401–450. 

O’Regan, J.K. (1992). Solving the  ‘real’ mysteries of visual perception: The world as an out‐side memory. Canadian Journal of Psychology, 46, 461–488. 

Page 209: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

209

Olausson, H., Cole, J., Rylander, K., McGlone, F., Lamarre, Y., et al. (2008). Functional role of unmyelinated tactile afferents in human hairy skin: sympathetic response and per‐ceptual localization. Experimental  Brain Research, 184, 135–140. 

Olausson, H., Lamarre, Y., Backlund, H., Morin, C., Wallin, B.G., et al. (2002). Unmyelinated tactile afferentssignal touch and project to insular cortex. Nature Neuroscience, 5, 900–904. 

Oliva, A., and Torralba, A.  (2001).  Identifying the perceptual dimensions of visual complexity of scenes. International Journal of Computer Vision, 42, 145‐175. 

Osgood, C.E., Suci, G.J., & Tannenbaum, P.H. (1957). The Measurement of Meaning. Univer‐sity of Illinois Press, Urbana.  

Özcan,  E. & Van  Egmond, R.  (2008). Product  sound design: An  interdisciplinary  approach. Proceedings of the 4th International Conference on Design Research. Sheffield: UK. 

Paivio, A. (1971). Imagery and verbal processes. New York: Holt, Rinehart, & Winston. Paivio, A. (2006). Mind and  its evolution; A dual coding theoretical interpretation. Mahwah, 

NJ: Lawrence Erlbaum Associates, Inc. Palmer, S.E. (1999). Vision Science: Photons to Phenomenology. MIT Press. Pan, Y., & Schmitt, B. (1996). Language and brand attitudes: Impact of script and sound matching 

in Chinese and English. Journal of Consumer Psychology 5, 263‐277. Park, C.W. & Young, M. (1983). Types and levels of involvement and brand attitude formation. 

In: P. Bagozzi & A. Tybout (Eds.). Advances in consumer research, 10, 320–324. Pearson, P.  (1970). Relationships between global and specific measures of novelty seeking. 

Journal of Consulting and Clinical Psychology 34, 199‐204. Peneau, S., Hoehn, E., Roth, H. R., Escher, F., & Nuessli, J. (2006). Importance and consumer 

perception of freshness of apples. Food Quality and Preference, 17, 9–19. Pfaffmann, C. (1960). The pleasures of sensation. Psychological Review, 67, 253–268. Phan, K.L., Wager, T., Taylor, S.F., & Liberzon,  I.  (2002). Functional neuroanatomy of emo‐

tion: a meta‐analysis of emotion activation studies  in PET and fMRI. Neuroimage, 16, 331–348. 

Pica, P., Lemer, C.,  Izard, V., & Dehaene, S. (2004). Exact and approximate arithmetic  in an Amazonian indigene group. Science 306, 499–503. 

Piggott,  J. R.  (1990). Relating  sensory and  chemical data  to understand  flavour.  Journal of Sensory Studies, 4, 261‐272. 

Pine, B.J., & Gilmore,  J.H.  (1998). Welcome  to  the experience economy. Harvard Business Review, 76, 97‐105. 

Pineau, N., Cordelle, S., Imbert, A., Rogeaux, M., & Schlich, P. (2003). Dominance Temporelle des Sensations – Codage et analyse d’un nouveau  type de données  sensorielles. XXXVème journée de statistiques, 2–6, 777–780. 

Pineau, N., Schlich, P.,  Cordelle, S.  Mathonnière. C., Issanchou, S., Imbert, A., Rogeaux, M., Etiévant. P., & Köster, E. (2009). Temporal Dominance of Sensations: Construction of  the TDS  curves and  comparison with  time–intensity. Food Quality and Prefer‐ence, 20, 450–455.  

Page 210: ANNA FENKO - TU Delft

REFERENCES 

 

210

Plassmann, H., O’Doherty, J., Shiv, B., & Rangel, A. (2008). Marketing actions can modulate neural  representations of experienced pleasantness. Proceedings of  the National Academy of Science USA, 105, 1050–1054. 

Ponterotto, J.G. (2005) Qualitative Research in Counseling Psychology: A Primer on Research Paradigms  and Philosophy of  Science.  Journal  of Counseling Psychology, 52  (  2), 126–136. 

Posner, M.I., Nissen, M.J., & Klein, R.M. (1976). Visual dominance: an information‐processing account of its origins and significance. Psychological Review, 83, 157‐171. 

Postrel, V. I. (2003). The substance of style: How the rise of aesthetic value is remaking commerce, culture, and consciousness. New York: Harper Collins, pp. 34‐65. 

Radeau, M.  (1994). Auditory–visual  spatial  interaction  and modularity.  In: Cahiers de Psy‐chologie Cognitive/Current Psychology of Cognition, pp. 3–51. 

