Contemporary Educational Psychology 32 (2007) 231–252

www.elsevier.com/locate/cedpsych

Academic goal orientations, multiple goal proWles, and friendship intimacy among early adolescents

Inbal Levy-Tossman ¤, Avi Kaplan, Avi Assor

Department of Education, Ben-Gurion University of the Negev, Beer Sheva 84105, Israel

Available online 7 September 2006

Abstract

This study investigated the relations between early adolescents’ academic motivational orienta-tions and an aspect of quality of friendship: intimacy. Two-hundred and three Jewish-Israeli seventhgrade students responded to surveys asking them about their academic achievement goals and aboutcharacteristics of their friendships. Variable-centered regression analyses suggested that masterygoals were positively associated with mutual sharing of diYculties, trust, and adaptive social prob-lem-solving between friends, whereas performance-approach goals were negatively associated withintimacy friendship. Moreover, both performance-approach and performance-avoidance goals wereassociated with mistrust, inconsideration, and tension between friends. A person-centered analysis,employing cluster analysis, suggested that proWles with a higher level of mastery goals relative toboth types of performance goals were associated with less mistrust among friends in comparison withproWles with a higher level of performance goals relative to mastery goals. The Wndings point to theconnection between academic motivation and social relationships in school.© 2006 Elsevier Inc. All rights reserved.

Keywords: Motivation; Intimacy; Friendship; Adolescence

1. Introduction

In the past two decades, achievement goal theory has become one of the most promi-nent frameworks for the conceptualization and investigation of students’ experiences inschool (Elliot, 1999; Pintrich, 1994). Most research in this framework, however, has

* Corresponding author.E-mail addresses: inbal-tossman@bezeqint.net (I. Levy-Tossman), akaplan@bgu.ac.il (A. Kaplan).

0361-476X/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.cedpsych.2006.06.001

232 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

focused on academic engagement such as cognitive and meta-cognitive strategies, grades,and school-related behavior and aVect (see reviews in Ames, 1992; Anderman & Maehr,1994; Brophy, 2004; Elliot, 1999; Kaplan, Middleton, Urdan, & Midgley, 2002; Pintrich,2000a; Urdan, 1997a). Only in the past few years have achievement goal theorists begun toexplore more broadly the relations between students’ academic motivational orientationsand other aspects of their school experiences such as social relationships (e.g., Anderman,1999; Kaplan, 2004; Patrick, Anderman, & Ryan, 2002) and overall well-being (e.g., Kap-lan & Maehr, 1999; Dykman, 1998). The Wndings from these investigations provide strongsupport to the notion that the academic and social domains are not distinct in students’school experiences (cf. Birch & Ladd, 1996; Blumenfeld, 1992; Urdan & Maehr, 1995; Van-del & Hembree, 1994). Moreover, this research suggests that diVerent academic motiva-tional orientations may be accompanied by characteristic qualities or styles of socialinteraction (e.g., Anderman & Anderman, 1999; Ryan, Hicks, & Midgley, 1997; Levy, Kap-lan, & Patrick, 2004).

The present study adds to this emerging literature by using both variable- and person-centered approaches to investigate the relations between academic goals orientations, stu-dents’ multiple goal proWles, and one important aspect of students’ social relationshipswith peers in school: Quality of friendship.

2. Achievement goal theory

Achievement goal theory concerns the purposes or goal orientations that students con-strue for action. These purposes represent comprehensive meaning systems of situations orcontexts that have cognitive, emotional and behavioral consequences (Ames, 1992; Dweck& Leggett, 1988; Kaplan & Maehr, 2002). Current research emphasizes the prevalence ofthree types of goal orientations among students: mastery goals, performance-approachgoals, and performance-avoidance goals (Elliot & Church, 1997; Middleton & Midgley,1997; Skaalvik, 1997).1 Mastery goals refer to an orientation towards developing compe-tence, and are often associated with a focus on understanding, developing skills, and mas-tering the task. Mastery-oriented students employ task-mastery and intrapersonal criteriaof improvement for evaluating progress and success and are not concerned with others’perceptions of their ability. In contrast, performance-approach goals refer to an orienta-tion towards demonstrating high ability. Performance-approach oriented students are con-cerned with the ability impressions that they make on others, and employ comparativestandards to evaluate their success in demonstrating superior ability. Performance-avoid-ance goals refer to an orientation towards avoiding demonstrating low ability. Perfor-mance-avoidance oriented students are also concerned with the ability impressions thatthey make on others; however, these students are oriented towards the possibility offailure—demonstrating low ability—and attempt to avoid it. These students also use

1 More recently, some achievement goal theorists suggested that a fourth achievement goal orientation, mas-tery-avoidance, conceptualized as an orientation to avoid deterioration of skills or missing opportunities forlearning, should be added to the list of the three goal orientations mentioned (Elliot, 1999; Pintrich, 2000b). Thisgoal orientation was suggested on a conceptual (i.e., theoretical symmetry) rather than an empirical basis, andthere is a wide perceptions that currently “there is much research to be done before mastery-avoid goals can beaccepted as valid or useful in goal theory” (Pintrich, 2003, p. 676). Therefore, the present study does not addressthis motivational orientation.

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 233

comparative standards to evaluate their success of not demonstrating low relative ability(Elliot & Church, 1997).

A large body of research over the last two decades established the positive relationsbetween mastery goals and adaptive academic outcomes such as deep learning strategies,self-regulated learning, adaptive coping with diYculties and with failure, and positive feel-ings toward learning tasks and toward school (Urdan, 1997a). Research Wndings are alsoquite clear about the negative outcomes associated with performance-avoidance goals(Elliot, 1999). These goals were found to be related to self-handicapping strategies, nega-tive feelings toward learning tasks, toward school and toward the self, low grades, and giv-ing up when meeting diYculties (Elliot, 1999). Findings regarding performance-approachgoals are not so consistent. Some studies found associations between this orientation andpositive outcomes such as positive feelings, investment, high self-eYcacy and high grades(Elliot, 1999). Other studies did not Wnd such relations (e.g., Middleton & Midgley, 1997),and some studies even found associations between performance-approach goals and nega-tive outcomes such as surface-level learning strategies, anxiety, unwillingness to cooperate,disruptive behavior and negative aVect (e.g., Kaplan, Gheen, & Midgley, 2002; Kaplan &Midgley, 1997; Levy et al., 2004; Skaalvik, 1997; for a review see Midgley, Kaplan, & Mid-dleton, 2001).

Originally, achievement goal theory treated the diVerent types of achievement goals sep-arately (e.g., Dweck, 1986; Nicholls, 1984). However, theorists soon acknowledged thatwhen considered as dispositions, mastery and performance goal orientations may be heldsimultaneously (Nicholls, Cheung, Lauer, & Patashnick, 1989). Since then, research indi-cated that students may hold both mastery and performance goals to varying degrees.Research conducted in diVerent countries and with samples of diVerent ages that examinedthe pattern of outcomes associated with adoption of multiple goals—mostly mastery andperformance-approach goals—suggests that proWles that include mastery goals (i.e., adominant mastery goals orientation or a combination of mastery and performance-approach goals) are associated with more adaptive outcomes than proWles that do notinclude mastery goals. In addition, holding a dominant performance-approach goal orien-tation is associated with more adaptive outcomes than not holding any achievement goals(see Harackiewicz, Barron, Pintrich, Elliot, & Thrash, 2002; Kaplan & Maehr, 2002 forrecent reviews).

