1

STUDENTS’ GOAL PREFERENCES,

ETHNOCULTURAL BACKGROUND AND THE

QUALITY OF COOPERATIVE LEARNING

IN SECONDARY VOCATIONAL EDUCATION

Daphne M. Hijzen

2

STUDENTS’ GOAL PREFERENCES,

ETHNOCULTURAL BACKGROUND AND THE

QUALITY OF COOPERATIVE LEARNING

IN SECONDARY VOCATIONAL EDUCATION

PROEFSCHRIFT

ter verkrijging van

de graad van Doctor aan de Universiteit Leiden

op gezag van de Rector Magnificus Dr. D.D. Breimer,

hoogleraar in de faculteit der Wiskunde en

Natuurwetenschappen en die der Geneeskunde,

volgens besluit van het College voor Promoties

te verdedigen op dinsdag 19 september 2006

te klokke 15:00 uur

door

Daphne Minette Hijzen

geboren te ’s Gravenhage

in 1975

3

Promotiecommissie

Promotores: Prof. dr. M. Boekaerts

Prof. dr. P. Vedder

Referent: Dr. R. Martens

Overige leden: Prof. dr. G. Kanselaar (Universiteit Utrecht)

Prof. dr. A. E. M.G. Minnaert (Universiteit Groningen)

Prof. dr. W. C. M. Resing (Universiteit Leiden)

Dit proefschrift werd mogelijk gemaakt met financiële steun van de Nederlandse Organisatie

voor Wetenschappelijk Onderwijs (NWO).

4

Voorwoord

Bij de totstandkoming van dit proefschrift zijn vele mensen betrokken geweest. Tijdens mijn

promotieperiode aan de Universiteit Leiden heb ik behalve een proefschrift, ook een aantal

bijzondere vriendschappen gemaakt. Het gaat daarbij allereerst om mijn kamergenoten

Michiel en Marie-José. Ik wil hen ontzettend bedanken voor al hun gezelligheid, hun morele

en natuurlijk intellectuele ‘support’ die ik de afgelopen jaren dagelijks mocht ontvangen. Ook

mijn andere collega’s van de afdeling Onderwijsstudies hebben door hun betrokkenheid en de

gezelligheid, niet in de laatste plaats tijdens de congressen die we samen bezochten, mijn tijd

in Leiden tot een hele fijne tijd gemaakt. Daarnaast verdienen alle proefpersonen mijn dank.

Ik wil van de gelegenheid gebruik maken om alle scholen met hun docenten en leerlingen te

danken voor hun medewerking en gastvrijheid. Deze scholen waren: het Koning Willem 1

College te Den Bosch, het Mondriaan College te Delft, Landstede en het Deltion College te

Zwolle, het Albeda College te Rotterdam, het ROC ASA Zorg en Welzijn Hoogsticht te

Utrecht, het Alfa College te Hardenberg, Hoogenveen en Groningen, het ROC Nijmegen te

Nijmegen en Boxmeer, en ten slotte het Horizon College te Heerhugowaard. Voorts ben ik

mijn eerste scriptiestudente Lineke Witteman dankbaar voor haar grote inzet tijdens de

dataverzameling. Natuurlijk hebben mijn familie en vrienden met hun vriendschap en steun

bijgedragen aan dit proefschrift. In het bijzonder dank ik mijn ouders voor hun hulp, adviezen

en vooral hun vertrouwen in mij. Met Constant bracht ik eindeloze schrijfsessies door, die het

schrijven bijna tot een ‘feestje’ maakten. Daniël bedank ik voor zijn medeleven en zijn

bijdragen aan de nodige ontspanning op zijn tijd. Ten slotte ben ik erg blij dat ik Bas heb

leren kennen tijdens deze periode en dank ik hem omdat hij altijd lief is, met mij meeleeft en

mij stimuleert als dat nodig is.

Daphne M. Hijzen

Den Haag

5

CONTENTS

1. Introduction………………………………………………………………. 1

Secondary vocational education: Aim, problems and challenges ………... 2

Problems in secondary vocational schools……………………………….. 2

The Dutch educational system……………………………………………. 4

Structure of secondary vocational schools in the Netherlands………........ 5

Cooperative learning settings in secondary vocational education………... 6

Motivation and the quality of cooperative learning………………………. 7

Contextual factors and the quality of CL..................................................... 8

Aim and design of the study………………………………….................... 8

Research questions………………………………………………............... 9

Structure of the thesis……………………………………………………... 11

2. The relationship between the quality of cooperative learning,

students’ goal preferences and perceptions on contextual factors

in the classroom…………………………………………………………. 15

Abstract…………………………………………………………………... 15

Introduction………………………………………………………………. 16

Method…………………………………………………………………… 23

Results……………………………………………………………………. 25

Discussion and recommendations…………………………….………….. 29

3. Exploring the links between students’ engagement in cooperative

learning, their goal preferences and perceptions of contextual

factors in the classroom………………………………………………… 53

Abstract………………………………………………………………….. 53

Introduction……………………………………………………………… 54

Research question..……………………………………………………… 58

Method………………………………………………………………….. 58

Results…………………………………………………………………... 62

Discussion and recommendations………………………………………. 69

4. Instructing cooperative learning; teacher related conditions

6

steering effective cooperative learning processes of students in

secondary vocational education………………………………………… 83

Abstract…………………………………………………………………… 83

Introduction………………………………………………………………. 84

Research question………………………………………………………… 88

Method…………………………………………………………………… 88

Results…..………………………………………………………………... 90

Discussion and recommendations.……………………………………….. 95

5. Explaining cooperative learning in multi-ethnic classes; the limited

role of students’ ethnocultural background…………………………… 108

Abstract…………………………………………………………………… 108

Introduction……………………………………………………………….. 109

Method……………………………………………………………………. 116

Results…..………………………………………………………………… 118

Discussion and recommendations………………………………………… 122

6. General conclusions and discussion…………………………………… 134

Answering the general questions………………………………………… 134

General implications for theory and practice……………………………. 140

Limitations of the study…………………………………………………. 143

Summary………………………………………………………………… 147

Samenvatting (Dutch summary)………………………………………. 151

Curriculum Vitae (in Dutch)…………………………………………… 156

7

Chapter 1

INTRODUCTION

Nowadays, having a diploma is almost a necessary condition for having access to any job.

Therefore, good education is essential for the future of each and every adolescent. Society has

a major interest in well educated adolescents, to enhance the prosperity of a country.

Unemployed people cost society a lot because of unemployment benefits, but also because

unemployment leads to an increased risk for delinquency and health problems, which in

addition lead to major expenses for society. In the Netherlands full-time education is

compulsory for adolescents up to the age of 16. Partial education is compulsory up to the age

of 18. Recently, the National Council of Education advised to extend compulsory education

up to the age of 23 (National Council of Education, October 2005), for youth who did not

achieve a so-called starter-qualification, which is the presupposed minimum level of

competency development required for entering the job market. Some local initiatives to

implement this advice already have been initiated. In the city of Rotterdam adolescents are

forced to follow an educational program up to 23 years. In Amsterdam, adolescents under the

age of 23 do not receive unemployment benefits, unless they go back to school, or are

prepared to enroll in an on-the-job-training program. The project is so successful that a

discussion was started to consider applying the project to adolescents up to the age of 27.

These examples illustrate the importance that the government and local authorities assign to a

qualified population. Schools hope to contribute to the development of competencies which

adolescents need to be successful in modern society. Secondary vocational education is one

of the places where students should be prepared for the challenges entailed in the modern

knowledge society. This knowledge society requires schools to adapt their traditional

knowledge transmission goals to goals as learning to cooperate, teaching students to learn to

use the knowledge they are taught and prepare them for life-long learning. These new

educational aims coincide with the introduction of many new educational methods.

Cooperative learning methods are part of these. One of the building blocks of many of these

methods is the development of communication skills and practicing all sorts of social skills

such as the skills for communicating and networking. While working in cooperative learning

settings, students learn to use the knowledge they are taught, and at the same time practice

social skills that are required for working in teams. This thesis aims to uncover leads for

future interventions directed at improving students preparation for the new society through

8

cooperative learning (henceforth CL) methods in accordance with changing educational

goals.

Secondary vocational education: Aim, problems and challenges

The aim of secondary vocational schools is to bridge the gap between formal learning in

school and practice, between working and learning, in order to realize an optimal form of

knowledge circulation (Ministry of Education, Culture and Science, 2005). Legislation for

this type of schools was established in 1996 [Education and Vocational Training Act] (WEB,

1996). This act emphasized the value of the implementation of a nationwide qualification

structure aimed at the development of curricula that are mainly vocationally oriented

(Rozendaal, 2002). The value of having an important practical component in the curriculum is

emphasized in this act (Slaats, 1999). This act highlighted the development of a new

educational and instructional approach that prepares students for the wide range of

requirements being set by employers. In that sense, the function of education has shifted from

knowledge transmission to teaching students how to use knowledge as a tool and how to self-

regulate their learning process.

Apart from emphasizing the practical component of learning in secondary vocational

schools, the Ministry of Education, Culture and Science (2005) has recently recognized the

importance of social skills and coping with emotions as important for personal development.

In the labor market such skills are highly valued. The labor market is asking for employees

who are emotionally stable, who are able to manage their own careers, are able to cooperate,

communicate in teams, and cope with changes and conflicts.

Problems in secondary vocational schools

Secondary vocational education has to cope with severe problems; high drop-out rates

(Batenburg, 1998), motivation problems and maladaptive social behavior (Neuvel, 2004).

Approximately 40 % of students starting a study program is secondary vocational education

do not finish their study program (School Inspectorate, 2002). Several of these students start a

new study program at the same school or at another school. The number of students that

actually leave school without a diploma is not exactly known, but estimations vary between

12 and 25 %. Research has shown that a lack of motivation is often the reason for students to

quit school. Voncken, Van der Kuip, Moerkamp, and Felix (2000) assessed ‘push and pull

9

factors’ for students to quit their study program in secondary vocational education. The three

reasons that were mentioned most frequently by students all referred to a lack of motivation,

namely ‘the study program is not interesting’, ‘I do not like to go to school any more’, and

‘the study program lacks the connection with the job’.

A part of the explanation for these problems in secondary vocational education can

possibly be found in socioeconomic, ethnic and cultural background factors. The background

of the students in secondary vocational education often is not very favorable for completing a

successful educational and professional career. Many students have an immigrant

background, an educational history with little successes, and parents with a low education

level (Angenent, 1997; van den Dool & Janssen, 1994). Several researchers have shown the

relationship between economic deprivation and problem behavior (Bradley & Corwyn, 2002).

Especially immigrant families have - on average - a lower social economic background than

national families. Moreover, due to a relatively high rate of unemployment, immigrant

families in the Netherlands have to cope more with poverty. Another explanation for the

occurrence of maladaptive social behavior is related to the developmental phase of the

students in secondary vocational education. Most students are in their late adolescence and

several researchers have shown that maladaptive social behavior peaks in this phase (e.g.,

Compas, Hinden & Gerhardt, 1995; Maughan & Rutter, 1998; Moffitt, 1993). Moreover, as

will be explained in the following section, the problem is reciprocal. The reason why more

immigrant students are enrolled in this type of schools is also a consequence of culturally

dependent career courses. Some immigrant students already enter the Dutch educational

system with a lower educational level – mainly due to language difficulties - and the selective

Dutch educational system determines that they only have access to the lower sections of

vocational education.

10

The Dutch educational system

Figure 1 represents a broad image of the Dutch educational system.

Figure 1: The Dutch Educational System

The Dutch educational system is divided into three sections: primary, secondary and tertiary

education. School starts with primary education, which lasts eight years and starts for children

at the age of four (voluntarily) or five (compulsory) and ends at the age of 12. At the end of

Primary Education

Pre-university education Senior general secondary education

Pre-vocational education

University Higher professional education

Secondary vocational education

11

primary school children are advised as to which type of secondary education they should

pursue. Based on this advice they continue with pre-vocational education, senior general

secondary education or pre-university education. Pre-vocational education, which is the

lowest level, lasts four years and leads on to secondary vocational education. Senior general

education is a five-year program and leads on to higher professional education. The duration

of a pre-university study program is six years and leads on to university. Interestingly, only a

pre-university diploma gives direct access to university. However, there are alternative routes

to get there. Pre-vocational education has four levels and only the highest level gives access to

senior general secondary education, from which students can enter pre-university education.

Students in the two highest levels of secondary vocational education may continue studies in

higher professional education and from there they move on to university. However these are

very long and difficult routes and not many students take these. Most students choose the

direct route. The complexity of diversity is due to the selective nature of the Dutch

educational system. In other words, students’ study program is determined, basically, at the

age of eleven or twelve, when they receive the recommendation about their future education.

As a result of language problems, students with an immigrant background often have had a

learning delay in primary school. This implies that at the time they were tested, their

ethnocultural background resulted in a low advice. This selectivity seems particularly

detrimental for immigrant children. As compared to national students, they start their school

careers with limited national language proficiency and limited competencies in other skill

domains relevant for success in primary school. In recent years they started catching up,

however due to the selectivity of the school system they continue their school career at

secondary levels; levels that may reflect their actual levels of performance but not necessarily

their potential or competency. It will cost these students a lot of willpower to undo the

negative effects of the selectivity of the system.

Structure of secondary vocational schools in the Netherlands

Most of the students in secondary vocational schools are between 16 and 21 years old.

Secondary vocational school begins for most students, after they completed a pre-vocational

school. Senior vocational school is divided into four levels. At the first level students are

trained to become assistants (6 to 12 months). At the second level they follow two to three

year courses for basic vocational training. At the third level students are enrolled in

professional training and at the fourth level they participate in middle-management training (3

12

to 4 years) or in a specialized training course (1 to 2 years). Apart from the distinction of

levels, also a distinction in study type is important. In vocational schools a distinction is made

between theoretical vocational training and theoretical apprenticeship training. In the former

type students spend between 20 to 60 percent of their time at the working place while in the

latter type they spend at least 60 percent of their time at the workplace. In the present study

we mainly focused on students enrolled in the second type of program.

Most students finishing secondary vocational school prefer to enter the labor market and do

not proceed with further studies (for further information on secondary vocational education in

the Netherlands, see Euridice database on education, 2004 website). Approximately 435.000

youngsters in the Netherlands choose a vocational education program (website Ministry of

Education, Culture and Science), which is the highest percentage of all students. The second

largest proportion consists of students in higher professional education.

Cooperative learning settings in secondary vocational education

Because of changing educational goals and changing student and teacher roles that

accompany these developments, students need to become self-regulated learners. Nowadays

students are expected to be able to work autonomously, provide social support to their peers,

take responsibility for their own learning processes, and share resources. CL instruction

methods are designed to promote these capacities. Cooperative learning refers to a set of

instruction principles that together describe how students might learn from and with each

other and, through working together, accomplish academic tasks. The term usually refers to

alternative ways of organizing classrooms that contrast with individualistic and competitive

classroom organizations (Webb & Palinscar, 1996). We use this broad definition of CL,

because it captures the broad range of settings of CL in secondary vocational schools. CL

settings may promote students’ involvement with and motivation for school and learning as

well as facilitate integration and prevent discrimination, by functioning as an activity setting

where students are able to connect with each other and learn from each other's abilities and

skills. This seems particularly important for immigrant students in vocational schools. An

earlier study in the Netherlands showed that lack of a sense of belonging characterized by a

poor relationship with teachers and fellow students is an important reason for immigrant

students to quit their study program (Hofman & Vonkeman, 1995; Voncken, Van der Kuip,

Moerkamp, & Felix, 2000).

13

In this thesis we will distinguish between four important components that a successful

cooperative learning situation requires. In the first place, students need a number of

cooperation skills, such as the skills to be able to express their own opinion, stimulate each

other, provide and receive help, listen to each other, and clarify their understanding of the task

(Cohen, 1994; Ros, 1994; Webb & Palincsar, 1996). A second important component of

successful cooperative learning is that students perceive some sense of group cohesion. They

need to feel part of their CL teams and feel at ease with each other. Chin, Salisbury, Pearson,

and Stollak, (1999) and Cohen (1994) pointed out that the activity level in the group is at its

best when students feel at home in the group. Thirdly, there has to be a sense of

interdependence within the group (Johnson & Johnson, 1994; Ros, 1994). Team members

have to feel responsible for each other's learning process and in some way have to depend on

each other. Students in a team should rely on each other and provide each other with academic

and emotional support. Fourthly, students’ attitudes towards CL instruction methods should

be positive. Students need to be motivated in order to be successful cooperators.

Motivation and the quality of cooperative learning

In the preceding section we defined the quality of CL. In this section we focus on variables

that affect the quality of CL. The focal point of this dissertation is formed by two foci in

exploring CL processes in secondary vocational schools, namely goal preferences and

contextual factors.

Students need to be motivated in order to be successful cooperative learners. Without

the appropriate motivation, failing cooperative learning processes may be the result instead.

For instance, students might continue to work individually instead of in teams (Vedder, 1985;

Veenman, Kenter, & Post, 1999), or some students might reduce effort by letting other team

members do all the work (Gagné & Zuckerman, 1999), or simply disturb each others’ learning

(Salomon & Perkins, 1998: Shanahan, 1998). Forms of failing CL processes are explained by

ineffective motivational self-regulation strategies. We view motivational processes as an

intricate part of students’ self-regulation, namely that part that is steered by students’ values

and goal preferences. It is generally assumed that students’ steer their behavior in the

direction of valued goals and away from non-valued goals (Boekaerts, Pintrich, & Zeidner,

2000). Therefore, we expect that students’ motivation depends on the connection between

students’ personal goal preferences and the school goals, a person incorporation or acceptance

of school goals will have a positive impact on the quality of CL in schools, whereas an

aversive relationship between personal and school goals will likely result in problematic

14

school adjustment. When we relate this to the aforementioned four components of successful

CL, it is expected that only with a positive connection between personal and school goals,

students will invest in building a sense of group cohesion, in being supportive to create

interdependence and they will invest in learning the appropriate CL skills. Generally, their

attitudes towards CL will be positive. Hence, we expect that to a certain extent, students’ goal

preferences predict the quality of cooperative learning processes.

Contextual factors and the quality of CL

We expect that contextual factors will also predict the quality of learning processes to an

important extent. The way CL is organized, the way the teacher guides the CL process and

the availability of social support from teachers and peers will affect the quality of CL. Also,

characteristics that are related to students’ ethnocultural background are expected to influence

the quality of CL processes. Even though the school achievements of immigrant students who

were born in the Netherlands (second generation immigrants) has considerably improved over

the last 15 years, they are nevertheless not leveling Dutch students’ performance and are

characterized by higher levels of grade repetition and drop-out (Glenn & De Jong, 1996).

Many studies report a relationship between school performance and students’ ethnocultural

background, but only few studies analyzed the underlying processes that explain immigrant

students’ educational position in the Netherlands (e.g., Boekaerts, 1998; Teunissen &

Mathijssen, 1996). The present study addresses this latter question by exploring to what

extent and through what processes students’ ethnocultural background influences the quality

of students’ learning in settings inviting them to get involved in cooperative problem solving.

The focal point of this dissertation is therefore the exploration of cooperative learning

processes in secondary vocational schools from the point of view of students’ goal

preferences and contextual factors. Here, we want to emphasize that the study is not an

intervention study, but a descriptive one. In our view, it is a prerequisite of an intervention to

have a thorough understanding of the complexity of factors that determine students’

engagement in cooperative learning.

Aim and design of the study

The main aim of this thesis is to define the role of goal preferences in the quality of CL

processes and to identify factors in the classroom context that teachers can manipulate in

15

order to promote effective CL processes and to prevent forms of misregulation. Special

attention will be paid to differences between students that are related to their gender, program

type, and ethnocultural backgrounds. By gaining more insight into these relationships we

hope to be able to provide some leads for future interventions that aim at improving students’

motivation for CL and the quality of their learning processes as well to help in preventing

drop-out in the long run.

The first wave of data-collection took place from January till May of 2002. The

second wave took place from October 2002 till January 2003, and the third wave started in

April and ended in June 2003. During each wave of data collection students completed several

questionnaires. During the first data-wave teachers also filled in a short questionnaire.

Eighteen CL teams were video-taped and interviewed during the second and third data-wave.

During the first data-wave students were halfway their first year. The second data-wave took

place halfway their second year, and the third at the end of their second year.

Research questions

Four broad questions are central to this study, namely:

1. What is the relationship between the quality of CL and students’ goal preferences and

contextual factors in the classroom?

Do goal preferences add to the prediction of the quality of cooperative learning, or are other

variables, such as the way CL is organized (e.g., the way the teacher instructs the CL process

and the availability of social support) more important variables in predicting the quality of

Cooperative Learning? What is the effect of gender and program type on these relationships?

We predict that belongingness, social support and mastery goals are positively and superiority

goal preferences negatively related to the quality of CL. Furthermore, we hypothesize that the

students’ perception of the quality of CL will be poor when they score low on the context and

social climate variables.

2. How can effective CL teams be distinguished from ineffective ones, and what

distinguishes them in terms of the students’ goal preferences and perceptions of

contextual factors in the classroom?

16

We expect that in effective CL teams, students’ social and mastery goals will be dominant.

We expect that effective CL teams perceive contextual factors as promoting mastery and

social goals as these are challenging, hands-on, and promoting interdependency. These teams

are preferably not too big so that effects of social loafing are minimal and team members get

along and support each other in a beneficial way. Team members of effective teams are also

expected to evaluate their teachers as specialists who are guiding the learning process

sufficiently, and they evaluate the school climate as transparent and supporting.

In ineffective teams we expect that team members’ social, superiority and affective

goals are more important than their mastery goals. These students are expected to be

preoccupied with their well-being and therefore less involved with the learning process.

Furthermore, we expect these team members to be less involved in goal planning, and less

conscious of their goals than effective team members. We anticipate that members of

ineffective teams evaluate the group task as boring and too theoretical. They might

experience difficulties in getting along and perceive teachers as controlling or lacking

compassion.

3. Which teacher related conditions are related to the quality of CL processes, and are

these relationships stable in the course of a year?

We expect that the extent that students were taught skills, knowledge and rules for CL best

predicts the quality at all three data-waves best. Also, we expect to find differences between

effective and ineffective cooperators with respect to whether the stability of particular teacher

related conditions for CL is beneficial or not to their CL.

4. Can we distinguish between separate profiles of person variables (Dutch language

proficiency and goal preferences) and context variables (social resources and school

belonging) that account for variations in the quality of CL and does ethnic

background play a role in explaining differences in these profiles and the quality of

CL?

We expect students that have profiles characterized by high values on social support,

belongingness, and mastery goals, negative or low scores on superiority goals, high values on

Dutch language proficiency, high scores on perceived availability of teacher and peer support,

high scores on school and peer identification and negative scores on school and peer

17

alienation, to report high quality of CL and vice versa for students who report low quality of

CL.

Structure of the thesis

The rest of this thesis consists of four studies that read as articles. In the final chapter we will

formulate general conclusions, based on the findings of the four studies. The chapters were

intended as independent articles. Therefore each chapter starts with an abstract and a short

introduction about the relevant theories. Each chapter also has its own conclusions and

discussion section and list of references. Some of the texts overlap because of this reason.

The first Chapter deals with the first research question, namely the relationship between

students’ goal preferences, their perceptions of contextual factors and the quality of CL. In

this Chapter we will also present the investigation of the role of students’ gender and their

program type in predicting the quality of CL. This Chapter functions as a broad context for

the other three Chapters, because it deals with all the variables (except for the role of

students’ ethnocultural background) that were expected to be related to the quality of CL, and

are explored into more detail in the rest of the thesis. In Chapter 3, an in-depth study on

profiles of effective and ineffective CL teams is presented. Students were interviewed and

took part in a stimulated recall setting where they were interviewed about their CL processes.

This study serves as a qualitative clarification of the quantitative findings of the first study.

Those findings are contextualized into more detail, by focusing on real-life observations and

stimulated recall sessions. In Chapter four, the third research question is answered. The study

has a longitudinal design and deals with conditions for effective CL, in the course of a year.

We attempt to identify conditions that predict effective CL processes and conditions that

predict failing CL processes, in the course of a year. In Chapter 5, the role of students’

ethnocultural background in the relation between goal preferences and the quality of CL is

investigated. Chapter 6 deals with general conclusions and implications that are based on the

conclusions from the four articles.

18

References

Angenent, H. (1997). Criminaliteit van allochtone jongeren [Criminal behavior of

immigrant youngsters]. Baarn: Intro.

Batenburg, TH. (1998). De loopbaan na het MBO van voortijdig schoolverlaters [Drop

outs’ careers after secondary vocational school]. Groningen:GION.

Bradley, R. H., & Corwyn, R. F., (2002). Socioeconomic status and child development.

Annual Review of Psychology, 53, 371-399

Boekaerts, M. (1998). Do culturally rooted self-construals affect students' conceptualization

of control over learning? Educational Psychologist, 33, 87-108.

Boekaerts, M., Pintrich, P., & Zeidner, M. (2000). Handbook of Self-Regulation.

San Diego: Academic Press.

Chin, W., Salisbury, D., Pearson, A., & Stollak, M. (1999). Perceived cohesion in

small groups. Small Group Research, 30, 751-766.

Cohen, E. G. (1994). Restructuring the classroom: Conditions for productive small

groups. Review of Educational Research, 64, 1-35.

Compas, B. E., Hinden, B., & Gerhardt, C. (1995). Adolescent development:

Pathways and processes of risk and resilience. Annual Review of Psychology,

46, 265-293.

Eurydice database on education (n.d.). Retrieved November, 5, 2004, from

http://www.eurydice.org.

Gagné, M., & Zuckermann, M. (1999). Performance and learning goal orientations as

moderators of social loafing and social facilitation. Small Group Research, 30,

524-541.

Glenn, Ch., & De Jong, E. (1996). Educating immigrant children: Schools and

language minorities in twelve nations. New York: Garland.

Hofman, R., & Vonkeman E. (1995). Condities voor adaptief onderwijs: De

Onderwijsmethode [Conditions for adaptive education: the instructional method].

Groningen: GION/RUG.

Johnson, D. W., & Johnson, R. T. (1994). Learning together and alone: Cooperative,

competitive, and individualistic learning. Englewood Cliffs, NJ: Prentice-Hall.

Maughan, B., & Rutter, M. (1998). Continuities and discontinuities in antisocial

behavior from childhood to adult life. In Advances in Clinical Child

Psychology (eds T. H. Ollendick & R. J. Prinz), pp. 1-47. New York: Plenum.

19

Ministry of Education, Culture and Science (2005). Retrieved November, 5,

2005, from http://www.minocw.nl/bve/informatiebo.html#organisatie

Moffitt, T. E. (1993). Adolescence-limited and life-course persistent antisocial

behavior: A developmental taxonomy. Psychological Review, 100, 674-701.

National Council of Education (2005). Retrieved November, 5, 2005, from

http://www.onderwijsraad.nl/pdfdocs/persbericht_werk_en_ervarings

onderwijs_voor_jongeren_tot_23_jaar.pdf

Neuvel, J. (2004). Veiligheidsmonitor 2004 ROC veilig [Safety monitor 2004 ROC safe]. Den

Bosch: Centrum voor Innovatie van Opleidingen (CINOP).

Ros, A. A. (1994). Samenwerking tussen leerlingen en effectief onderwijs [Cooperation

between students and effective education]. Dissertation,

Rijksuniversiteit Groningen.

Rozendaal, J. S. (2002). Motivation and information processing in innovative secondary

vocational education. Leiden University: Doctoral dissertation.

Salomon, G., & Perkins, D. N. (1998). Individual and social aspects of learning. In: P.D.

Pearson & A. Iran- Nejad (Eds.), Review of Research in Education (pp.1-24).

Washington: American Educational Research Association.

School Inspectorate (2002). Vroegtijdig schoolverlaten in het middelbaar beroepsonderwijs

[Drop-out in secondary vocational schools]. Utrecht: Inspectie van het Onderwijs.

Shanahan, T. (1998). On the effectiveness and limitations of tutoring in reading. In : P.D.

Pearson & A. Iran-Nejad (Eds.). Review of research in education, vol. 23 (1-24).

Washington AERA.

Slaats, A. (1999). Learning styles in secondary vocational education: disciplinary

differences. Learning and Instruction, 9, 475-492.

Teunissen, J., & Matthijssen, M. (1996). Stagnatie in onderwijsonderzoek naar de etnische

factor bij allochtone leerlingen [Stagnation in educational research on the ethnic

factor amongst immigrant students] Sociologische Gids, 43, 87- 99.

Van den Dool, P., & Janssen, A. (1994). Schoolloopbaanmanagement en – begeleiding in

het secundair beroepsonderwijs [School career management and – guidance in

secondary vocational schools]. In P. Stijnen, H. Münstermann, N. Deen & J. van

Kuijk (Red.), Management en begeleiding van schoolloopbanen. Alphen a/d Rijn:

Samsom, Tjeen Willink.

Vedder, P. (1985). Cooperative learning: A study on processes and effects of cooperation

between primary school children. Dissertation, Rijksuniversiteit Groningen.

20

Veenman, S., Kenter, B., & Post, K. (2000). Cooperative learning in Dutch primary

classrooms. Educational Studies, 26, 281-302.

Voncken, E., Van der Kuip, I., Moerkamp, T., & Felix, C. (2000). Je bent jong en je weet niet

wat je wilt: Een inventarisatie van push- en pullfactoren die leiden tot voortijdig

schoolverlaten in de BVE-sector [You are young and you do not know what you want:

an inventory of push- and pull factors that lead to drop-out in secondary vocational

education]. Amsterdam: SCO-Kohnstamm.

Webb, N. M., & Palincsar, A. S. (1996). Group processes in the classroom. In: D. C. Berliner

& R. C. Calfee (Eds), Handbook of Educational Psychology (pp. 841 - 873). New

York: Simon & Schuster MacMillan.

21

Chapter 2

THE RELATIONSHIP BETWEEN THE QUALITY OF COOPERATIVE

LEARNING, STUDENTS’ GOAL PREFERENCES AND PERCEPTIONS OF

CONTEXTUAL FACTORS IN THE CLASSROOM

1

Abstract

This study examined relationships between the quality of cooperative learning (CL) and

students’ goal preferences and perceptions of contextual factors in the classroom. Subjects

were 1,920 students in secondary vocational schools. The study focused on four different

types of goals: social support, belongingness, mastery, and superiority goals. It was found

that social support goals had the strongest relation with the quality of CL. Further we found

that the quality of CL was best predicted by a combination of social support goals,

evaluations of the extent that students were taught cooperation skills, perception of teacher

monitoring behavior, and the availability of academic and emotional peer support. Female

students’ preferences for mastery and social goals were stronger than those of male students,

whereas male students had a stronger preference for superiority goals. Program type

functioned as a moderator variable within the relation of students’ superiority/ individuality

goals and the quality of CL.

Key words: motivation, cooperative learning, contextual factors, vocational education.

1 This chapter is based on: Hijzen, Boekaerts and Vedder (2006). The relationship between the quality of cooperative learning, students’ goal preferences and perceptions of contextual factors in the classroom. Manuscript published in the Scandinavian Journal of Psychology, 47, 9 - 21.

22

INTRODUCTION

This paper reports a study into the relationship between motivational processes, contextual

factors and the quality of cooperative learning (CL) processes of adolescent students in

secondary vocational education in the Netherlands. We view motivational processes as an

intricate part of the students’ self-regulation process, namely that part that is steered by their

values and goal preferences. It is generally assumed that students steer their behavior in the

direction of valued goals and away from non-valued goals (Boekaerts, Pintrich & Zeidner,

2000). This is not to say that students are working with a clear goal dichotomy in mind;

personally valued and non-valued goals. Rather, our position is that many goals are located in

between these two extremes. Indeed, students are presented with multiple goals in the school

context. Some students might classify these goals in terms of desirable and undesirable ones

but for the majority of students the classification process might be more complex. Several

researchers, such as Deci and Ryan (1985) and Ryan and Deci (2000) argued that most

students will determine to what extent school goals are similar to – or might be combined

with – personally liked goals; they try to bridge the span between imposed and personal goals,

by “personalizing” the former type of goals (e.g., Lemos, 2002). We assume that students’

motivation levels at school largely depend on their perceptions of the connection between

their personal goals and the school goals.