Rakova, M. (2003) The Extent of the Literal. Houndmills, UK: Palgrave Macmillan, 2003. Ratchford, B.  (1987). New  insights about the FCB‐grid. Journal of Advertising Research, Au‐

gust/September, 24–38. Rein, M., & Schön, D.  (1977). Problem  setting  in policy  research.  In: Weiss, C.  (Ed.), Using 

Social  Research  in  Public  Policy Making.  Lexington Books,  Lexington, MA,  pp. 235–251. 

Renaud, S., &  Logeril, M. de.  (1992). Wine, alcohol, platelets, and  the French paradox  for coronary heart disease. Lancet, 339, 1523‐1526. 

Rennie, D. L. (Ed.). (2002). Qualitative research: History, theory and practice [Special  issue]. Canadian Psychology, 43 (3). 

Rensink, R.A., O’Regan, J.K., & Clark, J.J., (1997). To see or not to see: the need for attention to perceive changes in scenes. Psychological Science 8, 368–373. 

Rétiveau, A.N., Chambers, I.V.E., & Milliken, G.A. (2004). Common and specific effects of fine fragrances on the mood of women, Journal of Sensory Studies, 19, 373–394.

Richardson, A.  (1977). Verbalizer‐visualizer: a  cognitive  style dimension.  Journal of Mental Imagery, 1, 109‐125. 

Robertson, D., Davies,  I., & Davidoff,  J.  (2000). Color categories are not universal: Replica‐tions  and  new  evidence  from  a  stone‐age  culture.  Journal  of  Experimental  Psy‐chology: General 129, 369–398. 

Rock, I., & Victor, J. (1964). Vision and touch: An experimentally created conflict between the two senses. Science, 143, 594‐596. 

Rolls, E.T. (2005).  Emotion Explained. Oxford University Press, Oxford. Rolls, E.T. (2008). The Representation of Flavor in the Brain. In: The Senses: A Comprehensive 

Reference, pp. 469‐478.  Roos, G., & Wandel, M. (2005). “I Eat because I’m Hungry, because it’s Good and to Become 

Full”: Eating Voiced by Carpenters, Drivers and Engineers  in Contemporary Oslo. Food and Foodways, 13, 169–80. 

Rosch, E. (1978). Principles of categorization. In: Rosch, E. & Lloyd, B. (Eds.), Cognition and Catego‐rization. Lawrence Erlbaum Associates, Hillsdale, NJ, pp. 27‐48. 

Page 211: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

211

Ross, R.T. (1938). Studies in the psychology of the theatre. Psychological Record, 2, 127‐190. Rossiter, J., & Percy, L. (1991). A better advertising planning grid. Journal of Advertising Re‐

search, October, 11–21. Rozin, P. (1990). Development in the food domain. Developmental Psychology 26, 555–562.  Rozin, P. (1999). Food is fundamental, fun, frightening, and far‐reaching. Social Research, 66, 

9‐30. Rozin, P. (1999). Preadaptation and the puzzles and properties of pleasure. In: D. Kahneman, 

E. Diener and N. Schwarz (Eds.) Well‐being: The foundations of hedonic psychology. Russell Sage Foundation, New York, NY, pp. 109–133. 

Rozin, P.  (2005). The Meaning of “natural”: Process more  important  than content. Psycho‐logical Science, 16, 652‐658. 

Rozin,  P., & Royzman,  E.B.  (2001). Negativity  Bias, Negativity Dominance,  and  Contagion. Personality and Social Psychology Review, 5, 296‐320. 

Rozin, P., & Nemeroff, C.J. (1990). The laws of sympathetic magic: A psychological analysis of similarity and contagion. In J. Stigler, G. Herdt, & R.A. Shweder (Eds.), Cultural psy‐chology:  Essays  on  comparative  human  development  (pp.  205–232).  Cambridge, England: Cambridge University Press. 

Rozin, P., Ashmore, M. B., & Markwith, M.    (1996). Lay American conceptions of nutrition: Dose insensitivity, categorical thinking, contagion, and the monotonic mind. Health Psychology, 15, 438‐447. 

Rozin, P., Haidt, J., McCauley, C. R., & Imada, S. (1997). The cultural evolution of disgust.  In H. M. Macbeth (ed.).   Food preferences and taste: Continuity and change (pp. 65‐82).  Oxford, U.K.: Berghahn. 

Rozin, P.,  Spranca, M., Krieger,  Z., Neuhaus, R.,  Surillo, D.,  Swerdlin, A., & Wood, K.  (2004). Natural preference: Instrumental and ideational/moral motivations, and the contrast between foods and medicines. Appetite, 43, 147–154. 