3. Quality social relationship during early adolescence

Social relationships among students constitute a highly important component of stu-dents’ experiences in school. Research Wndings emphasize the importance of quality friend-ship for students’ well-being (Berndt, 1999). Friendships that are characterized by highlevels of caring and validation, support and help, ease of conXict resolution, mutual inti-mate exchange, and low levels of conXict and betrayal have been found to be an importantpredictor of students’ academic functioning (Anderman & Anderman, 1999; Ryan et al.,1997), as well as general adjustment and well-being (e.g., Parker & Asher, 1993; Shulman,Laursen, Kalman, & Karpovsky, 1997).

The importance of high quality friendship at school increases as students enter earlyadolescence. Early adolescence is a period where social relationships are particularlyimportant in deWning people’s sense of self (Kegan, 1982). As students begin to orient awayfrom parents and towards peers for support in personal issues (Youniss & Smollar, 1985),

234 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

the importance of interaction with friends as determinant of their self-deWnition, self-esteem, and coping abilities is greatly enhanced (Berndt, 1999).

One of the most important characteristics of quality friendship that develops duringadolescence is intimacy (Sullivan, 1953). Intimate friendship is distinguished from othertypes of friendship by the level and quality of mutual sharing of emotions and thoughtsbetween peers. Intimate friendship involves mutual openness, trust, honesty, communi-cation, self-disclosure, caring and respect (Paul & White, 1990; Prager, 1995; Sullivan,1953). The mutual trust in intimate friendship allows friends to expose vulnerabilities,fears, wishes, and hopes without risk of betrayal (Parker & Asher, 1993; Prager, 1995).Such interactions promote perspective taking, as friends try to understand each other.The honest exchange that involves evaluations of attributes and abilities is also very sig-niWcant in the construction of realistic self-perceptions and in identity formation (Savin-Williams & Brendt, 1990; Zarabatany, Conley, & Pepper, 2004). In contrast, lack of inti-mate friendship is related to negative outcomes such as sense of loneliness, aggression,depression, school dropout, crime, and substance abuse (Hussong, 2000; Parker & Asher,1987).

The factors that aVect students’ quality of friendship may vary widely and have yet to becomprehensively spelled-out by research (Berndt, 2004). In recent years, studies have inves-tigated such factors as gender, age, family structures, relationship with parents, and a vari-ety of personality attributes (e.g., Noack, Kretterk, & Walper, 2001; Weimer, Kerns, &Oldenburg, 2004). The present study aims to explore the relations between students’ aca-demic motivational orientations and the quality of their friendship relationships. Tradi-tionally, academic and social orientations in school were perceived to be in conXict. A highconcern with social relationships was thought to interfere with academic engagement.However, more recent research has shown that this assumption is unfounded, and that inmany cases, social relationships among students are actually conducive to academicengagement (Anderman, 1999; Berndt, 1999; Ryan, 2000).

Studies show, for example, that conversation among friends about school related issuesresult with shifts in students’ motivation towards the friends’ attitudes (e.g., Berndt, Lay-chak, & Park, 1990). Thus, when friends hold a positive attitude towards school, socialrelations may be associated with change in motivation towards more engagement in school(e.g., Kindermann, 1993). Urdan (1997b) found that adoption of both mastery goals andperformance-approach goals was related to having best friends who value school andlearning. However, for performance-approach goals, this Wnding was limited to highachieving students. In addition, Urdan found that having friends who de-value school wasrelated to students’ adoption of extrinsic goal orientations (being primarily concerned withgrades) and to a work-avoidance orientation.

4. Achievement goals and intimate relationships

The little research that examined the relations between achievement goals and socialprocesses among students suggests that these processes are indeed related (Anderman,1999). Overall, the Wndings suggest that mastery goals are associated with friends’ posi-tive orientation to school (Urdan, 1997b), but moreover, with the social goals of adher-ing to social norms and expectations and the desire to have friends (Anderman &Anderman, 1999). Performance goals were found to be related to the desire to havefriends and to be accepted by classmates (Anderman & Anderman, 1999). In addition,

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 235

however, performance-oriented students were found to be concerned with public dem-onstration of their self-worth; that is, a desire for social visibility, prestige, aYliationwith ‘the popular group,’ and establishing high social status (Anderman, 1999; Levyet al., 2004). This research seems to point to a general concern of performance-orientedstudents with their place in both the academic and social hierarchies of the classroom,and with aspects of self-presentation. In contrast, mastery-oriented students seem to beconcerned more with experiences of growth in both domains and less with impression-management.

However, whereas previous studies suggest that students’ achievement goals are associ-ated with their social goals of having friends and social status, research has not examinedyet the possible relations between students’ academic motivational orientations and thequality of students’ relationships with peers. For example, whereas Anderman and Ander-man (1999) found achievement goals to be associated with the general desire to havefriends (i.e., “I would like to get to know my school friends well,” “I want to be part ofthings that other kids are doing at school”), and Levy et al. (2004) found that achievementgoals were associated with diVerent criteria to have friends (e.g., “popularity” versus“learning together”) it is not clear what quality of friendship these students actually experi-ence. Therefore, the present study focuses on the relations of achievement goals and thismore speciWc aspect of social relationships: friendship intimacy.

The diVerent role that self-worth plays in mastery, performance-approach, and perfor-mance-avoidance goals (cf. Dweck, 1999; Nicholls, 1989) raises the hypothesis that theseachievement goals would be diVerentially associated with friendship intimacy. Intimateclose relations are characterized by mutual self-disclosure, trust, sensitivity and emotionalsupport by partners (e.g., Assor & AlW, 1996; Orlofsky, 1993; Paul & White, 1990; Prager,1995; Reis & Shaver, 1988). In intimate friendships, friends do not only spend timetogether, have common interests or show that they are happy to meet each other and spendtime together. Rather, they are able to share diYculties and weaknesses, seek support andadvice, and discuss conXicts and negative feelings toward each other. Moreover, such rela-tions are characterized by friends’ ability to listen to each other, to be sensitive to eachother’s concerns and feelings, and to oVer appropriate emotional and instrumentalsupport.

