Achievement goal theory

4Originally, two purposes, or achievement goal orientations, were the major focus of theory and research in achievement goal theory: mastery goals and performance goals (Ames, 1992; E.M. Anderman & Maehr, 1994; Dweck, 1986; Nicholls, 1992). Mastery goals refer to engagement with the purpose of developing competence (Dweck, 1986). When students are oriented to mastery goals, they look for ways to acquire new knowledge and skills. Such students engage in achievement tasks with a purpose of learning and of improvement relative to their past performance and with an orientation towards mastering the task (Ames, 1992). Mastery-oriented students are also less aware of or interested in differences between students and are not concerned with the way they are perceived by others (Ames, 1992; Maehr, 1984; Nicholls, 1984).

5In comparison, performance goals refer to engagement with the purpose of demonstrating competence (Dweck, 1986). When students are oriented to performance goals, they look for ways to earn a positive impression and avoid negative impression of their abilities (Dweck, 1986). Often, demonstration of competence depends on achieving better than others (Nicholls, 1984). Therefore, students who adopt performance goals engage in social comparison, and their self-esteem is contingent on the outcome of such comparisons (Ames, 1992; Dweck, 1999; Nicholls, 1984).

6About a decade ago, Elliot and his colleagues (Elliot, 1997, 1999; Elliot & Church, 1997; Elliot & Harackiewicz, 1996), as well as other researchers (e.g., Middleton & Midgley, 1997; Skaalvik, 1997), distinguished between two types of performance goals: performance-approach goals and performance-avoidance goals. Performance-approach goals refer to the orientation to demonstrate superior competence. Performance-avoidance goals refer to the orientation to avoid demonstrating lack of competence, and especially to avoid getting negative feedback from others (Elliot & Harackiewicz, 1996; Skaalvik, 1997). More recently, researchers also distinguished between mastery-approach and mastery-avoidance goals. Whereas mastery-approach goals refer to a purpose of developing competence, learning, understanding and improving, mastery-avoidance goals refer to a purpose of avoiding stagnation or deterioration of competence: not forgetting, not missing learning opportunities, and not losing skills (Elliot & McGregor, 2001; Pintrich, 2003).

7Research suggests that adoption of mastery-approach goals is related to positive outcomes such as use of deep learning strategies, self-regulated learning, adaptive coping with difficulties and with failure, positive feelings toward learning tasks and toward school, and general well-being (Urdan, 1997). Research findings are also quite clear about the negative outcomes associated with adoption of performance-avoidance goals (Elliot, 1999). These goals were found to be related to self-handicapping strategies, negative feelings toward learning tasks, toward school and toward the self, low grades and giving up when encountering difficulties (Elliot, 1999). Research findings regarding adoption of performance-approach goals are not so consistent. Some studies found associations between this orientation and positive outcomes such as positive feelings, investment, high self-efficacy and high grades (Elliot, 1999). However, other studies did not find such relations, and some studies even found associations with negative outcomes such as use of surface-level learning strategies, anxiety, disruptive behavior and negative affect (for reviews see Harackiewicz, Barron, Pintrich, Elliot & Trash, 2002; Midgley, Kaplan, & Middleton, 2001; Pintrich, 2003; Meece et al. 2006). Presently, relatively little research investigated the distinction between mastery-approach and avoidance goals, and the relevance of this distinction to processes in the educational context, particularly in elementary and secondary schools, is still unclear (Pintrich, 2003).

8Researchers also use mastery and performance goals to characterize motivational emphases in educational environments (Kaplan, Middleton, Urdan, & Midgley, 2002; Meece et al., 2006). Characteristics of the educational environment, such as type of tasks and evaluation practices, can emphasize to students a meaning of schooling as learning and development of skills – a mastery goals structure – or as evaluation of ability in comparison to others – a performance goals structure (Ames, 1992; Meece, 1991; Patrick, Anderman, Ryan, Edelin, & Midgley, 2001). Generally, students’ perceptions of their teachers as emphasizing mastery and performance goals structures were found to be related to their respective adoption of mastery and performance goals. Research also found that the pattern of relations between achievement goal structures and learning processes is similar to the pattern of relations between personal achievement goals orientations and these outcomes. Perceived emphasis on mastery goals was found to be associated with adaptive outcomes such as use of deep learning strategies, positive coping and positive well-being, whereas perceived emphasis on performance goals was found to be associated with maladaptive outcomes such as surface learning strategies, maladaptive coping, disruptive behavior, and negative affect (Kaplan et al., 2002; Meece et al., 2006).