Rozin, P.N., & Schulkin, J. (1990) Food selection. In: E.M. Stricker (Ed.) Neurobiology of food and fluid intake, 10, 297–328. 

Rudrauf, D. et al. (2003). From autopoiesis to neurophenomenology. Biological Research, 36, 27–66. 

Ruiz, J.H. (2006). The role of metaphor, metonymy, and conceptual blending in understand‐ing advertisements: The  case of drug‐prevention ads. Revista Alicantina de Estu‐dios Ingleses 19, 169‐190. 

Russell, J.A. (2003). Core affect and the psychological construction of emotion. Psychological Review, 110, 145–172. 

Sacks, O. (1995). An Anthropologist on Mars. Toronto: Vintage Canada, p. 122. Salminen, S., & Hiltunen, E. (1995). Cultural differences in occupational accidents. Part II: A case 

study of Finnish‐ and Swedish‐speaking workers in Finland. Presented at the 44th Nordic Meeting on Work Environment, Naantali, Finland. 

Scherer, K.R.  (2005). What are emotions? and how  can  they be measured? Social Science Information, 44 (4), 695‐729. 

Page 212: ANNA FENKO - TU Delft

REFERENCES 

 

212

Schifferstein,  H.N.J.,  &  Zwartkruis‐Pelgrim,  E.P.H.  (2008).  Consumer‐product  attachment: measurement and design implications. International Journal of Design 2, 1‐13. 

Schifferstein, H.N.J. & Spence, C. (2008). Multisensory product experience. In: Schifferstein, HNJ, Hekkert P. (Eds.). Product experience: Elsevier, p. 133‐162. 

Schifferstein, H.N.J. (2006). The relative importance of sensory modalities in product usage: a study of self‐reports. Acta Psychologica, 121, 41‐64. 

Schifferstein,  H.N.J.,  &  Cleiren,  M.P.H.D.  (2005).  Capturing  product  experiences:  a  split‐modality approach. Acta Psychologica, 118, 293‐318. 

Schifferstein, H.N.J., & Desmet, P.M. A. (2007). The effect of sensory impairments on product experience and personal well‐being. Ergonomics, 50, 2026‐2048. 

Schifferstein, H.N.J., & Verlegh, P.W.J. (1996). The role of congruency and pleasantness in odor‐induced taste enhancement. Acta Psychologica, 94, 87‐105. 

Schifferstein, H.N.J., Otten, J.J., Thoolen, F., & Hekkert, P. (2010). An experimental approach to assess sensory dominance in a product development context, Journal of Design Research, 8(2), 119‐144. 

Schiffman S., Robinson, D.E., Erickson, R.P.  (1977). Multidimensional‐scaling of odorants—examination  of  psychological  and  physiochemical  dimensions.  Chemical  Senses: Flavor, 2, 375–390.  

Schiffman, S.S. (1974). Physicochemical correlates of olfactory quality. Science, 185, 112–117.  Schleidt, M., Neumann, P., & Morishita, H. (1988). Pleasure and disgust: Memories and asso‐

ciations  of  pleasant  and  unpleasant  odours  in  Germany  and  Japan,  Chemical Senses, 13, 279–293. 

Schlosser, E.  (2001). Fast Food Nation: The Dark Side of  the All‐American Meal. Houghton Mifflin.  

Schmitt B., & Zhang S. (1998). Language structure and categorization: A study of classifiers in con‐sumer cognition, judgment and choice. Journal of Consumer Research 25, 108–22. 

Schmitt, B. (1999). Experiential marketing. Journal of Marketing Management, 15, 53‐67. Schmitt, B., Pan, Y., & Tavassoli, N. (1994). Language and consumer memory: The impact of 

linguistic differences between Chinese and English. Journal of Consumer Research 21, 419‐431. 

Schroeder, J. (2002). Visual consumption. London: Routledge. Schultz, T.J. (1978). Synthesis of social surveys on noise annoyance. Journal of the Acoustical 

Society of America; 74: 377–405. Seeger, A. (1975). The meaning of body ornaments: A Suya example. Ethnology, 14, 211‐224.  Sekuler, R. and Blake, R. (1994) Perception (3rd edn). Alfred A. Knopf, New York. Sekuler, R., Sekuler, A.B. &  Lau, R.  (1997). Sound alters visual motion perception, Nature, 

385, 308. Shams, L., Kamitani, Y., & Shimojo, S. (2000).  Illusions. What you see  is what you hear. Na‐

ture, 408, 788. Shams,  L.,  Kamitani,  Y., &  Shimojo,  S.  (2002).  Visual  illusion  induced  by  sound.  Cognitive 

Brain Research, 14, 147–152. 

Page 213: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

213

Shapiro, K.L.,  Egerman, B. & Klein, R.M.  (1984).  Effects of  arousal on human  visual domi‐nance. Perception & Psychophysics, 35, 547‐552. 