We expect that the students’ perception of the quality of CL depends to a large extent

on the goal preferences that they bring into the classroom. On the other hand we expect that

their perception of the classroom context itself, and more specifically the way they perceive

teacher behavior and the support they get from teacher and peers, determines the quality of

CL as well. For example, we expect that the students’ perceptions of the extent to which they

were taught cooperation skills (How information) and the social reasons they consider

important for CL (Why information) will affect how they appraise the quality of CL. For

future intervention purposes, attention to how students perceive the CL setting is of prime

importance. Our position is that, although students’ goal preferences have a large impact on

their perception of the quality of CL, it is difficult to influence their goal preferences in a

short period of time. By contrast, information about the contextual factors that influence

students’ perception of CL may provide researchers with useful information to hand down to

teachers and trainers. We realize that adaptations to classroom settings are much easier to

generate than changes in students’ goal preferences.

23

In this paper, we attempt to expand the focus of goal preferences from the

achievement domain to the social domain, acknowledging the large role played by students’

perception of the social context. The article is organized into three main sections. First, we

describe the quality of CL as the general beliefs students have about the reasons for learning

with and from each other and their awareness of how they have to go about learning in the CL

setting. Second, we describe the relation between goal preferences and CL. In the third section

we describe how goal preferences and perception of contextual factors is conceptualized in

the present study and report on the results.

---------------------------------------

Insert Figure 1 about here

---------------------------------------

The quality of cooperative learning

CL is not just a learning theory or a teaching method, it refers to a set of instructional

principles that together describe how students might learn from and with each other and,

through working together, accomplish academic tasks. Successful CL situations require in the

first place that students have positive beliefs about CL. In order to feel responsible for group

learning students also need to be aware of the skills that should be used and have easy access

to these skills. For example, students should make use of a number of cooperation skills,

including the skill to express their own opinion, stimulate each other, provide and receive

help, listen to each other and clarify their current understanding of the task (Cohen, 1994;

Ros, 1994; Webb & Palincsar, 1996). Furthermore, students need to feel responsible for each

other’s learning process and experience a sense of group cohesion and interdependence

(Johnson & Johnson, 1994; Ros, 1994; Webb & Palincsar, 1996). Chin, Salisbury, Pearson

and Stollak (1999) and Cohen (1994) pointed out that the activity level in the group is at its

best when students feel at home in the group. Based on a literature review we defined the

students’ perception of the quality of CL in terms of their perception of the group cohesion

and their own skills to participate successfully in CL.

It is easy to imagine that the shift of responsibilities from teacher to students that is

implied by CL may come with a variety of problems when students lack a positive attitude

towards CL or the skills to work together. For example, students may take the opportunity to

work alone instead of together (Vedder, 1985; Veenman, Kenter & Post, 2000), they may

24

disturb each other’s learning processes (Salomon & Perkins, 1998; Shanahan, 1998) or reduce

effort, resulting in lowered levels of engagement (Gagné & Zuckerman, 1999). We assume

that these and similar problems come about because the students perceive the quality of CL as

suboptimal.

Goal preferences and cooperative learning

We divided the studies that examined the relationship between students’ goal preferences and

the quality of the learning process into two categories. The first category examined the

relationship between one type of goal, namely achievement goals, and the quality of the

learning process. The second category focused on the relationship between multiple goals and

the quality of the learning process.

Several researchers took the mastery vs. performance dichotomy as their frame of

reference (e.g., Ames & Archer, 1988; Dweck, 1986; Urdan, 1999). These studies

documented that students who are mastery-oriented engage in tasks because they want to

acquire new knowledge and skills; their purpose is to develop competence. Performance

oriented students on the other hand, want to demonstrate competence relative to others.

Boggiano, Main, Flink, Barrett, Silvern and Katz (1989) and Dweck (1986) suggested that

students who have a strong preference for performance goals might easily run into problems

when they have to cooperate. These students might interpret unsolicited help and support as a

threat to their ego, leading to avoidance of CL situations. Functioning as a group member may

contrast with their wish to perform well at a more individual level (see also Schwartz & Bardi,

2001). By contrast, mastery goals are associated with high levels of performance on

personally challenging tasks in general. Students who pursue mastery goals are not focused on

out-besting their peers, they are academically oriented and want to learn something new, even

when it implies a lot of effort. Although many studies have been conducted on the

relationship between these two goal orientations and learning (for review, see Pintrich, 2000),

it is still unclear how these goal preferences are interrelated in CL contexts. In part this is due

to the fact that most of the reported studies dealt with learning in general rather than with

learning in CL settings.

A more complex perspective on goals preferences was adopted in the second category

of studies. Several researchers (Boekaerts, 1998, Dowson & McInerney, 2001; Ford, 1992;

Ford & Nicholls, 1991; Lemos, 1996; Wentzel, 1996) argued that students bring different

types of goals to the learning situation. In addition to achievement goals, students pursue

25

entertainment goals (e.g., I want to have fun at school), self determination goals (e.g., I want

to determine myself how I do things), working goals (e.g. I want to finish that task)

belongingness goals (e.g., I want to make many friends) and social support goals (e.g., I want

to provide help to peers).

Urdan and Maehr (1995) argued that social goals concern the social reasons for trying

to achieve in academic situations and consequently these goals play a crucial role in a CL

setting. Social goals are important to children of all ages (Ford, 1992), particularly to

adolescents who often consider these goals to be more important than academic learning goals

(Covington, 2000). In this respect, McInerney, Hinkley, Dowson and Van Etten (1998)

suggested that a combination of mastery (academic) and social goal orientations might be

more productive than mastery goals alone because feelings of belongingness and social

responsibility engendered by social goals provide added impetus for academic achievement.

Wentzel’s (1991) studies clarified the effect of social goals on learning. She showed that pro-

social behavior is positively associated with academic success (Ford, 1992; Wentzel, 1993;

1994) and that CL facilitates goal realization for those students who like to work in CL

settings and value group cohesion. Likewise, Connell, and Wellborn (1991) and Wentzel

(1994) suggested that a sense of belongingness facilitates the adoption of the goals that are

valued by the social group to which one belongs. The desire of individuals to achieve for the

sake of the group is a well-known phenomenon, and it forms the basis for much of the success

of CL (Hertz-Lazarowitz & Miller, 1992).

Goal preferences and perception of contextual factors in the present study

We adjusted Ford’s taxonomy (Ford, 1992; Ford & Nicholls, 1991) to measure a broad range

of intra- and interpersonal goals. In the present study, we limited the focus to four types of

goal preferences that are central to CL settings, namely mastery goals (e.g., I want to learn

about my future profession), superiority/individuality goals (e.g., I want to impress my peers),

social support goals (e.g., I want to help classmates with their tasks), and belongingness goals

(e.g., I want to make many friends). In line with Hickey and Granade (2004) and Urdan

(1997) we assume that the environment exerts a major influence on the salience of a particular

goal and its adoption. Boekaerts, De Koning, and Vedder (in press) reviewed studies that

examined the relationship between contextual variables and goal preferences. They listed the

context factors that play an important role in the classroom. We based our selection of

contextual factors on this review as well as on reviews of studies on CL (e.g. Cohen, 1994;

26

Webb & Palincsar, 1996). More specifically, in this study we concentrate on instructional

characteristics, such as the type of task, the type of evaluation/rewards, teacher instruction

behavior, teachers’ clarity on rules for CL, and students’ evaluations of the extent that they

were taught CL skills. We also measured the students’ perception of school climate, including

their perceptions of the availability of teacher’s academic and emotional support, and the

availability of peer academic and emotional support. Research on instructional characteristics

revealed that it is essential for effective cooperation that the task elicits positive

interdependence (see Cohen, 1994). This implies that students should perceive the task as

challenging, but not too complex, and that group assignments are structured in such a way

that each group member’s actions relate to and are required for task completion. The role of

reward in CL is not altogether clear yet. For example, Slavin (1995; 1996) concluded in a

meta-analysis that the effects of CL on students’ achievement are maximal and the risk at

social loafing are minimal, when a group reward is combined with individual accountability

for learning and learning outcomes. Other researchers (e.g., Cohen, 1994; Deci & Ryan, 1985)

suggested that a combination of a group reward and an individual reward will undermine the

group process.

Teacher (instruction) behavior has proved to be an important factor in several studies.

Teachers should facilitate students to complete the group assignments increasingly by

themselves. They also need to monitor their students’ learning process and intervene when

necessary to provide assistance or to model students’ social skills (see Johnson & Johnson,

1994), especially when students are not yet used to cooperating. Students prone to off-task

behavior should be monitored in particular. Drawing attention to the teacher’s role in CL

settings, Webb and Palincsar (1996) illustrated that in order to promote CL, teachers should

not only define the group assignment adequately; they should also be clear about the rules for

CL; i.e., teach the required concepts and strategies and give the criteria for success (see also

Johnson & Johnson, 2002). Webb and Palincsar described comprehensive programs of team

building and prosocial skill development that improve peer-to-peer interaction and through it

students’ social goals. Many other scholars (e.g., Gillies & Ashman, 1996; Hoek, Van den

Eeden, & Terwel, 1999; Webb & Farivar, 1994) have shown that explicit teaching of CL skills

coincides with an improvement of the quality of CL.

The quality of CL is also promoted by a social climate that is characterized by optimal

academic and emotional support from teacher and peers. Wentzel (1994) and Wentzel and

Wigfield (1998) showed that a supportive social climate promotes group cohesion, the use of

cooperation skills, and students’ attitude towards CL. In such a climate students feel respected

27

and supported when asking for help. It has also become increasingly clear that a sense of

relatedness with the teacher promotes pro-social behavior, particularly adaptive help-seeking

behavior (Brenner & Salovey, 1997; Newman & Schwager, 1993) and the pursuit of social

support goals. Students experiencing autonomy support and optimal structure were more

likely to be effortful and persistent while completing learning tasks. Our prediction is that

students’ perception of the availability of academic and emotional support from peers

crucially affects their perception of the quality of CL. A learning environment characterized

by social resources will give students confidence that they can rely on each other for support

with their school work. In this study we are dealing with adolescents in vocational education.

Adolescents must adjust to peer pressure and norms not only with respect to academic

performance but also in relation to interpersonal rules for help seeking, helping others, turn

taking, and sharing resources. Accordingly, we anticipate that peers will play a larger role

than teachers when it comes to turning for assistance.

Gender differences

Early studies (e.g., Gardner, 1993; Gardner, Mason & Matyas, 1989) suggested that girls

benefit more from cooperative classroom settings than boys. Several studies, among others

those of Anderman (1999), Charlesworth & Dzur (1987) and Cosden, Pearl, and Bryan (1985)

revealed that girls are more inclined to engage in behavior associated with successful CL,

such as helping others, verbal organization, and turn taking. A recent report from the

Organization for Economic Cooperation and Development (2003) also indicates that girls are

– in general – more interested in CL than boys. Eccles (1987) and Wentzel (1991) showed

that female students, as compared to their male contemporaries, prefer to learn in settings in

which they can combine mastery (understanding tasks) and social goals (being with friends,

supporting others, creating a sense of belonging and security). Voncken, van der Kuip,

Moerkamp, and Felix (2000) showed that the way female students experience school is

related to feelings of group cohesion. All these studies imply that female students rate the

quality of CL processes higher than male students. In contrast, school is perceived by male

students as a competitive arena, which makes social comparisons and peer pressure dominant

in their mental representation of the learning situation. Severiens and Ten Dam (1998)

conducted a meta analysis and reported that male students scored higher on a non-academic

orientation than female students and that male students scored higher than females on

superiority/ individuality goals and lower on both types of social goals. Based on the

28

literature, we expect female students to show higher scores on belongingness and social

support goals than male students. We also expect differences in the extent to which male and

female students pursue mastery goals. More specifically, we predict that male students will

have a lower rating of the quality of CL. In short, we expect gender to function as a moderator

variable in the relationship between students’ goal preferences and their perception of the

quality of CL.

Program type differences

As far as we know, no specific research has been done using different program types. Four

program types are represented in our sample, namely information and communications

technology (ICT) and engineering, retail and administration, health and welfare, and food and

tourism programs. It is important to note that male and female students are not equally

distributed over these program types and that this uneven distribution might lead to a program

type effect that masks an underlying gender effect. Therefore, we will explore program type

effects for male and female students separately. Learning how to take care of others is an

important aspect of the health and welfare program and students who enroll in this program

consider “care” as an important aspect of their future job image. It comes as no surprise that

girls are over-represented in this program. We expect that students enrolled in the health and

welfare program show a preference for both types of social goals. By contrast, ICT and

engineering students look forward to a professional career in a company where they are paid

well. They imagine their future in terms of an adventure in the world of bonuses and free

company cars.

Boys are over-represented in these program types. We expect these students to be

oriented more towards superiority/ individuality goals than to social goals. We did not have

clear expectations in relation to the two other program types, albeit that we expected food and

tourism students to report a higher preference for social goals than the ICT/engineering

students due to the alleged lower social orientation of the latter type of students. Health and

welfare students are expected to report higher perceptions of the quality of CL, because these

professional groups are seen as more socially oriented. We anticipate that ICT and

engineering students score lower on the quality of CL. In line with the expected differences in

the scores on the quality of CL and goal preferences, we expect the relationship between

students’ goal preferences and the quality of CL to differ between the program types,

particularly between the health and welfare and ICT/engineering programs.

29

In summary, we will explore the relationship between students’ goal preferences and

the quality of CL. We predict that belongingness, social support and mastery goals are

positively related and superiority/individuality goal preferences negatively related to the

quality of CL. Our second research question pertains to the relationship between the quality of

CL and perceptions of contextual factors in the classroom, including social climate. We

predict that the students’ perception of the quality of CL will be perceived as poor when

students score low on the context and social climate variables. We will also explore gender

and program type effects on the relationship between students’ goal preferences and the

quality of CL. We predict that female students score higher than male students on the quality

of CL and that female students report higher preferences for social support and mastery goals

whereas male students report higher preferences for superiority goals. As far as program type

effects are concerned, we hypothesized that health and welfare students and food and tourism

students score higher on the quality of CL and on social goals, particularly in comparison with

ICT and engineering students. We will examine whether gender and program type moderate

the relationship between goal preferences and perceptions of the quality of CL.

METHOD

Subjects

The present study is part of a larger project on motivational self-regulation in secondary

vocational high schools. Participants in the study were 1920 first-year students from 11

different secondary vocational schools in the Netherlands. The Netherlands has 42 regional

educational centers for secondary vocational education. They all received a letter in which we

explained the purpose and relevance of the study and invited them to participate. Eleven

schools responded positively. The other schools had a variety of reasons for not participating.

The most frequent reasons concerned the time investment of students and teachers and the

extra organizational burden of participation in a large research project. The eleven schools

that participated were spread evenly across the Netherlands. The students’ age ranged from 15

to 55 years and 5 months with an average of 18 years and 1 month (SD = 3.56 years). About

18% of the participating students had an immigrant background (defined in terms of either the

students’ own country of origin or the country of origin of at least one of the parents). Table 1

shows the distribution of participating students by gender and program type. Most students

were enrolled in health and welfare programs. More than three quarters of the health and

welfare students were female. Relatively few students were enrolled in engineering and ICT

and these students were predominantly male.

30

-----------------------------

Insert Table 1 and 2 about here

------------------------------

Instruments

Students were invited to complete several self-report questionnaires. Here, we focus on

students’ goal preferences, students’ perception of contextual factors in the classroom and the

quality of CL. Data collection took place in the second semester of the students’ first year.

Table 2 presents an overview of scales, sample items and Cronbach’s alphas of the different

scales used in this study. Students’ personal goals were assessed with the goal preference list

based on the Ford (1992) and Ford and Nichols (1991) taxonomy of broad goals. Students had

to report on the importance they attach to each of the goals by giving an indication of the

extent to which they want to achieve them. They were asked to choose from five response

categories ranging from “not at all” to “very much so”. Four goal domains were highly

relevant for the quality of CL: superiority and individuality goals, mastery goals,

belongingness goals and social support goals. The students’ perception of the quality of CL

was measured with the questionnaire for the Quality of Cooperative Learning (QCL).

Originally the list comprised four subscales, namely students’ perception of the quality of

group cohesion, which was made up of seven items, Cronbach’s alpha was 0.86. The second

subscale measured interdependence within the group, and had 7 items, Cronbach’s alpha was

0.83. The third subscale measured students’ perception of the quality of their cooperation

skills, this subscale was made up of 10 items and Cronbach’s alpha was 0.83. The fourth

subscale aimed to measure students’ attitude towards CL, it contained 8 items and Cronbach’s

alpha was 0.75. These subscales were highly correlated and were all part of the quality of CL.

A Principal Component Analysis on these four subscales resulted in one-factor solution. This

factor had an Eigenvalue of 1.8 and it explained 58% of the total variance. Sample items were

“I perceive myself as part of this group”, “When we work on a group task, we make sure that

all the team members understand the answers”, “I know when another person needs help” and

“Together you learn better than alone”. Students had to indicate on a four-point Likert scale to

what extent they agreed with each statement. Response categories ranged from “I disagree

very strongly” to “I agree very strongly”.

Students’ perceptions of contextual factors in the classroom were measured with a

questionnaire registering the Conditions for CL (CCL). This questionnaire measures students’

31

perception of the extent that teachers create or maintain conditions for the quality of CL.

Items are mainly based on reviews of studies on CL (e.g., Cohen, 1994; Webb & Palincsar,

1996). Four single items (response categories: 1= yes, 2 = no) concerned students’

perceptions of the type of task; these were about task difficulty, task challenge, the time for

the task and the need to consult group members in order to finish the task. One single item

concerned the type of reward students received after finishing the group task. The CCL

further measured the students’ evaluation of the extent that they were taught skills and

knowledge for CL at their present schools, rules for CL, and teacher behavior in relation to

CL. This latter scale consists of three subscales focusing on teachers’ monitoring behavior,

interventions and evaluations. Students had to report on a four-point Likert scale to what

extent they agreed with each item (1 = I completely disagree, 4 = I completely agree).

Four scales derived from the Questionnaire for Social Support (Boekaerts, 1987;

Vedder, Boekaerts & Seegers, 2005) measured students’ perceptions of the availability of

academic and emotional support from their teachers, and perceptions of the availability of

academic and emotional support from their peers. Students had to indicate how often their

teachers or peers provided them with emotional and academic support. Response categories

(4) ranged from “almost never” to “very often”.

Procedure

The questionnaires were administered during regularly scheduled lessons and the students

were instructed and supervised by the researchers. Each student received a personal code,

meaning that answers remained confidential. It took students two sessions of 45 minutes to

complete all the questionnaires. These sessions were spread over two different days, which

explains the different sample sizes. Many students failed to attend classes at both sessions.

Some students simply refused to cooperate with us the second time or to fill in the entire

questionnaires. The drop-out was therefore unsystematic.

RESULTS

Students’ goal preferences and the quality of cooperative learning

------------------------------------------------------------

Insert Table 3 about here

---------------------------------------------------------

32

Table 3 presents mean scores, standard deviations and correlation coefficients for the four

scales of the goal questionnaire, the scales for students’ perceptions of contextual factors in

the classroom and the scales for students’ perceptions of the social climate for the whole

sample. The table shows that students were quite optimistic about the quality of the CL

processes and that mastery was the most important goal domain among all students.

Belongingness and social support goals were also rated as important goal domains.

Superiority/individuality, however, was the least important goal domain.

We expected that students who value social and mastery goals perceive the quality of

CL as high and that students who value superiority/individuality goals perceive it as low.

Results presented in Table 3 indeed show that attaching importance to both mastery and social

goals relates positively to the quality of CL. Social support goals showed the highest

correlation coefficient (r 1339) = 0.33, p = .000). The correlation with belongingness goals (r

(1281) = 0.23, p = .000) and mastery goals (1262) = 0.23, p = .000) was slightly lower.

Superiority/individuality goals were not significantly related to the quality of CL, although in

contrast with our prediction, the correlation coefficient was not a negative one.

Students’ perceptions of contextual factors and the quality of CL

Apart from a positive relationship between students’ social and mastery goal preferences and

the quality of CL, we also predicted that the quality of CL would be related to students’

perceptions of contextual factors in the classroom, as defined by the type of task, reward

systems, students’ evaluations of the extent to which they were taught CL skills, teachers’

clarity on rules for CL, teachers instruction behavior, and aspects of social climate. Inspection

of significant correlation coefficients (above 0.20) between students’ perceptions of

contextual factors in the classroom revealed that students’ evaluation of the extent that they

were taught CL skills at their present schools was positively related to the quality of CL (r

(1465) = 0.35, p = .000) and so was the students’ perception of the teachers’ clarity on rules

for CL (r (1416) = 0.24, p = .000) and monitoring behavior (r (1453) = 0.20, p = .000). A

closer look at the social climate scales showed that our expectations were confirmed. Both

perceived availability of peer academic (r (1343) = 0.28, p = .000) and emotional support (r

(1336) = 0.30, p = .000) were related to the quality of CL. The scales for perception of the

availability of teacher support were also related to the quality of CL (r (1327) = 0.21, p =

.000).

33

Gender differences

---------------------------------

Insert Table 4 about here

----------------------------------

We expected gender and program type differences for students’ goal preferences and the

quality of CL. However, as explained in the Method section male and female students were

not evenly distributed over the different program types. For the following analyses we

excluded ICT/engineering students, because only 6% of these students were females. Table 4

shows means, standard deviations and F -values for the four goal preference subscales and the

quality of CL, for male and female students separately.

Univariate analyses showed that the main gender effects concern social support (F

[1,1340] = 35.61, p = .000, �2 = .03), superiority (F [1,1274] = 29.74, p = .000, � 2 = .02) and

mastery goals (F [1,1249] = 9.53, p = .002, � 2 = .01). As predicted, male students’ scores are

significantly higher in the superiority goal domain, whereas female students show higher

scores for social support goals and mastery goals. No significant differences were found for

belongingness goals. The difference for male and female students on the quality of CL

subscale is small but statistically significant (F [1, 1207] = 12.66, p = 0.000, � 2 = .01). Girls’

scores are slightly higher than boys’ scores. We examined the correlation coefficients that

were significant at the p < 0.01 level using Fisher’s z transformations of r’s. No significant

differences were found between the two samples. We conclude that gender does not moderate

the relationship between students’ goal preferences and the quality of CL. Next, we will

investigate program type differences.

Program type differences

The influence of program type differences were analyzed for male and female students

separately, because males and females were not evenly distributed across program types.

For female students we excluded comparisons that include ICT/engineering students since

female students were hardly represented in this program type. Univariate analyses showed

significant effects of program type for male students in three goal domains: social support

(F[3, 619] 11.47, p = .000, � 2 = .05), mastery (F[3, 570] = 4.09, p = .007, � 2 = .02) and

superiority goals (F[3, 588] = 4.47, p = .004, � 2 = .02). The significant program type effect

for female students concerned social support (F[3, 1007] = 2.71, p = .040, � 2 = .00). Table 5

34

presents means and standard deviations for the goal domains, the quality of CL subscales as

well as the results of post hoc multiple comparison tests (Scheffé; p < 0.05).

As can be seen in Table 5, health and welfare male students scored significantly higher

on social support goals than retail and administration and ICT/engineering male students did.

This confirmed our expectation that health and welfare students, including the male students,

show a preference for social support goals. Interestingly, the male health and welfare students

were more mastery-oriented than their male food and tourism peers. In line with our

expectations, male ICT/engineering students scored higher on superiority goals. Overall, the

ICT group seemed to be a special group compared to the other groups; they had the lowest

scores on social support and mastery goals. As expected, their scores differed most from the

health and welfare group, who had the highest scores on most goal domains and also on the

perceived quality of CL. We calculated the correlation coefficients between the quality of CL

and the four goal domains within each of the four program types separately using Fisher’s z-

transformations of r’s in order to test the significance of these differences. Significant

differences were only found for the relationship between superiority goals and the students’

quality of CL. In the ICT/engineering (male) subgroup this correlation coefficient was

positive (r (186) = 0.28, p = .000); in the health and welfare male subgroup no relation was

found (r (136) = -0.03, ns; Z = 2.71, p = .003), and in the food male subgroup the correlation

was negative (r (31) = -0.33, p = .063) (Z = 3.09, p = .001). The other correlation coefficients

did not differ significantly between the program types. Since the predicted moderator effect of

program type was limited to male students’ superiority goal preferences and the quality of CL,

we may conclude that we can hardly speak of a moderator effect of program type.

------------------------------

Table 5 about here

------------------------------

Unique contribution of goal preferences and perceptions of contextual factors to the quality

of CL

-----------------

Insert Table 6

------------------

35

In order to examine the unique contribution of each of the related goal preferences and

contextual variables to the quality of CL, hierarchical regression analyses were conducted

with the students’ perception of the quality of CL as a dependent variable. In the first step we

entered gender and program type, in the second step we entered the students’ perception of

contextual factors, including their evaluations of the extent that they were taught CL skills at

their present schools, their perceptions of teachers’ clarity on rules for CL, monitoring

behavior, perceived availability of peer academic support and peer and teacher emotional

support. In the third step we entered students’ goal preferences, including social support

goals, belongingness goals, and mastery goals. In the fourth step, two-way “gender × goals”

and “gender × context” interactions were entered and in the fifth step “program type × goals”

and “program type × context” interactions were entered into the equation. The analyses did

not yield significant interaction effects on step 4 and 5. Table 6 presents the results of the first

three steps. Gender significantly contributed to the explained variance in the quality of CL.

However, inclusion of contextual variables in the regression equation led to the disappearance

of the unique contribution of gender. Further inspection of step 2 shows that 21% of the total

variance was accounted for and that all contextual factors had a unique contribution to the

explained variance in the quality of CL, except for teachers’ clarity on rules for CL and

availability of teacher emotional support (see Table 6). When goal preferences were added in

step three, 25% of the variance was explained.

Hence, having information about students’ goal preferences explained 4% unique

variance in the quality of CL and this was mainly due to the students’ score on the social

support goals.

DISCUSSION

First of all we explored the relationship between students’ goal preferences and the quality of

CL. At the outset of the study, we predicted that belongingness, social support and mastery

goals would be positively relatively and superiority/ individuality goal preferences negatively

related to the quality of CL. The students in our sample gave most preference to mastery

goals, followed by social support goals and belongingness goals. Social support goals had the

strongest relationship with the quality of CL, again followed by mastery and belongingness

goals. Students who value helping and supporting each other, rated the quality of CL higher,

independent of their mastery and belongingness goals. We expected an overall negative

relationship between students’ superiority/individuality goals and the quality of CL. However,

36

superiority/individuality goals were not significantly related to the quality of CL, meaning

that whether students are high or low on preference for this type of goal is unrelated to their

perception of the quality of CL. This unexpected finding will be discussed later in relation to

the program type differences that we found.

Our second research question involved the relationship between the quality of CL and

perceptions of contextual factors in the classroom, including social climate. We assumed that

in order for students to cooperate well it is very important that they know how to cooperate in

the first place. In other words, we assumed that the quality of CL will be poor when students

indicate that they were not taught the necessary skills. Multiple regression analyses showed

that all contextual variables made a significant contribution to the variance explained in CL,

except the students’ perception of available emotional teacher support and the teacher’s

clarity of rules for CL. Important predictors were the students’ awareness that they had been

taught the necessary CL skills and that their teachers were monitoring their effective use of

these skills. Interesting for future research is to explore whether this relationship changes over

time. As predicted, perception of social climate was related to the quality of CL. Recall, that

the simple correlations showed all four social climate variables to be related to the perceived

quality of CL. In the regression analyses we noted that only perceived availability of

emotional and academic peer support were related to the quality of CL. It seems plausible,

therefore, that the higher students rate the availability of peer support the better they will

cooperate.

Thirdly, we explored gender and program type effects on the relationship between

students’ goal preferences and the quality of CL. Examination of gender differences in the

relationship between students’ goal preferences and the quality of CL revealed that, as

predicted, female students showed higher scores than male students on the quality of CL. Also

in line with our expectations, female students reported higher preferences for social support

and mastery goals whereas male students reported higher preferences for superiority goals.

The findings suggest that female students, more than their male peers, feel confident in CL

settings. These findings confirm previous findings (Eccles, 1987; Townsend & Hicks, 1997;

Voncken et al., 2000; Wentzel, 1991). We did not find a gender moderator effect but the

study revealed interesting program type effects.

Because of their social orientation, we predicted that health and welfare students and

food and tourism students score higher on the quality of CL, and on social goals, particularly

in comparison with ICT and engineering students. This prediction was partially confirmed.

Health and welfare students scored significantly higher on the perceived quality of CL and on

37

social support goals than ICT/engineering students, but also higher than retail and

administration students. These conclusions only pertain to male students. We explained

previously that health and welfare students are preparing for a career that requires them to

gain a favorable attitude toward and proficiency in social skills. We also found that health and

welfare students scored significantly higher in relation to this goal domain than food and

tourism students did. Another prediction was that ICT and engineering students report a

higher preference for superiority goals in anticipation of their future job in the world of

business. We found that these students scored higher on superiority goals than both the health

and the food sub-samples and that they had the lowest scores on social support goals and

mastery goals.

The program type moderator effect was limited to the relationship between

superiority/individuality goal preferences and the quality of CL. In the total sample the

correlation between CL and superiority goals was non-significant. A stronger correlation was

noted in the male samples of the ICT/engineering subgroup as compared to the correlations

in the health and welfare and food and tourism subgroups. Interestingly, the direction of the

relationship differed between program types as well. In the health and welfare sample no

relation was found, meaning that whether or not these students give preference to superiority

goals is independent of their perception of the quality of CL. In the food and tourism sample a

negative correlation was found, implying that food and tourism students, who want to impress

others and outperform their peers, report that the quality of CL is lower than peers who do not

have this tendency. In the ICT/ engineering group, superiority goals were positively related to

the quality of CL and this contradicted our predictions. However, in line with our discussion

on program type dependent goal orientations it is conceivable that this dominantly male

group, who scored significantly higher than the other groups on superiority goals, prefers

group assignments that invite them to compete with their peers and with other groups. In this

study we did not collect information on the nature of the curriculum or the group assignments

set to the students in the different program types. It might be that the group assignments set in

the socially oriented program types differ from those in program types that encourage students

to be more superiority minded. Several researchers (e.g., Boggiano et al., 1989; Dweck, 1986)

have argued that superiority goals are prevalent in traditional educational settings where

competition and achievement goals are a crucial part of the learning process. More research is

needed to study the underlying mechanisms of the program type effect.

Recommendations

38

Our main aim was to study the effect of goal preferences on CL and to identify factors that

teachers can manipulate to promote successful CL and to prevent forms of misregulation (e.g.

chatting, social loafing). Our findings to date are that the context plays a significant role in

predicting the quality of CL. This is very promising. Indeed, adjustments in the context are

much easier to bring about than changes in students’ goal preferences. Based on our findings

we are able to provide some guidelines for future interventions.

In the first place, it is important that teachers make students aware of what is required

for working in a CL setting and teach the necessary and sufficient skills explicitly. More

specifically, it is crucial that teachers teach their students how to listen to each other, to

evaluate the group process, to discuss, to support group members, to give an opinion, or to

solve group conflicts. Secondly it is important that teachers monitor the CL process, which

means that they need to walk around in the classroom, frequently check with the groups and

ask them how they are doing. Thirdly, teachers need to be aware that availability of peer

support is essential for effective CL, emotional as well as instrumental support. This implies

that teachers should not only encourage students to provide this type of support but also

encourage them to role-play this type of behavior.

The role of the teacher in providing support was less important than peer support.

However, this could be an artifact of the type of analyses that we conducted; several other

variables in the analyses referred to teacher behavior and these variables explained a large

portion of the variance in CL.