Thus, students’ academic achievement goals may join the multiple factors that aVect orare related to the level of intimacy that students would establish with their friends (Berndt,2004). Whereas the relation of academic motivational orientations with intimacy may notbe high, such relations may shed light on the possible Xow between students’ academic andsocial experiences in school. We hypothesize that friendship of either performance-approach or performance-avoidance oriented students may be characterized by a low levelof intimacy. Research suggests that, for many performance-oriented students, a salientaspect of friendship is its contribution to social status and social visibility (Levy et al.,2004). In addition, concern with social comparison and impression management may deterperformance-oriented individuals from exposing personal weaknesses and diYculties, andmight even encourage them to exaggerate and over-emphasize positive personal attributes.Moreover, the self-focus of performance-oriented students may undermine their capacityto listen attentively to their friends. As a result, the friends of performance-oriented stu-dents may act in ways that are less intimate and emotionally supportive toward them. Inaddition, the focus on impressing others can lead to the perception that sharing of vulnera-bilities and diYculties is risky. Thus, when students focus on creating a desirable public

236 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

image, the possibility of “betrayal” by friends may be highly salient and could lead tolower levels of trust, sharing, and therefore, intimacy (cf. Berndt & Keefe, 1996). This maybe more evident in performance-avoidance goals which focus on the risk of negative publicperceptions. Nevertheless, performance-approach goals may be associated with theseconcerns as well (cf. Kaplan, 2004).

In contrast, mastery goals are associated with little concern with impression manage-ment and with much more concern with personal development and growth (Kaplan &Maehr, 2002). Thus, in addition to having little concern with the social costs of the expo-sure of weakness, mastery-oriented students may be more likely to consult with theirfriends about ways of coping with personal diYculties that impede their development inareas that interest them. Nicholls’ (1992) notion that mastery goals are associated withthe belief that success and development is achieved through collaboration with others,learning from mistakes and diYculties, and exertion of eVort, may suggest that masterygoals would be positively associated with the tendency to trust one’s friends as partnerswho can help one to understand and overcome problems and weaknesses. This sugges-tion receives some support by research Wndings indicating that mastery-oriented stu-dents tend to choose friends on the bases of self-development and of enjoying beingtogether (Levy et al., 2004).

Research investigating the relations of multiple achievement goal proWles and socialrelations is extremely scarce. In one qualitative study, Levy et al. (2004) found that thesocial concerns of students with multiple goals (e.g., mastery and performance-approach)included both issues which theoretically are expected to be associated with mastery goalssuch as having friends, cohesion, social responsibility and issues which theoretically areassociated with performance goals such as social visibility and status. Previous researchraised the possibility that a simultaneous pursuit of high mastery and high performance-approach goals would be associated with high achievement, at a level similar to, and per-haps even higher than, a motivational proWle of high mastery goals and low performancegoals (Harackiewicz et al., 2002; Meece & Holt, 1993; Pintrich, 2000a). However, it mayvery well be that when friendship intimacy is concerned, the orientation to public impres-sion management that will still be associated with the high mastery and high performance-approach goals proWle may interfere with the willingness of such students to fully immersein intimate relationship, at least in comparison to those students who are really notconcerned with public self-worth.

5. Individual diVerences and intimate friendship: Gender and perceived eYcacy

Whereas students’ academic orientations are hypothesized to be related diVerentially tofriendship, there are other individual diVerences, sometimes associated with achievementgoals, which are commonly found to be related to friendship and therefore are importantto account for. Perhaps the most familiar of these characteristics is gender. Students’ gen-der has been continuously found to be related to level of friendship intimacy (Maccoby,1990). Whereas boys and girls do not necessarily diVer in number of friends or the timethey spend with friends, most research evidence suggests that girls are more likely thanboys are to have friendships characterized by mutual self-disclosure, sensitivity and emo-tional support, that is, intimate friendships.

Whereas girls’ friendships are more commonly found to be characterized by talking, car-ing, helping, conXict resolution, and intimacy, boys’ friendships are found to be characterized

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 237

by shared activities (Buhrmester & Furman, 1987; DuBois & Hirsch, 1993; Parker & Asher,1993; Shulman et al., 1997). This diVerence has been explained by socialization processes andsex roles (Maccoby, 1990). Thus, in investigating the relations between motivational orienta-tions and intimacy in friendship, it is important to take into account students’ gender.

Another characteristic whose relations with intimacy should be examined in investiga-tions concerning achievement goals is students’ perceived academic eYcacy, or students’beliefs about their ability to succeed in schoolwork. Perceived academic eYcacy is animportant variable in and of itself (Bandura, 1997; Schunk, 1991). High perceived aca-demic eYcacy is associated with positive cognitive, emotional and motivational outcomes(Schunk & Miller, 2002). High perceived academic eYcacy may be associated with higherintimacy among students as it lowers stress, reduces concerns with self-worth, andincreases a sense of well-being. And, when students feel better about themselves, they areable to react in a more caring and trusting ways toward their friends. Moreover, academicself-eYcacy is commonly moderately related to mastery goals and performance-approachgoals. Therefore, in assessing the relations of these motivational orientations with out-comes, students’ self-eYcacy should be controlled for.

6. Hypotheses

Based on the foregoing considerations, the following hypotheses were formulated:

1. Mastery goals will be positively associated with intimate friendship, whereas perfor-mance-approach goals and performance-avoidance goals will be negatively associatedwith intimate friendship, with performance-avoidance goals demonstrating strongernegative relations.

2. ProWles of high mastery goals and low performance-approach and performance-avoid-ance goals would be most highly associated with intimate friendship. ProWles of highmastery goals and high performance-approach goals would be associated with a higherlevel of intimate friendship than proWles with low mastery goals.

3. Perceived eYcacy and being a girl are expected to be positively associated with intimatefriendship.

7. Method

7.1. Participants

Participants were 203 seventh grade Jewish students (46% boys, 54% girls) from oneJunior High School in the center of Israel. The school serves students from several commu-nities and also includes a boarding facility. The communities vary in their socio-economicstatus, but most of them are composed of middle to upper-middle class families. Partici-pants constituted 75% of all the students in the cohort.

7.2. Instruments

The response scale for all the items in the survey was a 5 point Likert Scale ranging from1-“not at all true of me” to 5-“very true of me.” All items in the study appear in theAppendix A.

238 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

7.2.1. Personal achievement goalsThe scales assessing mastery, performance-approach, performance-avoidance goals and

perceived academic eYcacy were adapted from the Patterns of Adaptive Learning Survey(PALS) (Midgley et al., 2000). A sample item in the scale assessing mastery goals is “Animportant reason why I do my work in class is because I like to learn new things.” A sam-ple item in the scale assessing performance-approach goals is “I would feel successful inclass if I did better than most of the other students.” A sample item in the scale assessingperformance-avoidance goals is “One of my main goals is to avoid looking like I can’t domy work.” A sample item in the scale assessing academic eYcacy is “Even if the work ishard, I can learn it.”

The scales assessing achievement goals and perceived eYcacy were adopted from anestablished instrument. Therefore we conducted a conWrmatory factor analysis for thesescales. We used AMOS 5 (Arbuckle, 2003) to conWrm the structure of loadings suggestedby Midgley et al. (2000). The analysis provided adequate support for the structure of thescales (�2(210)D324.5, p < .001; TLID .89; CFID .91; RMSEAD .05, CI:0.42–0.64, pD .29).Range of item loadings on their respective scales was between 0.39 and 0.84. We con-structed four variables: mastery goals (5 items), performance-approach goals (6 items),performance-avoidance goals (6 items), and academic-eYcacy (5 items). Descriptive statis-tics and reliability of these variables are presented in Table 1.