Achievement goals, goal structures and social processes

9Most research in achievement goal theory examined the relations among the achievement goals structure, personal achievement goals, and individual processes related to engagement and experience (Kaplan et al., 2002). Only few studies investigated the relations between these motivational processes and the quality of social interactions among students (L.H. Anderman, 1999; Blumenfeld, 1992; Dweck, 1996; Urdan & Maehr, 1995). Research examined, for example, how achievement goals are related to students’ academic social behavior such as help-seeking (Karabenick, 2003; Ryan, Pintrich, & Midgley, 2001) and to quality of social motivation and social interaction such as intimacy, social responsibility, and conflict resolution strategies (L.H. Anderman, 1999; Darnon et al., 2006; Tossman, Kaplan, & Assor, 2007). In general, findings from this research suggest that mastery-oriented students are more likely to engage in adaptive help-seeking, adopt intimacy goals, and be open to opinions and arguments by others than are performance-oriented students.

10Clearly, the understanding of relations between academic motivation and social processes in educational settings is very important. Schools are social settings and students’ peers provide them important developmental resources (Ryan, 2001). Quality of social relations among peers is fundamental to adaptive development and well-being (Berndt, 1999, 2004). The importance of facilitative social relationships among students is highlighted further by the realization that schools are characterized by social diversity on multiple dimensions, including gender, ethnicity, immigrant status, and social class. This diversity plays an important role in students’ social identities, and can contribute to social tensions or be facilitative of well-being, depending on the quality of social relationships among students (Vivian, Hewstone, & Brown, 1997). In addition, schools are becoming more aware of the importance of high quality social cooperation for learning (Johnson, Johnson, & Smith, 2007; Slavin, 1995). Theory and research in the past few decades have been emphasizing the social nature of learning, and the facilitative cognitive and motivational processes that take place when students cooperate and collaborate on learning tasks (Johnson et al., 2007; Olson, 1980; Rogoff, 1998; Webb & Palincsar, 1996). Such understandings have led to the increasing use of cooperative and collaborative learning methods (Hertz-Lazarowitz, Kirkus, & Miller, 1992; Johnson et al., 2007; O’Donnell, 2006; Slavin, Hurley, & Chamberlain, 2003; Williams, 2007).

Achievement goals and attitudes towards cooperation

11Research suggests that quality cooperation between students depends on environmental, as well as on students’ personal, characteristics (Hertz-Lazarowitz et al., 1992; Johnson et al., 2007; Williams, 2007). For example, skills for working cooperatively are required for cooperation to result with desirable cognitive and affective outcomes (Granott & Gardner, 1994). In addition, however, students need to be motivated to cooperate with others (Hertz-Lazarowitz et al., 1992; Johnson et al., 2007). In a recent commentary on a set of reviews concerning the field of small-group learning, Palincsar (2007) highlighted the issue of students’ attitudes towards cooperation and collaboration as the major challenge facing theory and research.

12Students’ motivation to cooperate with peers on academic work may be related to various personal characteristics. For example, it may be that, similar to findings concerning other types of social-academic behavior such as adaptive help-seeking (Butler & Neuman, 1995; Newman, 1990; Ryan, Patrick, & Shim, 2005), students with high perceived academic efficacy would be more willing to cooperate with peers than students with lower perceived academic efficacy. However, beyond such characteristics, it seems that the purpose of engagement, and the concerns associated with such purpose (cf. Butler, 1998; Karabenick, 2003), would be related to students’ willingness or unwillingness to cooperate with peers on academic work.

13Because achievement goals refer to different goals for engaging in schoolwork (Elliot, 2005), attitudes towards cooperation could depend on the achievement goals that students perceive as emphasized in the educational environment, and on the achievement goals they adopt for engagement in the task (Kaplan, 2004). As can be derived from the definition of the achievement goals and achievement goal structures, in classrooms with mastery goals structure, and when adopting mastery goals, students focus on learning, understanding, and improvement, and would be willing to cooperate with peers when they perceive such interaction as contributing to their learning. These students would be less willing to cooperate with peers who they perceive as potentially interfering with their understanding and improvement. In comparison, in classrooms with a performance goals structure, students would be willing to cooperate with peers when such cooperation provides them an opportunity to demonstrate their ability. When adopting performance-approach goals, students focus on demonstrating high ability relative to others, and would be more willing to cooperate with peers when such cooperation provides them with the opportunity to demonstrate superior ability. When adopting performance-avoidance goals, students are concerned about demonstrating low ability, and would be more willing to cooperate with peers who provide them an opportunity for demonstrating that they are not less able than other students are. Both performance-approach and performance-avoidance oriented students would be less willing to cooperate with peers when they perceive the cooperation as posing a risk of making them look unable.