Shiv, B., Carmon, Z., & Ariely, D. (2005). Placebo effects of marketing actions: Consumers may get what they pay for. Journal of Marketing Research, 42, 383–393. 

Siminon, B.L., & Ruth, J.A. (1998). Is a company known by the company  it keeps? Assessing the  spillover  effects  of  brand  alliances  on  consumer  brand  attitudes.  Journal  of Marketing Research, 35(1), 30–42. 

Simner, J., & Ward, J. (2006). Synaesthesia: The taste of words on the tip of the tongue, Na‐ture, 444, 438. 

Simpson, J., & Weiner, E. (Eds.) (1989). The Oxford English Dictionary. Second edition. Oxford University Press, New York. 

Singer, G., & Day, R. N. (1969). Visual capture of haptually  judged depth. Perception & Psy‐chophysics, 6, 203‐205. 

Sloutsky, V.M., & Napolitano, A.C. (2003). Is a picture worth a thousand words? Preference for auditory modality in young children. Child Development, 74(3), 822‐833. 

Slovic, P., & Lichtenstein, S.  (1971). Comparison of Bayesian and  regression approaches  to the  study  of  information  processing  in  judgment.  Organizational  Behavior  and Human Performance, 6, 649‐744. 

Smets, G.J.F., & Overbeeke, C.  J.  (1995). Expressing  tastes  in packages. Design Studies, 16, 349–365. 

Snedecor, G.W., & Cochran, W.G.  (1980) Statistical Methods. Ames,  Iowa: The  Iowa State University Press.  

Soussignan, R., Schaal, B., Marlier, L., & Jiang, T. (1997). Facial and autonomic responses to bio‐logical and artificial olfactory  stimuli  in human neonates:  re‐examining early hedonic discrimination of odors. Physiology & Behavior, 62, 745–758. 

Spence, C., & Driver, J. (2004). Crossmodal space and crossmodal attention. Oxford, UK: Ox‐ford University Press. 

Spence, C., Nicholls, M. E. R., & Driver, J. (2001). The cost of expecting events in the wrong sensory modality. Perception & Psychophysics, 63, 330‐336. 

Stadtlander,  L.,  & Murdoch,  L.  (2000).  Frequency  of  occurrence  and  rankings  for  touch‐related  adjectives. Behavior Research Methods,  Instruments,  and Computers 32, 579‐587. 

Stein, B.E., and Meredith, M.A. (1993). The merging of the senses. The MIT Press, Cambridge, MA.  Stein, B.E., Meredith, M.A., Huneycutt, W.S., & McDade, L. (1989). Behavioral indices of mul‐

tisensory  integration:  orientation  to  visual  cues  is  affected  by  auditory  stimuli. Journal of Cognitive Neuroscience, 1, 12–24. 

Steiner,  J.E.  (1979). Human  facial  expressions  in  response  to  taste  and  smell  stimulation. Advances in Child Development and Behavior, 13, 257–295. 

Stemmler,  G.  (2003)  Methodological  Considerations  in  the  Psychophysiological  Study  of Emotion. In R.J. Davidson, K.R. Scherer and H. Goldsmith (eds) Handbook of the Af‐fective Sciences, pp. 225–55. New York and Oxford: Oxford  University Press. 

Page 214: ANNA FENKO - TU Delft

REFERENCES 

 

214

Stevenson, R.J., & Boakes, R.A. (2003) A mnemonic theory of odor perception. Psychological Review, 110 (2), 340‐364. 

Stratton, G.M. (1897). Vision without inversion of the retinal image. Psychological Review, 4, 341‐369, 463‐481. 

Strømnes, F.  (1973). A semiotic  theory of  imagery processes with experiments on an  Indo‐European and a Ural‐Altaic  language: Do  speakers of different  languages experi‐ence different cognitive worlds? Scandinavian Journal of Psychology 14, 291‐304. 

Strømnes, F. (1974). No universality of cognitive structures? Two experiments with almost‐perfect one‐trial learning of translatable operators in a Ural‐Altaic and IndoEuropean language. Scandinavian Journal of Psychology 15, 300‐309. 

Suzuki, M. & Gyoba, J. (2008). Visual and tactile cross‐modal mere exposure effects. Cogni‐tion & Emotion, 22, 147‐154. 

Suzuki, M., & Gyoba, J. (2001). Contrastive analysis of sensory‐relevance of factors affecting aesthetic impressions. Tohoku Psychologica Folia 60, 81‐92. 

Suzuki, M., Gyoba, J., & Sakuta, Y. (2005). Multichannel NIRS analysis of brain activity during semantic differential rating of drawing stimuli containing different affective polari‐ties. Neuroscience Letters 375, 53‐58. 