7.2.2. Intimacy in friendshipThe instrument assessing intimacy in friendship was adopted from Assor and AlW

(1996) self-report intimacy measure, which was based on Reis and Shaver (1988), as well ason White, Speisman, Costos, Jackson, and Batris (1989) and Paul and White (1990)conception of intimate relationship. The construct validity of the Assor and AlW (1996)self-report measure of intimacy was supported by its relations with several theoreticallyrelevant measures. SpeciWcally, participants’ scores on this self report scale were found tobe positively associated with the level of intimacy maturity they revealed in semi-structuredinterviews. Intimacy maturity was assessed via the method developed by White and hercolleagues (see White et al., 1989; and also Paul & White, 1990; White, Speisman, Jackson,Batris, & Costos, 1986) (rD .45, p < .01). Moreover, the intimacy self-report scale was alsopositively related to the Fidelity and Love subscales of the measure developed by Mark-strom-Adams, Sabino, Turner, and Berman (1994) to assess Eriksonian Ego strengths(rD .28 and .31, p < .01, respectively).

The intimacy self-report scale was modiWed to Wt seventh grade students. It includes itemsassessing readiness of self and of friends to expose diYculties, listen, and seek or give help, aswell as items tapping willingness to work out diYculties in the relationship. Sample items are:

Table 1Descriptive statistics of variables in the study

Variable name Number of items Reliability Mean Standard deviation Skewness

Mastery goals 5 0.75 3.48 0.76 0.03Performance-approach goals 6 0.85 3.14 0.95 ¡0.01Performance-avoidance goals 6 0.76 2.35 0.81 0.58Self-eYcacy 5 0.81 4.13 0.70 ¡0.76Intimacy 8 0.86 3.44 0.84 ¡0.24Mistrust 5 0.74 2.23 0.78 0.99

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 239

“When I encounter a problem in my schoolwork, I talk to my friends about it”, “When myfriends feel disappointed with themselves, they talk to me about it”, ” If I have diYculty withmy school work, I would rather that my friends not know about it” (reversed), “When I tryto tell my friends about a problem I have, they do not always listen to me” (reversed), “WhenI have a problem with my friends, I talk to them about it”; “When my friends are angry withme, they talk to me about it”, “When I talk with my friends about things that bother me inour friendship, they show no interest and even get angry” (reversed).

Exploratory factor analysis of the modiWed intimacy scale produced two factors. TheWrst factor included eight items assessing readiness of self and friends to expose diYcultiesand seek or give help, as well as to work out diYculties in the relationship, and onereversed item, which loaded negatively on this factor, and which subsequently wasremoved from all further analyses. We labeled this factor Intimacy. The second factorincluded six reversed items reXecting mistrust, inattention, carelessness, and lack of willing-ness to work out diYculties in the relationship. We labeled this factor Mistrust. The Inti-macy and Mistrust factors explained 43.24% of the variance in the items. The two factorswere modestly negatively correlated (rD¡.43, p < .01).

The factor loadings of items on their respective factors ranged from 0.46 and 0.78, andcross-loadings were equal or lower than 0.15. One item in the Mistrust scale was removedbecause it lowered the scale’s reliability. Table 2 presents the loadings of items on the Wnalfactors. Using AMOS 5, we compared the two-factor model with a model in which itemsassessing Intimacy and Mistrust load on a single intimacy factor. The analysis supportedthe structure of distinct factors (��2D 39.30, �dfD2, p < .01). Thus, we constructed twovariables: Intimacy (8 items) and Mistrust (5 items).

7.3. Procedure

Surveys were administered to students in their classrooms during regular school hours.Pairs of trained research assistants read instructions and survey items aloud to students.

Table 2Factor structure and loadings of intimacy items

Item Factor 1 Factor 2

When I have a problem with schoolwork, my friends listen to me and try to help me .78 ¡.01When my friends feel disappointed with themselves, they talk to me about it .73 ¡.03When I have a problem with my friends, I talk to them about it. .68 .04When I encounter a problem in my schoolwork, I talk to my friends about it .68 .00When my friends have diYculties in schoolwork, they talk to me about it .67 ¡.03When my friends and I are Wghting they try to understand my point of view .61 .04My friends help me decide what to do when I have problems with other friends .60 ¡.04When my friends are angry with me, they talk to me about it .57 ¡.02When I want to talk to my friends about a problem I have, they don’t always listen to me ¡.07 .76When I talk with my friends about things that bother me in our friendship, they show no

interest and even get angry¡.02 .65

My friends don’t always care about how I feel ¡.03 .57If I have diYculty with my schoolwork, I would rather that my friends not know about it ¡.15 .57Sometimes I feel that my friends talk too much about themselves .09 .46

Explained variance in items 33.65% 9.58%Eigenvalue 4.91 1.81

240 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

Students were told that the survey asked about how they felt about school, school work,and friendship, that the survey was not a test, that there were no right or wrong answers,and that the information in the survey would be kept conWdential. The survey was com-posed of two parts that were administered approximately a week apart from each other.The scales assessing personal achievement goals and academic perceived-eYcacy wereadministered in the Wrst part. The scales assessing intimacy in friendship were administeredin the second part.

8. Results

The results are presented in four parts. First, we present the descriptive statistics of thevariables. In the second section, we present the zero-order correlations among the vari-ables. In the third section, we present the results of hierarchical multiple regressions thattested the relations among personal achievement goals and students’ reports of intimacy infriendship. Finally, in the fourth section, we present the results of a MANOVA that testeddiVerences between diVerent proWles of multiple achievement goals, which were identiWedwith cluster analysis, on intimacy in friendship.

Descriptive statistics and reliability of the variables are presented in Table 1. All scalesexempliWed appropriate statistical characteristics, with good reliabilities, and approxima-tion to normal distributions.2 Perceived-eYcacy was the only scale with a relatively highmean, which is a common Wnding. Another common Wnding is the pattern in which Mas-tery Goals had the highest mean and Performance-Avoidance Goals the lowest meanamong the three goals. This may suggest that generally, students are either more masteryoriented than performance oriented, or, that students Wnd it easier to endorse mastery goalitems than they do performance goal items.

8.1. Correlations

Table 3 presents the correlations among the variables. Mastery goals were positivelyrelated to performance-approach goals, but were not related to performance-avoidancegoals. Performance-approach goals were positively related to performance-avoidancegoals. Self-eYcacy was positively related to mastery goals and negatively but weaklyrelated to performance-avoidance goals. Also, as expected, Intimacy was negatively relatedto Mistrust.