14However, the specific goals that students adopt for engagement may not be sufficient to explain their attitudes towards peers as cooperation partners. Social preferences and interactions are complex processes that are embedded in social meanings and relationships, and which involve considerations beyond the achievement goal (L.H. Anderman, 1999). For example, socially popular mastery-oriented students’ may have a general preference to cooperate with students who are more knowledgeable than they are, and thus promote their goal of understanding; however, they might be reluctant to cooperate with such students if they belong to unpopular social groups, in which case cooperation may be perceived to be detrimental to their social standing. Similarly, whereas performance-avoidance oriented students may have a general preference to cooperate with either lower ability peers (in order to avoid demonstrating low ability) or with higher ability peers (in order to belong to a high achieving group), they may be unwilling to cooperate with the former because of its implications to their social standing or with the latter because of the risk of being rejected (Levy, Kaplan & Patrick, 2004). These examples highlight the complex processes, some conscious and some less conscious, that operate in social settings and that can be related to preferences for peers. One theoretical perspective that has been very influential in explaining such social attitudes is Social Identity Theory (SIT; Tajfel & Turner, 1979, 1986).

Social Identity Theory

15Social Identity Theory (Tajfel & Turner, 1979, 1986) highlights the role of social groups in people’s self-concept and self-esteem. Social groups provide their members with a social identity, and therefore, have implications for people’s self-esteem: perceiving one’s group in positive terms contributes to the person’s sense of self-worth (Amiot, de la Sablonnière, Terry, & Smith, 2007; Tajfel & Turner, 1986; Vivian et al., 1997). SIT suggests that people’s desire for positive self-regard motivates them to perceive their in-group in positive terms and to use various strategies for achieving this goal. Often, these strategies involve comparison of one’s group with other groups in a way that is favorable to the in-group. Such a comparison may take the form of favoring the in-group, and under certain circumstances also of derogating the out-group (Tajfel & Turner, 1986; Hewstone, Rubin, & Willis, 2002). Research indicates that the mere salience of group membership – even a nominal one – is enough to instigate social identity processes that lead to such intergroup bias (Mackie & Smith, 1998). This bias manifests in a general preference towards others who are perceived to be similar on important social criteria over those who are perceived to be different (Tajfel & Turner, 1986).

16SIT also suggests that when self-worth is highlighted, people are motivated to employ social identity processes to enhance or protect their self-worth (Tajfel & Turner, 1986). Therefore, self-worth concerns – stemming either from personal dispositions or from contextual characteristics – are likely to manifest in intergroup bias such as in-group favoritism (Hogg & Abrams, 1990). Specifically, when people become concerned with their self-worth, they would categorize others into in-group (i.e., peers who are similar) and out-groups (i.e., peers who are different), compare the in-group and out-groups on subjectively important characteristics in ways that would present the in-group positively, and very often, prefer in-group members over out-group members (Taylor & Moghaddam, 1994). Importantly, people would classify others into in-group and out-groups, and compare these groups in ways that would favor their in-group, on the basis of a combination of subjective and socially salient (e.g., ethnicity, gender, popularity) criteria – those that would contribute most to the protection or enhancement of their self-esteem (Taylor & Moghaddam, 1994). Moreover, SIT suggests that self-worth oriented people would be most likely to engage in such intergroup bias when they perceive it to have a real impact on their self-worth – such as in socially meaningful situations (Hewstone et al., 2002).

Achievement goals in school and intergroup biases

17There is little question that schools constitute meaningful social situations for students. Whether schools are ethnically diverse or not, students tend to arrange themselves into social groups (Brown, 1989; Kinderman, 1993). Indeed, the importance of social groups in late childhood and early adolescence (Kegan, 1982) suggests that much, if not most, of social interactions among students is affected by their group membership, and therefore by social identity processes. Clearly, these processes are highlighted even more when students are asked to interact with each other, as happens in cooperative tasks.

18Recently, Kaplan (2004) proposed that different achievement goal orientations may be differentially associated with salience of social group membership and social discrimination towards people from the in-groups and the out-groups. Kaplan suggested that the self-worth concerns associated with performance goals may trigger social identity processes and lead to intergroup bias. More specifically, when oriented to performance-approach goals, and when perceiving the environment as emphasizing performance goals, people’s concerns with social comparison and self-worth may facilitate categorization of peers into in-group and out-groups on the basis of subjective similarity criteria, and promote the motivation to prefer in-group members over out-group members as means for protecting or enhancing self-esteem (Kaplan, 2004). In contrast, when oriented to mastery goals, and when perceiving the environment as emphasizing mastery goals, students would be less or not concerned with self-worth or social comparison. The belief that success is not dependent on others’ performance or perceptions but rather on learning and challenge focuses mastery-oriented students on the academic task at hand and away from self-worth. Thus, mastery goals would not be associated with intergroup biased and discrimination (Kaplan, 2004).