Suzuki, M., Gyoba,  J., Kawabata, H., Yamaguchi, H., & Komatsu, H.  (2006). Analyses of  the sensory‐relevance of  adjective pairs by  the modality differential method.  Shinri‐gaku Kenkyu 77, 464‐470. 

Sweetser, E.  (1990). From Etymology  to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge University Press, New York. 

Taft,  C.  (1997)  Color meaning  and  context:  comparisons  of  semantic  ratings  of  colors  on samples and objects. Colour Research and Application, 22, 40‐50. 

Takeshima, H.,  Suzuki,  Y.,  Sone,  T.  (1991).  Evaluation of  steady noise  from  a multidimen‐sional point of view.  Journal of Sound and Vibration; 151, 519–528. 

Taylor‐Clarke, M., Kennett, S., & Haggard, P.  (2004). Persistence of visual‐tactile enhance‐ment in humans. Neuroscience Letters, 354, 22–25. 

Teghtsoonian, R., & Teghtsoonian, M. (1970). Two varieties of perceived  length. Perception & Psychophysics, 8, 389‐392. 

Tinker, M.A. (1938). Effect of stimulus‐texture upon apparent warmth and affective value of colors. American Journal of Psychology 51,532‐535. 

Tomasello, M. (2003). Constructing a Language: A Usage‐based Theory of Acquisition. Cam‐bridge, MA: Harvard University Press. 

Turner, B.S. (1984) The Body and Society, Oxford: Basil Blackwell. Ullmann, S. (1963). Semantic universals. In: Greenberg, J. (Ed.) Universals of Language. MIT 

Press, Cambridge, MA, pp. 179‐207. Van Dale (2008). Van Dale Comprehensive Dictionary Dutch to English / Van Dale Groot Woor‐

denboek Nederlands / Engels, 4th edition, French & European Publs. Van der Helm, P.A. (2000). Simplicity versus  likelihood  in visual perception: From surprisals 

to precisals. Psychological Bulletin, 126, 770‐800. 

Page 215: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

215

Van Dijk, T.A., & Kintsch, W. (1983). Strategies of Discourse Comprehension. Academic Press, New York. 

Van Essen, D., Felleman, D., DeYoe, E., Olavarria, J., & Knierim, J. (1990). Modular and hierar‐chical organization of extrastriate visual cortex in the Macaque monkey. Cold Spring Harbor Symposia on Quantitative Biology 55, 679–696. 

Van  Ittersum, K., Pennings,  J.M.E., Wansink, B., & van Trijp, H.C.M.  (2007). The validity of attribute‐importance measurement:  A  review.  Journal  of  Business  Research,  60, 1177‐1190. 

Varela, F.J.  (1996). Neurophenomenology: A methodological  remedy  to  the hard problem. Journal of Consciousness Studies, 3 (4), 330–350. 

Verhagen, J.V., & Engelen, L. (2006). The neurocognitive bases of human multimodal food percep‐tion: Sensory integration. Neuroscience & Biobehavioral Reviews, 30(5): 613‐650. 

Verhoog, H., Matze, M., Lammerts van Bueren, E., and Baars, T. (2003) .The Role of the Con‐cept of  the Natural  (Naturalness)  in Organic Farming.  Journal of Agricultural and Environmental Ethics, 16 (1), 29‐49.  

Veryzer, R.W., & Hutchinson, J.W. (1998). The influence of unity and prototypicality on aes‐thetic  responses  to  new  product  designs.  Journal  of  Consumer  Research,  24(4), 374–394. 

Viberg, A. (1984). The Verbs of Perception: a Typological Study. In: Butterworth, B., Comrie, B., & Dahl, Ö. (Eds), Explanations for Language Universals. Mouton, Berlin, pp. 123–62. 

Villemure, C., Slotnick, B.M., & Bushnell, M.C. (2003). Effects of odors on pain perception: deci‐phering the roles of emotion and attention. Pain, 106, 101–108. 

Vroomen,  J., Bertelson, P.,   and de Gelder, B.  (2001). The ventriloquist effect does not de‐pend on the direction of automatic visual attention, Perception and Psychophysics, 63 (4), 651–659. 

Vygotsky, L.S. (1986). Thinking and Speech. New York: Plenum. Vygotsky, L.S. (1986). Thought and Language. Cambridge, MA: MIT Press. Waber, R.L., Shiv, B., Carmon, Z., & Ariely, D.  (2008). Commercial  features of placebo and 

therapeutic efficacy. Journal of Americal Medical Association, 299, 1016–1017. Wallace, M.T. & Stein, B.E.  (1997) Development of multisensory neurons and multisensory 

integration in cat superior colliculus. Journal of Neuroscience, 17, 2429–2444. Wallace, M.T.  (2004)  The  development  of multisensory  processes.  Cognitive  Processes,  5, 

69–83. Warde, A.  (1997). Consumption, Food and Taste: Culinary Antinomies and Commodity Cul‐

ture. London: Sage. Warde, A. (1999). Convenience Food: Space and Timing. British Food Journal 101(7): 518–27. Warren, D.H., & Cleaves, W.  (1971). Visual‐proprioceptive  interaction under  large amounts 

of conflict. Journal of  Experimental Psychology, 90, 206‐214. Weber S.T., Heuberger, E. (2008). The impact of natural odors on affective states in humans. 