Mastery goals were positively related to Intimacy and were not related to Mistrust. Per-formance-approach goals and performance-avoidance goals were negatively related toIntimacy and positively related to Mistrust. Self-eYcacy was positively related to Intimacyand negatively related to Mistrust.

8.2. Hierarchical regressions

In order to test the unique associations of each of the three achievement goals withIntimacy and Mistrust, and whether these associations are moderated by other achieve-

2 The variables of Intimacy and Mistrust were checked for outliers. Intimacy had none. Mistrust had 3 caseswith Z-scores over 3.2. All correlational analyses were conducted with as well as without these 3 cases. As the Wnd-ings were almost identical, the results are reported for the complete sample.

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 241

ment goals that students may be pursuing simultaneously, we ran two regression analy-ses with interaction terms: one with Intimacy and one with Mistrust as outcomevariables. In order to test the contribution of achievement goals to these variables overand beyond individual diVerences, we Wrst controlled for variance that might be associ-ated with gender or self eYcacy. Thus, in the Wrst step of the analysis, we entered stu-dents’ gender as a predictor, and in the second step we entered students’ academic self-eYcacy. Then, in the third step, we entered the three achievement goals variables. In thefourth step we entered three interaction terms that included all the conWgurations oftwo-way interaction between the three achievement goals. Finally, in the Wfth step, weentered one three-way interaction term of the three achievement goals variables. Fol-lowing the recommendations of Aiken and West, 1991, variables were centered beforeforming the interaction terms in order to reduce the level of multicollinearity. Theresults are presented in Table 4.

The Wndings indicated that being a girl and having high sense of academic eYcacy isassociated with Intimacy. Independently, Mastery goals were also positively associatedwith Intimacy and Performance-approach goals were negatively related to Intimacy. Thesetwo achievement goals added 5% (p < .05) to the explained variance in Intimacy. Terms of

Table 3Correlations among variables in the study variables

Note. N D 203 (NIntimacy D 175).a p < .05.b p < .01.

2 3 4 5 6 7

1. Mastery goals 0.21b 0.12 0.36b 0.02 0.20a ¡0.032. Performance-approach goals 1.00 0.57b 0.02 ¡0.27b ¡0.20a 0.33b

3. Performance-avoidance goals 1.00 ¡0.16a ¡0.30b ¡0.17a 0.41b

4. Self-eYcacy 1.00 0.06 0.24b ¡0.24b

5. Gender 1.00 0.38b ¡0.22b

6. Intimacy 1.00 ¡0.38b

7. Mistrust 1.00

Table 4Hierarchical multiple regressions results with achievement goals predicting intimacy

a p < .05.b p < .01.c p < .001.

Intimacy Mistrust

Step 1 Step 2 Step 3 Step 1 Step 2 Step 3

Gender .38c .37c .35c ¡.22b ¡.21b ¡.10Self-eYcacy .23c .18a ¡.24c ¡.21b

Mastery goals .19a ¡.03Performance-approach goals ¡.21a .19a

Performance-avoidance goals .05 .25b

R2 .14 .20 .25 .05 .10 .24�R2 .06c .05a .05c .14c

242 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

interaction between the goals did not add a signiWcant amount of variance to the regres-sion model.3

The Wndings from the second regression suggested that a sense of self-eYcacy was nega-tively associated with Mistrust. Mistrust was unrelated to mastery goals, but positivelyassociated with performance-approach and Performance-avoidance goals, which togetherexplained additional 14% (p < .001) of the variance in this variable. Initially, genderemerged as a signiWcant predictor of Mistrust, suggesting that being a boy is associatedwith Mistrust. However, once achievement goals were entered into the regression, genderwas no longer signiWcant, suggesting that the association between gender and Mistrust isrelated to diVerent patterns of achievement goals among boys and girls. No interactionterms were signiWcant in this analysis and the inclusion of the interaction terms did not adda signiWcant amount of explained variance.

8.3. Intimacy among proWles of multiple achievement goals

Testing the possible moderation of achievement goals on the relations of other achieve-ment goals and intimacy provides one way of investigating multiple goals. However, thisvariable-centered approach has several limitations. First, this approach favors the role ofmain variable eVects, and thus potentially masks possibly more subtle eVects of variablecombinations. Second, this approach assumes linearity of relations and may mask non-lin-ear and possible complex eVects of various levels of variable combinations. Third, whereasinteraction terms of two variables are rather easily interpreted, as the combinations ofmore variables are investigated the complexity of interpretation increases while its reliabil-ity decreases. Finally, while maintaining the power of continuous variables, interactionterms do not refer to the actual motivational proWles that students in the sample report(Pastor, Barron, Miller, & Davis, 2004). Thus, in order to supplement the variable-centeredapproach, we employed a person-centered approach which characterizes students’ multiplegoals on the basis of their actual reports.

We used a hierarchical cluster analysis to form groups of students with diVerent proWlesof multiple achievement goals. The three achievement goal variables—mastery, perfor-mance-approach, and performance-avoidance—were used as the basis for the cluster anal-ysis. Cluster analysis has been recommended for purposes of identifying patterns orproWles in a sample (Hair, Anderson, Tatham, & Black, 1998). Cluster analysis, using Wardmethod with Euclidean distance, was chosen over other methods of construction of proWles(e.g., median split) because it is better in creating groups that are relatively homogenous onthe variables of interest. It has also been used previously with success in creating achieve-ment goal proWles (e.g., Bembenutty, 1999; Hodge & PetlichkoV, 2000; Meece & Holt,1993). The main concern in employing cluster analysis to form groups is the lack of strictcriteria for making a decision about the number of clusters (Hair et al., 1998). Weemployed two criteria: (1) parsimony—since the aim of the analysis is data reduction, we

3 The analysis suggested that there was a signiWcant interaction between mastery goals and performance-avoid-ance goals. Further analyses indicated that the signiWcance of this interaction was a part of a suppression eVectcaused by the simultaneous inclusion of the two interaction terms involving mastery goals: masterygoals £ performance-approach and mastery goals£ performance-avoidance. Whereas this is a potentially inter-esting Wnding to explore, the non-signiWcant F test did not warrant pursuit of such an interpretation (Pedhazur,1982, pp. 375–376).

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 243

considered a smaller number of clusters better than a higher number; (2) a signiWcant“jump” in the agglomeration, or fusion coeYcient (Aldenderfer & BlashWeld, 1993). Thefusion coeYcient indicates the amount of information that is lost whenever cases or clus-ters are combined. Cluster analysis provides such a coeYcient for each step. This begins inthe Wrst step with the merging of two cases in the sample into one cluster, and ends with theconstruction of one cluster that includes all participants. An examination of a pattern inthe size of coeYcients during the last ten steps of cluster construction (i.e., 10 to 9, 9 to 8, 8to 7ƒ) indicated that the shift from seven clusters to six clusters was marked by a distinctincrease in the size of coeYcients, with coeYcients prior to this shift being much smallerthan coeYcients following this shift. This suggested the six clusters as the best solution(Aldenderfer & BlashWeld, 1993, p. 57). Table 5 presents the characteristics of the six clus-ters on the achievement goals variables, gender, self-eYcacy, and the intimacy variables.