19A recent study provided support for these hypotheses. In this study, Levy, Kaplan and Patrick (2004) used interviews to test the relations of 5^th grade students’ personal achievement goals, social identity processes, and biased preferences towards peers as cooperation partners. The findings indicated that, generally, mastery-oriented students had little concern with group-membership and social status, and tended to prefer to cooperate with peers who contributed to their learning, regardless of their group membership or social status. In contrast, performance-approach and performance-avoidance oriented students were highly concerned with group membership and social status, and tended to evaluate peers as cooperation partners along the implications of such cooperation for their own social status in the class. Interestingly, to these students, the implications for their social status seemed more important in evaluating peers as cooperation partners than were contributions to demonstration of ability or avoiding demonstrating low ability (Levy et al., 2004). This study supported the hypotheses relating students’ personal goal orientations and social identity processes. However, it did not address the possibility that contextual characteristics that highlight different achievement goals – for example, teachers’ practices that emphasize mastery or performance goals – would also elicit biased preferences for peers as cooperation partners.

The present study

20The present study set to investigate the role of perceived achievement goal structures in school in intergroup biases, and more particularly, in students’ biased preferences for peers as cooperation partners. The study focuses specifically on in-group favoritism – students’ preferences to cooperate with peers who they perceive to be similar to them – and on out-group rejection – students’ preferences not to cooperate with peers who they perceive to be different from them. Support for the role of achievement goal structure in students’ biased preferences for peers as cooperation partners may highlight the importance of the academic context in students’ social attitudes and relations. Such an understanding may enhance the tools available to educators and other practitioners in achievement settings who attempt to improve cooperation among students from different social groups.

21Supporting the unique role of achievement goal structure in students’ biased preferences for peers required the ability to distinguish this effect from the role of students’ personal dispositions towards certain achievement goals in these biases. Investigation of this unique effect can be conducted through simultaneous multiple regression, in which both personal achievement goals and achievement goal structure are independent variables. However, whereas students’ personal achievement goals reflect their dispositions to adopt certain achievement goals, they are also quite strongly related to the concurrent achievement goal structure (e.g., Kaplan & Maehr, 1999). Therefore, the inclusion of personal achievement goals that were assessed concurrently with the achievement goal structure in the regression may mask relevant variance shared by the achievement goal structure and the biased preferences for peers. In order to confront this challenge, the present study used a longitudinal design which included a transition between educational environments. Employing students’ personal achievement goals from a prior learning environment provided an opportunity to assess personal dispositions towards achievement goals that have little relation to the achievement goal structure in the learning environment after the transition. Moreover, this longitudinal design allowed controlling for students’ prior dispositions towards biased preferences for peers, as well as their perceptions of the achievement goal structure from the previous school, and thus provided a more stringent test of the relations between achievement goal structure and students’ biased preferences for peers as cooperation partners in the new learning environment. The current study also included another important motivational characteristic that is commonly associated with students’ achievement goals and outcome processes in schools – academic efficacy. The inclusion of efficacy in this study was meant to increase confidence that the variance shared by academic goal structure and biased preferences for peers is not due to students’ sense of efficacy in school.

Hypotheses

22A perceived performance goals structure is hypothesized to be positively related to students’ biased preferences towards peers as cooperation partners, over and beyond students’ personal achievement goals and academic efficacy.

23A perceived mastery goals structure is hypothesized to be unrelated to students’ biased preferences towards peers as cooperation partners.

Participants

24The study followed 100 Israeli-Jewish students (46% boys, 54% girls) as they made the transition from sixth grade in elementary school to seventh grade in a large Junior High school. These 100 students constituted 87% of all the students in their cohort at sixth grade, and 37% of all the students in their cohort at seventh grade. The elementary school was the only school in the students’ community. The community was of upper-middle class and had a small number of immigrants. The school is highly regarded in the community and beyond, and is considered to prepare its students well, as is indicated by high average rates on standardized tests. The terms of the study as approved by the Israeli Ministry of Education prevented collection of such achievement data for individual students. The Junior High school was the regional school in the area and as students made the transition to this school they joined other students from several communities. This regional Junior High school is similar in characteristics such as population, size, accreditation of teachers, and success rate of students on standardized tests, to other regional Junior High school that serve semi-agricultural communities in central Israel.