Finally, we want to remark that the relatively weak link between student goal preferences and

the quality of CL may be due to the fact that not all students are aware of the multiple goals

they pursue in the classroom and of the relationships between their multiple goals and aspects

of CL. Currently we are conducting a follow-up stimulated-recall study where we assess the

significance that students attach to different types of goals while working on specific tasks in

a CL setting. After their group working sessions, groups of students are invited to provide

information about their goal preferences and their actual perception of the quality of CL.

Preliminary findings indicate that students do not spontaneously reflect on the link between

their goal preferences and the quality of CL. Discussing personal goals in order to make

students aware of the role these goals play in the learning process might be an important step

towards more successful CL. Teachers need to invite their students to think about their own

goals and about the links between personal goals and the goals presented to them by teachers,

39

course books, and other students. Such reflection might help them to adopt teacher-set

learning goals and self-regulate their learning more efficiently (see Boekaerts & Corno,

2005). Also, teachers need to create a classroom environment where peer support is promoted

and valued. At the same time, this type of environment will stimulate students to pursue their

social support goals, which are also crucial for successful CL.

40

References

Ames, C., & Archer, J. (1988). Achievement goals in the classroom: Students’ learning

strategies and motivation processes. Journal of Educational Psychology, 80, 260–267.

Anderman, L. H. (1999). Expanding the discussion of social perceptions and academic

outcomes: Mechanisms and contextual influences. In M. L. Maehr & P. R. Pintrich

(Series Eds.) and T. Urdan (Vol. Ed.), Advances in motivation and achievement: Vol.

11. The role of context (Vol. 11, pp. 303–336). Stamford, CT: JAI.

Boekaerts, M. (1987). Situation-specific judgments of a learning task versus overall measures

of motivational orientation. In E. De Corte, H. Lodewijks, R. Parmentier & P. Span

(Eds.), Learning and instruction: European research in an international context (Vol.

1, pp. 169–179). Oxford: Pergamon.

Boekaerts, M. (1998). Coping in context: Goal frustration and goal ambivalence in relation to

academic and interpersonal goals. In E. Frydenberg (Ed.), Learning to cope:

Developing as a person in complex societies (pp. 175–197). Oxford: Oxford

University Press.

Boekaerts, M., & Corno, L. (2005). Self-regulation in the classroom: A perspective on

assessment and intervention. Applied Psychology: An International Review, 24, 199–

231.

Boekaerts, M., De Koning, E., & Vedder, P. (in press). Goal-directed behavior and contextual

factors in the classroom: an innovative approach to the study of multiple goals.

Educational Psychologist.

Boekaerts, M., Pintrich, P., & Zeidner, M. (2000). Handbook of self-regulation. San Diego:

Academic Press.

Boggiano, A. K., Main, D. S., Flink, C., Barrett, M., Silvern, L., & Katz., P. A. (1989). A

model of achievement in children: The role of controlling strategies in helplessness

and affect. In R. Schwarzer, H. M. van der Ploeg & C. D. Spielberger. (Eds.),

Advances in test anxiety research (Vol. 6, pp. 13–26). Amsterdam: Swets &

Zeitlinger.

Brenner, E., & Salovey, P. (1997). Emotional regulation during childhood: Developmental,

interpersonal and individual considerations. In P. Salovey & D. J. Sluyter (Eds.),

Emotional development and emotional intelligence: Educational implications (pp.

168–192). New York: Basic Books.

41

Charlesworth, W., & Dzur, C. (1987). Gender comparisons of preschoolers behavior and

resource utilization in group problem-solving. Child Development, 58, 191–200.

Chin, W., Salisbury, D., Pearson, A., & Stollak, M. (1999). Perceived cohesion in small

groups. Small Group Research, 30, 751–766.

Cohen, E. G. (1994). Restructuring the classroom: Conditions for productive small groups.

Review of Educational Research, 64, 1– 35.

Connell, J. P., & Wellborn, J. G. (1991). Competence, autonomy, and relatedness: A

motivational analysis of self-esteem processes. In M. R. Gunnar & L. A. Sroufe

(Eds.), Minnesota Symposium on Child Psychology (pp. 43–77). Hillsdale, NJ:

Erlbaum.

Cosden, M., Pearl, R., & Bryan, T.H. (1985). The effects of cooperative and individual goal

structures on learning disabled and nondisabled students. Exceptional Children, 52,

103–114.

Covington, M. V. (2000). Goal theory, motivation, and school achievement: An integrative

review. Annual Review of Psychology, 51, 171–198.

Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and

self-determination in human behavior. New York: Plenum Publishing Co.

Dowson, M., & McInerney, D. (2001). Psychological parameters of students’ social and work

avoidance goals: A qualitative investigation. Journal of Educational Psychology, 35–

42.

Dweck, C. (1986). Motivational processes affecting learning. American Psychologist, 41,

1040–1048.

Eccles, J. (1987). Gender roles and women’s achievement-related decisions. Psychology of

Women Quarterly, 11, 135–72.

Ford, M. E. (1992). Motivating humans: Goals, emotions, and personal agency. Newbury

Park, CA: Sage.

Ford, M. E., & Nichols, C. W. (1991). Using a goal assessment to identify motivational

patterns and facilitate behavorial regulation and achievement. In M. Maehr & P. R.

Pintrich (Eds), Advances in motivation and achievement: Vol. 7. Goals and self-

regulatory processes (pp. 61–84). Greenwich, CT: JAI Press.

Gardner, A., Mason, C., & Matyas, M. L. (1989). Equity, Excellence and Just Plain Good

Teaching. The American Biology Teacher, 55, 72-77.

Gagné, M., & Zuckermann, M. (1999). Performance and learning goal orientations as

42

moderators of social loafing and social facilitation. Small Group Research, 30, 524–

541.

Gardner, H. (1993). Multiple intelligences: The theory in practice. New York: Basic Books.

Gillies, R., & Ashman, A. (1996). Teaching collaborative skills to primary school children in

classroom-based workgroups. Learning and Instruction, 6, 187–200.

Hertz-Lazorowitz, R., & Miller, N. (1992). Interaction in cooperative groups. New York:

Cambridge University Press.

Hickey, D. T., & Granade, J. (2004). The influence on sociocultural on our theories of

engagement and motivation. In D. M. McInerney & S. Van Etten (Eds.), Big theories

revisited on sociocultural influences on motivation and engagement (Vol. 4, pp. 223–

247).Greenwich, CT: Information Age Publishing.

Hoek, D., Van den Eeden, P., & Terwel, J. (1999). The effects of integrated social and

cognitive strategy instruction on mathematics achievements in secondary education.

Learning and Instruction, 9, 427–448.

Johnson, D. W., & Johnson, R. T. (1994). Learning together and alone: Cooperative,

competitive, and individualistic learning. Englewood Cliffs, NJ: Prentice-Hall.

Johnson, D. W., & Johnson, R. T. (2002). Learning together and alone: Overview and meta-

analysis. Asia Pacific Journal of Education, 22 , 95-105.

Lemos, M. (1996). Students’ and teachers’ goals in the classroom. Learning and Instruction,

2, 151–171.

Lemos, M. (2002). Social and emotional processes in the classroom setting: A goal approach.

Anxiety, Stress & Coping, 15, 383–400.

McInerney, D. M., Hinkley, H., Dowson, M., & Van Etten, S. (1998). Children’s beliefs

about success in the classroom: Are there cultural differences? Journal of

Educational Psychology, 90, 621–629.

Newman, R., & Schwager, M. (1993). Students’ perceptions of the

teacher and classmates in relation to reported help seeking in

math. Elementary School Journal, 94, 1, 3–18.

Organization for Economic Co-operation and Development (2003). Education at a Glance,

OECD Indicators 2003. Retrieved at February, 2004 at Website: www.oecd.org/home,

ISCED: http://www.uis.unesco.org/uisen/act/act_p/isced.html.

Pintrich, P. (2000). Multiple goals, multiple pathways: The role of goal orientation in learning

and achievement. Journal of Educational Psychology, 92, 545–555.

Ros, A. A. (1994). Samenwerking tussen leerlingen en effectief onderwijs [Cooperation

43

between students and effective education]. Dissertation, Rijksuniversiteit Groningen.

Ryan, R. M., & Deci, E. (2000). Self-determination theory and the facilitation of intrinsic

motivation, social development and well-being. American Psychologist, 55, 68–78.

Salomon, G., & Perkins, D. N. (1998). Individual and social aspects of learning. In: P. D.

Pearson & A. Iran-Nejad (Eds.), Review of Research in Education, 23 (pp. 1–24).

Washington: American Educational Research Association.

Schwartz, S. H., & Bardi, A. (2001). Value hierarchies across cultures: taking a similarities

perspective. Journal of Cross-Cultural Psychology, 32, 268–290.

Severiens, S., & ten Dam, G. (1998). A multi-level meta-analysis of gender differences in

learning orientations. British Journal of Educational of Psychology, 68, 595–608.

Shanahan, T. (1998). On the effectiveness and limitations of tutoring in reading. In P. D.

Pearson & A. Iran-Nejad (Eds.), Review of Research in Education, 23 (pp. 217–234).

Washington: American Educational Research Association.

Slavin, R. E. (1995). Cooperative learning: Theory, research, and practice (2nd ed). Boston:

Allyn and Bacon.

Slavin, R. E. (1996). Research on cooperative learning and achievement: What we know,

what we need to know. Contemporary Educational Psychology, 21, 43–69.

Urdan,T. C., & Maehr, M. L. (1995). Beyond a two-goal theory of motivation and

achievement: A case for social goals. Review of Educational Research, 65, 213-

243.

Urdan, T. C. (1997). Examining the relations among early adolescent students’ goals and

friends’ orientations toward effort and achievement in school. Contemporary

Educational Psychology, 22, 165–191.

Urdan, T. (1999). (Ed.), Advances in motivation and achievement, Vol. 11: Motivation in

context. Stanford, CT: JAI Press.

Vedder, P. (1985). Cooperative learning: A study on processes and effects of cooperation

between primary school children. Dissertation, Rijksuniversiteit Groningen.

Vedder, P., Boekaerts, M., & Seegers, G. (2005). Perceived social support and well being in

schools; the role of students’ ethnicity. Journal of Youth and Adolescence, 34, 235–

246.

Veenman, S., Kenter, B., & Post, K. (2000). Cooperative learning in Dutch primary

classrooms. Educational Studies, 26, 281–302.

Voncken, E., Van der Kuip, I., Moerkamp, T., & Felix, C. (2000). Je bent jong en je weet niet

44

wat je wilt: Een inventarisatie van push- en pullfactoren die leiden tot voortijdig

schoolverlaten in de BVE-sector [You are young and you do not know what you want:

an inventory of push- and pull factors that lead to drop-out in secondary vocational

education]. Amsterdam: SCO-Kohnstamm.

Webb, N. M., & Farivar, S. (1994). Promoting helping behavior in cooperative small groups

in middle school mathematics. American Educational Research Journal, 31, 369–395.

Webb, N. M., & Palincsar, A. S. (1996). Group processes in the classroom. In D. C. Berliner

& R. C. Calfee (Eds.), Handbook of educational psychology (pp. 841–873). New

York: Simon & Schuster MacMillan.

Wentzel, K. R. (1991). Relations between social competence and academic achievement in

early adolescence. Child Development, 62, 1066–1078.

Wentzel, K. R. (1993). Motivation and achievement in early adolescence: The role of multiple

classroom goals. Journal of Early Adolescence, 13, 4–20.

Wentzel, K. R. (1994). Relations of social goal pursuit to social acceptance, classroom

behavior, and perceived social support. Journal of Educational Psychology, 86, 173–

182.

Wentzel, K. R. (1996). Social goals and social relationships as motivators of

schooladjustment. In J. Juvonen & K. R. Wentzel (Eds.), Social motivation:

Understanding children’s school adjustment. New York: Cambridge University

Press.

Wentzel, K., & Wigfield, A. (1998). Academic and social motivational influences on

students’academic performance. Educational Psychological Review, 10, 1555–1575.

45

Figure 1

Model of research

Goal Preferences

Perception of contextual factors in the Classroom

Quality of Cooperative Learning

Gender/ program type

46

Table 1

Sample Characteristics

Program Type N % Female

ICT/ Engineering 347 6.05

Retail & Administration 355 52.1

Food & Tourism 96 55.2

Health & Welfare 1122 83.2

Total 1920 62.4

47

Table 2

Categories, Sample Items, Number of Items and Cronbachs’ Alpha Coefficients

Category Sample item # items Alpha

Students’ Goal

Preferences

Superiority/

individuality

I want to impress others 9 .93

Mastery I want to learn more about my profession 6 .92

Belongingness I want to get along with my peers 6 .86

Social Support I want to help others in case they need help 7 .91

Perceived

Quality of CL

Quality of CL I perceive myself as part of this group 29 .90

Conditions for

CL

48

Task difficulty Most group members think the task is too difficult 1 -

Task challenge Most group members think the task is challenging 1 -

Task time We have sufficient time for finishing the task 1 -

Task consulting Students need to consult each other in order to finish

the group task

1 -

Type of reward After finishing the task, we receive an 1) individual

reward, 2) group reward 3) both

1 -

Cooperation skills

and knowledge

At this school we learned how to have a good quality

group discussion

8 .86

Rules for CL Before we start to work on the group task, teachers

explain us how to plan

9 .87

Teachers

Monitoring

Behavior

Teachers walk around the classroom when we

cooperate

5 .83

Teacher

Interventions

If we are too noisy while we cooperate, teachers

intervene

5 .77

49

Teacher

Evaluations

After finishing the group task, teachers explain what

went well during CL and what needs improvement

4 .80

Social Climate

Academic support

teacher

When I do not understand the lesson, I get support from

my teacher

7 .80

Academic support

peers

When I do not understand the lesson the I get support

from my peers

7 .82

Social support

teachers

When I am sad my teacher supports me 6 .82

Social support

peers

When I am sad my peers support me 6 .89

50

Table 3

N, Mean Scores, Standard Deviations and Correlation Coefficients of the quality of CL, Goals, Contextual factors, and Social Climate

N

M

SD

Sup

Mas

Bel.

Soc

Task

chall.

Task

diff

Task

consult

Task

time

Rew

Skill

Rules

Mon

Int

Eva

T ac

T em

P ac

P em

Quality of CL

(QCL)

1526 2.8 .28 .02 .23** .23** .33** -

.16**

.10** -.10** -.06* .03 .35** .24** .20** .19** .19** .17** .21** .28** .30**

Superiority (Sup) 1816 3.11 1.01 .14** .26** .09** -.03 -

.08**

.05 .05 .03 -.01 .01 .05 .09** .04 -.01 .05 -

.02**

-

.01**

Mastery (Mas) 1784 4.27 .64 .46** .66** -.07* .03 -.10** -

.08**

-.04 .09** .01 .05* .02 -.03 .16** .12** .13** .17**

Belongingness

(Bel.)

1821 4.15 .68 .49** -.03 .04 -.08 .01 -

.07*

.13** .05 .05 .09** .04 .17** .17** .16** .22**

Social Support

(Soc)

1915 4.18 .72 -.04 .04 -.07** -.06* -.02 .12** .01 .04 .04 -.02 .15** .15** .19** .33**

Task challenge 1704 1.57 .49 .05 .03 -.04 .01 - - - - - - - -.03 .01

51

(Task chall) .18** .16** .18** .15** .19** .09** .10**

Task difficulty

(Task diff)

1740 1.87 .34 -.05 .11** -.02 .03 -.03 .03 .00 .02 .02 .02 .02 .05

Task consulting

( Task consult)

1745 1.10 .30 .04 -.04 -

.08**

-.02 -.05 .01 .06** -.03 -.05 -.05* -.07*

Task time

(Task time)

1731 1.37 .48 .06* -

.14**

-.16* -.12* -

.07**

-

.12**

-

.08**

-

.14**

-

.12**

-

.16**

Type of reward

(Rew)

1636 1.83 .89 .06* .08** .05 .03 .03 -.03 -.02 -.02 -.00

Coop. Skills

(Skill)

1908 2.65 .45 .52** .34** .29** .36** .23** .23** .09** .13**

Rules for CL

(Rules)

1933 2.48 .46 .47** .40** .52** .23** .26** .01 .01

Teacher Monitoring

(mon)

1893 2.52 .53 .45** .50** .28** .25** .09** .08**

Teacher Intervening

(int)

1924 2.44 .56 .51** .22** .24** .04 .04

52

Teacher Evaluation

(eva)

1932 2.30 .60 .21** .24** .04 .00

Teacher Aca

Support (Taca)

1951 2.8 .63 .62** .38** .29**

Teacher Emo

Support (T emo)

1895 2.09 .64 .31** .47**

Peer Academic

Support (P ac)

1926 2.8 .60 .67**

Peer Emotional

Support (P emo)

1909 2.64 .80

*p <.01, ** p < .001,

53

Table 4

Means and Standard Deviations for the Goal Preference Subscales and the quality of CL Subscale,

for Male (N = 402) and Female (N = 1171) Students

M SD F p η 2

Mastery Learning

Goals

Male 4.21 .63 9.53 .002 .00

Female 4.33 .60

Belongingness Male 4.13 .70 .62 .43 .00

Female 4.16 .67

Social Support Male 4.07 .71 35.61 .000 .03

54

Female 4.33 .64

Superiority Male 3.30 .99 29.74 .000 .02

Female 2.93 1.03

Quality of CL Male 2.76 .26 12.66 .000 .01

Female 2.83 .28

55

Table 5

Means and Standard Deviations for Goal Preferences and the quality of CL by Program Type and Gender

ICT/ eng

Health

Retail

Food

Multiple Comparisons

(Scheffé, p <.05)

η 2

Mastery Goals Male 4.13 (.76) 4.31 (.57) 4.15 (.67) 3.93 (.67) health > food

.02

Female - 4.33 (.60) 4.33 (.60) 4.29 (.61) .00

Social Support Male 3.81 (.79) 4.23 (.62) 3.93 (.79) 3.94 (.65) health > ICT/

engineering,

retail

.05

Female - 4.33 (.62) 4.24 (.69) 4.38 (.70) .00

Belongingness Male 4.15 (.73) 4.17 (.67) 4.15 (.73) 3.85 (.63) .01

Female - 4.16 (.67) 4.19 (.64) 4.02 (.76) .00

Superiority Male 3.43 (.80) 3.19(1.03) 3.49 (.93) 3.05(.94) .02

Female - 2.91(1.02) 2.98(1.09) 3.05(1.07) .00

56

Quality of CL Male 2.76 (.33) 2.78 (.24) 2.75 (.29) 2.75 (.22) .01

Female - 2.83 (.27) 2.79 (.30) 2.86 (.22) .01

57

Table 6

Summary of Hierarchical Regression Analyses for Background, Contextual Factors and

Goal Preferences Predicting the Quality of CL

step predictor ß R� F

1 Back ground .01 5.40**

Gender .09*

Program type .04

2 Back ground .21 31.44***

Gender .05

Program type -.03

Perceptions of Contextual Factors in the

Classroom

Students evaluations of the extent that they

were taught CL skills

.24***

Rules for CL skills .07

Teachers monitoring behavior .07*

Perceptions of the Social Climate

Perceived availability of emotional peer

support

.20***

Perceived availability of academic peer

support

.12**

Perceived availability of emotional teacher

support

-.02

3 Back ground

Gender .02

58

Program type -.04

Students Goal Preferences .25 28.67***

Social Support .20***

Belongingness .00

Mastery .03

Perceptions of Contextual Factors in the

Classroom

Students evaluation of the extent that they

were taught CL skills

.23***

Rules for CL skills .07

Teachers monitoring behavior .08**

Perceptions of the Social Climate

Perceived availability of emotional peer

support

.13**

Perceived availability of academic peer

support

.13***

Perceived availability of emotional teacher

support

-.02

*p<.05, **p<.01, ***p�.001

59

Chapter 3

EXPLORING THE LINKS BETWEEN STUDENTS’ ENGAGEMENT IN

COOPERATIVE LEARNING, THEIR GOAL PREFERENCES AND PERCEPTIONS

OF CONTEXTUAL FACTORS IN THE CLASSROOM

2

Abstract

In this in-depth study we questioned effective and ineffective cooperative learning (CL) teams

about their goal preferences, the quality of CL, and perceptions of contextual factors in the

classroom in a stimulated-recall setting. Mastery and social responsibility goals – together

with ‘learning for a certificate’ goal- tended to be more prevalent in effective teams, while

learning for a certificate and entertainment goals were dominant in ineffective teams.

Students’ belongingness goals were negatively related to socially oriented task engagement in

ineffective teams. Task characteristics, group composition, and teacher behavior were

mentioned as reasons for effective or ineffective CL.

Key words: goal preferences, engagement, contextual factors, cooperative learning

2 This chapter is based on: Hijzen, Boekaerts and Vedder (2005). Exploring the links between students’ engagement in cooperative learning, their goal preferences and perceptions of contextual factors in the classroom. Manuscript (re)submitted for publication in: Learning and Instruction.

60

INTRODUCTION

In a study on the quality of cooperative learning in senior vocational high schools in the

Netherlands one of the students commented during an interview: “It is just dead boring at

school, I don’t think I actually learn anything at all, I hope I will at least learn something in

my traineeship”. This statement indicates a mismatch between students’ goal preferences and

the goals that are set by the school, which may result in low levels of student engagement and,

eventually, in a state of disengagement for large groups. This comment points at a common

problem in senior vocational high schools where the drop out rate is alarmingly high. About

37 percent of the first year students quit school (School Inspectorate, 2002). Many of these

dropouts reported that they had experienced their study as uninteresting, too theoretical, too

difficult and non-supportive for their future career (Voncken, Van der Kuip, Moerkamp, &

Felix, 2000). Dropping out of school often is an outcome of underlying motivational problems

that students experienced during their school history. In the present in-depth study we will

focus on students’ engagement levels in cooperative learning (CL) settings in secondary

vocational education. CL is a common instructional method in senior vocational high schools

in the Netherlands. Of the first–year students in secondary vocational schools, one third work

less than 45 minutes per week on CL, and one third between 45 to 90 minutes per week,

whereas the other one third work between 45 and 90 minutes per day in CL teams (Hijzen,

Boekaerts, & Vedder, 2005). Often CL methods involve hands-on tasks, in which students

can develop their professional skills together. In CL settings students depend on each other

for learning and the conditions may encourage them to attend school regularly. CL methods

are also believed to foster a mastery goal orientation which is related to positive educational

outcomes (Midgley & Urdan, 1992). Unfortunately, placing students in CL groups does not

automatically create favorable conditions for learning. Many students do not actually work

together when asked to collaborate. They are engaged in task-irrelevant behavior instead. In

this study we will approach this problem from the perspective of motivational self-regulation.

Three types of engagement

Motivation strategies are a crucial aspect of students' self-regulation in academic and non-

academic activities. We define motivation strategies as processes that are governed by goal

preferences. Elliot and Sheldon (1998) view goal preferences as the “personally meaningful

objectives that individuals pursue in their daily lives”. We would like to add that these goal

61

preferences steer and direct, or in other words self regulate student behavior in the classroom.

Self-regulation refers to a multi-level, multi component process where students target their

cognitions, feelings and actions in the service of their goals (Boekaerts & Corno, 2005).

Students are expected to regulate their own learning behavior appropriately, showing higher

levels of engagement when their goal preferences are in line with the goals they are expected

to achieve in school. Engagement refers to the intensity and quality of students' involvement

in initiating and carrying out learning activities (Wellborn, 1991). Ryan (2000) made a

distinction between motivation and engagement; the former concerns cognitions underlying

involvement in one’s schoolwork (e.g., motivational beliefs) and the latter refers to students’

actual involvement in their schoolwork (concrete actions, engagement). Hence, engagement

can be perceived as an indicator of students’ motivation.

In this study we will focus on students’ actual involvement or engagement in CL.

Following Chapman (2003) we distinguished between three types of engagement that may

occur during CL, namely task-relevant engagement, task-irrelevant engagement and socially

oriented task engagement. The behavior of students who are actively engaged in the task is

characterized as task-relevant behavior. These students are active, focus on the task and

persist when obstacles occur. The behavior of disaffected students is characterized as task-

irrelevant behavior. These students do not try hard, give up easily in the face of challenges

and engage in task-irrelevant behavior like chatting or disturbing others (Chapman, 2003).

Socially oriented task engagement is related to the first type of engagement. Students who are

socially engaged are primarily involved in social activities such as providing help and

emotional support.

In this study we will attempt to distinguish CL teams that predominantly show (social)

task-relevant engagement during CL (i.e. effective CL teams), from teams that show task-

irrelevant engagement most of the time (ineffective CL teams), in terms of their goal

preferences and their perceptions of contextual factors in the classroom.

Previous studies

In a previous study (Hijzen, Boekaerts, & Vedder, 2006), we focused solely on students’

general cognitions underlying their perception of the quality of CL. Generally weak

relationships were found between students’ goal preferences and the quality of CL. Several

reasons were given for this relatively weak link. First, students might not be aware of their

goal preferences and therefore do not consider them when reflecting on the quality of CL. It is

62

also plausible that students’ general goals, as measured with the goal questionnaire, differ

from goal preferences in a context (see for example Lemos, 1996). It was found that students’

scores on the quality of CL were most strongly associated with their social support goals (e.g.,

“I want to support my peers”), followed by belongingness goals (e.g., I want to get along

with my peers) and mastery goals (e.g., “I want to learn new things”). No relationship was

found with students’ superiority or individuality goals (e.g., “I want to be the best student of

my class”).

In this study we explored students’ goal preferences and their perception of the quality

of CL using stimulated-recall. The quality of CL was analyzed using a measure of

engagement during CL. In order to discover which types of goals were most salient in the CL

setting, we did not limit the range of goals to the traditional mastery and performance

distinction (e.g., Ames & Archer, 1988; Dweck 1986; Urdan, 1999), or to the four goals, that

were the object in our previous research. Instead, we explored the full range of goal

preferences that students spontaneously mentioned in combination with the quality of CL. We

reasoned that several goal preferences may underlie students’ engagement in the classroom.

With the exception of a few studies (e.g., Downson & McInerney, 2001; Lemos; 1996),

students’ subjective goal definitions have not been the object of research. In this in-depth

study, we combine students’ self-reports on a-priori goal preferences with the goals that they

spontaneously mention while reflecting on their own activities in actual CL settings.

We reasoned that the relatively weak link between students’ goal preferences and the

quality of CL in our previous study is largely due to students’ perceptions of contextual cues.

We assume that the context plays an important role in shaping the quality of CL. Earlier

research supports that environmental features trigger specific goal preferences, thus

influencing the quality of CL (e.g., Ames, 1992; Ames & Archer, 1988, Blumenfeld, 1992;

Hijzen, Boekaerts, & Vedder, 2006). We will pay close attention to students’ perceptions of

contextual factors in the classroom that may or may not promote goal preferences that might

shape the conditions for effective CL. In the present study a multi-method, context-bound

approach was used in order to capture students’ thinking about the type of goal preferences

and the type of context that stimulate their CL processes best.

Engagement, goal preferences, and contextual factors in effective CL teams

By comparing participating CL teams that have reported to cooperate very well with teams

that have reported to cooperate poorly, we will gain insight into the salient goal preferences of

63

these reflective groups, and into the nature of the relationship between perceptions of the

members of the teams of contextual factors and the reported quality of CL. Based on findings

of previous studies we expect that in effective CL teams, students’ social and mastery goals

will be dominant. For example, McInerney, Hinkley, Dowson and Van Etten (1998)

suggested that a joint mastery and social goal orientation is more productive than mastery

goals alone because a sense of belongingness and social responsibility provoked by social

goals provides an added drive for academic achievement. Indeed, the foundation for much of

the success of CL is the well-known phenomenon that individuals show goal striving for the

sake of the group (Hertz-Lazarowitz & Miller, 1992) Wentzel (1989) found that high-

achieving students and low-achieving students displayed specific unique social goal patterns.

High-achievers tended to focus on getting things done on time and being successful,

responsible, and dependable students, whereas making friends and having fun were less

valuable goals. Low-achieving students on the other hand reported that the latter two goals

were important. In line with these results, we expect that students in effective CL teams

combine salient mastery and social goals. We also expect that these team members are more

conscious of their goal preferences than students in ineffective teams. Earlier research

suggested that formulating mastery goals facilitates students’ intrinsic motivation (e.g.,

Cantor, Norem, Niedenthal, Langston, & Brower, 1987; Sheldon & Kasser, 1995; Sheldon &

Elliot, 1997) and at the same time boosts their socially oriented task engagement in CL.

Based on reviews of conditions for CL (e.g., Cohen, 1994; Webb & Palincsar, 1996),

we expect that effective CL teams perceive contextual factors as promoting mastery and

social goals as these are challenging, hands-on, and promoting interdependency (Webb &

Palincsar, 1996). These teams are preferably not too big so that effects of social loafing are

minimal and these team members get along and support each other (e.g., Chin, Salisbury,

Pearson, & Stollak, 1999; Sharan & Sharan, 1992) in a beneficial way. Team members of

effective teams are also expected to evaluate their teachers as specialists who are guiding the

learning process sufficiently, but not in a too strict way. In other words, they view their

teachers as coaches who make appropriate use of scaffolding methods (Brown & Palincsar,

1989; Oortwijn, Boekaerts, & Vedder, in preparation) and they evaluate the school climate as

transparent and supporting.

Engagement, goal preferences and contextual factors in ineffective CL teams

64

In ineffective teams we expect that team members’ social and affective goals are more

important than their mastery goals. They are expected to target their cognitions, feelings and

actions in the service of task-irrelevant engagement, such as chatting, making friends or just

having fun. These students are expected to be preoccupied with their well-being and therefore

less involved with the learning process. Furthermore, we expect these team members to be

less involved in goal planning, and less conscious of their goals than effective team members.

Because superiority goals (i.e. having a performance orientation) are likely to conflict with

goals that are salient in CL settings, we expect that students who hold these goals to

experience the CL setting as a threat to their wish to perform at a more individual level (e.g.,

Schwartz & Bardi, 2001). Lepper and Hodell (1989) found that students who try to get good

grades, because they want to comply or to obtain a certificate are not intrinsically motivated

and they may disengage from a task when they judge that they might not achieve this goal

We expect that members of ineffective teams evaluate the group task as boring and too

theoretical. They might experience difficulties in getting along and perceive teachers as

controlling or not involved.

Research question

Students’ perspectives on their goal preferences were taken as a starting point and their task-

relevant, task irrelevant and socially oriented task engagement levels were observed during

CL and further examined in a stimulated-recall setting. CL teams that function well will be

compared with teams that function poorly or in fact are not cooperating at all. The main

research question in this study is: “How can we distinguish effective CL teams from ineffective

CL teams, and characterize them in terms of students’ goal preferences and perceptions of

contextual factors in the classroom?” We will investigate whether our findings lend support

to models of CL that were presented on the basis of previous research and whether

conceptualizations used in earlier studies adequately cover processes that bring about or

substantiate effective CL.

METHOD

Subjects

Participants in this in-depth study were 57 second year-students from five different schools

for secondary vocational education in the Netherlands. Secondary vocational school starts for

65

most students at the age of 16 after they completed a junior vocational school. Senior

vocational school is divided into four levels. At the first level students train to become

assistants (6 to 12 months). At the second level they have two to three year courses for basic

vocational training. At the third level students are enrolled in professional training and at the

fourth level they participate in middle-management training (3 to 4 years) or in a specialized

training course (1 to 2 years). Most students finishing secondary vocational school prefer to

enter the labor market and do not proceed with further studies (for further information on

secondary vocational education in the Netherlands, see Euridice database on education, 2003

website). Secondary vocational schools offer a broad range of program types. Subjects in the

present study were enrolled in three different program types; ICT, health and welfare and

retail and administration programs, these were level 3 and level 4 studies. Five teachers, who

often used CL instruction methods, and their classes were willing to participate. In the five

classes 18 CL teams participated. From these 18 teams we selected four teams (nine males;

eight females in total) that reported to cooperate very well and four teams (five males; nine

females in total) that said to cooperate very badly or that did not cooperate at all. Two of the

teams that said to cooperate very well were enrolled in retail and administration, one in ICT,

and one in health and welfare programs. In the teams that were unsatisfied with their CL one

team was enrolled in retail and administration, two in ICT and one in health and welfare

programs. The eight teams were spread over four different classrooms.