A MANOVA that tested diVerences between the clusters on the three achievement goalswas signiWcant (FD69.38, p < .001, �2D0.63) as were the univariate analyses for each of theachievement goals (mastery goals: FD 36.09 p < .001, �2D .48; performance-approachgoals: FD 151.342, p < .001, �2D .79; performance-avoidance goals: FD 79.76, p < .001,�2D .67). Post hoc Tukey HSD was used to identify the pairs of clusters that are statisti-cally diVerent from each other on each of the three achievement goals. These Wndings arepresented in Table 5.

In order to provide a clearer characterization of the clusters in terms of relative endorse-ment of achievement goals, we employed a combination of three criteria: (1) the absolute levelof the response scale, with scores between 1.00 and 2.39 indicating low level of endorsement,scores between 2.40 and 3.59 indicating medium level of endorsement, and scores between 3.60and 5.00 indicating high level of endorsement; (2) standard scores of the achievement goalsvariables, with Z-scores higher than .49 considered to be high on the variable, Z-scoresbetween ¡.49 and .49 considered as medium, and Z-scores lower than ¡.49 considered as lowon the variable; and (3) the signiWcant diVerences found between the clusters in endorsement ofthe goals. This analysis led to the characterization of the clusters along Wve levels of endorse-ment of the achievement goals: high (H), medium-high (M/H), medium (M), medium-low (M/L), and low (L). This characterization is presented in the parenthesized letters adjacent to themeans in Table 5: cluster 1 was characterized as medium-low on mastery and performance-approach goals and as low on performance-avoidance goals (M/L, M/L, L), cluster 2 wascharacterized as medium in mastery goals and low in both performance goals (M, L, L), cluster3 was characterized as medium-low in mastery goals and medium in both performance goals(M/L, M, M), cluster 4 was characterized as medium-high in mastery goals and high in bothperformance goals (M/H, H, H), cluster 5 was characterized as high in mastery goals, mediumin performance-approach goals, and medium-low in performance-avoidance goals (H, M,M/L), and cluster 6 was characterized as high in mastery and performance-approach goalsand medium in performance-avoidance goals (H, H, M).4

4 Characterization of the proWles on levels of achievement goals mostly paralleled the signiWcant diVerences oneach achievement goal variable, with a few exceptions: the level of mastery goals in cluster 4 (M/H) was not sig-niWcantly diVerent from that of cluster 2 (M) on the one hand and from that of clusters 5 (H) and 6 (H) on theother; the level of performance-avoidance goals in cluster 1 (L) was not signiWcantly diVerent from that of cluster5 (M/L); the levels of performance-approach goals of clusters 3 and 5 (M) were signiWcantly diVerent from eachother; and the levels of performance-avoidance goals of clusters 1 and 2 (L) were signiWcantly diVerent from eachother.

244I. L

evy-Tossm

an et al. / Contem

porary Educational P

sychology 32 (2007) 231–252

st at p < .05.

Medium-High; M, Medium; M/L, Medium-Low; L,

31) Cluster 5 (N D 31) Cluster 6 (N D 23) Mean

.45) (H) 4.13cW(0.86) (H) 4.24dgj(1.00) 3.487) (M) 2.95dhkmn(¡0.20) (H) 4.30eiln(1.22) 3.144) (M/L) 1.99gjkm(¡0.44) (M) 2.48dhlm(0.16) 2.35

4.42a 4.32 4.1313 10 9418 13 1093.67 3.42 3.441.92de 2.20f 2.23

Table 5Analyses of variance between clustered motivational proWles in achievement goals and in intimacy

Note: CoeYcients in the same row that share a superscript letter are signiWcantly diVerent from each other at leaNumbers in parentheses are Z-scores.Letters in parentheses near means indicate characterization on level of achievement goals with H, High; M/H,Low.

Cluster 1 (N D 43) Cluster 2 (N D 25) Cluster 3 (N D 50) Cluster 4 (ND

Mastery goals (M/L) 3.06abcd(¡0.55) (M) 3.46aefg(¡0.03) (M/L) 2.89ehij(¡0.78) (M/H) 3.82bh(0Performance-approach goals (M/L) 2.57abcde(¡0.60) (L) 1.62afghi(¡1.60) (M) 3.31bfjkl(0.18) (H) 4.25cgjm(1.1Performance-avoidance goals (L) 1.87abcd(¡0.59) (L) 1.39aefgh(¡1.19) (M) 2.64beij(0.36) (H) 3.60cWkl(1.5Self-eYcacy 4.12 4.14 3.93a 3.99Gender boys 16 4 30 21

Girls 27 21 20 10Intimacy 3.46 3.72 3.19 3.29Mistrust 2.04a 1.84bc 2.46bd 2.92acef

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 245

Table 5 presents results of tests investigating diVerences among the clusters on the out-come variables. An ANOVA testing the diVerences among the clusters on self-eYcacy indi-cated signiWcant diVerences (FD2.55, p <.05, �2D .06). A post hoc Tukey HSD test indicatedthat students in cluster 3 (M/L, M, M) reported signiWcantly less eYcacy than students incluster 5 (H, M, M/L). A cross-tabulation test suggested that the clusters were also diVerentfrom each other by gender (�2(203,5)D20.48, p < .001). Girls seem to be more highly repre-sented in proWles with low performance-approach and avoidance goals, whereas boys aremore highly represented in proWles with high performance-approach goals. The proWles inwhich both mastery goals and performance-approach goals were relatively high had approx-imately similar numbers of boys and girls. Finally, a MANOVA that tested diVerencesbetween the six clusters on Intimacy and Mistrust found signiWcant diVerences (FD4.09,p <.001, �2D0.11). Individual univariate analyses with a Tukey post hoc indicated that theclusters did not diVer signiWcantly on level of Intimacy (FD1.23, p > .1, �2D0.05), but thatthere were signiWcant diVerences among the clusters in reported level of Mistrust (FD7.91,p <.001, �2D0.20): students in clusters 1 (M/L, M/L, L), 2 (M, L, L), 5 (H, M, M/L), and 6 (H,H, M) reported signiWcantly less Mistrust than did students in cluster 4 (M/H,H,H); and stu-dents in clusters 2 (M, L, L) and 5 (H, M, M/L) also reported signiWcantly less Mistrust thandid students in cluster 3 (M/L, M, M). These Wndings suggest a pattern in which proWles withlevel of endorsement of Mastery goals relatively higher than the endorsement of both Perfor-mance goals are associated with less Mistrust than are proWles with level of endorsement ofPerformance goals relatively higher than the endorsement of Mastery goals.