Chemical Senses, 33, 441–47. 

Page 216: ANNA FENKO - TU Delft

REFERENCES 

 

216

Weimer, W.B.  (1977). A conceptual  framework  for cognitive psychology: motor theories of mind.  In: R. Shaw,  J. Bransford  (Eds), Perceiving Acting and Knowing: Toward an Ecological Psychology. Lawrence Erlbaum Associates, Hillsdale, NJ, pp. 267–311. 

Welch, R.B. (1999). Meaning, attention, and the "unity assumption" in the intersensory bias of  spatial  and  temporal  perceptions.  In  G.  Aschersleben,  T.  Bachmann,  &  J. Müsseler (Eds.). Cognitive contributions to the perception of spatial and temporal events. Amsterdam: Elsevier, pp. 371‐387. 

Welch, R.B., & Warren, D.H. (1986). Intersensory Interactions, in: K.R. Boff, L. Kaufmann, J.P. Thomas (Eds.), Handbook of Perception and Human Performance, Wiley, New York, pp. 4–5. 

Westerink, J., & Kozlov, S. (2004). Freshness in oral care: Attributes and time‐dependency of a multidimensional, dynamic concept. Journal of Sensory Studies, 19, 171–192. 

Whittlesea, B.W.A. (1993). Illusions of familiarity. Journal of Experimental Psychology: Learn‐ing, 19, 1235–1253. 

Whorf,  B.L.  (1956).  Language,  Thought  and  Reality:  Selected  Writings  of  Benjamin  Lee Whorf, ed. J.B. Carroll. MIT Press, Cambridge, MA. 

Willander, J., & Larsson, M. (2006). Smell your way back to childhood: Autobiographical odor memory. Psychonomic Bulletin & Review, 13, 240‐244. 

Williams,  J.  (1976). Synaesthetic adjectives: A possible  law of  semantic  change.  Language, 52, 461‐418. 

Williams,  L.E., & Bargh,  J.A.  (2008).  Experiencing physical warmth promotes  interpersonal warmth. Science, 322, 606–607. 

Willis, J., & Todorov, A. (2006). First impressions: making up your mind after a 100‐ms expo‐sure to a face. Psychological Science, 17, 592–598. 

Wilson, E.O. (1984). Biophilia: The human bond with other species. Cambridge, MA: Harvard University Press. 

Wilson, T.D., & Dunn, E.W.  (2004). Self‐knowledge:  its  limits, value and potential  for  improve‐ment. Annual Review of Psychology, 55: 493–518. 

Wojciszke, B., et al. (1998).  On the dominance of moral categories in impression formation. Personality & Social Psychology Bulletin, 24, 1245–1257. 

Wood, R.C. (1995). The Sociology of the Meal. Edinburgh: Edinburgh University Press. Wright, B. (1962). The influence of hue, brightness, and color on apparent warmth. American 

Journal of Psychology , 75, 232‐241. Wright, B., & Rainwater, L. (1962). The meaning of color. Journal of General Psychology, 67, 89‐99. Wundt, W.  (1874). Principles of physiological psychology. Titchener, E.B.,  trans.  London: Allen. 

Translation of Wundt, 1874. [New York, 1904] Yellott, J.I., & Kaiwi, J.L. (1979). Depth inversion despite stereopsis: The appearance of random‐

dot stereograms on surfaces seen in reverse perspective. Perception, 8, 135‐142.  Yeshurun, Y. & Sobel, N. (2010). An odor is not worth a thousand words: From multidimensional 

odors to unidimensional odor objects, Annual Review of Psychology, 61, 219–241. 

Page 217: ANNA FENKO - TU Delft

SENSORY DOMINANCE IN PRODUCT EXPERIENCE 

 

217

Zajonc, R.B.  (1968). Attitudinal effects of mere exposure.  Journal of Personality and Social Psychology, 9, 1–27. 

Zampini, M., &  Spence,  C.  (2004).  The  role  of  auditory  cues  in modulating  the  perceived crispness and staleness of potato chips. Journal of Sensory Studies, 19, 347‐363.  

Zampini, M., & Spence, C. (2005). Modifying the multisensory perception of a carbonated bever‐age using auditory cues. Food Quality and Preference, 16, 632‐641.  