Group composition

The CL settings differed from class to class. In the ICT class we selected three teams of

students that worked on a (computer) project for an eight week period in teams that consisted

of four individuals. Students were expected to work on the project for three days a week.

Teachers were present to assist when help was required. All courses that students attended

during the rest of the week were related to the project.

Two CL teams of health and welfare students were selected and were observed during

social science classes. The teams consisted of six persons. Students in these classes worked on

a variety of group assignments. For example, students were asked to prepare a role play or to

cooperatively work out a treatment plan for an imaginary person with a certain type of

behavioral problem. For the assignments the students had to work together.

Two retail and administration classes participated. In the former class (two teams were

selected) modes of direct instruction were combined with group work. This was the only

classroom where direct teaching was dominant. In the latter (one team was selected) we

66

observed students during a simulation project. A team of six students had to organize their

own virtual company and their aim was to make it profitable. The project lasted for almost

one school year. Each student had a specific role in the company, with the attached

responsibility (e.g., director, secretary, assistant). Different aspects of what they were taught

during class were integrated in the project, e.g., working out a business plan. We observed

during the weekly meetings when these six students had to evaluate, plan and divide tasks.

Procedure and instrumentation

In each classroom, we made three video-registrations of the CL teams, with two weeks

intervals. The first video-recording started at the beginning of the (CL) projects and each

recording took 10 minutes. The students also participated in a self-report study and completed

questionnaires on their goal preferences and the Quality of Cooperative Learning. After the

last video-registration we confronted students with a video-recording of their CL team and

interviewed them about their goal preferences, the quality of CL, influences on their

engagement levels and goal conflicts.

Questionnaires

Before students were observed and interviewed they participated in a survey. They completed

the Goal Importance and Facilitation Inventory (GIFI) which is based on Fords’ taxonomy of

broad goals (Ford, 1992; Ford & Nichols, 1991). Students indicate how salient each of the

seven broad goal domains are in their current life, namely students’ mastery, affective,

belongingness, social support, self-determination, material gain, and superiority/ individuality

goals. Response categories ranged from "I disagree very strongly” (1) to “I agree very

strongly” (5).

Observational studies

In order to rate engagement levels of the individual students in CL, we developed an 18-item

rating list. Items were based on a study by Skinner and Belmont (1993) and the Quality of CL

questionnaire (Hijzen, Boekaerts, & Vedder, 2006). The rating list contained three subscales.

Items in the first subscale measured students' task-irrelevant engagement. A sample item is

"this student chats", Cronbach’s alpha was .89. A second subscale measured students' task-

relevant engagement. A sample item is "this student concentrates on the task", Cronbach’s

alpha was .95. The third subscale aims to measure students’ socially oriented task

67

engagement. A sample item is “this student offers team members help”, Cronbachs’ alpha

was .90. Response categories (4) of the items varied from "almost never" to "very often".

In order to optimize interrater reliability, a scoring guide was developed before the

actual rating process started. Examples were listed for each item in the rating list and two

raters were trained to identify them. They rated the engagement levels of the 57 students, on

one video-tape using the rating list. This video-recording was the same as the one used in the

stimulated recall session. Cohens’ Kappa indicated an acceptable level of agreement between

the two raters, K = .66, p < .01.

Interviews/ stimulated-recall

Semi-structured group interviews were conducted with the observed teams of students. The

videotape formed the basis for the interviews which contained 16 questions. We asked

students about their goal preferences, their perceptions of contextual factors in the classroom

and the quality of the cooperation in their team. Teams were asked to explain motives for

task-relevant and task-irrelevant behavior during CL. On average an interview took 45

minutes. The interviews were taped and transcribed and the interview protocols were

restructured by subdividing the statements into broad categories. We distinguished two

categories of statements, namely those about students’ goal preferences and those about

students’ perceptions of contextual factors in the classroom. Interview protocols were

analyzed, using an inductive content analysis approach, based on the work of Lemos (1996)

and Dowson and McInerney (2001). Table 1 presents the scoring categories.

-------------------------------

Insert Table 1 about here

------------------------------

A scoring guide was developed with examples of each category and statements about goals

were counted and categorized, by four raters. Interrater agreement was 83 % ((100/ n

statements) x n agreements). An agreement meant that at least three out of four raters agreed

about the classification of a particular statement (see Lemos, 1996). Interrater agreement (two

raters) for assigning statements about contextual characteristics (see Table 1) was 94 % ((100

/ n statements) x n agreements).

Quantitative analysis

Quantitative analysis of the self-reports and observed scores for engagement consisted of

calculating mean scores, univariate analyses of variance and correlation analyses. In line with

68

Lemos (1996) quantitative analyses of the interview statements consisted of frequency

analysis and analysis of the profiles’ congruency (Serafini coefficient C). The congruence

coefficient is a measure of the relative distance between two profiles. A profile considers

several variables at the same time and C summarizes the different distances between the

profiles. C values range from 0 to 1:

From 0.90 to 1.00: Perfect congruence

From 0.70 to 0.89: High congruity

From 0.40 to 0.69: Moderate congruity

From 0.20 to 0.39: Low congruity

From 0.00 to 0.19: No congruity

For more information on calculating congruence coefficients see Lemos (1996). Goal profiles

were derived by calculating the percentage of students’ statements on each goal. We counted

the proportion of each particular goal type as compared to the total number of goal statements.

RESULTS

In this section, we will compare goal profiles and perceptions of contextual factors of teams

that reported in the interviews that the quality of CL in their team was very good with teams

that were highly unsatisfied with the quality of CL. An additional selection criterion was that

the mean scores on task-relevant and socially oriented task engagement in the effective teams

should be above 3, and in the ineffective teams below 3. By contrast, task-irrelevant

engagement should be higher in the ineffective teams as compared to the effective teams.

Table 2 presents these teams engagement scores

----------------------------------

Insert Table 2 about here

----------------------------------

As predicted, observed engagement scores were dissimilar in the teams. In ineffective teams

task-relevant and socially oriented task engagement tended to be lower and task-irrelevant

engagement higher than in effective teams. Univariate analyses showed that the teams scored

significantly different on socially oriented task engagement (F (1, 20) = 9.44, p = .006, �² =

.32). We will conclude this section with two case studies in order to illustrate in more detail

why some teams were successful in CL and others were not.

69

----------------------------------

Insert Table 3 about here

----------------------------------

Goal preferences and engagement in CL

Table 3 presents effective and ineffective team members’ goal preferences as obtained by

their GIFI-self reports. A closer look at these teams’ goal preferences showed that effective

teams’ most prevalent goal preferences were affective, social support goals and self-

determination goals, whereas ineffective team member’s most prevalent goal preferences

were belongingness, affective and self-determination goals. The most striking difference

concerned students’ belongingness and social support goals. Effective team members tended

to report slightly lower scores on belongingness goals than ineffective team members, while

the reversed pattern is observed for social support goals. However, no significant differences

between goal preferences for effective and ineffective CL teams were found. In order to

explore how the different goals are related to the three engagement types we calculated

Pearson correlation coefficients for effective and ineffective teams for task-relevant, task-

irrelevant and socially oriented task engagement, and the goal domains of the GIFI. We

expected students’ mastery and social goals to be related to task-relevant and socially oriented

task engagement. Remarkably, only one goal domain was significantly related to students’

engagement scores, namely students’ belongingness goals. This relationship only concerned

the ineffective teams. The direction, in ineffective teams, of the relationship was not in line

with what we predicted. That is to say, preferences for belongingness goals were negatively

related to their socially oriented task behavior (r = -.75, p < .05), implying that ineffective

team members who report that belongingness goals are salient in their current life

demonstrate less socially oriented task behavior than those report that these goals are not

important in their life. Inspection of Table 3 reveals that in the group of ineffective teams,

belongingness goals were dominant over mastery and social support goals, while this pattern

is reversed in the effective teams. This finding suggests that wanting to feel at home in the

group (belong) may hinder rather than facilitate socially oriented task engagement.

Goal profiles

After the categorization of students’ statements on all goal preferences, it became apparent

that the aforementioned goal preferences were not the only ones that were prevalent in the CL

70

setting. A profile was elaborated for the two types of teams in order to account for the relative

importance of each goal within the total set of goal preferences. We counted all the statements

that referred to goal preferences per subgroup, and calculated the proportion of each particular

goal type as compared to the total number of goal statements. Figure 1 presents percentages of

students’ statements in effective and ineffective CL teams referring to their goal preferences.

----------------------------------

Insert Figure 1 about here

----------------------------------

Figure 1 shows that in effective teams the most prevalent goals were mastery, learning for a

certificate, social responsibility and entertainment goals. Interestingly, these goals are largely

the same type of goals that were found to be positively related to the quality of CL in previous

studies (Hijzen, Boekaerts, & Vedder, 2006). In ineffective teams, entertainment, learning for

a certificate, mastery and work avoidance goals rated high. The most considerable, and

expected, difference was the strong prevalence of entertainment goals in ineffective CL

teams, they were as salient as the ‘learning for a certificate’ goal. In the effective CL teams,

mastery goals were most salient and revealed learning for certificate goals for importance.

Interestingly in this group, social responsibility goals scored as high as entertainment goals.

Social responsibility goals were not mentioned at all in ineffective CL teams. Another striking

difference concerned the percentage of statements on work avoidance goals, in ineffective

teams the percentage of statements on this goal domain was much higher than in effective

teams.

The congruence coefficient (C) (Serafini, 1981 in Lemos 1996) allowed comparisons

between the two profiles. The goal profiles of students in effective and ineffective CL teams

showed a moderate congruity (coefficient C between the two profiles = .55), meaning that the

profiles of effective and ineffective teams are only moderately comparable. Moreover, as

expected, students in the ineffective teams were less conscious of their goals; fewer

statements referred to goal preferences (27 statements) than in effective teams (54

statements). Also, the range of different goal preferences mentioned was smaller for

ineffective teams than for the effective teams.

Compared to students self-reports about their overall goal preferences, many of the

same goals were mentioned when they were asked to reflect on the actual CL setting (e.g.,

mastery, social support, belongingness and affective goal preferences). Surprisingly, self-

determination, superiority/ individuality and material gain were not mentioned at all in the CL

context. Conceivably, these goal preferences are not important in the CL setting. Furthermore,

71

it was noted that mastery goals were more prevalent in the actual learning situation compared

to the students’ scores reported on the GIFI, a finding that was also detected by Lemos. She

found that “The confrontation with the real classroom activities strengthens the focus of

students’ behavior within the academic dimension” (Lemos, 1996, p. 167).

Contextual factors and engagement in CL

Important to note is that goal preferences were rarely spontaneously mentioned while

reflecting on task-relevant, task-irrelevant, or socially oriented task engagement during CL

sessions. Groups pointed far more often at the context for explaining their CL. We asked all

teams to explain instances of effective and ineffective CL. Results presented here are limited

to the four ineffective and four effective CL teams.

Statements that explained successful CL in the effective teams referred to the group

composition (17 statements). The size of the team (“In a large team it is harder to keep your

promises”), and the types of relationships among the team members were mentioned as

reasons for cooperating successfully (“We are friends, therefore we cooperate better”. “You

dare to say more, for example when somebody did not complete his part of the project”. “We

understand each other very well and our communication is exceptionally good, I think that is

an important part of the success”). Students alluded to particularly motivated persons in the

group who contributed to successful CL (two teams), to improvements in teacher behavior in

the sense that the teacher became less dominant and to students’ improvement of cooperation

skills. Improvement of school climate was also mentioned as an explanation for improvement

in CL (more transparency). Several statements referred to the fact that the team was just good

at working independently.

The ineffective teams explained their ineffective CL in terms of concerns about task

characteristics (Some teams worked before the period of data-collection at tasks that were not

real group tasks), teacher behavior (according to some of the participants, some teachers were

just not capable of teaching; their lessons were unstructured, unclear and chaotic), absence of

rewards and school context (school structure was experienced as chaotic).

More negative statements referring to contextual factors were made in the ineffective

CL teams than in the effective teams. Thirteen statements referred to the curriculum. Many

tasks were not real group tasks, they could easily be accomplished as individual tasks and no

consultations were required to complete the task. The group tasks were often experienced as

boring, non-supportive for a future career, too easy or unstructured. “I think that the lessons

72

de-motivate me. The tasks are just dead boring. Often I do not understand the purpose of the

lessons”. Teacher behavior was mentioned as a reason for ineffective CL twelve times.

Complaints varied from too little coaching “I would like to get more support from the teacher.

I mean … When I have a question and this person just gives you a book and says “It is in

here” and walks away. What kind of help is this”, to chaotic lessons “I really would like to

know what the purpose is, I think that when she (the teacher) changes her teaching method all

of a sudden, that she can not expect that the class will take her seriously”, and boring

teaching methods “She is reading the book, not knowing other ways to explain what we

should do and how”. It was found that in case of too much autonomy, students were of the

opinion that more teacher involvement was required. This was the situation in the ICT-class.

Students were unsatisfied with the quality and quantity of help they received from teachers.

They were unable to get support from the teacher when they were looking for help. For these

students autonomy tended to correspond to teacher negligence. Group composition was

mentioned seven times as an explanation of task-irrelevant engagement. Problems ranged

from regular absenteeism of one of the group members or the dominance of group members

who were absolutely not motivated. One team mentioned that they were rewarded

individually while they worked on a group project. This was perceived as demotivating. Six

students blamed the chaotic and unstructured school climate for ineffective CL instances. A

statement referring to this issue is “The organization at school is dreadful, you never know

what to do, nothing gets settled, and I get so annoyed by that”. Some teams acknowledged

that they chatted too much or were not serious enough (6 statements). Two of the ineffective

CL teams admitted that the most important reason for ineffective CL was that they were just

not capable of working independently.

Case study of an effective CL team

In order to illustrate the role of goal preferences and students’ perceptions of contextual

factors in the classroom on the quality of CL, we will conclude the result section with a case

study of a team (n = 6) that displayed high task-relevant engagement (M = 3.4, SD = .90) and

low levels of task-irrelevant behavior (M = 1.5, SD = .77). The mean observation score of

socially oriented task engagement was 3.2, SD = .86. We will compare this team with an

ineffective CL team. The second team (n = 3) displayed low levels of task-relevant

engagement (M = 1.4, SD = .72), high levels of task-irrelevant engagement (M = 3.5, SD =

.07) and low levels of socially oriented task engagement (M = 1.5, SD = .50). This team

73

initially consisted of four male ICT-students. One group member cuts classes on a regular

basis and was therefore omitted from this in-depth study.

The effective team consisted of retail and administration students who were involved

in the virtual company (see method section). The team consisted of five male students and one

female. Interesting to note is that interviews with these students revealed that their goal

preferences were very similar. All team members stated they learned to get a certificate (To

get the certificate that is my only goal”). However, most of the time certificate goals were

mentioned in combination with mastery and social responsibility goals (“I want to obtain my

certificate and do it with the team. You will be a lot more experienced after this”). Social

support goals were mentioned as well (“We think that everybody in the team deserves the

certificate, so you’ll try to help and motivate each other. You do want him or her to receive

his or her certificate as well”). Inspection of the GIFI goal profile showed that, as expected,

social support (M = 4.7, SD = .39), and academic learning (M = 4.6, SD =.36) were the most

popular goal domains in this team.

The group composition was mentioned as one of the main reasons for an improvement

of the quality of cooperative learning processes (“The group is smaller now, the weakest

persons left the team and we are now willing to sacrifice more for each other. Help each

other with things. Like “Can I help you?” And we feel far more responsible for each other.

You help each other and it is a win/win situation”). At the time the project started ten team

members were involved, the least motivated students left the team in the course of the school

year. Task-irrelevant engagement was common in the beginning (“We just sat together for

three hours and actually worked for 10 minutes at the most, the rest of the time we were

chatting”). Conflicts were also common at that time. The following statement referred to the

behavior of a person that had left the team (“He never kept his promises. He did not do

anything. Time and again he would promise to do things, but he never did. If we said

something about his behavior, he got angry”). A person who is still a member of the team and

showed little effort in the beginning of the project said “Well, I said my life was very busy

then, but others said you simply had other priorities”. At the time that the team got smaller

and the team members started acting more responsibly this person’s behavior improved

significantly. In the beginning this team was not very good at solving conflicts (“We were too

afraid to start a major dispute, which could spread discord in the team”). When the team got

smaller and students got used to each other, conflicts were solved more easily. At the end they

knew each other much better and comprehended how to treat each other and what to expect of

each team member. Their fear of saying something about a group members’ task-irrelevant

74

engagement decreased over time. Another influence on the QCL was teacher behavior. In this

particular case the teacher initially gave them too little autonomy which made them passive.

One team member said: “He guided too much, he was very dominant”. Later on, the teacher

got less involved and the QCL improved. The team members increasingly took up their own

responsibilities.

Case study of an ineffective CL team

Goals that students in the ineffective CL team mentioned most frequently were academic/ to

get a certificate goals, statements that captured this goal are “The only reason why I am here

is to get my certificate, I am not interested in what we learn, because that is just dead boring”

Academic/ to prepare for a future education goals were also frequently mentioned. As one of

the team members puts it: “It’s about getting my certificate, not about the stuff I learn here,

because that is just not very interesting, I want to go to another program type after this

study”. Inspection of the GIFI goal profile showed that, in line with previous findings,

belongingness (M = 5, SD = .00) and affective goals were most prevalent (M = 4.9, SD = .10)

for the team.

The high level of task-irrelevant engagement in this team was strongly related to the

negative behavior of the group member who was often absent. A statement referring to this

issue is: “T. really isn’t motivated, he does not show up a lot and yes…that influences the

group as a whole, it demotivates the group as well”. It was very difficult to plan group

sessions in this team, because of T’s high rate of absenteeism. In the end, they just decided to

work without T. A second reason for high task irrelevant engagement scores in this group was

the behavior of another group member, who -although physically present- showed very low

engagement during CL. His task-irrelevant behavior can be explained by the fact that he

actually wants to become an actor, and admitted he chose the wrong school. Again teacher

support was mentioned as an important factor that influenced engagement levels during CL.

This team (as was the whole class) was not satisfied with the quality and quantity of their

teachers’ support. Teachers were often unavailable when their help was required. Also the

type of project that they were working on was not a real group project. According to these

students (and other ICT students), many aspects of the project had to be prepared individually

and they obtained individual grades afterwards, whereas the project was introduced as a group

project. Furthermore, they were not interested in the curriculum itself, they said that it was not

75

challenging and interesting. A statement illustrating this is “Last year I started this course.

When you ask me now what I have learned so far, I must say that it is close to nothing”.

DISCUSSION

Students’ goal preferences in effective and ineffective CL teams

Effective CL processes seem to be associated with students’ goal preferences as well as with

their perception of contextual aspects and the appropriate context. We predicted task-relevant

engagement to be dependent on a combination of social goals and mastery goals. We found

that the difference between effective and ineffective CL teams lies precisely in the

combination of these goals. We also found that it is essential to distinguish between different

types of social goals. Comparison of effective and ineffective team members’ self-reports

suggested that belongingness goals might be more popular in ineffective teams. A negative

relationship between students’ belongingness (GIFI) goals and socially oriented task

engagement in the ineffective teams was found. On first sight this finding may seem strange

but it actually corroborates the findings presented by Wentzel (1989), namely that low-

achievers consider making friends and having fun as valuable goals and that pursuing these

goals might interfere with task-relevant engagement. It is important to keep in mind that this

result only concerns ineffective CL teams. Perhaps these students are more concerned with

their well-being and therefore preoccupied with the social environment and cues that signal

threats to friendships. The analysis of the goal profiles supports this notion; it showed that in

ineffective teams, entertainment and work-avoidance goals are popular goals, while in

effective teams mastery goals are equally important as certificate goals. Another valuable

difference regarded the prevalence of social responsibility goals in the effective teams.

Students in these teams reasoned that in their future career people will also expect them to be

able to cooperate, while ineffective teams did not mention these goals it at all. Effective CL

teams seem to be engaged in learning because they have a genuine interest in what they learn

and they understand the value of CL. This finding seems in line with findings of a study of

Levy, Kaplan, and Patrick (2004) who found that mastery oriented students’ evaluated

cooperation in the first place in terms of its contribution to their academic goals and were less

concerned with social relationships.

Students in ineffective teams seem less conscious of their goal preferences than

students in effective teams. Their goals were very broad. Although getting a certificate was in

76

both type of teams one of the most popular goals, how and why they wanted to achieve that

goal seemed to be of inferior importance in the ineffective group. A significant finding of this

study is that not all students devote much thinking to choosing their goals and think about

their goal systems (Conti, 2000). Earlier research suggested that formulating goals, especially

goals that connect with and well-represent one’s sense of self, can facilitate students’ intrinsic

motivation (Cantor et al, 1987; Sheldon & Kasser, 1995; Sheldon & Elliot, 1997). Hence,

formulating goals in the classroom ought to be put on the teacher’s agenda.

Contextual factors and engagement in CL

One of the major points made in this article is that students’ engagements cannot be

understood separately from the environment. In short, the type of task, the teacher, and the

group composition were the most frequently mentioned factors for explaining effective or

ineffective instances of CL. Effective and ineffective teams clearly had different appraisals of

contextual factors that impacted on their quality of CL.

Ineffective teams explained their task-irrelevant engagement as a result of the group

tasks. According to these students, many tasks were not genuine group tasks. They

complained that many group tasks were boring, not supportive for a future career, too easy

and not challenging enough.

Ineffective team members had many complaints about their teachers, who were often

not there in case they needed help. A healthy balance between teacher guidance and student

autonomy seems crucial for CL. Skinner and Belmont (1993) argued that student engagement

is optimized when the social context fulfills children’s basic psychological needs. These

include the needs to be competent, autonomous, and related to other people. The feeling of

competence is influenced by the amount of structure the teacher provides by communicating

clearly about expectations and responding consistently and predictable to students’ questions.

Students in ineffective teams expected more assistance while cooperating. They also

experienced a lack of explanations on the purpose of the lessons. Teachers’ sloppiness in this

respect was a recurring complaint. Students often did not know how to proceed because

teachers were not clear enough about the purpose of the tasks, before they started. The need

for autonomy is promoted when students experience autonomy support. However, more

teacher involvement was required in a situation of too much autonomy as in the ICT-class.

Teachers need to economize on autonomy generating learning situations, reserving that type

of learning environment for situations where students are used to cooperate. At such a point

77

they can decrease guidance and increase students’ responsibility in the learning process (see

also Boekaerts & Martens, in press).

Students’ need for relatedness (Skinner & Belmont, 1993) is associated with the level

of involvement, which refers to ‘the quality of the interpersonal relationship with teachers and

peers”. The group composition was very important in explaining successful CL. Students in

effective teams usually work longer in the same teams and feel more at ease with each other.

Negative effects on the quality of cooperative learning may also arise when the group is too

cohesive. Under such conditions group members might conform to group norms that are

adverse to learning, in which case erroneous or incomplete solutions for problems or a lack of

intersubjectivity may go unnoticed (e.g., Kanselaar & Van der Linden, 1984). However, it is

a long way before cohesion becomes a problem. This study primarily points at a lack of

cohesion as a motive for problematic CL. In ineffective teams, students had to deal with

absent or strongly demotivated group members more often.

The quality of CL also depends on students’ general abilities to cooperate. Students in

ineffective teams mentioned the fact that they were just not good at working independently.

Also these teams mentioned that they had chatted too much, meaning that for them

entertaining and belongingness goals were more important than mastery goals. In effective

teams students explained the successful CL, by the fact that they were simply good at CL,

while in the ineffective teams students admitted that they were not capable of dealing with the

independence that comes with CL. This finding implies that it is important to teach students

the skills and knowledge to cooperate and for teachers to guide the CL process along (Gillies

& Ashman, 1996; Hoek, Van den Eeden, & Terwel, 1999; Oortwijn, Boekaerts, & Vedder, in

preparation; Webb & Farivar, 1994).

Interesting to note is that ineffective and effective teams came from the same

classrooms and were therefore enrolled in the same CL settings, therefore one contextual

factor can not solely explain the quality of CL. We assume that the combination of the above

mentioned contextual factors is crucial in explaining the quality of CL. For example the tasks

may be boring at some point, but with a motivated teacher and team members, students may

still be capable to self-regulate the motivation process (Boekaerts, 2005).

Recommendations

The goals that were identified in this study were spontaneously mentioned student goals.

They differ somewhat from the goals that were specified prior to the research in the GIFI.

78

Lemos (1996) suggested that students bring a set of general goals to the classroom and

implement these when faced with the real classroom setting. We think that the general goals

that students bring to the classroom are rather abstract goals and that they still need to adapt

them to the local CL conditions. The interviews and the stimulated recall sessions allowed us

to register context-sensitive goals because, the setting possibly gave a type of context and

relevance to thinking about goals that made students more aware of their goals and their

importance in the specific context of the CL. If indeed reflecting on and formulating goals is

important, than the stimulated recall setting situation in this study might be seen as a proto

type approach to encouraging it throughout the school program in order to make students

more aware of their goals. Ideally, reflection on students’ goal preferences should be a

recurring element of the curriculum, especially in secondary vocational education where the

drop-out rate is very high. As explained in the introduction, dropping out of school often is an

outcome of underlying motivational problems that students have experienced, which might

have been caused by a mismatch between students’ goal preferences and school goals. Goal

reflection can be a means for enhancing students’ motivation, so that they become more aware

of the goals they want to achieve and think of strategies how to achieve these. Discussing goal

preferences in order to make students aware of the role these goals play in the learning

process might be an important step towards more successful CL (Hijzen, Boekaerts, &

Vedder, 2006). Students should be invited to think about their own goals and about the links

between goal preferences and the goals presented to them by teachers, instruction methods,

course books, and other students. Such reflection might help them to adopt teacher-set

learning goals and self-regulate their learning more efficiently (see Boekaerts & Corno,

2005).

CL may be a way to enhance students’ motivation for learning, provided that students

are aware of their goal preferences and CL is well implemented. In order to predict the quality

of students’ CL processes more accurately, more (longitudinal) research on contextual factors

is needed. Important to note is that the quality of CL differed within settings, therefore a

combination of personal goal preferences and contextual factors might explain the quality of

CL best. Some teams are able and willing to cope with obstacles, distractions and draw backs,

while others are not. As we discussed previously, this highly depends on what goal

preferences they had in the first place. If, for example, students’ most important goal is to

have fun, they will easily get distracted by ill designed group tasks and poor teacher behavior,

while students whose mastery goals are salient might try harder to cope with ill-structured CL

settings. Nevertheless, a well-designed CL setting is crucial, because it triggers, promotes,

79

stimulates or hampers certain goal preferences. Although the comment that we started the

article with “It is just dead boring at school, I don’t think I actually learn something at all, I

hope I will at least learn something in my traineeship” reflected a lot of dissatisfaction and

seemed very negative, it also gives reason for hope. It means that students are in school

because they want to learn something. Schools are no unchangeable institutions and this study

yielded some interesting and specific leads, perceived and stated by students themselves,

which may promote better CL and increase students’ general motivation for school.

80

References

Ames, C. (1992). Classrooms: Goals, structures, and student motivation. Journal of

Educational Psychology, 84, 261-271.

Ames, C., & Archer, J. (1988). Achievement goals in the classroom. Students’ learning

strategies and motivation processes. Journal of Educational Psychology, 80, 260-267.

Blumenfeld, P. C. (1992). Classroom learning and motivation: Clarifying and expanding goal

theory. Journal of Educational Psychology, 84, 272-281.

Boekaerts M. (2005). Self-Regulation: with a focus on the regulation of motivation and effort.

In W. Damon & R. Lerner (Series Editors) Handbook of Child Psychology, Vol 4.,

Child Psychology in Practice (I. E. Sigel & K. A. Renninger, volume editors). New-

York: Wiley.

Boekaerts, M., & Corno, L. (2005). Self-regulation in the classroom: A perspective on

assessment and intervention. Applied Psychology: An International Review.

Boekaerts, M., & Martens, R . (in press) Motivated learning: What it is and How it can be

Enhanced. In L., Verschaffel, F. Dochy, M. Boekaerts, M, & S. Vosniadou (Eds).

Instructional psychology: Past, present and future trends. Handbook in the Advances

in Learning and Instruction Series, Amsterdam: Pergamon.

Brown, A., & Palincsar, A. (1989). Guided, cooperative learning and individual knowledge

acquisition. In L. B. Resnick (Ed.), Knowing, learning and instruction: Essays in honor

of Robert Glaser (pp. 393-451). Hillsdale, NJ: Lawrence Erlbaum.

Cantor, N., Norem, J., Niedenthal, P., Langston, C., & Brower, A. (1987). Life tasks, self-

concept ideal, and cognitive strategies in a life transition. Journal of Personality and

Social Psychology, 53, 1178-1191.

Chapman, E. (2003). Alternative approaches to assessing student engagement rates. Practical

Assessment, Research & Evaluation, 8(13). Retrieved June 8, 2005 from

http://PAREonline.net/getvn.asp?v=8&n=13.

Chin, W., Salisbury, D., Pearson, A., & Stollak, M. (1999). Perceived cohesion in small

groups. Small Group Research, 30, 751-766.

Cohen, E. G. (1994). Restructuring the classroom: Conditions for productive small groups.

Review of Educational Research, 64 (1) 1-35.

Conti, R. (2000). College goals: Do self-determined and carefully considered goals predict

intrinsic motivation, academic performance, and adjustment during the first semester?

Social Psychology of Education, 4, 189-211.

81

Dowson, M., & McInerney, D. M. (2001). Psychological parameters of students’ social and

work avoidance goals: A qualitative investigation. Journal of Educational Psychology,

93, 35-42.

Dweck, C. (1986). Motivational processes affecting learning. American Psychologist, 41,

1040-1048.

Elliot, A. J., & Sheldon, K. M. (1998). Avoidance personal goals and the personality-illness

relationship. Journal of Personality and Social Psychology, 75, 1282-1299.

Eurydice database on education (n.d.). Retrieved November, 5, 2004, from

http://www.eurydice.org.

Ford, M. E. (1992). Motivating humans: goals, emotions, and personal agency.

Newbury Park, CA: Sage.

Ford, M. E., & Nichols, C. W. (1991). Using a goal assessment to identify motivational

patterns and facilitate behavorial regulation and achievement. In M. Maehr & P. R.

Pintrich (Eds), Advances in motivation and achievement: Vol 7. Goals and self-

regulatory processes (pp. 61-84). Greenwich, CT: JAI Press.

Gillies, R., & Ashman, A. (1996). Teaching collaborative skills to primary school

children in classroom-based workgroups. Learning and Instruction, 6, 187-200

Hertz-Lazorowitz, R., & Miller, N. (1992). Interaction in cooperative groups.

New York: Cambridge University Press.

Hijzen, D., Boekaerts, M., & Vedder, P. (2006). The relationship between the quality

of cooperative learning, students’ goal preferences and perceptions of contextual

factors in the classroom. Scandinavian Journal of Psychology, 47, 9-21.

Hijzen, D., Boekaerts, M., & Vedder, P. (submitted). Instructing cooperative learning:

teacher related conditions steering effective cooperative learning processes of

students in secondary vocational schools

Hoek, D., Van den Eeden, P., & Terwel, J. (1999). The effects of integrated social and

cognitive strategy instruction on mathematics achievements in secondary education.

Learning and Instruction, 9, 427,-448.

Kanselaar, G., & Van der Linden, J. (1984). Leren door samenwerken. In J. Lowyck and H,H.

Tillema (Eds.). Vormen van leren en onderwijzen in de klas [Learning through

collaborative learning]. ( pp 39-59). Lisse: Swets & Zeitlinger.

Lemos, M. S. (1996). Student’s and teacher’s goals in the classroom. Learning and

Instruction, 6, 151-171.

Lepper, M. R., & Hodell, M. (1989). Intrinsic motivation in the classroom. In C. Ames & R.

82

E. Ames (Eds.), Research on motivation in education (Vol. 3, pp. 73-105). New York:

Academic Press.