Instruments

25Variables in this study were assessed with surveys. The response scale for all the items in the survey was a 5 point Likert Scale ranging from 1-”not at all true of me” to 5-”very true of me.”

26Perceived achievement goal structures: The scales assessing perceived mastery and performance goals structures (see Appendix) were from the Patterns of Adaptive Learning Survey (PALS; Midgley et al., 2000). Items in the performance goals structure scale focused on teacher’s emphasis on demonstrating high ability. A sample item assessing perceived mastery goals structure is “My teacher wants us to understand our work, not just memorize it.” A sample item assessing perceived performance goals structure is “My teacher lets us know which students get the highest scores on a test.”

27Personal achievement goals and academic efficacy: The scales assessing mastery, performance-approach, performance-avoidance goals and academic efficacy (see Appendix) were also from the PALS (Midgley et al., 2000). A sample item in the scale assessing mastery goals is “An important reason why I do my work in class is because I like to learn new things.” A sample item in the scale assessing performance-approach goals is “I would feel successful in class if I did better than most of the other students.” A sample item in the scale assessing performance-avoidance goals is “One of my main goals is to avoid looking like I can’t do my work.” A sample item in the scale assessing academic efficacy is “Even if the work is hard, I can learn it”.

28Biased preferences for cooperation partners: The measures of biased preference for peers as cooperation partners were constructed for this study. In line with the intergoup relations literature (e.g., Tajfel & Turner, 1986), items were constructed to tap students’ personal subjective construction of social dimensions defining peer groups. For this purpose, rather than indicating social criteria for group membership, items described peers as “similar” to or as “different” from the responding student, leaving the meaning of similarity and difference to respondents’ interpretation. In addition, the emphasis on preferences for peers as partners in cooperative learning highlighted the issue of tolerance and intolerance of different opinions and ways of thinking. Therefore, items were also constructed to emphasize similarity and difference in thinking. Five items were constructed for this scale, two assessing preference for peers perceived to be similar as cooperation partners and three assessing rejection of peers perceived to be different as cooperation partners. The items assessing preference for similar peers were: “If the teacher would pair us up, I want to work only with someone who thinks like me” and “If the teacher would pair us up, I want to work only with someone who is similar to me in many ways.” The items assessing rejection of peers perceived to be different were: “If the teacher would pair us up, I wouldn’t want to work with someone who is different from me,” “If the teacher would pair us up, I wouldn’t want to work with someone who thinks differently from me,” and “When working in groups, I prefer not to work with students who are different from me.”

29Preference to cooperate with peers with different levels of ability: In order to validate the assumption that the criteria that students used to define similarity and difference when responding to items assessing biased preference to cooperate with peers were not only peers’ academic ability, we constructed items assessing preference to cooperate with peers who have different levels of ability. Two items assessed preference to cooperate with peers with higher academic ability (“When working in groups, I prefer to work with students who are better than I am in school,” “If the teacher would pair us up, I prefer to work only with someone who is better than I am in school”), two items assessed preference to cooperate with peers with similar level of academic ability (When working in groups, I prefer to work with students who have the same level of ability as I do in school,” “If the teacher would pair us up, I prefer to work only with someone who has the same level of ability as I do in school”), and two items assessed preference to cooperate with peers with lower level of academic ability (When working in groups, I prefer to work with students who are less good than I am in school,” “If the teacher would pair us up, I prefer to work only with someone who is less good than I am in school”).

Procedure

30All participating students received permission to participate from their parents or guardians. Surveys were administered to students in their classroom during regular school hours once in sixth grade and once in seventh grade. In both years, administration of surveys was conducted in late winter (March), after students had the opportunity to become familiar with the teachers and peers in their learning environment. Teachers were present in the classroom but were asked to avoid interacting with the students and were not involved in survey administration. Pairs of trained research assistants read instructions and survey items aloud to students. Students were told that the survey asked about how they felt about school and school work, and were asked to think about their home-room teacher and her lessons when responding to the items. Students were told that the survey was not a test and that there were no right or wrong answers. Students were asked to write their names on the survey, but were assured that the purpose was for matching surveys from different administrations and that the information in the survey would be kept completely confidential. The survey was composed of two parts that were administered approximately a week apart from each other. The scales assessing personal achievement goals and academic self-efficacy were administered in the first part. Perceived teacher’s emphasis on the different achievement goals and biased preferences for cooperation partners were administered in the second part. Administration of each part of the survey took approximately 40 minutes.