9. Discussion

Research in the past several years has been establishing the understanding that students’social relationships and interactions are related to their attitudes towards schoolwork(Anderman, 1999). More recently, some research has started to look at the possible role ofstudents’ academic motivation in the type and quality of their social interactions (Ander-man & Anderman, 1999; Kaplan, 2004; Levy et al., 2004). The present study adds to thiswork by focusing on one important aspect of the quality of students’ social relationships:the intimate nature of their friendships. Intimacy in friendship is a complex phenomenonthat draws on multiple personal and social psychological resources (Berndt, 2004). Never-theless, our hypothesis that academic motivation would be related, albeit weakly, to qual-ity of intimate friendships among early adolescents in school, was supported. Overall, theWndings suggest that mastery goals are likely to be associated with friendship that is char-acterized by intimacy. Students who reported that when doing schoolwork they focus ondevelopment, personal growth, interest and learning were also likely to report that theirrelationship with friends involves trust, positive social problem-solving, and mutual shar-ing of diYculty and weaknesses. In contrast, performance-approach goals and perfor-mance-avoidance goals are likely to be associated with friendship that is characterized bylack of intimacy. Students, who reported a concern with appearing more able than othersor with not appearing less able than others, were also likely to report that their relationshipwith friends involves concealment of problems and diYculty, lack of willingness to listen,and inconsideration. These relations were found above and beyond diVerences in intimacythat was explained by students’ gender and perceived academic eYcacy.

It appears, then, that when demonstrating smartness and superior ability is more impor-tant than learning and understanding, students may pay a signiWcant interpersonal price.

246 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

When schooling and schoolwork are construed as a continual competitive struggle inwhich self-worth depends on performance relative to others, students may be less able orwilling to develop high quality friendships. Indeed, it is logical that students who perceivetheir engagement in academic tasks as involving a risk for demonstrating low ability arelikely to be more suspicious of their friends and to avoid exposing weaknesses, sharingproblems, or consulting. These students may also be prone to the perception that theirfriends are not so interested in them for who they are. Interestingly, students who engage inschoolwork with the purpose of demonstrating smartness seem also to harbor doubtsabout the trustworthiness of their friends and to avoid exposure of diYculties, arguably, asthese may hamper their aspirations for a positive public image.

In contrast, when students understand school and schoolwork as opportunities for self-development and learning, they seem to hold similar perspectives towards their interper-sonal relationships. Findings from the variable-focused analysis suggest that, in and ofitself, such an orientation may be related to more willingness to engage in honest exchangeof feelings and mutual help-seeking. This tendency, however, seemed to have manifesteddiVerently in the person-centered analysis.

When proWles of multiple achievement goals were examined, the Wndings suggested thatit was motivational proWles in which performance-approach and performance-avoidancewere higher than mastery goals that were associated with the negative characteristics ofmistrust and inconsideration. Indeed, more so than the absolute level of goal orientations,it seemed to be the relative level of pursuit of these motivational orientations that wasrelated to diVerences in quality of friendship, with a higher level of mastery goals relative toperformance goals—particularly performance-avoidance goals—being associated with lessmistrust between friends. Thus, when mastery goals are considered together with theperformance goals, the Wndings seem to suggest that its role may be less in facilitatingintimacy and more in moderating, at least to some degree, the detrimental eVects of perfor-mance goals on quality of friendship.

When mastery and performance-approach goals were pursued at similar high or lowlevels (with lower levels of performance-avoidance goals), quality of friendship was notsigniWcantly diVerent from that associated with proWles in which mastery goals werehigher than the two performance goals. And yet, it was proWles in which mastery goalswere higher than performance-approach goals that were found to be signiWcantlydiVerent from both proWles with highest scores of mistrust, suggesting that in terms ofintimate aspects of relationships, having higher mastery goals relative to performancegoals may be somewhat more desired than having equal levels of mastery and perfor-mance-approach goals.

The relative strength of mastery goals and performance goals may be said to represent abalance that students strike between the desire to develop competence and grow and theconcern with the implications of one’s competence to self-worth (Kaplan & Flum, 2005).These two general desires arguably exist in any individual (Erikson, 1968). The desires’absolute strength, particularly in speciWc domains such as school, may diVer among indi-viduals. However, their relative strength may represent a person’s more or less adaptiveresolution of their potentially competing pulls: towards growth versus towards establishingor protecting self-worth. Erikson suggested that in such psychosocial conXicts, the adaptiveresolution is a constructive balance in which the positive pole dominates over the threaten-ing pole—which is, nevertheless, adaptively present. Thus, it may be that a motivationalproWle in which mastery goals are higher than performance goals—particularly perfor-

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 247

mance-avoidance goals—represents such a constructive balance between these two desiresthat involves a secure sense of self-worth, and a resource for coping with the task of estab-lishing close friendships during early adolescence. In contrast, proWles in which perfor-mance goals dominate over mastery goals may represent less adaptive balances in whichself-worth concerns overshadow growth, involve a less secure sense of self-worth, andinterfere with developmental tasks such establishing intimacy among friends (Flum &Kaplan, 2006; Kaplan & Flum, 2005).

The data used for this study were correlational and were collected at one point of time.These limitations suggest that no conclusion can be made with regard to the eVect of aca-demic motivational orientations on quality of friendship. Indeed, it may be that those stu-dents who develop intimate friendships become also more mastery oriented and lessperformance-approach and avoidance oriented. However, the literature and the person-centered results do not really support this suggestion. The Wndings from research con-cerned with friends’ inXuence on motivation as well as the relatively high intimacy charac-terizing students who seem to be somewhat disengaged from school (i.e., relatively low onall three achievement goals) suggest that one can have good and intimate friends and yetnot be mastery oriented. Nevertheless, rather than suggesting that the causal direction ofrelations always Xows from academic motivational orientations to interpersonal relations,a more pertinent hypothesis would be that there is an underlying process involving self-worth concerns (Crocker & Wolfe, 2001; Steele, 1988; Tesser, 2001), which relates pro-cesses in the academic and interpersonal domains (Kaplan, 2004). Thus, a student who ispredisposed to sense a threat to self-worth may manifest both performance goals in theacademic domain and low quality of friendship in the interpersonal domain. Still, such anorientation to self-worth may be instigated by environmental conditions, for example thosewhich emphasize performance goals in the classroom (cf. Kaplan, 2004; Levy-Tossman,Kaplan, & Assor, 2005). Thus, particularly when moving to a new context in which friend-ships are not established, it very well may be that an environmental emphasis on perfor-mance goals poses a threat to students’ sense of self worth, which in turn may lead to lowerquality of friendship among students. These are of course empirical questions and shouldbe investigated in future research.

The Wndings of the present study highlight the relations between academic motivationalorientations and social relationships (Anderman, 1999). Schools perhaps are Wrst and fore-most social contexts. Learning in school is always embedded in social relationships, andacademic motivation is likely to have social consequences (Kindermann, 1993). Moreover,it seems that the type of academic orientations could be systematically related to quality ofstudents’ social relationships (cf. Levy et al., 2004). Such a perspective should be exploredfurther in order to inform attempts to reform educational environments in ways that con-tribute simultaneously to better academic learning as well as to more positive relationshipsamong students.