Levy, I., Kaplan, A., & Patrick, H. (2004). Early adolescents' achievement goals, social

status, and attitudes towards cooperation with peers. Social Psychology of Education,

7, 127-159.

McInerney, D. M., Hinkley, H., Dowson, M., & Van Etten, S. (1998). Children’s beliefs

about success in the classroom: Are there cultural differences? Journal of Educational

Psychology, 90, 621-629.

Midgley, C., & Urdan, T. C. (1992). The transition to middle level schools: Making it a good

experience for all students. Middle School Journal, 24, 5-14.

Oortwijn, M., Boekaerts. M., & Vedder, P. (in preparation). The impact of teacher-guidance

on elementary aged students during a cooperative learning curriculum in math.

Ryan, A. M. (2000). Peer groups as a context for the socialization of adolescents’

motivation, engagement, and achievement in school. Educational Psychologist, 35,

101-111.

School Inspectorate (2002). Vroegtijdig schoolverlaten in het middelbaar beroepsonderwijs

[Drop-out in secondary vocational schools]. Utrecht: Inspectie van het Onderwijs.

Schwartz, S. H., & Bardi, A. (2001). Value hierarchies across cultures: taking a similarities

perspective. Journal of Cross-Cultural Psychology, 32, 268-290.

Sharan, Y., & Sharan, S. (1992). Expanding Cooperative Learning Through Group

Investigation. New York: Teachers College Press.

Sheldon, M., & Kasser, T. ( 1̀995). Coherence and congruence: Two aspects of personality

integration. Journal of Personality and Social Psychology, 68, 531-543.

Sheldon, M., & Elliot, A. (1997). Not all personal goal goals are personal: Comparing

autonomous and controlled reasons for goals as predictors of effort and attainment.

Personality and Social Psychology Bulletin, 24, 546-557.

Skinner, E., & Belmont, M. (1993). Motivation in classroom; Reciprocal effects of teacher

behavior and student engagement across the school year. Journal of

Educational Psychology, 85, 571-581.

Voncken, E., Van der Kuip, I., Moerkamp, T., & Felix, C. (2000). Je bent jong en je weet

niet wat je wilt: Een inventarisatie van push- en pullfactoren die leiden tot voortijdig

schoolverlaten in de BVE-sector [You are young and you do not know what you want:

an inventory of push- and pull factors that lead to drop-out in secondary vocational

education] . Amsterdam: SCO-Kohnstamm.

83

Webb, N. M., & Farivar, S. (1994). Promoting helping behavior in cooperative small

groups in middle school mathematics. American Educational Research Journal, 31,

369-395.

Webb, N. M., & Palincsar, A. S. (1996). Group processes in the classroom. In: D.C.

Berliner & R. C. Calfee (Eds), Handbook of Educational Psychology (pp. 841 - 873).

New York: Simon & Schuster MacMillan.

Wellborn, J. G. (1991). Engaged vs. disaffected action: Conceptualization and measurement

of motivation in the academic domain. Unpublished Dissertation, Graduate School of

Human Development and Education, University of Rochester, Rochester, NY.

Wentzel, K. R. (1989). Adolescent classroom goals, standards for performance, and academic

achievement: An interactionist perspective. Journal of Educational Psychology, 81,

131-142.

84

Table 1

Student Statements referring to Goal Preferences

Goal preferences Sample from the interview

Achievement/ mastery

Academic Learning “Yes, I like to learn new things all the time”. “I am here to learn stuff, so I will

be able to do it myself later in my career”.

Achievement/ performance

Academic/ to get a certificate “It want to engage in school tasks, because getting my certificate is very

important, just to have something in my pocket, eh well yes it is just handy and

easy. I don’t know yet what I want to do later”.

Superiority/ Individuality ‘I want to be a better student than others”

Academic/ to prepare for a

future education

“Before this school I was in Pre Vocational Education, and eventually I want to

go to higher vocational education. So this is just a stepping stone”.

Academic/ complying “I will engage in the group task, otherwise I get kicked out of the classroom”

(see Lemos, 1996).

Academic/ working “So that I don’t have to do homework” (see Lemos, 1996).

Work avoidance “Minimize the effort, I just try to do as little as possible” (See Dowson &

McInerney, 2001).

Well-being goal domain

Entertainment “I want to enjoy myself (laugh) as much as possible. I would like to say that I

enjoyed my time at school” (compare enjoyment goals of Ford, 1992; Lemos,

1996).

Affective Goals “I want to feel at ease and self-confident, I express myself better than”

Social Goal domain

Social Responsibility “It’s all about cooperation, alone you are nothing. Therefore it is important to

learn how to do it” (Ford, 1992).

Social Support “It is important to support each other, ask questions like “Can I help you?”

Belongingness “I think it is very important that you like your group members”

Self-Determination

Self-Determination “I like to use my creativity and think for myself. If I already know that we have

to write up some boring things and I can see no challenge at all, then I lose my

motivation”.

Context

Type of Task “The task is not really a group task, otherwise we would have to ask each other

questions, and answer them. This is just answering questions from a book and

85

that is not what I consider group work”

Teacher Behavior “I want more guidance; I think we have too much freedom now. She (the

teacher) was never there”

Group Composition Since we got to know each other better, the quality of our cooperation is much

better. In this group we all like each other”

Reward “I think it is good that with a group task, everybody still receives an individual

reward, so you know exactly who did what and when”. “We work much harder

when we receive a reward afterwards”

86

Table 2

Effective teams’ (N = 4, 17 persons ) and ineffective teams’ (N = 4, 14 persons) engagement

scores

CL Engagement (observation) Effective teams Ineffective teams

Task-relevant engagement 3.46 .87 2.98 1.24

Task-irrelevant engagement 1.57 .86 1.92 1.01

Socially oriented task engagement 3.18* .88 2.02 .80

87

Table 3

Effective (N = 18) and ineffective teams members’ GIFI goal preferences (N = 14)

Goal Preferences Effective teams Ineffective teams

Affective goals 4.48 .72 4.30 .69

Social Support 4.42 .73 3.86 .73

Mastery 4.29 .60 3.91 .85

Self Determination 4.43 .53 4.25 .62

Belongingness 3.80 1.07 4.52 .74

Material Gain 4.05 .77 3.94 .66

Superiority 3.50 .96 3.16 .92

88

master

y

soc r

esp.

learn

ing fo

r a ce

rtifica

te

enter

tainm

ent

learn

ing fo

r fut.

edu.

belon

gingn

ess

affec

tive

supe

riority

aca w

orkin

g goa

l

socia

l sup

port

work a

voida

nce

0%

10%

20%

30%

40%

50%

effective CL

ineffective CL

Figure 1

Goal Profiles (Interviews) of Effective CL Teams (N = 4) and Ineffective CL Teams (N = 4)

89

Chapter 4

INSTRUCTING COOPERATIVE LEARNING; TEACHER RELATED CONDITIONS

STEERING EFFECTIVE COOPERATIVE LEARNING PROCESSES OF STUDENTS

IN SECONDARY VOCATIONAL SCHOOLS 3

Abstract

In this study, we examined the relationship between the quality of cooperative learning (CL)

processes and students’ perception of conditions for the quality of CL steered by teachers, in

the course of a year. Subjects were 1016 students in secondary vocational schools. Two

questionnaires were used. The first measured students’ perceptions on the extent that they

have been taught skills and knowledge for CL, teachers clarity on rules for CL and teachers

monitoring, intervention and evaluation behavior. The second questionnaire measured

students’ perceptions on the quality of CL. It was found that the extents that students were

taught skills and knowledge for CL and teachers’ clarity on rules for CL were highly related

to the quality of CL, during all three waves. The quality of CL was evaluated most favorably

at its best during wave II when scores on all teacher related conditions were also highest. We

distinguished between a group of ineffective and effective cooperators that completed

questionnaires at all three waves and found that effective cooperators had higher scores on

all scales, in particular on the extent that they were taught skills and knowledge for CL.

Key words: teacher behavior, cooperative learning, secondary vocational schools

3 This chapter is based on: Hijzen, Boekaerts and Vedder. Instructing cooperative learning: teacher related conditions steering effective cooperative learning processes of students in secondary vocational education: Manuscript submitted for publication in: Teachers and Teaching: Theory and Practice

90

INTRODUCTION

Cooperative Learning (CL) methods are generally acknowledged to promote deep level

processing, positive peer relationships, social skills, positive attitudes and increased

motivation towards school subjects, as well as interracial acceptance (e.g., Cohen, 1994;

Slavin, 1995; Webb & Palincsar, 1996). Made enthusiastic by this resourcefulness, many

secondary vocational schools experiment with and introduce a wide range of cooperative

instructional methods. Unfortunately, in daily classroom practice many attempts to stimulate

students to engage in CL fail. During CL, students can easily hide in the group and free-ride

with the more active group members (Cohen, 1994; Salomon & Globerson, 1989), resulting in

a reduction of effort. Although placed in groups, in reality students often work individually

instead of together (Vedder, 1985; Veenman, Kenter, & Post, 2000; Witteman, 2003).

Furthermore, some students perceive working in groups mainly as an opportunity to chat and

disturb each other’s learning processes (Shanahan, 1998).

In the present study we examined the learning conditions that teachers in secondary

vocational education create in order to promote students’ CL processes. The study is

embedded in a larger project on motivational self-regulation strategies of students’ enrolled in

Dutch secondary vocational education. In previous studies (Hijzen, Boekaerts, & Vedder,

2006) it was found that the quality of CL was related to students’ goal structure.

Belongingness goals (e.g., I want to get along with my peers), mastery goals (e.g., I want to

learn more about my profession) and social support goals (e.g., I want to help others in case

they need help) were found to be related to the way students cooperate. Especially students

that gave a high rating for these latter goals perceived a high quality of their CL processes.

Furthermore, the availability of emotional and academic peer support predicted the quality of

CL. In the same study we explored classroom context variables, such as task characteristics,

the group composition, reward structures and students’ perceptions on teacher behavior. The

study showed that -apart from students’ goal preferences- contextual factors that were related

to teacher behavior predicted the quality of CL. Students indicated that the extent that they

were taught knowledge and skills for effective CL, as well as teachers’ clarity on rules for CL

were important variables determining the quality of CL. A previous in-depth study (Hijzen,

Boekaerts, & Vedder, in press) also revealed the importance of teacher behavior in

influencing students’ engagement levels during CL. For example, students reported that their

motivation to engage in CL was strongly determined by teachers’ instructional behavior. A

typical statement that referred to this issue was: “I would definitely like to get more teacher

91

support. I mean … that when I have a question and this person just gives you a book and says

here it is and walks away, what kind of help is that?” In our opinion the role of the teacher

has been underestimated in many studies; the focus has been too much on the assumption that

students consciously regulate and steer their learning processes themselves, independent of

what teachers do.

The present study was set up to further explore the relationship between students’ perceptions

on teachers’ behavior in relation to the quality of CL in the course of a year. In previous

studies (Hijzen, Boekaerts, & Vedder, 2006) we focused on these relationships in a cross-

sectional design. In the present study we investigate whether changes or stability in students’

perceptions of teacher related conditions for CL over time coincide with changes or stability

in the quality of CL. Insight into the relationship between teachers’ instructional behavior and

the quality of CL may be helpful for determining leads for future intervention purposes,

directed at improving CL in secondary vocational schools. Various studies (e.g., Boekaerts,

de Koning, & Vedder, 2006; Cohen, 1994; Hijzen, Boekaerts, & Vedder, 2006; Slavin 1995,

1996; Webb & Palincsar, 1996) investigated and reviewed effects of particular conditions on

the quality of CL. In earlier studies we found that the types of teacher related conditions for

CL related to the quality of CL, concerned students’ perceptions on teacher control behavior

(monitoring, intervention and evaluation instances) and instructional behavior (the extent that

students were taught rules for CL and the extent that they were taught knowledge and skills

for CL).

In the present study, students’ perceptions on teacher related conditions for CL and the

quality of CL will be measured three times in the course of a year. Conceivably, some types

of teacher related conditions are less important in a later stage than at the beginning of

working in CL teams. For example, we hypothesize that monitoring behavior will become

less significant to students after they become more experienced in CL. Next, we will describe

the (expected) relationships between the quality of CL and teachers’ control behavior and the

extent that they taught skills and rules for CL. We will conclude this section with our

expectations and hypotheses on stability and change of these relationships.

Teachers’ control behavior

In the last two decades the classroom setting in senior vocational schools in the Netherlands

has gradually changed from an exclusively, traditional, competitive, and individualistic

92

educational setting to a setting with more attention on cooperative learning that requires and

stimulates both cognitive and motivational self-regulation skills (Boekaerts & Minnaert,

2003). Cooperative learning refers to ‘a set of instructional methods in which students are

encouraged and required to work together on academic tasks’ (Slavin, 1987). The term

usually refers to alternative ways of organizing classrooms that contrast with individualistic

and competitive classroom organizations (Webb & Palinscar, 1996).

While learning activities are always performed by students, teachers can engage in a

wide range of behaviors to facilitate and regulate students’ behavior in the classroom

(Brophey & Good, 1986; Resnick, 1989; Vermunt & Verloop, 1999). The way the teacher

instructs and guides the CL process influences students’ engagement in CL and academic

achievements. Students’ CL processes and performance are positively affected when teachers

encourage high level cognitive interactions (i.e. discussions, explaining and reflecting on

problem solving strategies) (Chinn, O’Donnell, & Jinks, 2000; Webb, Troper, & Fall, 1995).

It is suggested that teachers should only provide help on request. Direct supervision during

CL processes is not desirable; students’ feelings of responsibility will decrease (Cohen, 1994).

Ames (1992) found that a mastery goal orientation will be promoted when teachers allow

students to participate in decision-making, share responsibility and grant them increasing

independence with regard to the learning process. A mastery goal orientation is associated

with high levels of engagement, responsibility and academic achievements (Deci & Ryan,

2000).

It has also been found that teachers should monitor students learning and intervene to

provide assistance (Johnson & Johnson, 2002). More specifically, teachers need to increase

students’ group skills, as long as students lack experience with CL methods and prevent off-

task behavior. Before students start working on the group task, the teacher should clearly

specify the objectives and rules for the lesson both with respect to social and academic

aspects of the task. Furthermore, it is also important that teachers evaluate the group

processes with the CL teams afterwards. Finally teachers should diminish the level of

coaching in the course of learning processes, so that students increasingly take responsibility

for their own learning processes.

Teachers’ instructional behavior

Students need a number of cooperation skills, such as the skills to express their own opinion,

stimulate each other, provide and receive help, listen to each other and clarify their current

93

understanding of the task (Cohen, 1994; Ros, 1994; Webb & Palincsar, 1996). It is important

to acknowledge that students' early attempts at regulating their own work may not always be

successful. Effective CL requires practice. Comprehensive programs of team building and

prosocial skills development seem to improve peer-to-peer interaction skills (Webb &

Palincsar, 1996). Many other scholars have shown that explicit teaching of CL skills supports

an improvement of the quality of CL (e.g., Gillies & Ashman, 1996; Hoek, Van den Eeden, &

Terwel, 1999; Webb & Farivar, 1994). Gillies (2003) mentioned the following skills: actively

listening, considering the other person’s perspective on issues, stating ideas freely, being

responsible for one’s own behavior and constructively critiquing the ideas presented as

interpersonal skills that facilitate communication. Teachers need to model positive

interpersonal skills, have students practice the skills, and encourage the students to process

how effectively they are performing the skills (Phipps, Phipps, Kask, & Higgins, 2001).

Stability of conditions and the quality of CL

The study has a longitudinal design in order to investigate the stability of conditions and the

quality of CL over time. We hypothesized that it is crucial that teachers keep paying attention

to teaching skills, knowledge and rules for CL during all three data-waves. In order to become

independent cooperators it is essential that students need to know how to cooperate (e.g.,

Webb & Palincsar, 1996). However, with regard to teachers’ control behavior we expect that

decreasing control promotes effective CL over time. We expect that students become more

independent and skilled at working in CL settings, over time. Therefore, we argue that

students need more extrinsic stimulation (as in teachers monitoring, and interventions)

initially, but that their stimulation should be decreased once they have become sufficiently

skillful (e.g., Johnson & Johnson, 2002). We expect differences between effective and

ineffective cooperators; we expect that ineffective cooperators need to be stimulated more

than effective cooperators. The latter need their teachers to monitor their CL processes and

intervene during a prolonged period.

Research question

The main question of the present study is:

94

“Which teacher related conditions are related to the quality of CL processes, and are these

relationships stable in the course of a year?” We expect that a stability of teaching students’

skills and knowledge and rules for CL, and a decrease in teachers’ control behavior during the

second and third data wave correspond to an increase in the quality of CL from the first to the

third wave.

METHOD

Subjects

Participants in the study were 1016 students from 11 different schools for secondary

vocational education in the Netherlands. The Netherlands have 42 regional educational

centers for secondary vocational education. They all received a letter in which we explained

the purpose and relevance of the study and invited them to participate. Non-participating

schools had a variety of reasons. The most frequent reasons concerned the time investment of

students and teachers and the extra organizational burden of participation in yet another

research project. The eleven schools that participated were spread evenly across the

Netherlands. Schools for secondary vocational education offer a variety of programs,

preparing students for particular professional careers. Secondary vocational school starts for

most students at the age of 16 after they completed a junior vocational school. Senior

vocational school delivers educational programs for four broad competency levels. At the first

level students train to become assistants (6 to 12 months). At the second level they have two

to three year courses for basic vocational training. At the third level students are enrolled in

professional training and at the fourth level they participate in middle-management training (3

to 4 years) or in specialized training courses (1 to 2 years). Most students finishing secondary

vocational school prefer to enter the labor market and do not proceed in further studies4. We

distinguish engineering, ICT, retail and administration, food and tourism and health and

welfare programs. Participating students were students that worked at least 45 minutes per

week in CL settings. Table 1 presents the distribution of participating students by program

type and level.

--------------------------

Insert Table 1 here

--------------------------

4 For further information on secondary vocational education in the Netherlands, see Euridice database on education, 2003 website.

95

A closer look at Table 1 shows that most students were enrolled in health and welfare and

level 4 programs. Only a few students were enrolled in level 1 and food and tourism

programs. For exploring longitudinal data we selected students that participated in all three

data-waves. Only 260 students completed all questionnaires at all measurement points. The

loss of students between measurement points was considerable, partly due to students’

decisions not to participate in all rounds of data collection and partly because students

changed schools or program type or were absent for other reasons. However, since students

from different program types and levels were not equally distributed (for example, only three

ICT-students completed all three questionnaires), we limited analyses to the group that was

most representative to the total sample and which was also the largest group; the health and

welfare students (N = 120).

Instruments

Table 2 presents an overview of scales, sample items and Cronbachs’ alphas of the different

scales used in this study.

--------------------------

Insert Table 2 here

--------------------------

The quality of CL was measured with the questionnaire for the Quality of Cooperative

Learning (QCL). Originally the list comprised four subscales, namely students’ perception of

the quality of group cohesion, students’ perception of interdependence within the group,

students’ perceptions of the quality of their cooperation skills and students’ attitude towards

CL. Items in these subscales were “I feel part of my group”, “In this group we support each

other”, “I know how to support my group members, when they need help”, and “I prefer to

work in a team over working individually”. Students had to indicate on a four-point Likert

scale to what extent they agreed with each statement. Response categories ranged from "I

disagree very strongly" to “I agree very strongly”. These subscales were highly correlated and

together measure the quality of CL. A Principal Component Analysis on these four subscales

resulted in a one-factor solution, representing an overall QCL score. This factor had an

Eigenvalue of 2.4 and it explained 59 % of the total variance.

The questionnaire for the Conditions for CL (CCL) aimed to measure students’

perceptions of the extent that teachers create or maintain conditions for the quality of CL.

Items were based on review studies on the conditions for productive small-groups (e.g.,

96

Cohen 1994; Ros, 1994; Slavin 1995; Webb & Palincsar, 1996). The questionnaire consisted

of five scales. A first scale captured students’ evaluations on the extent that they have been

taught cooperation skills and knowledge at their present schools skills. A second subscale

aimed to measure students’ perceptions on the clarity of the rules for effective CL. Items in a

third scale aimed to assess teacher monitoring behavior. The fourth scale assessed teachers’

interventions in cases of off-task behavior. The fifth scale aimed at measuring teachers’

evaluation methods. Sample items are presented in Table 2. Students were asked to evaluate

the conditions during the last four weeks. Response categories (4) ranged from "I disagree

very strongly" (1) to “I agree very strongly” (4).

Procedure

Questionnaires were administered three times, once during the students’ first year and twice

during their second year, with intervals of six months. Researchers assisted students while

administering the questionnaires. It took students one hour on average to complete the

questionnaires.

RESULTS

First we will present an overview of means and standard deviations of the time spent on CL,

interrater agreement, and the instructional subscales, and the relationship with the quality of

CL as reported for the whole sample at three data-waves (wave I: N = 1016, wave II: N =619,

wave III: N = 424).

Organization characteristics

-----------------------------

Insert Figure 1 here

----------------------------

Amount of time spent on CL

Figure 1 presents the amount of time spent on CL of students in the total sample.

Most students (33 percent) spent approximately 90 minutes per week in CL teams and 23

percent spent at least 45 minutes per day in CL teams.

Earlier we reported about the quality of the instruments. Another aspect concerning

the quality of the data is the interrater agreement, after all the way this study is designed

97

allows exploring to what extent students agree on their evaluations of teachers control and

instructional behavior. We calculated intra-class agreements (ICC) on all three waves, for all

classes. The ICC score (Rho score) is an index of reliability for measurements of the same

phenomenon. With the ICC score we can investigate whether several raters independently

measured the same phenomenon. An instrument with Rho score 1 will produce the same

measurement each time it is used, while a Rho score of 0 will produce different scores each

time it is used (Hayes, Walton, Szomor, & Murrell, 2001). Reliability coefficients ranged

from .80 to 1.00. Interestingly, students within the same classes highly agreed with each

other, only five out of 32 classes had intra-class agreement coefficients that were below .99.

This means that individual students’ scores are indicative of what happens at class level in

terms of conditions for and the regulation of CL.

Students’ perceptions on teacher behavior and the quality of CL at three data waves

------------------------------

Insert Table 3 here

-----------------------------

Table 3 presents mean scores, standard deviations, and correlation coefficients of the five

teacher related conditions subscales and the quality of CL, of the total sample at three waves.

Inspection of Table 3 shows that, students were generally positive on the quality of their CL

processes. The highest scores were found during wave II. Scores on the five subscales of

teachers’ instructional behavior were somewhat lower and seem quite stable over time;

generally scores are higher during wave II and lower during wave I. Scores on teachers’

evaluation behavior during wave III, were quite low.

In order to explore the relationship between the teacher related conditions and the

quality of CL, correlation coefficients between the scales of the conditions and the quality of

CL at three measurement points, were calculated. As described in the introduction during

wave I, we expected positive correlations between all scales of teachers’ instructional

behavior with the quality of CL. We expected that during wave II and wave III the association

between the scales that measured teachers’ control behavior and the quality of CL would be

less significant than during wave I. Because of the large sample all subscales measuring

teacher related conditions were related to the quality of CL. However, we will limit the

discussion to correlation coefficients above .20. In line with our expectations, the quality of

CL was related to the subscales that measured students’ perceptions on the extent that rules

for CL were clearly explained by their teachers and their evaluations on the extent that they

98

were taught CL skills and knowledge. This was the case for all three waves. During wave II

teachers’ monitoring, intervention and evaluation behavior were also related to the quality of

CL, which was in contrast with our expectation that teachers’ control behavior would be

negatively related to the quality of CL at a later stage of CL. In line with our expectations,

during wave III only the scales that measured students’ perceptions on the extent that they

were taught skills and knowledge for CL, and teachers’ clarity on rules for CL were related to

the quality of CL. However, comparisons of correlation coefficients of teachers control

behavior and the quality of CL between wave III and the other two waves, resulted in only

one significant difference, namely the relationship between teachers monitoring behavior was

significantly less strong during wave III than during wave II (Z = -2, 65, p = .003).

The analyses described in the latter two paragraphs function as a framework for

understanding longitudinal results of the group health and welfare students that completed the

questionnaires at the three data points. In order to explore whether students’ perceptions on

teacher control and instructional behavior were stable in the course of a year, we will explore

the longitudinal data next.

Stability and changes of conditions for effective cooperative learning

--------------------------------

Insert Table 4 about here

------------------------------

In order to investigate whether the students that completed the questionnaire at all three data-

points (N = 120) were stable in their ratings on the five subscales in the course of a year, we

conducted GLM repeated measures analyses separately for the five subscales, as within-

subjects factors at three points in time. Table 4 presents an overview of the scores.

A main effect was found for rules for CL for measurement time (Wilks’ F [2, 136] =

14.98, p = .000, �² = .18). Measurement time also (weakly) co-varied with monitoring (Wilks’

F [2, 139] = 3.28, p = .040, �² = .04) and with the evaluation scores (Wilks’ F (2, 136) = 3.07,

p =.050, �² =.04). Interestingly, scores on all three subscales were significantly higher during

wave II. Stability of scores was found for the two other scales, namely students’ perceptions

on the extent that they were taught skills and knowledge for CL and teachers’ intervention

behavior.

We also conducted a repeated measures analysis on the quality of CL. A weak time

effect was found (Wilks’ F [2, 118] = 3.08, p = .050, �² = .05). Inspection of the within

subject contrasts table showed that the quality of CL was significantly higher at wave II than

99

it was at wave I. These higher scores parallel the earlier reported scores for all students

participating in wave II. In order to analyze the combination of changes in conditions and

changes in the quality of CL, we looked for patterns in the quality of CL first.

Longitudinal patterns of the quality of CL and teacher related conditions

We conducted cluster analysis in order to identify groups of students with different

longitudinal patterns of the quality of CL. The k-means method was used, because this

method is sensitive to decisions on the preferred number of clusters and the values for the

initial cluster centers. Two clusters were used, based on the interpretability of the resulting

clusters and the desire to have a reasonable number of students within the clusters.

The first profile captured students with overall high scores on the quality of CL, with

somewhat higher scores on the second data-wave, and the lowest scores on the first data-

wave. The second profile grouped students with overall low scores on the quality of CL and

similar scores on the three data waves. Sixty-one students were grouped in profile 1 and 59

students were represented by profile 2. The distinction of these two profiles implies that we

did not find a group of students that clearly changed their quality of CL in the one-year

period. This finding suggests that whatever teachers do in terms of creating conditions for CL,

they do not affect students to the extent that ineffective cooperators become effective.

We hypothesized that ineffective cooperators are more dependent on the teachers’

control behavior during the second and third data wave than effective cooperators, who were

expected to rely more on the extent that they were taught skills and knowledge and rules for

CL. Hence, we expected interaction effects between teacher related conditions and the

profiles and hoped that these interactions would reveal the teachers’ capacity to adapt

conditions for CL to the needs of particular groups of students depending on their repeated

quality of CL. A repeated measures analysis was conducted using the five subscales at the

three measurement points as within-subject factors and the quality of the students’ CL profile

as the between subject factor. No interaction effects were found and thus no support was

found for our hypothesis. We found a within-subjects factor effect for time though (Wilks’ F

[10, 88] = 2.68, p =.006, �² = .23) and a main between-subject factor effect for quality of CL

profile (Wilks’ F [5, 93] = 4.77, p = .001, �² = .20). Figure 2.1 to 2.5 graphically represent the

relationships.

100

------------------------------

Insert Figure 2.1 to 2.5 here

---------------------------------

The time effect concerned the subscale that measured students’ perceptions of the extent to

which they were taught skills and knowledge for CL (see Figure 2.1), and the extent that

students were taught rules for CL (see Figure 2.2). Posthoc tests showed that scores for both

students with the weak and high quality of CL profiles on the quality of CL was at its highest

at wave II, confirming our previous findings. For whatever reasons teachers have adjusted

their instruction as regards CL. The findings suggest however that adjustment was hardly for

adaptive reasons (e.g., compensating for lacking skills or reported low quality of CL). The

effect that we found resembles a self-efficacy effect (i.e. teachers feel more in control and

more positive about themselves when interacting with students who perform and behave the

way they want them to. It might be the reason why these students get more attention.

Inspection of the plots shows that at all times students with profile 1 (i.e., the high quality

profile) scored higher than students with profile 2 on the scales that measured teachers control

behavior and teachers’ instruction behavior and that the extent that students were taught skills

and rules was highest (Figure 2.1 and 2.2) at wave II. Inspection of the plots also shows that

the difference between effective and ineffective cooperators was highest for the scale that

measured students’ perceptions of the extent that they were taught skills for CL (see Figure

2.1); effective cooperators scores were considerably higher than ineffective cooperators

scores.

Despite the fact that we were unable to find significant interaction patterns, inspection

of the plots showed some interesting trends that are in line with our hypotheses that teachers

need to improve their adaptivity. First of all, inspection of Figure 2.3 shows that the weak

cooperators perceived a major decrease in teachers’ monitoring behavior after wave II.

Secondly, inspection of Figure 2.4 shows a similar trend in relation to teachers’ intervention

behavior. In contrast, the effective cooperators perceived almost no change in teachers’

monitoring and intervention behavior after wave II. Moreover, Figure 2.5 shows that for the

effective cooperators scores on the teachers’ evaluation behavior between wave I and wave II

highly increased, whereas this pattern was not that obvious for the ineffective cooperators.

In short effective cooperators can be characterized by high scores on all teacher related

conditions. Especially their high scores on the extent that they have been taught skills and

knowledge for CL distinguishes them from ineffective cooperators. Ineffective cooperators,

101

on the other hand, can be characterized by low scores on all teacher related conditions,

particularly their scores on the extent that they were taught skills and knowledge for CL were

low. Moreover, they reported that teachers’ control behavior highly decreased after wave II.

DISCUSSION AND RECOMMENDATIONS

Our primary objective in this research was to explore the relationship between teachers’

control and instructional behavior and the quality of CL. The stability of scores on teacher

related conditions and the relationship with the quality was analyzed.

We expected optimal CL processes when students were explicitly taught rules,

knowledge and skills for CL and when teachers’ regularly monitor the CL process, intervene

when the situation required it and evaluate the group process regularly. Moreover, a decrease

of teacher monitoring and intervention behavior and a stability of the other variables was

expected to predict the best CL results over time. In line with previous findings (Hijzen,

Boekaerts & Vedder, 2006), we found that the scales that measured students’ evaluations of

the extent that they were taught skills and knowledge for CL and rules for CL were most

strongly related to the quality of CL at wave I. Complementary findings are that these scales

were also related during the other two waves. In other words, we can add to previous findings

(Hijzen, Boekarts & Vedder, 2006) that -also in the long run- it is important that teachers

make students aware of what is required for working in a CL setting. They should explicitly

and sufficiently teach them the necessary skills. More specifically, it is crucial that teachers

teach their students how to listen to each other, to evaluate the group process, to discuss, to

support group members, to give an opinion, or to solve group conflicts, and explain what is

expected of them in terms of goals and the evaluations of CL processes.

We also expected that the relationships between teachers’ control behavior and the

quality of CL would diminish over time, following either an improvement of the QCL or the

maintenance of an already high level of the QCL. However, despite a positive relationship of

intervention behavior at wave I, our expectations could not be confirmed. Teachers’ control

behavior and the quality of CL were also positively related at wave II, while we expected that

a decrease in teachers control behavior would predict the quality of CL at a later stage. Of

course students need to be monitored when learning how to cooperate, but an unexpected

finding is that this scale still contributed to the model during wave II that took place in the

students’ second school year. Perhaps students were more motivated and independent

cooperators in the beginning of their second year as compared to first year students, but why

102

did they still feel the urge to receive teacher guidance and feedback and why did this urge

disappear at the time of the third round of data collection? Lack of a strong relationship

between the quality of CL and teachers’ monitoring, intervention and evaluation processes

during wave III may indicate that the students in our sample became indeed more independent

cooperators and relied more on their own skills and capabilities.