Zampini, M., Guest, S., & Spence, C. (2003). The role of auditory cues in modulating the percep‐tion of electric tooth brushes. Journal of Dental Research, 82, 929‐932.  

Zarzo, M.  (2008). Psychological dimensions  in  the perception of everyday odors: pleasant‐ness and edibility. Journal of Sensory Studies, 23, 354–376. 

Zellner, D. A., & Durlach, P.  (2002). What  is  refreshing? An  investigation of  the  color and other sensory attributes of refreshing foods and beverages. Appetite, 39, 185–186. 

Zellner, D. A., & Kautz, M. A. (1990). Color affects perceived odor  intensity. Journal of Experi‐mental Psychology: Human Perception and Performance, 16, 391‐397. 

Zellner, D. A., Bartoli, A. M. & Eckard, R. (1991). Influence of color on odor identification and liking ratings. American Journal of Psychology, 104, 547‐561. 

Zellner, D.A., & Durlach, P. (2003). Effect of color on expected and experienced refreshment, intensity, and liking of beverages. American Journal of Psychology, 116, 633‐647. 

Zhang, S., & Schmitt, B. (1998). Language‐dependent classification: The role of classifiers  in cognition, memory and ad evaluation. Journal of Experimental Psychology: Applied 4, 375–85. 

Zhang, S., & Schmitt, B.  (2001). Creating  local brands  in multilingual  international markets. Journal of Marketing Research 38, 313‐325. 

Zhang, S., & Schmitt, B.  (2002). Activating sound and meaning  in brand name evaluations: The  role  of  language  proficiency.  In:  Bilinguals’ Differential  Processing. Working paper, UCLA Marketing Studies Center. 

Zhang. S., & Sood, S. (2002). “Deep” and “surface” cues: brand extension evaluations by chil‐dren and adults. Journal of Consumer Research, 29(1), 129–141. 

Zlatev, J. (2006). What’s in an image schema? Bodily mimesis and the grounding of language. In: Hampe, B. (Ed.), From Perception to Meaning: Image Schemas in Cognitive Lin‐guistics. Mouton de Gruyter, Berlin, pp. 313–342. 

Page 218: ANNA FENKO - TU Delft
Page 219: ANNA FENKO - TU Delft

219 

 

 

 

 

 

 

 

 

 

ACKNOWLEDGEMENTS  

PhD research is a long project that requires the help of many people. I would like to thank everybody who contributed to the success of my work.  

First of all, I am grateful to my promoter Paul Hekkert and co‐promoter Rick Schif‐ferstein. They invited me to work on this project and they always expressed confi‐dence  in  its success.  I  learned a  lot  from both of them. Paul  inspired me to think strategically, to focus on the big picture and to ask important theoretical questions. Rick  taught me  to pay attention  to methodological details and  to appreciate  the beauty of  clear definitions, well designed experiments  and precise conclusions.  I would  like  to  thank our dean Cees de Bont  for his help; without  it  this defense would have not taken place. I am grateful to Zoltan Rusak, Eelco van Stokkom, and Tanja Herbert for their advices and support.  

I did not do this research alone. Many students and employees of Industrial Design helped me to run experiments and collect the data.  I want to thank my graduate student Tse‐Chia (Joyce) Huang for her commitment to the “freshness” study. I am grateful  to Agnes Tan  for her help  in  the organization of the “naturalness” study, and to Dirk‐Jan Oudshoorn, Christiaan Maats and Reinoud Valks for their contribu‐tion to data collection and analysis for this study. I want to thank Roderick Huiijgen, Cees Jan Stam, Jeroen van der Liende, Merel Pick, and Charlotte Marsman for their contribution to data gathering  for the “warm” study, and Alexandra Bochaver  for gathering Russian data for the cross‐cultural study. I want to thank Anna Bernard, Kjille  Schutten,  Peter  van  den Herik, Alex Maliepaard  and Dennis Willemsen  for 

Page 220: ANNA FENKO - TU Delft

ACKNOWLEDGEMENTS 

 

220

their contribution to data gathering for the “noisy” study, and Mariska Rooijakkers, Katja Leuschner, Gaspard Bos, and Rein Pas for their help with preliminary experi‐ments. 

I am grateful to all my colleagues at the Department of Industrial Design who con‐tributed  to  the  creative atmosphere at  the  faculty. During  these  four years  I en‐joyed  sharing  the office with Kamiel, Gineke, Thomas, Bea, Gael, and Valentijn.  I want to thank all my roommates for the nice working environment they created. I am grateful to all members of Design Aesthetics section for our lively meetings on Monday mornings which filled me with energy for the whole week. I want to thank Cha Joong, Jaap, Sonia, Jackie and Liz for the great time we had together while or‐ganizing the PhD day. Special thanks to Gael for sharing recipes of delicious French dishes,  to Cha  Joong  for practical  tips on Dutch housing and  tax  systems, and  to Olga, my former colleague from Ukraine, for all the fun we had together. 