Psychometric properties of variables in the study

32The items employed in the current study appear in the Appendix. Scales assessing achievement goal structures, personal achievement goals, and perceived academic efficacy were adopted from established scales. In order to increase the confidence that the variance investigated was of items that students interpreted in both waves as assessing the target variables, we used exploratory factor analyses, with maximum likelihood and oblique rotation. These analyses indicated that only three items in each of the scales assessing performance goals structure, the three personal achievement goals, and academic efficacy were equivalent in loadings in both waves [1]. Since reliability of each of these scales with three items was satisfactory, variables were constructed with these items in order to maintain equivalence in their meaning across the data waves. Along the purpose of the study as presented above, and the study’s hypotheses, the analyses employ the achievement goal structures from both waves of data, the personal achievement goals variables from the data collected in 6^th grade, and the efficacy variable from the data collected in 7^th grade. Table 1 presents the descriptive statistics of the variables in the study.

33Items assessing the biased preferences for peers as cooperation partners were also subjected to exploratory factor analysis, with maximum likelihood and oblique rotation. In both waves, items assessing preference for peers perceived to be similar loaded together on one factor with items assessing rejection of peers perceived to be different. In order to validate this scale, we conducted a Smallest Space Analysis (SSA) with the items in the scale and the items assessing students’ preferences to cooperate with peers of different levels of ability. SSA is a non-metric multi-dimensional scaling method, which employs the rank-ordered correlations between each pair of items to create a visual representation of the relations among all the items in the analysis (Shye, Elizur, & Hoffman, 1994). Visual proximity between items indicates their correlation relative to the correlations of all other items in the analysis. The geometric representation provided allows the examination of the meaning of items and groups of items relative to the other items in the analysis (Guttman & Greenbaum, 1998). We expected items assessing biased preference towards similar peers (and rejection of different peers) to appear closer to items assessing preference to cooperate with peers with similar ability than to items assessing preference to cooperate with peers with higher or lower ability. We still expected items assessing biased preferences to be somewhat distinct, or perhaps to include but not to be restricted to, items assessing preference for peers with similar ability.

34Figure 1 presents the SSA results with data collected when students were in 6^th grade. Figure 2 presents the SSA results with data collected when students were in 7^th grade. In both analyses, clear distinctions can be seen between items assessing preference for peers with lower ability, items assessing preference for peers with similar ability, and items assessing preference for peers with higher ability. Moreover, the five items assessing biased preferences for similar peers and rejection of different peers indeed appear closer to the items assessing preference for peers with similar ability. However, these items do not overlap, indicating similar but not identical meanings. We used the two items assessing preference for peers with similar ability to construct a variable that will be used to control for this preference when testing the study’s hypotheses.

Figure 1

SSA results of items assessing preferences for peers with different levels of ability and biased preferences for similar and different peers in sixth grade.

Figure 2

SSA results of items assessing preferences for peers with different levels of ability and biased preferences for similar and different peers in seventh grade.

35Table 1 presents the descriptive statistics of the variables in the study. The distribution of all the variables in the study approximated normality. Also, all the variables manifested satisfactory reliability, except for performance-avoidance goals, which was somewhat low. The overall pattern of the variables’ means was similar to other studies among elementary and early Junior-High schools samples. The mean of mastery goals structure was higher than the mean of performance goals structure both in 6^th and in 7^th grade. Similarly, the means of personal mastery goals and performance-approach goals were higher than the mean of performance-avoidance goals. The mean of perceived academic efficacy in 7^th grade was high. The means of biased preferences for cooperating with similar and not different peers was moderate in both years.

36Table 2 presents the correlations among the variables in the study. Mastery goals structure in both waves was weakly associated with performance goals structure in 6^th grade (rs=-.20 & -.23) but not related to performance goals structure in 7^th grade. Personal mastery goals were not related to personal performance-approach or personal performance-avoidance goals. Mastery goals structure in both waves was positively related to personal mastery goals in 6^th grade (rs=.46 & .31), and performance goals structure in both waves was positively related to personal performance-approach goals in 6^th grade (rs=.32 & .19), but, interestingly, not to personal performance-avoidance goals in 6^th grade. Academic efficacy in 7^th grade was weakly and positively related to personal mastery goals in 6^th grade (r=.27) but not to personal performance goals or to the achievement goal structure variables. Biased preferences for peers in 6^th grade were negatively related to mastery goals structure in both waves (rs=-.33 & -.32). These biased preferences in 7^th grade were associated negatively with mastery goals structure in 7^th grade (r=-.37). Personal mastery goals in 6^th grade were negatively related to biased preferences for peers in 6^th grade (r=-.36). Biased preferences for peers in each wave were positively related with the corresponding performance goals structure (rs=.26 & .41, respectively) and with performance-approach goals in 6^th grade (r=.32). Preferences to cooperate with peers with similar ability in 7^th grade were strongly and positively related to biased preferences for peers in both waves (rs=.57 & .70). Preferences to collaborate with peers with similar ability in 7^th grade were negatively associated with mastery goals structure in 6^th grade (r=-.32) and positively with performance goals structure in 7^th grade (r=.21).