We expected that a stability of teaching students’ skills and knowledge and rules for

CL, and a decrease in teachers’ control behavior during wave II and wave III correspond to an

increase in the quality of CL from wave I to wave III. Cluster analysis on the quality of CL of

the health and welfare students that completed the questionnaires of CL at all three waves

showed that we could distinguish between students that had generally high scores on the

quality of CL and students that had generally low longitudinal scores on the quality of CL.

Inspection of these two groups’ scores on the five teacher related conditions showed that

teacher related conditions indeed make the difference, at all stages. Students with high scores

on the quality of CL scored higher on all teacher behavior subscales. Again the importance of

the extent that students were taught skills and rules for CL is illustrated by the large difference

on scores between the high and low quality cooperators.

103

References

Ames, C. (1992). Classrooms: Goals, structures, and student motivation. Journal of

Educational Psychology, 84, 261-271.

Boekaerts, M., & Minnaert, A. (2003). Assessment of students’ feelings of autonomy,

competence, social relatedness; a new approach to measuring the quality of the

learning process through self-assessment. In M. S. R. Segers, F. J. R. C. Dochy &

E. C. Cascallar (Eds.), Optimizing New Methods of Assessment: in Search of

Quality and Standards (pp. 225-246). Dordrecht, The Netherlands: Kluwer Academic

Publishers.

Boekaerts, M., De Koning, E., & Vedder, P. (2006). Goal-directed behavior and

contextual factors in the classroom: an innovative approach to the study of

multiple goals. Educational Psychologist,41, 33-51.

Brophey, J. E., & Good, T. L. (1986). Teacher behavior and student achievement. In M.C.

Wittrock (ed.), Handbook of research on teaching, 3, 328-375. New York: MacMillan.

Chinn, C. A., O'Donnell, A. M., & Jinks, T. S. (2000). The structure of discourse in

collaborative learning. Journal of Experimental Education, 69, 77-97.

Cohen, E. G. (1994). Restructuring the classroom: Conditions for productive small

groups. Review of Educational Research, 64, 1-35.

Deci, E. L., & Ryan, R. M. (2000). The "what" and "why" of goal pursuits: Human needs and

the self-determination of behavior. Psychological Inquiry, 11, 227-268.

Gillies, R., & Ashman, A. (1996). Teaching collaborative skills to primary school children in

classroom-based workgroups. Learning and Instruction, 6, 187-200.

Gillies, R. (2003). Behaviors, interactions, and perceptions of junior high school students

during small-group learning. Journal of Educational Psychology, 95, 137-147.

Hayes, K., Walton, J. R., Szomor, Z. L., & Murrell, G. (2001).Reliability of five methods for

assessing shoulder range motion. Australian Journal of Physiotherapy, 47, 289-294.

Hijzen, D., Boekaerts, M., & Vedder, P. (2006). The Relationship between

Students' Goal Preferences, perceptions of Contextual Factors in the Classroom

and the Quality of Cooperative learning. Scandinavian Journal of Psychology,

47, 9-21.

Hijzen, D., Boekaerts, M., & Vedder, P. (submitted). Exploring the links between

students’ engagement in cooperative learning, their goal preferences and

perceptions on contextual factors in the classroom.

104

Hoek, D., Van den Eeden, P., & Terwel, J. (1999). The effects of integrated social and

cognitive strategy instruction on mathematics achievements in secondary education.

Learning and Instruction, 9, 427- 448.

Johnson, D. W., & Johnson, R. T. (2002). Learning together and alone: Overview and meta-

analysis. Asia Pacific Journal of Education, 22 , 95-105.

Phipps, M., Phipps, C., Kask, S., & Higgins, S. (2001). University students’ perceptions of

cooperative learning: implications for administrators and instructors, The Journal

of Experiential Education, 24, 14-21.

Resnick, L. (Ed.). (1989). Knowing, learning, and instruction. Hillsdale, NJ: Lawrence

Erlbaum Associates.

Ros, A. A. (1994). Samenwerking tussen leerlingen en effectief onderwijs [Cooperation

between students and effective education]. Dissertation, Rijksuniversiteit

Groningen.

Salomon, G., & Globerson, T. (1989). When teams do not function the way they

ought to. International Journal of Educational Research, 13, 89-99.

Shanahan, T. (1998). On the effectiveness and limitations of tutoring in reading. In :

P.D. Pearson & A. Iran-Nejad (Eds.). Review of research in education, 23 (1-

24). Washington AERA. Macmillan.

Slavin, R. E. (1987). Developmental and motivational perspectives on cooperative

learning: A reconciliation. Child Development, 58, 1161-1167.

Slavin, R. E. (1995). Cooperative learning: Theory, research, and practice (2nd ed.).

Boston: Allyn and Bacon.

Slavin, R. E. (1996). Research on cooperative learning and achievement: What we know,

what we need to know. Contemporary Educational Psychology, 21, 43-69.

Vedder, P. (1985). Cooperative Learning: A Study on Processes and Effects of Cooperation

Between Primary School Children. Netherlands: Rijksuniversiteit Groningen.

Veenman, S., Kenter, B., & Post, K. (2000). Cooperative learning in Dutch primary

classrooms. Educational Studies, 26, 281-302.

Vermunt, J. D., & Verloop, N. (1999). Congruence and friction between learning and

teaching. Learning and Instruction 9, 257-280.

Webb, N. M., & Farivar, S. (1994). Promoting helping behavior in cooperative small

groups in middle school mathematics. American Educational Research Journal,

31, 369-395.

Webb, N. M., & Palincsar, A. S. (1996). Group processes in the classroom. In: D.C.

105

Berliner & R.C. Calfee (Eds), Handbook of Educational Psychology (pp. 841 -

873). New York: Simon & Schuster MacMillan.

Webb, N. M., Troper, J. D., & Fall, R. (1995). Constructive activity and learning in

collaborative small groups. Journal of Educational Psychology, 87, 406-423.

Witteman, L. (2003). Samenwerken of samen werken [Cooperating or working

together alone]. Unpublished dissertation. Universiteit Leiden.

106

Table 1

Sample characteristics

Program Level

Program Type Level 1 Level 2 Level 3 Level 4 Total

Engineering 3 7 63 73

ICT 10 21 91 123

Retail & Administration 10 22 41 108 181

Food & Tourism 1 13 5 4 23

Health & Welfare 43 61 177 355 616

Total 54 109 251 602 1016

107

Table 2

Categories, Sample Items, Number of Items and Cronbachs’ Alpha Coefficients

Category Sample item ��

������

Alpha

Perceived Quality

of CL

Quality of CL I perceive myself as part of this group 31 -

Conditions

Subscales

Rules for CL Before we start to work on the group task, teachers explain

us how to plan

9 .87

Cooperation skills At this school we learned to listen to each other during

group work

8 .86

Teacher:

monitoring

While working on the group task the teacher evaluates the

progress

5 .83

Teacher:

intervention

When we are very noisy during group work, teachers

intervene

5 .74

Teacher: evaluation After we finish the group task, the teacher explains what

went well and what needs improvement

4 .80

05

10152025303540

45 minp/w

90 minp/w

45 minp/d

90 minp/d

> 90min p/d

Time spent on CL data wave1Time spent on CL data wave2Time spent on CL data wave3

Figure 1:

Time Spent on CL

109

Table 3

Pearson Correlation Coefficients between the Quality of CL and Conditions for CL

Data wave 1 Data wave 2 Data wave 3

N M SD r

QCL

N M SD r

QCL

N M SD r

QCL

Quality of CL 1135 2.83 .29 - 619 2.91 .29 - 434 2.88 .32 -

Rules for CL 1170 2.46 .44 .25** 615 2.55 .46 .33** 434 2.46 .41 .28**

CL knowledge

and skills

1171 2.67 .41 .34** 616 2.74 .43 .41** 434 2.71 .41 .39**

Teacher:

monitoring

1170 2.54 .52 .18** 616 2.61 .49 .28** 432 2.46 .52 .12

Teacher:

intervention

1166 2.45 .49 .20** 614 2.51 .54 .25** 432 2.36 .53 .18**

Teacher:

evaluation

1168 2.35 .57 .18** 614 2.40 .67 .20** 432 2.19 .59 .14*

** p < .001, *p < .05

110

Table 4

Mean scores and standard deviations of contextual factors at three data-waves

wave I wave II wave III

M SD M SD M SD Test of within contrasts

( p <.05)

Rules for CL 2.35 .44 2.53 .39 2.44 .45 wave 2>1,3

CL knowledge and skills 2.69 .41 2.76 .41 2.68 .42

Teacher: monitoring 2.44 .52 2.54 .47 2.47 .49 wave 2>1,3

Teacher: intervention 2.41 .52 2.46 .51 2.39 .51

Teacher: evaluation 2.21 .54 2.34 .59 2.19 .63 wave 2>1,3

Quality of CL 2.83 .30 2.91 .28 2.89 .27 wave 2>1

111

1 2 3

time

2,50

2,60

2,70

2,80

2,90

3,00

Estim

ated M

argina

l Mea

nsCluster Number of Case

profile 1profile 2

Estimated Marginal Means of CL skills

Figure 2.1:

Strong and weak cooperators’ mean scores on the scale that measured students’ evaluations of the

extent that they were taught skills and knowledge for CL, at three data waves

1 2 3

time

2,30

2,40

2,50

2,60

Estim

ated M

argin

al Me

ans

Cluster Number of Case

profile 1profile 2

Estimated Marginal Means of rules for CL

Figure 2.2:

Strong and weak cooperators’ mean scores on the scale that measured students’ evaluations of

teachers’ clarity on CL rules, at three data waves

112

1 2 3

time

2,30

2,35

2,40

2,45

2,50

2,55

Estim

ated

Mar

gina

l Mea

ns

Cluster Number of Case

profile 1profile 2

Estimated Marginal Means of monitoring

Figure 2.3:

Strong and weak cooperators’ mean scores on the scale that measured students’ evaluations of

teachers’ monitoring behavior, at three data waves

1 2 3

time

2,25

2,30

2,35

2,40

2,45

2,50

Estim

ated M

argina

l Mea

ns

Cluster Number of Case

profile 1profile 2

Estimated Marginal Means of intervention

Figure 2.4:

Strong and weak cooperators’ mean scores on the scale that measured students’ evaluations of

teachers’ intervention behavior, at three data waves

113

1 2 3

time

2,05

2,10

2,15

2,20

2,25

2,30

2,35

2,40

2,45

Estim

ated

Mar

gina

l Mea

ns

Cluster Number of Case

profile 1profile 2

Estimated Marginal Means of evaluation

Figure 2.5:

Strong and weak cooperators’ mean scores on the scale that measured students’ evaluations of

teachers’ evaluation behavior, at three data waves

114

Chapter 5

EXPLAINING COOPERATIVE LEARNING IN MULTI-ETHNIC CLASSES; THE

LIMITED ROLE OF STUDENTS’ ETHNOCULTURAL BACKGROUND 5

Abstract

This study explored cooperative learning in multi-ethnic classes. We explored what combination

of student characteristics and student appraisals of the school context was related to higher levels

of the quality of cooperative learning and whether these varied by ethnocultural group. Eighteen

hundred students were clustered into four profiles specifying the characteristic levels of language

proficiency, goal preferences, social resources and school/ peer identification/alienation. Four

student profiles were identified; a school-disaffected, a weak communication/school bonding, a

school-adjusted and a frustrated profile. Students that were grouped in the school-adjusted

profile showed the highest scores on CL. This profile was characterized by clear goals, high

perceived availability of social support and high scores on school and peer identification. The

school-disaffected profile had the lowest scores on CL. This profile was characterized by no clear

goals, a lack of social resources and peer/school identification. Students from different

backgrounds were disproportionately distributed. The weak communication profile was

characteristic of Caribbean students and the school-adjusted profile was characteristic of the

Dutch.

Key words: ethnocultural background, goal preferences, quality of cooperative learning

5 This chapter is based on: Hijzen, Boekaerts and Vedder (2006). Explaining cooperative learning in multi-ethnic classrooms; the limited role of students’ ethnocultural background. Manuscript submitted for publication in: Applied Psychology: An International Review.

115

INTRODUCTION

Even though school achievements of immigrant students who were born in the Netherlands

(second generation immigrants) have considerably improved over the last 15 years, they still do

not match Dutch students’ performance and are characterized by higher levels of grade repetition

and drop-out (Mulder, Roeleveld, Van der Veen, & Vierke, 2005). Many studies have reported a

relationship between school performance and students’ ethnocultural background, but only a few

studies have analyzed the underlying processes and student characteristics that explain immigrant

students’ educational position in the Netherlands (e.g., Boekaerts, 1998; Teunissen & Mathijssen,

1996). The present study addresses this latter question by exploring what processes and what

student characteristics influence the quality of students’ learning. We particularly focus on

learning in settings that invite students to get involved in cooperative problem solving. Once we

have a better picture of the processes and characteristics that underlie the quality of cooperative

learning we will explore whether and to what extent these processes and characteristics vary by

ethnocultural group. We hope to identify groups of students that allow for a group wise approach

in education that goes beyond a general class wide educational approach or an approach adapted

to individual students; a class wide approach may lack adaptability to the needs of particular

groups of students, whereas an individualized approach is hardly feasible, at least in the Dutch

secondary vocational schools that participated in the present study.

In this study we focus on the quality of cooperative learning (CL) of students in secondary

vocational schools. The choice of CL is related to the suggestion by some scholars (e.g., Cohen,

1994; Slavin, 1995; Webb & Palincsar, 1996) that CL settings may promote students’

involvement with and motivation for school and learning; as well as facilitate integration and

prevent discrimination, by functioning as an activity setting where students are able to connect

with each other and learn from each other's abilities and skills. This seems particularly important

for students in vocational schools in the Netherlands, because many of them drop out of school

for motivational reasons and communication problems (e.g., Voncken, Van der Kuip, Moerkamp,

& Felix, 2000).

116

In the last two decades the classroom setting in senior vocational schools in the

Netherlands has gradually changed from an exclusively, traditional, competitive, and

individualistic educational setting to a setting with more attention for cooperative learning

requiring and stimulating both cognitive and motivational self-regulation skills (Boekaerts &

Minnaert, 2003). Cooperative learning refers to ‘a set of instructional methods in which students

are encouraged or required to work together on academic tasks’ (Slavin, 1987). This study deals

with three important components that a successful CL situation requires. In the first place,

students’ attitudes towards CL methods should be positive. They need to acknowledge the value

of CL, in order to be prepared to engage in CL in the first place. Secondly, students need a

number of cooperation skills, such as the skills to express their own opinion, stimulate each other,

provide and receive help, listen to each other and clarify their current understanding of the task

(Cohen, 1994; Ros, 1994; Webb & Palincsar, 1996). A third important component of a successful

CL setting is interdependency. Students need to feel responsible for each other's learning process

and experience a sense of group cohesion (Johnson & Johnson, 1994; Ros, 1994). Chin,

Salisbury, Pearson, and Stollak, (1999) and Cohen (1994) pointed out that the activity level in the

group is at its best when students feel at home in the group.

Person-related explanations for variation in the quality of CL

Dutch language proficiency

Dutch language proficiency is crucial in CL. Learning in CL teams may promote students’

language competence. However, for effective CL, students need a number of cooperation skills in

the first place (Hijzen, Boekaerts, & Vedder, 2006). These skills heavily depend on their language

competences; in order to provide team members with sufficient support, or to be a good listener

and debater students have to be fluent in the Dutch language.

Students’ goal preferences

Whether students engage in CL depends on the personal significance they attach to the tasks and

the context in which it is embedded. The extent to which a task triggers personal significance

117

depends on the type of goal preferences that students have (Baumeister & Heatherton, 1996). In

previous studies we explored the extent to which students endorse social support, belongingness,

mastery and superiority goals in close relation to the quality of their CL. Studies by Eccles (1987)

and Wentzel (1991) suggest that cooperative learning is an effective learning setting for students

preferring a combination of mastery (understanding tasks) and social goals (being friends with

and, supporting others). In such a setting they can create a sense of belongingness and security.

Mastery goals are associated with high levels of performance on personally challenging tasks in

general (Ford, 1992). A combination of social and mastery goals should be most conducive to the

quality for CL (McInerney, Hinkley, Dowson, & Van Etten, 1998). It is exactly this combination

with mastery goals that is likely to predict the quality of CL.

Also, and in contrast to this goal profile Wentzel and Wigfield (1998) showed that

students who prefer superiority and individuality goals have lower scores on the quality of CL

than students who prefer social goals. Functioning as a group member may contrast with students’

wish to perform well at a more individual level.

Context related explanations for variation in the quality of CL

Social resources, school and peer identification, and alienation

Apart from students’ goals we expect students’ perceptions of the social environment to be

important to the CL process (Vedder & Boekaerts, in press). Samdal, Nutbeam, Wold, and

Kannas (1998) found that students’ sense of security and feelings of being treated fairly and

supported by their teachers were crucial in predicting their general satisfaction with school.

Feelings of well-being and school identification are important to the quality of the learning

process. Feelings of alienation on the other hand will negatively influence students’ learning

processes (Deci & Ryan, 2000)

A number of studies showed that perceived availability of social support is a better

predictor of well-being than actually received support (Wethington & Kessler, 1986); therefore

we will focus on students’ perceptions of emotional and academic support, in relation to their

identification with and alienation from school and peers.

118

In this study we explore what combination of student characteristics and student appraisals

of the school context is related to higher levels of the quality of cooperative learning. Particular

combinations will be referred to as profiles.

We hypothesize that students who have profiles characterized by good language

proficiency, a preference for social support, belongingness, and mastery goals, negative or low

scores on superiority goals, high scores on perceived availability of teacher and peer support and

on school and peer identification, and negative scores for school and peer alienation to report high

quality of CL and vice versa for students who report low quality of CL.

The role of students’ ethnocultural background

Research findings as regards the question whether immigrant students profit more or less from,

and perform better or worse in CL settings than national students are inconclusive. This is logical

since the two categories of students compared, i.e., national versus immigrant students, are

unspecified, which makes it impossible to compare whatever in-group variation with whatever

between-group variation. In this particular study the national students are Dutch adolescents with

white west European born parents and the immigrant students are either from the Caribbean

(Surinamese or Antilleans) or from the Mediterranean (Moroccan or Turkish).

Nationals and immigrants in the Netherlands

With almost 10% of its 16 million population born elsewhere, the Netherlands has a modest

immigrant population in comparison to other Western countries (Berry, Phinney, Sam, & Vedder,

2006). The population’s ethnic origin is 94% Dutch, which includes about 125,000 immigrants

from the Netherlands Antilles and Aruba (the Antilleans), who have Dutch citizenship (because

these are dependent overseas territories), as well as an unknown portion of 295,000 immigrants

from Surinam (which is a former overseas territory). The main origins of other immigrants are

Morocco, Turkey, and former Yugoslavia.

In this study we focus on second-generation immigrants, defined as those students that

were born in the Netherlands with at least one parent who came to the Netherlands as an

immigrant. We distinguish Mediterranean students, who are from Turkey and Morocco, and

119

Caribbean students who are from the Netherlands Antilles, Aruba and Surinam. Dutch students

clearly differ from immigrant students in educational attainment levels, with Turkish and

Moroccan students having the lowest levels, Dutch the highest, while Antillean and Surinamese

students are in between (Van Ours & Veenman, 2001).

The migration history of the Caribbean community clearly differs from that of the Mediterranean.

The former have a history of contacts with the Dutch, who were the colonizers of these

communities. The cultural distance is rather small. This is clearest with respect to language

proficiency. The language of instruction in most schools in Surinam and in the Netherlands

Antilles is still mainly Dutch, meaning that they grow up learning Dutch, before they arrive in the

Netherlands. Moreover, the legal systems, the health care system, and religious institutions are all

rooted in a common colonial history. This is not to say they are similar, but they look very much

the same. This is completely different with Turkey and Morocco, which population is mainly

Muslim. The cultural distance between the Netherlands and these two countries is large, even

larger because Turkish and Moroccan immigrants did not come from the more Westernized large

cities, but came and come from rural areas.

The possible impact of ethnocultural background on cooperative learning, student characteristics

and student appraisals of the school context

CL methods are based on socio-constructivist theories that strongly emphasize students’

autonomy. Palfreyman (2001) suggested that autonomy is contradictory to concepts as tradition,

authority and non-Western culture. He clarified that non-Western students are low on feelings of

autonomy and therefore encounter problems in Western schools. However, students from more

collectivistic cultures than the Dutch, such as Moroccans (Pels, 1998) are more used to participate

in group behavior, which might make them better equipped for learning in CL settings than Dutch

students. In particular their cooperation skills may be better developed than those of Dutch

students. A recent study on the effects of educational settings and instructional approaches that

emphasize cooperative learning, discussion, and justifying particular problem solving strategies in

mathematics showed that, for disadvantaged students (mainly immigrant students), “modern”

settings that stress the role of social interaction are less effective in terms of achieving teacher

120

desired learning outcomes than more structured and teacher guided approaches (Timmermans,

2005). Wittebrood and Keuzekamp (2000) suggested that Turkish and Moroccan students, unlike

Surinamese and Antillean students experience difficulty when working with modern, interactive

instruction methods. At the same time, Matthijssen (1993) suggested that immigrant students have

difficulty in attaching relevance to the knowledge and skills that they are supposed to learn in

Dutch schools. Moreover, these students tend to be docile towards teachers, whom they see as

authorities. The first characteristic hinders learning and requires additional support, whereas the

second impedes asking for help. In fact, these students need greater security and self confidence

and they should be encouraged to develop self initiative and autonomy in learning. CL may be

quite appropriate in such a situation.

The uncertainty about the possible blessings or risks of CL for immigrant students led us

to explore the relationship between such aspects of the quality of CL as perceived group cohesion,

cooperation skills or attitudes towards CL and students’ ethnocultural background.

For future intervention purposes, ethnocultural background in itself, although interesting,

is not a variable that can be manipulated to improve CL processes in secondary vocational

education. We will therefore focus on changeable individual and contextual characteristics that

can account for ethnic differences in the quality of CL processes, represented by students’ Dutch

language proficiency, their motivation or goal preferences, and students’ identification with

school and peers.

The language assimilation model posits that immigrant students’ proficiency in the national

language is a better predictor of students’ academic achievements and social participation than

proficiency in the home language (e.g., Vedder & Virta, 2005). For second language acquisition,

the actual oral and written contacts between two ethnic groups are important (Bialystok, 2001).

Blom and Severiens (2000) found that in the Netherlands immigrant students’ vocabulary was

significantly smaller than that of Dutch students. In a CL setting students use language to

understand the learning tasks, to ask for help and support other students. Some researchers (e.g.,

Boers, 2001; Hajer, 1996; Kirchmeyer; 1993) found that immigrant students participated little in

collaboration and processes of decision making, because of limited language proficiency.

Students who are less fluent in Dutch can hardly participate at the same level as their Dutch

classmates do. As mentioned previously, Mediterranean students encounter more language

121

difficulties than Antillean and Surinamese students because the latter groups are more familiar

with Dutch customs and language due to historical bonds with the Netherlands.

Earlier studies on goal preferences and students’ ethnocultural background yielded

contradictory findings. McInerney, Roche, McInerney, and Marsh (1997) studied Anglo,

Australian Aboriginal and native American students and found that these groups hardly differed

in the appreciation of goals that were related to learning and achievement. Vedder and Boekaerts

(2006) also found comparable goal structures between Dutch students and students living in

Curacao in the Caribbean. However, Dutch students were just a little more oriented towards

superiority than the Caribbean students. Other studies (e.g., Suarez-Orosco, 1998; Valdez, 1998)

suggested that differences in appreciation of culture-bound values did influence students’ goal

preferences.

In this study we will investigate whether students with different cultural backgrounds

differ with regard to their goal preferences and whether these differences have consequences for

CL.

In terms of students’ feelings of school identification and alienation earlier research reported that

particular groups of immigrant students in the Netherlands switch schools so frequently that this

impacts on their school identification. They change schools because their parents frequently move

into different neighborhoods or even between cities (Mulder, Roeleveld, Van der Veen, & Vierke,

2005). This is even more worrying because another study (Vedder, Boekaerts, & Seegers, 2005)

found that young Dutch adolescent students rely more on parental academic support than

immigrant students, who rely more on instructional support from the teacher. No differences were

found for the perceived availability of parental emotional support or the perceived availability of

peer support between different ethnic groups. Pels (1998) showed that Moroccan students were

brought up with less personal parental attention than their Dutch peers. Also, Distelbrink and Pels

(2000) found that Turkish and Moroccan students felt less academically supported by their

parents than by their teachers and peers. In short immigrant students in the Netherlands in matters

of learning and school tend to depend more on the teachers than on their parents.

In the present study we will explore whether students from different ethnic groups differ in their

appraisal of the availability of academic and emotional support from teachers and peers, and in

the degree to which they identify with or alienate from school and peers, and what the

122

consequences are for the quality of CL. We hypothesize that immigrant students will rely more on

teacher and peer support than Dutch students do. If we find support for this hypothesis this may

indicate an additional risk for the immigrant students. Students’ higher desire for teacher and peer

support may lead to higher feelings of frustration, when their desire for support from teacher and

peers is not satisfied.

Based on the evidence presented thus far we expect to find differences between national,

Mediterranean, and Caribbean students living in the Netherlands, both as regards the quality of

cooperative learning and the profiles representing particular combinations of student

characteristics and student appraisals of the school context.

METHOD

Participants

Participants in this study were 1806 students from 11 different secondary vocational schools that

were spread evenly across the Netherlands. Students were enrolled in different study programs,

namely ICT, engineering, retail and administration, food and tourism, and health and welfare

programs. They were predominantly of Dutch origin (N = 1599), and their mean age was 17.11,

SD = 3.52 during the first data collection period. About 12 percent of the participating students

had an immigrant background, determined by their parents’ birthplace. The Mediterranean group

(Moroccan and Turkish) consisted of 119 students, while 88 students had a Caribbean

(Netherlands Antilles or Surinam) background.

Instruments

Students’ ethnocultural background: Participants were asked in which country they were born

and in which country their parents were born.

Four scales were administered. Table 1 presents an overview of scales, sample items and

Cronbachs’ alphas of the different scales used in this study. Structural equivalence of the scales in

the three ethnic groups was assessed using exploratory factor analyses followed by a test of

123

factorial agreement. Tucker phi coefficients higher than .90 are seen as evidence for factorial

agreement (Ten Berge, 1986). Values of these tests are also presented in Table 1.

------------------------------

Insert Table 1 about here

------------------------------

The quality of CL (QCL) comprised four subscales, namely; students’ perceptions of the quality

of group cohesion, students’ perceptions of the quality of interdependence within the group,

students’ perception of the quality of their cooperation skills and a subscale that measured

students’ attitudes towards CL. All questions referred to the group learning situations that the

students had participated in during the previous four weeks. Students had to indicate on a four-

point Likert scale to what extent they agreed with each statement. Response categories ranged

from "I disagree very strongly" to “I agree very strongly” (Hijzen, Boekaerts, & Vedder,

2006).These subscales were highly correlated and were all part of the quality of CL. A Principal

Component Analysis on these four subscales resulted in a one-factor solution. This factor had an

Eigenvalue of 1.8 and it explained 58% of the total variance.

Students’ personal goals were assessed with the goal preference list based on the Ford

(1992) and Ford and Nichols (1991) taxonomy of broad goals. Students had to report on the

importance they attach to each of the goals by giving an indication of the extent to which they

want to achieve them. They were asked to choose from five response categories ranging from “not

at all” to “very much so”. Four goal domains related to the quality of CL, represented by four

scales, were selected from the list and used in this study, namely superiority and individuality,

mastery, belongingness, and social support goals (For more details see Hijzen, Boekaerts, &

Vedder (2006)).

Students’ language competence was measured with one subscale assessing students’

proficiency in the Dutch language (Kwak, 1990). Students had to indicate how well they did in

speaking, writing, understanding and reading Dutch. The scale items were measured on a four-

point Likert-type scale (1= not at all, 4= very good). The scale for Perceived Availability of

Social Support (e.g., Vedder, Boekaerts, & Seegers, 2005) measures students’ appraisals of the

availability of academic and emotional support from their teachers and peers. Students had to

indicate on a four-point Likert scale how often they get support. Answering categories ranged

from “almost never” to “very often”. Four scales were from the Relational and School Identity

124

Scale (Meeus, 1996) and measured students’ identification with their peers and school, students’

feelings of alienation from peers and their school. Response categories (4) ranged from “I

disagree very strongly” to “I agree very strongly”.

Reliability coefficients of all measures were good, Cronbachs’ alpha coefficients ranged

from .73 to .93. Tucker’s phi comparisons for each pair of countries were very high (0.97 – 1.00),

indicating overall good structural equivalence for all measures.

Procedure

The scales were administered during regular classroom activities. Researchers were present to

assist the teachers, and to answer any questions. All participants were informed that participation

was voluntary, and that responses were anonymous. It took students two sessions of 45 minutes to

complete all the scales.

RESULTS

The relationship between individual and context related characteristics and the quality of CL

Pearson correlation coefficients were calculated in order to explore the relationships between

individual and context related characteristics and the quality of CL. Because of the large sample,

almost all correlations were statistically significant. The associations were generally in the

expected direction, all person-related (Dutch language proficiency and goal preferences) and

context-related (social resources, belongingness and alienation) appraisals were significantly

correlated with the quality of CL. Especially students’ mastery (r = .23, p = .000) and social

support (r = .29, p =.000) goal preferences and their appraisals of the availability of emotional (r

= .27, p =.000) and academic support (r = .25, p =.000) and identification with peers (r = .19, p =

.000) and school (r = .23, p =.000) were interrelated with the quality of CL. Contrary to our

expectations, peer alienation was not significantly related to the quality of CL. We therefore

excluded this subscale from further analyses.

Profiles

125

A person-oriented approach was used to gain insight into student profiles. Cluster analyses were

conducted with the following variables: language proficiency, goal preferences, academic and

emotional peer and teacher support, school and peer identification, and school alienation scales.

We used the k-means method, because this method is sensitive to decisions as to the preferred

number of clusters and the values for the initial cluster centers. Based on the fit with the dominant

theoretical notions guiding the study and on the interpretability of the resulting clusters, we found

four clusters. Figures 1.1, 1.2, 1.3 and 1.4 present the four clusters.

--------------------------------------

Insert Figures 1.1, 1.2, 1.3 and 1.4 about here

---------------------------------------

Cluster 1 can be described as the school disaffected profile. This highly negative profile groups

students who score low on the goal domains, meaning that they had no clear goals. They were

slightly positive about their proficiency in the Dutch language, but dissatisfied with the

availability of academic and emotional support from teachers and peers. These students also

lacked a sense of school and peer identification. However, they scored slightly negative on school

alienation.

A second, more diffuse cluster shown in Figure 1.2, can be defined as the weak

communication / weak school bonding profile. Scores on Dutch language proficiency were

extremely low. Students with this profile scored relatively high (and positive) on school

alienation, meaning that compared to students in the other clusters, these students were highly

unsatisfied with their study program. The scores on the availability of academic and emotional

peer and teacher support were positive, meaning that students in this profile were satisfied with

the amount of support they received, but their goal preferences were diffuse, implying that they

pursued no clear goals. Interestingly, superiority was the most valued goal domain. The marked

discrepancy between being satisfied with social resources but feeling alienated at the same time

implies that these students’ wish to change school program might be related to personal

characteristics, such as their lack of Dutch language proficiency.

Cluster 3 is the positive school adjusted profile. Scores on superiority goals were

negative. Students in this cluster scored positively on the other goal domains and they were also

positive about the availability of support. They felt attached to school and peers and had no

126

intention at all of changing peers or school. Students with this profile enjoyed school and focused

on learning.

We labeled cluster 4 ‘the frustrated profile’ (Figure 1.4). Interesting about this profile is

the overall frustration that it expresses. Students with this profile seem to have good intentions but

their resources - or in other words their environment - do not match their needs. They score

positively on all goal preferences, are satisfied with their level of Dutch language proficiency, but

dissatisfied with the availability of academic and emotional support, especially from teachers

(possibly caused by a mismatch between students’ social goal preferences and the perceived lack

of academic and emotional support from persons in their environment). They were slightly

negative on identification with peers and school and were considering changing their school

environment. Unlike cluster 2 that grouped students who seem frustrated due to personal

characteristics, this cluster expresses frustration as a consequence of a lack of resources. A

relatively high proportion of students fell in this cluster. Indeed cluster 4 (see Figure 1.4) was the

largest cluster consisting of 704 students, followed by the first and third cluster (see Figure 1.1

and 1.3) that consisted of 652 and 654 students, respectively. The second cluster consisted of 395

students.