I am grateful to my family members, especially to Philipp, who took over my duties back home and made it possible for me to finish this project. I would like to thank my Russian friends, Julia, Tanya, Mark, and Inga, who helped me to stay connected with my cultural roots.  I also want to  thank my  friends  in  the Netherlands, Duke, Tanja, Pieter, Ron, and Erwin, who made  this country  feel  like home. Thank you, Roeslan, for your love and support, for your positive life attitude and for delicious dinners you cooked when I was working late. I am so happy we are together! 

 

ANNA FENKO 

Page 221: ANNA FENKO - TU Delft

221 

 

 

 

 

ABOUT THE AUTHOR  

Anna Fenko was born  in Moscow, Russia. She graduated as psychologist  (diploma with honors) and received a “candidate of psychology” degree from Moscow State University. Before moving to the Netherlands she was teaching consumer psychol‐ogy  at Moscow University  and  consulting  at  a  local marketing  agency.  She  com‐pleted several  research projects  funded by  international scientific  foundations.  In 2003‐2004 she was a Fulbright visiting scholar at  Jones School of Management at Rice University, Houston, USA. In August 2006 she started the PhD research at the Department of  Industrial Design of Delft University of Technology. The  results of this  project  are  presented  at  various  international  conferences  and  published  in Applied  Ergonomics,  Food  Quality  &  Preference, Materials  &  Design,  Journal  of Pragmatics, and Metaphor & Symbol. Her scientific interests include cognitive and affective drivers of  consumer behavior, multisensory  integration and  cultural dif‐ferences in product experience, and symbolic meaning of everyday products. 

 

Page 222: ANNA FENKO - TU Delft
Page 223: ANNA FENKO - TU Delft

223 

 

 

 

 

 

 

 

 

 

 

LIST OF PUBLICATIONS  

Fenko, A., Otten, J.J., & Schifferstein, H.N.J. (2010). Describing product experience in different languages: The role of sensory modalities. Journal of Pragmat‐ics, 42, 3314–3327.  

Schifferstein, H.N.J., & Fenko, A. (2010).  Investigating sensory dominance  in prod‐uct experience. Proceedings of the 7th Conference on Design & Emotion, 4‐7 October 2010, Chicago, Il, USA. 

Fenko, A., Schifferstein, H.N.J., & Hekkert, P. Noisy products: Does appearance mat‐ter? International Journal of Design, submitted. 

Fenko,  A.,  Schifferstein, H.N.J., & Hekkert,  P.  (2010)  Looking  hot  or  feeling  hot: What determines the product experience of warmth? Materials & Design, 31, 1325–1331.  

Fenko, A., Schifferstein, H.N.J., & Hekkert, P.  (2010). Shifts  in  sensory dominance between  various  stages  of  user‐product  interactions.  Journal  of  Applied Ergonomics, 41, 34–40.  

Fenko, A.,  Schifferstein, H.N.J., & Hekkert,  P.  (2010).  Can we  design  pleasurable products by combining pleasurable sensory properties? Proceedings of the International  Conference  on  Kansei  Engineering  and  Emotion  Research 2010, 2‐4 March 2010, Paris, France. 

Bochaver, A., & Fenko, A.  (2010). Metaphors  in happy and unhappy  life stories of Russian adults. Metaphor and Symbol, 25, 243–262. 

Page 224: ANNA FENKO - TU Delft

PUBLICATIONS 

 

224

Fenko, A., Schifferstein, Huang, T.‐C., & Hekkert, P.  (2009). What makes products fresh: The smell or the color? Food Quality and Preference, 20, 372–379. 

Fenko, A., Schifferstein, & Hekkert, P. (2009). Which senses dominate the product experiences of warmth and  freshness? Paper presented at  the 8th Pang‐born Sensory Science Symposium, 26‐30 July 2009, Florence, Italy. 

Fenko, A., Schifferstein, H.N.J., & Hekkert, P. (2008) Factors contributing to product experience: The cases of ‘warmth’ and ‘freshness’. Proceedings of the 6th Conference on Design & Emotion, 6‐9 October, 2008, Hong Kong. 

Fenko, A.,  Schifferstein, H.N.J., & Hekkert, P.  (2008) Which  senses dominate  the different stages of product experience. Proceedings of the 4th Design Re‐search Society Conference, 16‐19 July, 2008, Sheffield, UK.  

Fenko, A. (2008) Sensory dominance in product experience: the paradox of theory and practice. Paper presented at  the 7th Nordcode Seminar, 28‐30 May, 2008, Lund, Sweden.