Table 1

Variable name # of Reliability Mean Standard Skewness items deviation Mastery goals 6th grade 3 .79 3.46 1.01 -.38 Performance-app. goals 6th grade 3 .81 3.34 1.09 -.22 Performance-avd. goals 6th grade 3 .57 2.87 .85 .14 Mastery goal str. at 6th grade 5 .79 3.73 .82 -.79 Mastery goal str. at 7th grade 5 .82 3.53 .85 -.23 Performance goal str. at 6th grade 3 .67 2.78 1.09 .66 Performance goal str. at 7th grade 3 .75 2.77 1.04 .11 Biased Preferences 6th grade 5 .85 2.41 .91 -.02 Biased Preferences 7th grade 5 .88 2.54 .93 .10 Academic efficacy 7th grade 3 .68 4.27 .68 -.58 Preference for peers with similar ability 2 .85 2.72 1.16 .29 7th grade Note. N=100. app.=approach. avd.=avoidance

Descriptive statistics of variables in the study.

Table 2

2 3 4 5 6 7 8 9 10 11 1. Mastery goals 6th gr. .06 .14 .46 .31 -.04 .23 -.36 -.19 .27 -.27 2. Perf-app. goals 6th gr. 1.00 .42 .01 -.15 .32 .19 .32 .13 .10 -.08 3. Perf-avd. goals 6th gr. 1.00 .08 -.14 .20 .05 .22 -.07 .08 -.02 4. Mastery struc. 6th gr. 1.00 .32 -.20 .08 -.33 -.12 -.05 -.32 5. Mastery. struc. 7th gr. 1.00 -.23 .11 -.32 -.37 .05 -.20 6. Perf. struc. 6th gr. 1.00 .38 .26 .23 -.14 .09 7. Perf. struc. 7th gr. 1.00 .16 .41 -.10 .21 8. Biased Preferences 1.00 .35 -.08 .57 6th gr. 9. Biased Preferences 1.00 -.17 .70 7th gr. 10. Efficacy 7th gr. 1.00 -.14 11. Prefer similar ability peers 1.00 7th gr. Note. N=100. gr.=grade. Perf.-app.=Performance-approach. Perf.-avd.=Performance-avoidance. struc.=structure. Coefficients equal to or greater than .24 are significant at least at p<.05.

Pearson correlations among variables in the study

Multiple regressions

37Due to the relatively small sample size, we decided to conduct analyses on mastery goal structure and performance goal structure separately. Therefore, the study’s hypotheses were tested with two multiple regression analyses. One analysis tested the relations between mastery goals structure and biased preferences for similar and not different peers as cooperation partners. The second analysis tested the relations between performance goals structure and these biased preferences for peers. In each analysis, the first predictor was biased preferences for peers from 6^th grade. This was meant to control for any biased preferences that students may have come with as they moved from elementary school to Junior High school. In a second step, the respective achievement goal structure from 6^th grade was entered into the regression as well as the respective personal achievement goals. This was meant to control for students’ predisposition to perceive an emphasis on the goal structure, and allowed to estimate the relations of the difference in achievement goals structure between 6^th and 7^th grade on the dependent variable. The inclusion of the personal achievement goals was meant to control for students’ predispositions that may be related to their biased preferences for peers. In the third step, students’ academic efficacy in 7^th grade was entered into the regression with the attempt to control for variance that this characteristic may have on students’ biased preferences for peers as cooperation partners. Finally, in the fourth step, the focal variable of the achievement goal structure in 7^th grade was entered into the regression.

38Tables 3 and 4 present the results of the regression analyses testing for the relations of mastery goals structure and performance goals structure on biased preferences for peers respectively. Mastery goals structure in 7^th grade was significantly and negatively related to biased preferences for peers as cooperation partners in 7^th grade, over and beyond all other variables in the regression. Mastery goals structure in 7^th grade added 7% to the explained variance in biased preferences above the variance explained by the other variables. Performance goals structure in 7^th grade was significantly and positively related to biased preferences for peers as cooperation partners in 7^th grade, over and beyond all the other variables in the regression. Performance goals structure in 7^th grade added 15% to the explained variance in biased preferences for peers.