The relationship between student profiles and the quality of CL

In order to investigate the relationship between student profiles, the quality of CL and

ethnocultural background a four (profiles) by three (ethnocultural groups) ANOVA was

conducted with the quality of CL as the dependent variable.

Since students with a disaffected profile had negative scores on social and mastery goals,

weak scores on language proficiency, negative scores on the availability of support, a lack of

school and peer belonging and a wish for changing school and peer group, we expected that this

profile would express the lowest quality of CL. In the school adjusted profile we expected the

highest quality of CL, since the scores on social and mastery goals were high, and the Dutch

language proficiency satisfactory (also appraisals of the availability of social support were

positive, and scores on belongingness to school and peers were high, whereas scores on alienation

were low).

127

The analysis yielded a main effect of cluster only (F [3, 1442] = 13.81, p = .000, �² = .03).

Neither an ethnicity main effect nor an interaction effect was found. Contrary to our expectations,

Dutch (M = 2.83, SD = .29), Mediterranean (M = 2.86, SD=.33) and Caribbean students (M =

2.83, SD = .31) hardly differed on the quality of CL. Posthoc tests showed that, as expected,

students with a disaffected profile (M = 2.73, SD = .21), together with students with a weak

communication/school-bonding profile (M = 2.76, SD = .27) had the lowest scores on the quality

of CL. Students with a school adjusted profile had the highest scores on the quality of CL (M =

2.93, SD = .29). This finding confirmed our predictions namely that positive scores on social

goals and mastery goals, negative scores on superiority goals, positive perception of the

availability of support, identification to school and peers and no wish for changing school and

peers are related to perceptions of high CL. The frustrated profile (Figure 1.4) was characterized

by intermediate scores for the quality of CL (M = 2.84, SD = .29).

The fact that students from different ethnic background had similar scores on the quality

of CL does not rule out the possibility that a comparable quality of CL is realized on the basis of

resources and conditions for CL that vary between ethnocultural groups. We analyzed whether the

proportions of students in each profile varied in terms of the students’ ethnocultural background.

--------------------------------

Insert Figure 2 about here

--------------------------------

Figure 2 presents the distribution of ethnic groups over the four clusters. The distribution of

clusters clearly differed by group (�² (6, N = 1791) = 20.80, p = .002). Inspection of Figure 2

shows that cluster 2 - the weak communication profile-, is clearly more characteristic of the

Caribbean group than of the Dutch and Mediterranean groups. This contradicts our expectation

that Dutch and Caribbean students would have a higher language proficiency than their

Mediterranean peers. Moreover, profile 3 - the school adjusted profile, is most representative of

the Dutch group and is least represented in the Caribbean group. Furthermore, profile 4 - the

frustrated profile is most characteristic of the Mediterranean students. This latter finding

corroborates earlier reported findings that Mediterranean students feel less supported whereas

they actually need more support than Dutch students.

128

DISCUSSION AND RECOMMENDATIONS

This study explored what combination of student characteristics and student appraisals of the

school context was related to higher levels of the quality of cooperative learning and whether

these vary by ethnocultural group. We expected to find differences between national,

Mediterranean, and Caribbean students living in the Netherlands, both as regards the quality of

cooperative learning and the profiles representing particular combinations of student

characteristics and student appraisals of the school context. Cluster analyses resulted in four

different profiles, namely a 1) school disaffected profile, 2) weak communication/ school bonding

profile, 3) school adjusted profile and 4) frustrated profile.

The relationship between student profiles and the quality of CL

We expected a relationship between the different profiles and the quality of CL. In line with

Wentzel and Wigfield (1998) and McInerney, Hinkley, Dowson, and Van Etten (1998), we

anticipated that students with high and positive scores on social and mastery goals would express

high quality of CL. Furthermore, we expected that positive scores on Dutch language proficiency,

appraisals of the availability of academic and emotional support from teachers and peers, and

identification with school and peers, and negative scores on school and peer alienation would be

associated with the quality of CL. As expected, we found that students in the school adjusted

profile had the highest scores on the quality of CL, while -in line with our expectations- students

in the school disaffected and the weak communication/ school bonding profiles had the lowest

scores on the quality of CL. Students with a frustrated profile had intermediate scores on the

quality of CL. Inspection of the distribution of students over the different clusters showed that

most students had a frustrated profile. While these students did not display low scores on the

quality of CL, many students felt unsupported by their environment and alienated from their

schools. Future interventions in the schools should explicitly focus on strengthening the social

environment. It is vital that schools shape an environment where it is logical that students support

each other, for example by investing in social skill development and improving group attachment.

Teachers need to create a classroom atmosphere where emotional and academic support is

promoted and valued. At the same time, this type of environment will stimulate students to pursue

129

their social support goals, which are crucial for successful CL as well (Hijzen, Boekaerts, &

Vedder, 2006). Many students (profile 1 and profile 2) had no clear goal preferences. Having

clear goals is very important for successful learning (e.g., Conti, 2000). Hence, suggestions for

interventions concern goal setting. Teachers should invite students to think about their personal

goals and the way these goals connect to school goals. Talking about personal goals in order to

make students conscious of the role these goals play in the learning process might be a significant

move towards more successful CL. Fortunately, a high number of students were in cluster 3, the

school adjusted profile, that was associated with the highest quality of CL and satisfaction with

the availability of support.

Ethnocultural background and the quality of CL

Dutch and immigrant students were not evenly distributed over the different profiles. Only the

disaffected profile comprised comparable proportions of Dutch, Caribbean and Mediterranean

students. As expected, the weak communication/ school bonding profile was more typical of

immigrant students than of the Dutch. However, the difference concerned only the Caribbean

students, while we expected more language difficulties for Mediterranean students. This cluster

was characterized by negative scores on Dutch language proficiency and school and peer

identification and positive scores on school alienation. Besides, students from this cluster had no

specific goal preferences; superiority was their most outspoken goal preference. This finding

contradicts our expectations that immigrant students would value superiority goals less than

Dutch students and it contradicts an earlier finding reported by Vedder and Boekaerts (2006) that

showed that Antillean students had lower scores in the superiority goal domain than Dutch

students. These researchers measured Antillean students’ goal preferences in Antillean schools,

while our research took place in the Dutch context. Perhaps Antillean immigrant students are

more prone to prove themselves and show off their abilities in a Dutch environment than their

peers who are living in the Caribbean. The high scores on school alienation seem to confirm our

prediction that immigrant students feel less attached to their school environment than Dutch

students. This finding may have been caused by several factors, such as differences in cultural

background, or in other words, a gap between the home cultural background and the school

culture, frequently changing schools (Mulder, Roeleveld, Van der Veen, & Vierke, 2005), or

130

language difficulties. A comparison revealed that students in both the school disaffected profile

and the frustrated profile lack social resources. Students in the former cluster report low

emotional support from peers and students in the latter cluster report low social support from their

teacher. The main differences between the two clusters is having or not having clear goals and

reported language proficiency. Obviously having clear goals and language proficiency are

important in learning, particularly in CL. As such it deserves to be given special attention in

students’ curriculum and class organization. Actually, in senior vocational high schools in the

Netherlands generally, students’ Dutch language proficiency is taken for granted. Suggestions for

future intervention programs therefore concern enhancing communication competence and paying

attention to differences between Dutch and immigrants’ language proficiency in class. For

example in composing CL teams.

The frustrated profile was most representative of Mediterranean students and least of the

Dutch group, which confirms our expectation that more Mediterranean students feel alienated and

unable to identify with school and peers, compared to Dutch and Caribbean students.

131

References

Baumeister, R., & Heatherton, T. (1996). Self-regulation failure: An overview. Psychological

Inquiry, 7, 1-15.

Berry, J., Phinney, J., Sam, D., & Vedder, P. (2006). Immigrant youth in cultural transition:

Acculturation, identity and adaptation across national contexts. Mahwah, NJ: Lawrence

Erlbaum Ass.

Bialystok, E. (2001). Bilingualism in development: Language, literacy and cognition. New York:

Cambridge University Press.

Blom, S., & Severiens, S. (2000). Zelfstandig leren van autochtone en allochtone leerlingen in het

Studiehuis [Independent learning of Dutch and immigrant students in the Studiehuis].

PMVO; Amsterdam.

Boekaerts, M. (1998). Do culturally rooted self-construals affect students' conceptualization of

control over learning? Educational Psychologist, 33, 87-108.

Boers, F. (2001). Remembering figurative idioms by hypothesising about their origin. Prospect

16(3): 35-42. BVE monitor-deelnemersmonitor 2003-2004, website bij minocw.

Chin, W., Salisbury, D., Pearson, A., & Stollak, M. (1999). Perceived cohesion in small

groups. Small Group Research, 30, 751-766.

Cohen, E. G. (1994). Restructuring the classroom: Conditions for productive small groups.

Review of Educational Research, 64, 1-35.

Conti, R. (2000). College goals: Do self-determined and carefully considered goals predict

intrinsic motivation, academic performance, and adjustment during the first semester?

Social Psychology of Education, 4, 189-211.

Deci, E. L., & Ryan, R. M. (2000). The "what" and "why" of goal pursuits: Human needs and the

self-determination of behavior. Psychological Inquiry, 11, 227-268.

Distelbrink, M., & Pels, T. (2000). Opvoeding in het gezin en integratie in het onderwijs

[Upbringing in the family and integration at school]. In: Pels, T. (ed.). Opvoeding en

integratie. Een vergelijkende studie van recente onderzoeken naar gezinsopvoeding en

pedagogische afstemming tussen gezin en school (pp. 114-139). Assen: Van Gorcum.

Eccles, J. (1987). Gender roles and women’s achievement-related decisions. Psychology of

Women Quarterly, 11, 135–72.

132

Ford, M. E. (1992). Motivating humans: goals, emotions, and personal agency. Newbury

Park, CA: Sage.

Ford, M. E., & Nichols, C. W. (1991). Using a goal assessment to identify motivational patterns

and facilitate behavorial regulation and achievement. In M. Maehr & P. R. Pintrich (Eds),

Advances in motivation and achievement: Vol 7. Goals and self-regulatory processes (pp.

61-84). Greenwich, CT: JAI Press.

Hajer, M. (1996). Leren in een tweede taal. Interactie in vakonderwijs aan een meertalige mavo-

klas [Learning in a second language. Interaction in didactical education with a multi

language mavo-class]. Groningen: Wolters Noordhoff.

Hijzen, D., Boekaerts, M., & Vedder, P. (2006). The relationship between students' goal

preferences, perceptions of contextual factors in the classroom and the quality of

cooperative learning. Scandinavian Journal of Psychology, 47, 9-21.

Hijzen, D., Boekaerts, M., & Vedder, P. (submitted). Exploring the links between students’

engagement in cooperative learning, their goal preferences and perceptions on contextual

factors in the classroom.

Johnson, D. W., & Johnson, R. T. (1994). Learning together and alone: Cooperative,

competitive, and individualistic learning. Englewood Cliffs, NJ: Prentice-Hall.

Kirchmeyer, C. (1993). Multicultural task groups. An account of the low contribution level of

minorities. Small Group Research, 24, 127-148.

Kwak, K. (1990). Second language learning in a multicultural society: A comparison between

the learning of a dominant language and a heritage. PhD thesis Queen's University,

Kingston, ON, Canada.

Matthijssen, M. (1993). Lessen in orde voor immigranten: leerling perspectieven van allochtonen

in het voortgezet onderwijs [Lessons alright for immigrants: students’ perspectives of

immigrants in vocational schools].Amersfoort: Academische Uitgeverij.

McInerney, D. M., Hinkley, H., Dowson, M., & Van Etten, S. (1998). Children’s beliefs

about success in the classroom: Are there cultural differences? Journal of Educational

Psychology, 90, 621-629.

McInerney, D. M., Roche, L. A., McInerney, V., & Marsh, H. W. (1997). Cultural perspectives

on school motivation; the relevance and application of goal theory. American Educational

Research Journal, 34, 207-236.

133

Meeus, W. (1996). Studies on identity development in adolescence: An overview of research and

some new data. Journal of Youth and Adolescence, 25, 569-598.

Mulder, L., Roeleveld, J., Van der Veen, I., & Vierke, H. (2005).Onderwijsachterstanden tussen

1988 en 2002: ontwikkelingen in basis- en voortgezet onderwijs [Educational backlog

between 1988 and 2002: developments in primary- and secondary education]. ITS

Radboud Universiteit, Nijmegen.

Palfreyman, D. (2001). The socio-cultural construction of learner autonomy and learner

independence in a tertiary EFL institution. Unpublished PhD thesis. Canterbury:

Canterbury Christ Church University College, Department of Language Studies.

Pels, T. (1998). Opvoeding in Marokkaanse gezinnen. De creatie van een nieuw bestaan

[Upbringing in Moroccan families. The creation of a new existence]. Assen: Van Gorcum.

Ros, A. A. (1994). Samenwerking tussen leerlingen en effectief onderwijs [Cooperation

between students and effective education]. Dissertation, Rijksuniversiteit

Groningen.

Samdal, O., Nutbeam, B., & Kannas, L. (1998). Achieving health and educational goals through

schools-study of the importance of the school climate and the students’ satisfaction

with school. Health Education Research, 13, 383-397.

Slavin, R. E. (1987). Developmental and motivational perspectives on cooperative learning: A

reconciliation. Child Development, 58, 1161-1167.

Slavin, R. E. (1995). Cooperative learning: Theory, research, and practice (2nd ed.).

Boston: Allyn and Bacon.

Suarez-Orosco, M. (1998). Commentary. Paper presented at Transnationalism and the Second

Generation, Harvard University.

Ten Berge, J. M. F. (1986). Rotatie naar perfecte congruentie en de multiple groep methode

[Rotation to perfect congruence and the multiple group method]. Nederlands Tijdschrift

voor de Psychologie, 41, 218-225.

Teunissen, J., & Matthijssen, M. (1996). Stagnatie in onderwijsonderzoek naar de etnische factor

bij allochtone leerlingen [Stagnation in educational research on the ethnic factor amongst

immigrant students]. Sociologische Gids, 43, 87-99.

Timmermans, R. (2005). Addition and substraction strategies: assessment and instruction.

Dissertation: Radboud University Nijmegen.

134

Valdez, G. (1998). The world outside and inside school: Language and immigrant children.

Educational Researcher, 27,4-18.

Van Ours, J., & Veenman, J. (2001). The Educational Attainment of Second Generation

Immigrants in the Netherlands, IZA Discussion Papers 297, Institute for the Study of

Labor (IZA).

Vedder, P., & Boekaerts, M. (in press). Motivation in senior secondary vocational education; a

cross-cultural comparison. Cross Cultural Psychology.

Vedder, P., Boekaerts, M., & Seegers, G. (2005). Perceived social support and well being in

schools; the role of students' ethnicity. Journal of Youth and Adolescence, 34, 269-

278.

Vedder, P., & Virta E. (2005). Language, ethnic identity, and the adaptation of Turkish

immigrant youth in the Netherlands and Sweden. International Journal of Intercultural

Relations, 29, 317-337.

Voncken, E., Van der Kuip, I., Moerkamp, T., & Felix, C. (2000). Je bent jong en je weet

niet wat je wilt: Een inventarisatie van push- en pullfactoren die leiden tot voortijdig

schoolverlaten in de BVE-sector [You are young and you do not know what you want: an

inventory of push- and pull factors that lead to drop-out in secondary vocational

education]. Amsterdam: SCO-Kohnstamm.

Webb, N. M., & Palincsar, A. S. (1996). Group processes in the classroom. In: D.C. Berliner

& R.C. Calfee (Eds), Handbook of Educational Psychology (pp. 841 - 873). New York:

Simon & Schuster MacMillan.

Wentzel, K. R. (1991). Relations between social competence and academic achievement in early

adolescence. Child Development, 62, 1066-1078.

Wentzel, K. R., & Wigfield, A. (1998). Academic and social motivational influences on

students' academic performance. Educational Psychological Review, 10, 1555-1575.

Wethington, E., & Kessler, R. (1986). Perceived Support, Received Support, and

Adjustment to Stressful Life Events. Journal of Health and Social Behavior 27, 78-89.

Wittebrood, K., & Keuzekamp, S. (2000), Rapportage Jeugd 2000. Trajecten van jongeren naar

zelfstandigheid [Trajectories of youngsters to independence]. Den Haag: Sociaal en

Cultureel Planbureau.

135

Table 1

Categories, Sample Items, Number of Items, Cronbachs’ Alpha Coefficients and Tucker Phi Coefficients

Category Sample item(s) #

items

Alpha Tucker’s Phi Coefficients of Factorial

Agreement

Students Goal Preferences Dutch vs.

Mediterranean

Dutch vs.

Caribbean

Caribbean vs.

Mediterranean

Superiority/ individuality I want to impress others 9 .93 1.00 .99 .99

Mastery I want to learn more about my profession 9 .92 1.00 .99 .99

Belongingness I want to get along with my peers 6 .86 1.00 .99 .99

Social Support I want to help others in case they need help 7 .91 1.00 1.00 1.00

Quality of Cooperative

Learning

I perceive myself as part of this group”, “When we work on

a group task, we make sure that all the team members

understand the answers”, “I know when another person

needs help” and “Together you learn better than alone”.

29 .90 1.00 1.00 1.00

Language Competence

Dutch language competence I speak (e.g. read) Dutch well 4 .87 1.00 1.00 1.00

Social Climate

136

Academic support teacher When I do not understand the lesson, I receive support

from my teacher

7 .80 .99 .99 .98

Academic support peers When I do not understand the lesson the I receive support

from my peers

7 .82 1.00 .98 .98

Emotional support teachers When I am sad my teacher supports me 6 .82 1.00 1.00 .99

Emotional support peers When I am sad my peers support me 6 .89 .97 .99 .99

School and Peer

Identification & Alienation

Identification with peers My friends give me self confidence 5 .91 1.00 1.00 1.00

Identification with school My education gives me self confidence 5 .88 1.00 .99 .99

Alienation from peers I regularly think about finding new friends 3 .93 1.00 1.00 1.00

Alienation from school I regularly think about finding another education 3 .91 1.00 1.00 1.00

137

Figure 1.1:

The school disaffected profile

-1,6

-1,4

-1,2

-1

-0,8

-0,6

-0,4

-0,2

0

0,2

0,4

0,6 social support

belongingness

mastery

superiority

Dutch language proficiency

academic support teacher

academic support peers

emotional support teachers

emotional support peers

peer identification

school identification

school alienation

Figure 1.2:

The weak communication/ weak school bonding profile

-1

-0,8

-0,6

-0,4

-0,2

0

0,2

0,4 social support

belongingness

mastery

superiority

Dutch language proficiency

academic support teacher

academic support peers

emotional support teachers

emotional support peers

peer identification

school identification

school alienation

138

-0,4

-0,2

0

0,2

0,4

0,6

0,8

1

social supportbelongingnessmasterysuperiorityDutch language proficiencyacademic support teacheracademic support peersemotional support teachersemotional support peerspeer identificationschool identificationschool alienation

Figure 1.3:

The school adjusted profile

-0,6

-0,4

-0,2

0

0,2

0,4

0,6

0,8 social support

belongingness

mastery

superiority

Dutch language proficiency

academic support teacher

academic support peers

emotional support teachers

emotional support peers

peer identification

school identification

school alienation

Figure 1.4: The frustrated profile

139

0

5

10

15

20

25

30

35

40

profile 1 profile 2 profile 3 profile 4

DutchMediterraneanCaribbean

Figure 2:

Percentages of Dutch, Mediterranean and Caribbean students by profile

140

Chapter 6

GENERAL CONCLUSIONS AND DISCUSSION

This final chapter has two main aims. Firstly, we will discuss the conclusions of the four

separate studies and draw general conclusions. Secondly, we will exemplify what the

implications of these conclusions are for theory and practice, illustrated by an additional

study on teacher perceptions of the quality of CL in their classrooms.

Answering the general questions

The main aim of this study was to investigate the role of students’ goal preferences in CL

settings and to determine factors in the classroom context that teachers can manipulate in

order to promote successful CL processes. Special attention was paid to differences between

students related to their gender, program type and ethnocultural background. By gaining

insight into these relationships we intended to uncover leads for future interventions aimed at

improving students’ motivation for CL and the quality of their learning processes as well as

preventing drop-out in the long run. Four broad, explorative questions were central to this

dissertation. Namely:

1. What is the relationship between students’ goal preferences, contextual factors in the

classroom and the quality of CL?

2. How can effective CL teams be distinguished from ineffective ones, and what

distinguishes them in terms of the students’ goal preferences and perceptions of

contextual factors in the classroom?

3. Which teacher related conditions coincide with effective CL processes and

which conditions are related to failing CL processes, in the course of a year?

4. Can we distinguish between separate profiles of person variables (Dutch language

proficiency and goal preferences) and context variables (social resources and

school belonging) that account for variations in the quality of CL and does ethnic

background play a role in explaining differences in these profiles and the quality of

CL?

These questions were dealt with in four studies, presented in Chapters two, three, four and

five of the thesis. In Chapter two we described the relationship between students’ goal

141

preferences, contextual factors and the quality of CL processes. Findings of the study

described in this chapter functioned as a framework for the other studies. In Chapter three we

presented an in depth study exploring how effective CL teams could be distinguished from

ineffective teams in terms of students’ goal preferences and their perceptions of contextual

factors. In Chapter four we presented a longitudinal study on the relationship between teacher

steered conditions for CL and the quality of CL. In Chapter five we described a study that

investigated the role of students’ ethnocultural background and the quality of CL.

What is the relationship between students’ goal preferences, contextual factors in the

classroom and the quality of CL?

As to the relationship between students’ goal preferences and the quality of CL we predicted

that belongingness, social support and mastery goals are positively and superiority/

individuality goal preferences negatively related to the quality of CL. We found that social

support goals were strongly related to the quality of CL. Also students’ belongingness and

mastery goals were related – although slightly less- to the quality of CL. This confirms

previous findings of studies by McInerney, Hinkley, Dowson, and Van Etten (1998) and

Wentzel (1993), suggesting that a combination of social and mastery goals is preferable in

social learning settings, such as CL. Unexpectedly, students’ superiority/ individuality goals

were not significantly related to the quality of CL, while we expected that students, who are

superiority minded, to be less capable of working in CL settings since they were expected to

outperform others and less cooperatively minded.

As to the relationship between the quality of CL and perceptions of contextual factors

in the classroom we found that students’ evaluation of the extent that they were taught CL

skills at their present schools was most strongly and positively related to the quality of CL.

This finding confirms previous findings (e.g., Gillies & Ashman, 1996; Hoek, Van den Eeden,

& Terwel, 1999; Webb & Farivar, 1994). Also related to the quality of CL were students’

perception of the teachers’ clarity on rules for CL, teacher monitoring behavior, perceived

availability of peer academic and emotional support, and perceived availability of teacher

support.

In line with several other studies, (e.g., Anderman, 1999; Charlesworth & Dzur, 1987;

Cosden, Pearl, & Bryan 1985; Eccles, 1987; Wentzel, 1991) which showed that female

students are more inclined to engage in behavior associated with successful CL, such as

helping others, we found that females showed overall higher scores than male students on the

142

quality of CL. They also reported higher preferences for social support and mastery goals

whereas male students reported higher preferences for superiority goals. Consequently, these

findings suggest that female students, more than their male peers, feel secure in CL settings.

Our study in secondary vocational schools dealt with students enrolled in ICT/

engineering, retail and administration, health and welfare, and food and tourism programs.

Male and female students were not equally distributed over these program types and this

uneven distribution might have led to a program type effect that masks an underlying gender

effect. Therefore, we explored program type effects for male and female students separately.

Interestingly, we found a positive relationship between students’ superiority goals and the

quality of CL in the ICT/ engineering (male) subgroup, whereas in the health and welfare male

subgroup no such relationship was found. In the food male subgroup a negative relationship

was found.

We concluded that the quality of CL was best predicted by a combination of students’

social support goals, their evaluations of the extent that they were taught cooperation skills,

teachers’ monitoring behavior, and the availability of academic and emotional peer support. A

salient predictor in the classroom context was the students’ perception of the extent that they

had been taught the necessary CL skills.

How can effective CL teams be distinguished from ineffective ones, and what

distinguishes them in terms of the students’ goal preferences and perceptions of

contextual factors in the classroom?

In the study described in Chapter three we distinguished effective CL teams that

predominantly show (social) task-relevant engagement (being concentrated and active) during

CL from ineffective teams that show task-irrelevant engagement (being distracted, and

chatting most of the time), in terms of their goal preferences and perceptions of contextual

factors in the classroom.

The goal questionnaire revealed that effective teams’ most prevalent goal preferences

were affective, social support and self-determination goals, whereas ineffective team

members’ most prevalent goal preferences were belongingness, affective and self-

determination goals. Hence, the most remarkable distinction between these teams concerned

their belongingness and social support goals. Effective team members were inclined to report

somewhat lower scores on belongingness goals than ineffective team members, while the

reversed pattern was observed for social support goals. In effective teams mastery goals were

143

more important than belongingness goals. Remarkably, only one goal domain was related to

students’ engagement scores, namely their belongingness goals. A negative relationship was

found between students’ belongingness goals and task-relevant engagement in the ineffective

teams. Inspection of students’ interview statements –where the full range of possible goal

preferences was considered- showed that mastery and social responsibility goals, together

with ‘learning for a certificate’ goal, tended to be more prevalent in effective teams, while

learning for a certificate and entertainment goals were dominant in ineffective teams. The

most substantial dissimilarity in goal preferences was the strong prevalence of entertainment

goals in ineffective CL teams. In line with Sheldon and Kasser (1995) and Sheldon and Elliot

(1999) we found that students in ineffective teams seem less conscious of their goal

preferences than students in effective teams. Their goals were very broad. Important to note is

that goal preferences were not often specifically mentioned in explaining task-relevant, task-

irrelevant, or social task related engagement during CL sessions. Groups pointed at the

context far more often to explain their CL. Task characteristics, group composition, and

teacher behavior were often mentioned as reasons for effective or ineffective CL. Ineffective

teams explained their task-irrelevant engagement as a result of the group tasks. According to

these students, many tasks were not genuine group tasks. Also they complained that the tasks

were boring, too simple and not challenging enough. This confirms findings of Webb and

Palincsar (1996) and Cohen (1994) suggesting that for effective CL students perceive the

tasks as challenging, as hands-on, and promoting interdependency. In line with Sharan and

Sharan (1992) and Chin, Salisbury, Pearson, and Stollak (1999) we found that students in

effective teams usually worked longer in the same teams and felt more at ease with each other

than team members in ineffective teams, who often had to deal with absent or highly

unmotivated team members. Ineffective team members had many complaints about their

teachers, who were often not there in case they needed help. Furthermore, they mentioned the

fact that they were just not good at working independently.

Which teacher related conditions coincide with effective CL processes and which

conditions are related to failing CL processes, in the course of a year?

In the study presented in Chapter 4, we explored the conditions that teachers in secondary

vocational education created in order to promote students’ CL processes. We found that the

extent that students were taught skills and knowledge for CL and teachers’ clarity on rules for

CL were highly related to the quality of CL, during all three waves. In other words, findings

144

of the study presented in Chapter 2 can be complemented by stating that -also in the long run-

it is important that teachers explicitly teach the skills, knowledge and rules for CL. The

quality of CL was at its best during the second wave when scores on all teacher related

conditions were also highest. Hence, teachers’ control behavior and the quality of CL were

also positively related at the second data point, while we expected that a decrease in teachers

control behavior would predict the quality of CL at a later stage. However, during the third

data-wave only the scales that measured students’ perceptions on the extent that they were

taught skills and knowledge for CL, and teachers’ clarity on rules for CL were highly related

to the quality of CL, whereas the scales that measured students’ perceptions of their teachers’

control behavior were indeed less related to the quality of CL.

When we distinguished between the group of ineffective and effective cooperators we

noticed that the teacher related conditions made the difference: at all stages effective

cooperators had higher scores on all the scales, in particular on the extent that they were

taught skills and knowledge for CL. Furthermore, we signaled a tendency that the weak

cooperators perceived a major decrease in teachers’ monitoring behavior after the second

data-wave. A similar trend was found in relation to teachers’ intervention behavior. The

effective cooperators perceived almost no change in teacher monitoring and intervention

behavior after the second data-wave.

Can we distinguish between separate profiles of person variables (Dutch language

proficiency and goal preferences) and context variables (social resources and school

belonging) that account for variations in the quality of CL and does ethnic background play

a role in explaining differences in these profiles and the quality of CL?

Four student profiles were identified; a school-disaffected, a weak communication/school

bonding, a school-adjusted and a frustrated profile. Not surprisingly, students who were

grouped in the school-adjusted profile showed the highest scores on CL. This profile was

characterized by goal preferences for social and mastery goals, low scores on superiority

goals, high perceived availability of social support and positive scores on school and peer

identification. The school-disaffected profile had the lowest scores on CL. This profile was

characterized by no clear goals, a lack of social resources and peer/school identification.

Students from different backgrounds were disproportionately distributed. The weak

communication profile was characteristic of Caribbean students. Students with this profile

were dissatisfied with their Dutch language proficiency. Also, they scored relatively high on

145

school alienation. The scores on the availability of academic and emotional peer and teacher

support were high, meaning that students in this profile were satisfied with the amount of

support they received, but their goal preferences were diffuse. Interestingly, superiority was

the most valued goal domain in this profile. The frustrated profile was characterized by clear

goals, but dissatisfaction with the availability of academic and emotional support, especially

from teachers. Furthermore, they were slightly low on identification with peers and school

and were considering changing their school environment. Somewhat disappointing, the

highest proportion of students fell in this cluster. The lowest proportion of students fell in the

weak communication cluster. Contrary to our expectations we had to conclude that students’

ethnocultural background had neither a direct effect on the quality of CL nor on students’ goal

preferences. In summary, we were unable to confirm the hypotheses that specific ethnic

groups were better equipped for CL (Palfreyman, 2001; Wittebrood & Keuzekamp, 2000).

However, we were able to determine ethnicity-related variables (Dutch language proficiency,

school alienation) that are likely affected by educational interventions.

As was shown in Chapter two, three, and five, we concluded that with respect to students’

goal preferences, social support and mastery goals were most vital in predicting the quality of

CL. Having clear goals in the first place was also noteworthy in predicting effective and

ineffective CL processes. All studies showed that the classroom context was -even more-

crucial in predicting the quality of CL, especially the extent to which students were taught the

appropriate knowledge, skills and rules for CL was found to be a crucial -and lasting-

precondition of successful CL. The fact that some teams were able and willing to cope with

hindrances, distractions and obstacles while working in CL settings, whereas others were not,

highly depends on what goal preferences they had in the first place. A well-designed CL

setting is decisive, because it elicits, promotes, or hinders certain goal preferences.

Furthermore, as was found in Chapters two and five, background variables such as

ethnocultural background, gender and program type had no direct effect on the quality of CL.

Instead, they indirectly influenced the quality of CL.

CL may be a way to enhance students’ motivation for learning, provided that students

are aware of their goal preferences and CL is well implemented as an instructional process.

These conclusions are promising for finding leads for future intervention purposes; stable

factors, such as students’ ethnocultural background, gender and program type, as well as their

goal preferences are difficult to change. These factors have been found to be less essential in

146

predicting the quality of CL than perceptions of the classroom context, which is likely to be

much more susceptible for intervention purposes.

GENERAL IMPLICATIONS FOR THEORY AND PRACTICE

Students’ goal preferences and the quality of CL

With respect to the relationship between students’ goal preferences and the quality of CL, this

dissertation underlined the need for promoting students’ social and mastery goals. A

significant finding of this study is that not all students devote much thinking to choosing their

goals, whereas formulating goals, can facilitate students’ intrinsic motivation (Sheldon &

Kasser, 1995; Sheldon & Elliot, 1999). Hence, our first recommendation concerns the

importance of discussing personal goals in order to make students sensitive of the role these

goals play in the learning process. The goals that students bring into the classroom are often

abstract goals, which still need to be adapted to the CL setting. Especially the study that was

described in Chapter three suggested that inviting students to talk about their goals could be

an important step towards more effective CL; perhaps the stimulated-recall setting provided a

favorable context for thinking about one’s goals; it made students more aware of their goals

and their importance in the specific context of CL. In other words, the stimulated-recall

setting might be considered as an ideal educational intervention to make students more

sensitive of their goals. Ideally, this should be a recurring element of the curriculum,

especially when we consider the high drop-out rates in secondary vocational schools.

Dropping out of school often is an outcome of underlying motivational problems that students

have experienced. Students should be invited to think about their own goals and about the

links between their personal goals and school goals. This reflection might facilitate adoption

of teacher-set learning goals and self-regulation of their own learning goals (see Boekaerts &

Corno, 2005). Moreover, since social support goals were found to be important in predicting

the quality of CL, teachers should create a classroom context where students are stimulated to

rely on each other for help.

Perceptions of contextual factors and the quality of CL

The studies reported in this thesis suggest that, more important than students’ goals, the

classroom context was crucial for the quality of CL. We would therefore like to end this

147

dissertation with additional findings that underline the importance of paying attention to the

classroom setting and especially to the teachers’ role in CL.

A number of teachers completed an abridged version of the questionnaire that aimed

to measure conditions for CL and also participated in an interview about CL in their

classrooms. Comparisons of teacher and student scores yielded some interesting findings that

together point toward a call for improvement of adaptability of teacher behavior during CL.

Teachers’ scores on all teacher behavior items was (much) higher than students’ scores. This

finding indicates that -according to the teachers- they were far more involved and active

during CL than evaluated by the students. For example, teachers were more positive about the

cooperation skills that they taught their students than the students themselves. This finding is

of special importance when we consider findings of the in-depth study described in Chapter

two; students in ineffective CL teams mentioned that they were not good at cooperating and

that they missed the skills for effective CL. When we contextualize this finding in the overall

results of the study (the strong relationship between the quality of CL and the extent that

students were explicitly taught CL skills), we may conclude that students perceive their

teachers as less monitoring and intervening than the teachers themselves. Remember that the

study presented in Chapter four showed that after a one year period, ineffective cooperators

reported lower levels of teacher instruction and teacher monitoring.

These same teachers also participated in interviews about the way CL was realized in

their classes, their ideas on the usefulness and feasibility of working with CL methods and the

quality of their students CL skills. More than half of the teachers answered that students

cooperatively completed tasks in their lessons, however the way this took place was often less

organized and structured than CL settings are meant to be designed by educationalists. Some

of the statements suggested that teachers do not always set up highly structured CL settings

with real group tasks and clear notions about group compositions: “Simply make an

assignment together” or “They don’t get a real group task but they have to make assignments

together. They can complete these in couples or in larger groups, as long as they are engaged

and learn something...”.

Teachers were asked about their attitudes towards CL. Most teachers were positive

about the general usefulness of CL. Most teachers mentioned that practising CL skills is

important, because students need these skills in their future career: “Improving social skills is

very important. Being able to cooperate in the future, that is what it is all about, they learn to

solve a problem together, and in their future professions they also have to solve tasks

together”. However, only a small percentage of the teachers thought that students learn more

148

in CL settings than in traditional settings. An often mentioned disadvantage of CL was the

time aspect. Coaching different teams and solving conflicts takes a lot of time. One teacher

stated “When they enter the classroom, they have to get their books and pencils and so on. I

have to check who is absent. This already takes ten minutes and then you have to start

explaining the topics, compose groups, well yes there is too little time”. Many felt to have too

little instruction time and that in the remaining time they insufficiently covered the prescribed

curriculum. They compared this to more traditional instructional settings and were of the

opinion that more instruction time and curriculum coverage would be realized using more

traditional approaches. This suggests a bad prospect for CL, because it may be an early signal

of re-lapse and may reflect their unwillingness to invest in well structured CL methods.

Many teachers were ambiguous about the quality of their students’ CL skills. A

teacher stated: “I notice that students are not really capable of CL. They don’t take any

initiative. I have to push them and even then it still goes wrong often. I think CL is very

important, but also really difficult to make it work.” Another teacher said: “On the one hand

I think it is important that students learn to cooperate, because they need the skills for their

future career. However, I think that nowadays students cooperate too often and it demotivates

them. I think that the individual student receives less attention”.

Teachers mentioned that most students are extrinsically motivated, and have difficulty

working independently in CL settings. Relevant statements were “I noticed that you have to

set up a strong extrinsic motivation, control them a lot” or “I don’t notice that students get

more independent by CL. The assignments they hand in are of a very low quality. This is

often caused by the fact that they receive a mark that does not seriously add up to their total

grade”. These findings nicely illustrate the gap between student and teacher perceptions on

conditions for CL. The teachers are of the opinion that they instruct and control CL processes

sufficiently and more or less hold their students responsible for ineffective CL process. Even

though teachers believe that they invest sufficient time in teaching CL skills and knowledge,

and that they control CL processes, it may be necessary to encourage them to invest more

time in the preparation and reflection on CL lessons and the reasons why this form of learning

is beneficial for students.

Hence, our most central recommendation concerns the teachers’ role during CL. They

should carefully think about how to organize the CL setting in terms of group composition,

task characteristics and rewards. Teachers should pay ample attention to and explicitly teach

the necessary skills for learning in CL teams. More specifically, as mentioned in Chapters

two, three, and four, it is crucial that teachers teach their students how to listen to each other,

149

to evaluate the group process, to discuss, to support group members, to give an opinion, or to

solve group conflicts. Secondly, it is important that teachers monitor the CL process carefully,

which means that they need to walk around in the classroom, frequently check with the

groups and ask them how they are doing. Thirdly, teachers need to be aware that availability

of peer support is essential for effective CL, emotional as well as instrumental support.

Teachers should encourage students to provide this type of support, and create conditions to

facilitate this support.

Finally, although the study in Chapter five showed that ethnocultural background had

no direct relationship with the quality of CL, students’ communication skills are very

important in CL setting. Therefore we think that improving students’ communication skills

should be given special attention in the curriculum and in class organization. Actually, in

senior vocational high schools in the Netherlands generally, students’ Dutch language

proficiency is taken for granted, as something that was given sufficient attention in students’

preceding school career. Suggestions for future intervention programs therefore concern

enhancing communication competence and paying attention to differences between Dutch and

immigrants’ language proficiency in class.

LIMITATIONS OF THE STUDY

This thesis concludes with some suggestions for future research related to particular

limitations of the present study. The most important limitations of this study concern a

sampling bias favoring girls and the significant loss of subjects in the course of data-

collection. Boys are underrepresented at all data-waves. This was due to the fact that

relatively many programs of health and well-being participated. These programs are

traditionally favored by girls. Furthermore, students enrolled in cognitively less demanding

programs were underrepresented, while students enrolled in cognitively more demanding

programs were overrepresented. Last but not least, only 260 students completed the

questionnaires at three data-waves. With this high attrition rate we cannot exclude the

possibility that our findings are only generalizable to students that attend class frequently and

persistently and do not drop out. Perhaps our findings paint a more positive picture in terms of

students’ motivation and the quality of cooperative learning than warranted for the whole

population of senior vocational high schools. It would be interesting to find out more about

the non-participating students. Did they drop out and if so were their reasons for drop-out

related to a mismatch between their personal goal preferences and the school imposed goals?

150

Were they dissatisfied with particular aspects of the classroom context. Perhaps we ended up

with a selection of the highly motivated students, while actually having more insight in the

behavior and reasons of the less motivated students would be particularly interesting for this

type of research. After all, we wanted to find out more about the role of students’ motivation,

as represented by students’ goal preferences for students’ learning and school adjustment.

More in general it would be interesting to replicate the study with a larger and more evenly

distributed group of subjects at all data-waves.

151

References

Anderman, L. H. (1999). Expanding the discussion of social perceptions and academic

outcomes: Mechanisms and contextual influences. In M. L. Maehr & P. R. Pintrich

(Series Eds.) and T. Urdan (Vol. Ed.), Advances in motivation and achievement: Vol.

11. The role of context (Vol. 11, pp. 303–336). Stamford, CT: JAI.

Boekaerts, M., & Corno, L. (2005). Self-regulation in the classroom: A perspective on

assessment and intervention. Applied Psychology: An International Review, 24, 199–

231.

Charlesworth, W., & Dzur, C. (1987). Gender comparisons of preschoolers behavior and

resource utilization in group problem-solving. Child Development, 58, 191–200.

Chin, W., Salisbury, D., Pearson, A., & Stollak, M. (1999). Perceived cohesion in small

groups. Small Group Research, 30, 751–766.

Cohen, E. G. (1994). Restructuring the classroom: Conditions for productive small groups.

Review of Educational Research, 64, 1– 35.

Cosden, M., Pearl, R., & Bryan, T.H. (1985). The effects of cooperative and individual goal

structures on learning disabled and nondisabled students. Exceptional Children, 52,

103–114.

Eccles, J. (1987). Gender roles and women’s achievement-related decisions. Psychology of

Women Quarterly, 11, 135–72.

Gillies, R., & Ashman, A. (1996). Teaching collaborative skills to primary school children in

classroom-based workgroups. Learning and Instruction, 6, 187–200.

Hoek, D., Van den Eeden, P., & Terwel, J. (1999). The effects of integrated social and

Learning and Instruction, 9, 427–448.

McInerney, D. M., Hinkley, H., Dowson, M., & Van Etten, S. (1998). Children’s beliefs

about success in the classroom: Are there cultural differences? Journal of

Educational Psychology, 90, 621–629.

Palfreyman, D. (2001). The socio-cultural construction of learner autonomy and learner

independence in a tertiary EFL institution. Unpublished PhD thesis. Canterbury:

Canterbury Christ Church University College, Department of Language Studies.

Sharan, S., & Sharan, Y. (1992). Expanding Cooperative Learning through Group

Investigation. New York:Teachers College Press.

Sheldon, K. M., & Elliot, A.J. (1999). Goal striving, need-satisfaction, and longitudinal well-

being: The Self-Concordance Model. Journal of Personality and Social Psychology,

152

76, 482-497.

Sheldon, K. M., & Kasser, T. (1995). Coherence and congruence: Two aspects of personality

integration. Journal of Personality and Social Psychology, 68, 531 543.

Webb, N. M., & Farivar, S. (1994). Promoting helping behavior in cooperative small groups

in middle school mathematics. American Educational Research Journal, 31, 369–395.

Webb, N. M., & Palincsar, A. S. (1996). Group processes in the classroom. In D. C. Berliner

& R. C. Calfee (Eds.), Handbook of educational psychology (pp. 841–873). New

York: Simon & Schuster MacMillan.

Wentzel, K. R. (1991). Relations between social competence and academic achievement in

early adolescence. Child Development, 62, 1066–1078.

Wentzel, K. R. (1993). Motivation and achievement in early adolescence: The role of multiple

classroom goals. Journal of Early Adolescence, 13, 4–20.

Wittebrood, K., & Keuzekamp, S. (2000), Rapportage Jeugd 2000. Trajecten van jongeren

naar zelfstandigheid [Trajectories of youngsters to independence]. Den Haag: Sociaal

en Cultureel Planbureau.

153

Summary

The research presented in this thesis is an exploration of the relationship between students’

motivation, represented by students’ personal goals and the quality of cooperative learning

(henceforth CL) processes of first and second year students enrolled in secondary vocational

schools. Special attention has been paid to contextual factors and their influence on the

quality of CL, and to differences between students that are related to their ethnocultural

background. Cooperative learning refers to all those learning methods, where students work

on assignments cooperatively in situations that allow or stimulate cooperation.

Four questions were central to this thesis, namely. 1) What is the relationship

between students’ goal preferences, contextual factors in the classroom and the quality of CL?

2) How can effective CL teams be distinguished from ineffective ones, and what distinguishes

them in terms of the students’ goal preferences and perceptions of contextual factors in the

classroom? 3) Which teacher related conditions coincide with effective CL processes and

which conditions are related to failing CL processes, in the course of a year? 4) Can we

distinguish between separate profiles of person variables (Dutch language proficiency and

goal preferences) and context variables (social resources and school belonging) that account

for variations in the quality of CL and does ethnic background play a role in explaining

differences in these profiles and the quality of CL?

The Netherlands has 42 regional educational centers for secondary vocational education. They

all received a letter in which we explained the purpose and relevance of the study and invited

them to participate. Eleven schools evenly spread over the Netherlands participated. The

study had a longitudinal design with three data-waves. Data were gathered at three data

points, between December 2001 and May 2003. During the first data-wave students were

halfway in their first year. The second data-wave took place halfway in their second year, and

the third data-wave at the end of the second year. During our study we had to deal with a large

decline of participating students; during the first data-wave 1920 completed our

questionnaires, at data wave two 897 and at data wave three only 639 students. Senior

vocational school delivers educational programs for four broad competency levels. The first

competency level is the lowest level and level four is the highest. Students enrolled at all four

levels participated. Students in secondary vocational schools can choose for different program

types. We distinguished engineering and ICT, retail and administration, food and tourism, and

health and welfare programs.

154

Students’ perceptions on the quality of their CL processes were measured in this study

as well as contextual factors (the extent that they were taught knowledge, skills and rules for

CL, teacher monitoring, intervention and evaluation behavior), students attitudes towards CL

and their Dutch and general language proficiency. Furthermore, students completed

questionnaires on their goal preferences and the social climate in school.

With respect to the relationship between students’ goal preferences and the quality of CL we

concluded that social support goals had the strongest relationship with the quality of CL,

followed by mastery and belongingness goals. Regarding the relationship between contextual

factors and the quality of CL we can conclude that students’ perceptions on the extent that

they were taught skills, knowledge and rules for CL and teachers monitoring the learning

process, were related to the quality of CL. Also, the availability of peer and teacher support

were related to the quality of CL. The quality of CL was best predicted by a combination of

students’ social support goals, their evaluations of the extent that they were taught

cooperation skills, teachers’ monitoring behavior and the availability of academic and

emotional peer support. An interesting finding was that students’ goal preferences only added

a little to the explanation of the quality of CL. Context appraisals were much more important

in explaining the quality of CL. Furthermore, we were able to confirm previous findings

suggesting that female students had higher scores on the quality of CL. They also valued

social support and mastery goals more than male students who often had higher scores on

superiority goals.

An important question in the study was how to distinguish between effective CL teams

and ineffective teams. The in-depth study with a stimulated-recall method showed that a

remarkable difference between these groups was related to students’ belongingness and social

support goals. Students in ineffective CL teams preferred belongingness goals over their

social support goals, whereas the reversed pattern was found in effective CL teams. Important

to note is that only one relationship was found between students goal preferences and their

engagement levels, which was precisely between students’ belongingness goals and task

related engagement. This was a negative relationship in the ineffective CL teams. Also we

found that mastery and social responsibility goals – together with ‘learning for a certificate’

goal - tended to be more prevalent in effective teams, while learning for a certificate and

entertainment goals were dominant in ineffective teams. The most substantial dissimilarity in

goal preferences pertained to the strong prevalence of entertainment goals in ineffective CL

155

teams. A last interesting difference between the CL teams was that students in ineffective

teams seemed less conscious of their goal preferences than students in effective teams. Both

groups pointed at the context far more often to explain their CL, than to their goal

preferences. Task characteristics, group composition, and teacher behavior were often

mentioned as reasons for effective or ineffective CL.

Results of the longitudinal study showed that the extent that students were taught skills

and knowledge for CL and teachers’ clarity on rules for CL was highly related to the quality

of CL, during all three waves. Effective cooperators had higher scores on all scales at all three

data-waves (teacher’ monitoring, intervention and evaluation behavior, rules and skills for

CL), in particular as regards the extent that they were taught skills and knowledge for CL.

Wave 2 showed the highest scores on the quality of CL and the scores on all teacher related

conditions were highest at that same time. Furthermore, we signaled a tendency that the weak

cooperators perceived a major decrease in teachers’ monitoring and intervention behavior

after the second data-wave. In contrast, the effective cooperators perceived almost no change

in teachers’ monitoring and intervention behavior.

Four student profiles were identified; a school-disaffected, a weak

communication/school bonding, a school-adjusted and a frustrated profile. Students that were

grouped in the school-adjusted profile showed the highest scores on CL. This profile was

characterized by clear goal preferences for social and mastery goals, low scores on superiority

goals, high perceived availability of social support and high scores on school and peer

identification. Students in the school-disaffected profile had the lowest scores on CL. This

profile was characterized by no clear goal preferences, a lack of social resources and

peer/school identification. Students from different backgrounds were disproportionately

distributed. The weak communication profile was characteristic of Caribbean students, their

scores on Dutch language proficiency were extremely low. Students with this profile scored

relatively high (and positive) on school alienation. Superiority was the most valued goal

domain in this profile. The frustrated profile was characterized by clear goals, but

dissatisfaction with the availability of academic and emotional support, especially from

teachers. Furthermore, they were slightly negative on identification with peers and school and

were considering changing schools. Somewhat disturbingly, the highest proportion of all

students fell in this cluster. Interestingly, the school-adjusted profile was characteristic of the

Dutch students.

156

Concluding, students’ goal preferences contribute just weakly to the explanation of the quality

of CL, whereas students’ perceptions on contextual factors were important predictors. Social

support and mastery goals were most vital in predicting the quality of CL. Especially the

extent that students were taught the appropriate knowledge, skills and rules for CL was found

to be a crucial -and lasting- precondition of successful CL. Also the social climate in the

classroom was important for effective CL. Moreover, gender, program type, and ethnocultural

background had no direct effect on the quality of CL.

In our view these results are promising. Whereas students’ goal preferences are

difficult to change and gender and ethnocultural background are stable characteristics,

changes in the classroom context are much easier to bring about. Cooperative learning can

be a means to motivate students and prevent drop-out to a certain extent, when teachers pay

more attention to stimulating students’ reflections on their goal preferences, stimulating

students’ social and mastery goals, stimulating students to reflect on the link between their

personal goals and school goals. Furthermore this research underlined the importance of

teaching students the appropriate skills and knowledge and rules for CL explicitly and paying

attention to stimulating language proficiency. The teacher should better monitor the CL

process and intervene when necessary. Finally, the teacher should create a social climate

where students are invited to provide and receive support.

157

Samenvatting

Het onderzoek dat in dit proefschrift wordt gepresenteerd is een verkenning van de

samenhang tussen de motivatie, gerepresenteerd door persoonlijke doelen, en de kwaliteit van

het samenwerkend leren (KSL) van eerstejaars en tweedejaars leerlingen die het Middelbaar

Beroepsonderwijs volgen. Speciale aandacht wordt geschonken aan contextuele factoren die

de KSL beïnvloeden en aan verschillen tussen leerlingen die samenhangen met hun culturele

achtergrond, geslacht en studierichting. Onder samenwerkend leren worden al die vormen van

leren verstaan, waarbij leerlingen gezamenlijk aan een opdracht werken.

Vier vragen stonden centraal in deze dissertatie. 1) Wat is de relatie tussen

persoonlijke doelen, contextuele factoren en de kwaliteit van het samenwerkend leren? 2) Hoe

kunnen effectieve van niet effectieve samenwerkingsgroepen onderscheiden worden, en wat

onderscheidt deze groepen in termen van hun persoonlijke doelen en hun percepties ten

aanzien van contextuele factoren in de klas? 3) Welke docent-gerelateerde factoren gaan

samen met effectieve samenwerkprocessen en welke gaan samen met ineffectieve

samenwerkprocessen, in de loop van een jaar? 4) Kunnen we profielen onderscheiden van

persoonlijke doelen (bijv. Nederlandse taalvaardigheid) en contextuele factoren (bijv. sociale

bronnen en schoolidentificatie) die verschillen verklaren in de KSL en speelt de culturele

achtergrond van leerlingen een rol bij de verklaring van verschillen tussen deze profielen en

de kwaliteit van het samenwerkend leren?

Alle Regionale Opleiding Centra in Nederland (42) ontvingen een brief met daarin een uitleg

over het onderzoek en een verzoek tot medewerking. Elf scholen, verspreid over Nederland

namen uiteindelijk deel aan het onderzoek. Het design van de studie was longitudinaal met

drie meetmomenten. Het eerste meetmoment vond plaats in de tweede helft van het eerste

leerjaar van de leerlingen, het tweede moment vond plaats in januari en februari van het

tweede jaar en het laatste meetmoment vond plaats aan het eind van het tweede leerjaar

(mei/juni). Data werden verzameld in de periode van december 2001 tot mei 2003. Een

aanzienlijk gedeelte van de leerlingen viel uit; tijdens het eerste meetmoment vulden 1920

leerlingen de vragenlijsten voor dit onderzoek in, tijdens het tweede meetmoment waren dat

er nog 897 en tijdens het derde meetmoment 639.

Het MBO kent vier niveaus, waarbij niveau 1 het laagste niveau is en niveau 4 het

hoogste niveau. Leerlingen van alle niveaus participeerden in het onderzoek. Leerlingen op

het MBO kunnen verschillende studierichtingen volgen, we onderscheiden leerlingen van

158

Zorg en Welzijn, Horeca en Toerisme, Handel en Administratie, en ICT en Techniek

opleidingen.

In dit onderzoek werden percepties (beoordelingen) van leerlingen ten aanzien van de

kwaliteit van het samenwerken (mate van groepscohesie, mate van onderlinge

afhankelijkheid, vaardigheden voor samenwerkend leren en de attitude ten opzichte van het

samenwerkend leren), percepties van leerlingen ten aanzien van contextuele factoren (de

organisatie van het samenwerken, mate van aanleren van regels en kennis voor samenwerken

en docentgedrag), attitudes van leerlingen ten aanzien van samenwerken, en de

communicatievaardigheid van leerlingen gemeten. Daarnaast vulden leerlingen een vragenlijst

in over de mate waarin zij verschillende doelen wilden bereiken en hun percepties ten aanzien

van het sociale klimaat op school.

Met betrekking tot de relatie tussen persoonlijke doelen van leerlingen en de KSL, konden we

concluderen dat sociale hulp doelen de sterkste relatie met de KSL hebben, gevolgd door

leerdoelen en “ergens bij willen horen” doelen. Wat betreft de relatie tussen de percepties van

leerlingen van de leercontext en de KSL concludeerden we dat de sterkst gerelateerde

variabelen zijn de mate waarin leerlingen vaardigheden, kennis en regels voor samenwerkend

leren hebben aangeleerd, en de mate waarin docenten het leerproces in de gaten houden.

Tevens bleken percepties van leerlingen ten aanzien van de mogelijkheid tot het krijgen van

hulp van medeleerlingen en de docent samen te hangen met de KSL. Een combinatie van

sociale hulp doelen, de mate waarin leerlingen vaardigheden en kennis voor het samenwerken

krijgen aangeleerd, de mate waarin de docenten leerprocessen in de gaten houden, en de

mogelijkheid tot het krijgen van academische en emotionele hulp van medeleerlingen,

voorspelden de KSL het best. Een interessante bevinding is dat doelvoorkeuren van leerlingen

maar voor een klein gedeelte bijdragen aan de voorspelling van de KSL. De context bleek

belangrijker te zijn om de KSL te voorspellen. Verder vonden we bevestiging voor een

eerdere onderzoeksbevinding dat vrouwelijke studenten hoger scoren op de kwaliteit van het

samenwerken. Daarnaast scoorden zij hoger op sociale hulpdoelen en leerdoelen dan mannen,

terwijl mannen hoger scoorden op superioriteitsdoelen.

Een belangrijke vraag binnen dit onderzoek was hoe we het samenwerken van

effectieve van niet-effectieve groepen kunnen onderscheiden. Uit een dieptestudie waarbij een

‘stimulated-recall’ methode werd gebruikt, konden we concluderen dat een belangrijk verschil

tussen deze groepen te maken heeft met de ‘ergens bij willen horen’ doelen en sociale

ondersteuningsdoelen. In de groep effectieve samenwerkende leerlingen vonden leerlingen

159

het minder belangrijk om ergens bij te horen en belangrijker om elkaar te helpen en hulp te

ontvangen, terwijl het omgekeerde patroon gevonden werd bij de groep die gekenmerkt werd

door ineffectief samenwerken. Belangrijk op te merken is dat er maar één relatie werd

gevonden met de KSL en deze betrof precies deze ‘ergens bij willen horen’ doelen. Deze

doelen hingen negatief samen met taakgerelateerde betrokkenheid tijdens het samenwerken,

in de groepen die ineffectief samenwerkten. Uit interviews bleek dat leerlingen in effectieve

groepen leerdoelen, sociale doelen en het behalen van een diploma het belangrijkst vonden.

In de ineffectieve groepen waren leren voor een diploma en ‘ontspanning’ doelen het meest

populair. Deze sterke voorkeur voor ontspanningsdoelen, dus het lol hebben op school, is een

zeer opvallend verschil tussen de groepen. Een ander opmerkelijk verschil tussen de groepen

lag in het feit dat leerlingen in ineffectieve teams zich minder bewust waren van hun doelen

dan leerlingen in effectieve teams. Opgemerkt moet worden dat de leerlingen frequenter

refereerden aan de manier waarop docenten het samenwerken organiseerden, als een

belangrijke reden voor effectief of niet effectief samenwerken, dan aan hun doelvoorkeuren.

Beide groepen noemden het type taak, de groepssamenstelling en het gedrag van de docent als

belangrijke redenen voor het al dan niet slagen van het samenwerken. Een ander opvallend

verschil is dat leerlingen in ineffectieve groepen aangaven dat ze gewoon niet goed konden

samenwerken, met andere woorden deze leerlingen misten de kennis en vaardigheden om

goed te kunnen samenwerken.

Uit resultaten van een longitudinale studie naar de relatie tussen docentgerelateerde

condities voor het samenwerken en de KSL, konden we concluderen dat ook op de lange

termijn de mate waarin leerlingen kennis, vaardigheden en regels voor effectief samenwerken

kregen aangeleerd, het sterkst gerelateerd was aan de KSL. Leerlingen die effectief

samenwerkten scoorden te allen tijde hoger op de mate waarin zij deze aangeleerd kregen,

maar ook beoordeelden zij de mate waarin de docent hun leerprocessen in de gaten hield,

ingreep in het leerproces en evalueerde hoger dan leerlingen die niet goed samenwerkten. De

KSL was voor alle leerlingen het hoogst ten tijde van het tweede meetmoment, toen zij ook

alle docentgerelateerde condities het hoogst waardeerden. Verder constateerden we een trend

waarbij de leerlingen die niet goed samenwerkten een sterke daling waarnamen in de mate

waarin docenten hun leerproces in de gaten hielden en de mate waarin zij ingrepen, terwijl

leerlingen die goed samenwerkten deze daling niet rapporteerden.

We konden vier profielen met betrekking tot persoongerelateerde variabelen

(communicatievaardigheid en doelen) en context gerelateerde variabelen (de mate van de

aanwezigheid van hulp van docenten en leerlingen, de mate van identificatie met school en

160

met medeleerlingen en de mate van het voornemen van school te willen veranderen)

onderscheiden. Een eerste profiel werd gekarakteriseerd door een gebrek aan doelvoorkeuren,

een gebrek aan sociale bronnen en een gebrek aan identificatie met school en met

medeleerlingen. Leerlingen met dit profiel hadden de laagste scores op de KSL. Een tweede

(schoolaangepast) profiel werd gekarakteriseerd door duidelijke doelvoorkeuren voor sociale

en leerdoelen, hoge scores op de aanwezigheid van sociale hulp en hoge scores op

identificatie met school en medeleerlingen. Leerlingen met dit profiel scoorden het hoogst op

de KSL. Een derde profiel was het “gefrustreerde profiel” dat kan gekarakteriseerd worden

door duidelijke doelvoorkeuren, maar juist lage scores op de aanwezigheid van sociale

bronnen en lage scores op de mate van peer en schoolidentificatie. Ook overwogen deze

leerlingen van school te veranderen. Alarmerend is dat de meeste leerlingen in dit profiel

vielen. Een vierde profiel kan omschreven worden als een ‘zwak communicatieprofiel”.

Leerlingen met dit profiel scoorden erg laag op Nederlandse taalvaardigheid, hun doelen

waren onduidelijk, hun scores op de aanwezigheid van sociale ondersteuning waren redelijk

hoog, maar scores op hun voornemen om eventueel van school te veranderen waren redelijk

hoog. Tevens was het opmerkelijk dat deze leerlingen vrij hoog scoorden op

superioriteitdoelen. We concludeerden dat etniciteit geen directe invloed heeft op de kwaliteit

van het samenwerkend leren, ook vonden we geen verschillen tussen Nederlandse en

allochtone leerlingen die te maken hadden met hun doelvoorkeuren. Wel vonden we dat de

verdeling van Nederlandse en allochtone leerlingen over de profielen niet gelijk was. De

meeste Caribische leerlingen vielen binnen het zwakke communicatieprofiel en de meeste

Nederlandse leerlingen vielen binnen het schoolaangepaste profiel, het profiel met de hoogste

scores op de KSL.

Concluderend kunnen we constateren dat de doelen van leerlingen slechts voor een klein

gedeelte bijdroegen aan de voorspelling van de KSL, terwijl percepties van contextuele

factoren belangrijke voorspellers bleken te zijn. Sociale ondersteuning- en leerdoelen waren

het meest belangrijk voor de kwaliteit van het samenwerkend leren. Daarnaast bepaalde de

mate waarin leerlingen kennis, vaardigheden en regels voor het samenwerkend leren

aangeleerd kregen voor een groot deel de KSL. Geslacht, studierichting en etnische

achtergrond hadden als zodanig geen directe invloed op de KSL. De bevindingen zijn in onze

optiek hoopgevend omdat doelvoorkeuren moeilijk zijn te beïnvloeden, geslacht en etnische

achtergrond zijn geheel onbeïnvloedbaar. Veranderingen in de context van de klas

daarentegen zijn veel gemakkelijker te bewerkstelligen.

161

Samenwerkend leren kan een manier zijn om de motivatie van leerlingen te verhogen

en drop-out tot op zeker hoogte te voorkomen, mits docenten meer aandacht besteden aan het

stimuleren van het reflecteren op doelen, het stimuleren van sociale en leerdoelen en de link

tussen persoonlijke doelen en schooldoelen verduidelijken. Daarnaast laat dit onderzoek zien

dat het van belang is dat er meer aandacht wordt besteed aan het aanleren van

samenwerkingsvaardigheden, de regels voor het samenwerken en de

communicatievaardigheid van studenten. De docent moet daarbij het leerproces beter in de

gaten houden en ingrijpen wanneer dit nodig is. Voorts zullen docenten een sociaal klimaat in

de klas moeten creëren waarbij leerlingen aangemoedigd worden elkaar hulp te bieden en het

normaal is hulp te ontvangen. Tevens zullen docenten meer aandacht moeten schenken aan

het verbeteren van de communicatievaardigheid van studenten in het MBO.

162

Curriculum vitae

Daphne Minette Hijzen werd geboren op 3 november 1975 te ’s Gravenhage. Het Voortgezet

Wetenschappelijk Onderwijs genoot zij op het Haags Montessori Lyceum te ’s Gravenhage.

Zij behaalde haar diploma in 1995. Vervolgens studeerde zij Pedagogiek aan de Universiteit

Leiden. In 2000 legde zij haar doctoraal examen af met de specialisatie Leerproblemen. Haar

aio-periode liep van februari 2001 tot en met juni 2005. Vanaf oktober 2005 is zij aangesteld

als onderzoeker bij het Nederlands Onderzoekscentrum Arbeidsmarkt & Allochtonen te

Amsterdam. Haar belangrijkste taak is de ontwikkeling van een instrument om

probleemgedrag op het VMBO in kaart te brengen en de ontwikkeling van een

onderwijsmethode om dit gedrag te verminderen.