Table 3

Dependent variable: Biased preferences for peers at 7th grade Step 1 Step 2 Step 3 Step 4 Biased Preference 6th grade .33b .31a .31a .22 Mastery goals structure 6th grade .07 .01 .07 Personal mastery goals 6th grade -.12 -.04 .00 Academic efficacy 7th grade -.17 -.17 Mastery goals structure 7th grade -.30a R2 .11 .12 .15 .22 ?R2 .01 .03 .07 Note: N=77, ap<0.05, bp<0.01

Results of multiple regression analysis testing the relations of mastery goals structure and biased preferences for peers as cooperation partners.

Table 4

Dependent variable: Biased preferences for peers at 7th grade Step 1 Step 2 Step 3 Step 4 Biased Preference 6th grade .33b .30a .30a .29a Performance goals structure 6th grade .20 .19 .02 Personal performance-app. goals 6th grade .00 .01 -.02 Personal performance-avd. Goals 6th grade -.19 -.16 -.14 Academic efficacy 7th grade -.16 -.14 Performance goals structure 7th grade .43b R2 .11 .17 .19 .34 ?R2 .06 .02 .15 Note: N=77, ap<0.05, bp<0.01

Results of multiple regression analysis testing the relations of performance goals structure and biased preferences for peers as cooperation partners.

39Whereas the SSA analysis indicated that while Biased Preferences for similar peers are related, but not identical, to preferences for peers with similar ability, we conducted two additional regressions in order to test whether, indeed, achievement goal structures are related to preferences for similar peers along students’ subjective criterion of similarity and difference, beyond that which is based on peers’ perceived ability. The regressions reported above were run with an additional step: students’ preferences for peers with similar level of ability were entered before entering the focal variable of achievement goal structure. As expected, students’ preference for peers with similar ability added a significant amount to the explained variance in biased preferences for peers (38% and 45% in the mastery goals structure and performance goals structure analyses respectively). However, the inclusion of this variable in the regression did not change at all the magnitude of the negative relations of mastery goals structure in 7^th grade and biased preferences for peers. Whereas the inclusion of preferences for peers with similar ability did reduce the magnitude of the relations of performance goals structure in 7^th grade and biased preferences for peers, the relation was still significant explaining additional 4% of the variance in biased preferences for peers as cooperation partners. This analysis, however, also manifested a suppression effect, indicated by a significant negative relation between personal performance-approach goals in 6^th grade and biased preferences for peers in 7^th grade.

Limitations of the study

44Whereas the study employed longitudinal data, and controlled for personal characteristics, interpretation of the findings is compromised by the small number of participants. The sample size did not allow the investigation of a comprehensive model that takes into account perceiving multiple goal structures in the educational environment (cf. Harackiewicz et al., 2002). Also, the focus on learning with the home-room teacher, rather than on a specific subject-domain, qualifies the findings in light of the suggestion that students may hold different achievement goals in different subject domains. Over the transition to Junior High, the role of the home-room teacher changes as are the subjects that she or he teaches. This is likely to have an effect on students’ motivation and social relationships. Additionally, the present study did not assess the mediating processes between achievement goal structures and students’ biased preferences for peers as cooperation partners. Future research should address the question concerning the effect of multiple achievement contexts, with different emphases, on the social relationships of students. Future research should also investigate the specific processes that are elicited when students perceive their learning environment as emphasizing different achievement goals and which affect their social attitudes and preferences for peers from different social groups (see Kaplan, 2004).

Conclusion

45The findings of this study support the relationship between the educational environment’s achievement goal structure and students’ biased preferences for peers as cooperation partners. Specifically, the findings suggest that when teachers emphasize demonstrating superior ability and besting others, students’ preferences for cooperation partners may be biased by criteria of peers’ similarity and difference. In contrast, when teachers emphasize learning, understanding, and personal improvement, students would be more likely to agree, and perhaps even prefer, to cooperate with peers from different social groups. These findings join other research that highlights the important role of the academic culture of schools in students’ social life (L.H. Anderman, 1999; Blumenfeld, 1992). They also highlight the possible contribution of school cultures that emphasize competition and social comparison to negative social relationships among students, and the promise that emphasizing meaningful learning and understanding, personal improvement, and collaboration (cf. Ames, 1992) may have in alleviating such negative social relationships.