Appendix A

A.1. Mastery goals

1. One of my goals in the class is to learn as much as I can.2. An important reason why I do my class work is because I like to learn new things.3. I like class work that I’ll learn from even if I make a lot of mistakes.

248 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

4. I like class work best when it really makes me think.5. An important reason why I do my class work is because I want to get better at it.

A.2. Performance-approach goals

1. I would feel successful in class if I did better than most of the other student.2. I would like to show my teacher that I’m smarter then the other students in my class.3. I want to be better than other students in my class.4. I would feel really good if I were the only one who could answer the teachers’ questions

in my class.5. I want to do better than other students in my class.6. Doing better than other students in my class is important to me.

A.3. Performance-avoidance goals

1. The reason I do my class work is so my teacher doesn’t think I know less than theothers.

2. It’s very important to me that I don’t look stupid in my class.3. An important reason I do my class work is so that I won’t embarrass myself.4. The reason I do my math work in this class is so others won’t think I’m dumb.5. One reason I might not participate in my class is to avoid looking stupid.6. One of my main goals in my class is to avoid looking like I can’t do my work.

A.4. Academic self-eYcacy

1. Even if the work is hard, I can learn it.2. I’m certain I can master the skills taught in class this year3. I can do even the hardest work in this class if I try.4. I can do almost all the work in my class if I don’t give up.5. I’m certain I can Wgure out how to do even the most diYcult class work.

A.5. Intimacy

1. When I encounter a problem in my schoolwork, I talk to my friends about it2. When I have a problem with schoolwork, my friends listen to me and try to help me.3. When my friends feel disappointed with themselves, they talk to me about it.4. When I have a problem with my friends, I talk to them about it.5. When my friends have diYculties in schoolwork, they talk to me about it.6. My friends help me decide what to do when I have problems with other friends.7. When my friends are angry with me, they talk to me about it.8. When my friends and I are Wghting they try to understand my point of view.

A.6. Mistrust

1. When I want to talk to my friends about a problem I have, they don’t always listen tome.

2. My friends don’t always care about how I feel.

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 249

3. If I have diYculty with my schoolwork, I would rather that my friends not know aboutit

4. When I talk with my friends about things that bother me in our friendship, they showno interest and even get angry.

5. Sometimes I feel that my friends talk too much about themselves.

References

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA:Sage Publications.

Aldenderfer, M. S., & BlashWeld, R. K. (1993). Cluster analysis. Newbury Park, CA: Sage.Ames, C. (1992). Achievement goals and the classroom motivational climate. In D. Schunk & J. Meece (Eds.), Stu-

dent perceptions in the classroom (pp. 327–348). Hillsdale, NJ: Erlbaum.Anderman, E. M., & Maehr, M. L. (1994). Motivation and schooling in the middle grades. Review of Educational

Research, 64, 287–309.Anderman, L. H. (1999). Expanding the discussion of social perceptions and academic outcomes: mechanisms

and contextual inXuences. In T. C. Urdan (Ed.), Advances in motivation: The role of context (Vol. 11, pp. 303–336). Greenwich, CT: JAI Press.

Anderman, L. H., & Anderman, E. M. (1999). Social predictors of changes in students’ achievement goal orienta-tions. Contemporary Educational Psychology, 25, 21–37.

Arbuckle, J. L. (2003). Amos 5. Chicago. IL: Smallwaters.Assor, A., & AlW, O. (1996). Identity formation and intimacy in young married couples: a diVerentiated dimensional

approach. Paper presented at the International Society for the Study of Behavioral Development, Quebec,Canada.

Bandura, A. (1997). Self-eYcacy: The exercise of control. New York: Freeman.Bembenutty, H. (1999). Sustaining motivation and academic goals: the role of academic delay of gratiWcation.

Learning and Individual DiVerences, 11, 233–257.Berndt, T. J. (1999). Friends’ inXuence on students’ adjustment to school. Educational Psychologist, 34, 15–28.Berndt, T. J. (2004). Children’s friendships: shifts over a half-century in perspectives on their development and

their eVects. Merrill-Palmer Quarterly, 50, 206–223.Berndt, T. J., & Keefe, K. (1996). Friends’ inXuence on school adjustment: a motivational analysis. In J. Juvonen

& K. R. Wentzel (Eds.), Social motivation: Understanding children’s school adjustment (pp. 248–277).Cambridge, MA: Cambridge University Press.

Berndt, T. J., Laychak, A. E., & Park, K. (1990). Friends’ inXuence on adolescents’ academic achievement motiva-tion: an experimental study. Journal of Educational Psychology, 82, 664–670.

Birch, S. H., & Ladd, G. W. (1996). Interpersonal relationships in the school environment and children’s earlyschool adjustment: the role of teachers and peers. In J. Juvonen & K. R. Wentzel (Eds.), Social motivation:Understanding children’s school adjustment (pp. 199–225). Cambridge, MA: Cambridge University Press.

Blumenfeld, P. C. (1992). Classroom learning and motivation: clarifying and expanding goal theory. Journal ofEducational Psychology, 84, 272–281.

Brophy, J. (2004). Motivating students to learn (2nd ed.). Mahwah, NJ: Erlbaum.Buhrmester, D., & Furman, W. (1987). The development of companionship and intimacy. Child Development, 58,

1101–1113.Crocker, J., & Wolfe, C. T. (2001). Contingencies of self-worth. Psychological Review, 108, 593–623.DuBois, D. L., & Hirsch, B. J. (1993). School/non school friendship patterns in early adolescence. Journal of Early

Adolescence, 13, 102–122.Dweck, C. S. (1986). Motivational processes aVecting learning. American Psychologist, 41, 1040–1048.Dweck, C. S. (1999). Self-theories: Their role in motivation, personality, and development. Philadelphia, PA:

Psychology Press.Dweck, C. S., & Leggett, E. L. (1988). A social-cognitive approach to motivation and personality. Psychological

Review, 95, 256–273.Dykman, B. M. (1998). Integrating cognitive and motivational factors in depression: initial tests of a goal-orienta-

tion approach. Journal of Personality and Social Psychology, 74, 139–158.Elliot, A. J. (1999). Approach and avoidance motivation and achievement goals. Educational Psychologist, 34,

169–189.

250 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

Elliot, A. J., & Church, M. A. (1997). A hierarchical model of approach and avoidance achievement motivation.Journal of Personality and Social Psychology, 72, 218–232.

Erikson, E. (1968). Identity: Youth and crisis. New York: Norton.Flum, H., & Kaplan, A. (2006). Exploratory orientation as an educational goal. Educational Psychologist, 41, 99–110.Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis (5th ed.). Upper Sad-

dle River, NJ: Prentice-Hall.Harackiewicz, J. M., Barron, K. E., Pintrich, P. R., Elliot, A. J., & Thrash, T. M. (2002). Revision of achievement

goal theory: necessary and illuminating. Journal of Educational Psychology, 94, 638–645.Hodge, K., & PetlichkoV, L. (2000). Goal proWles in sport motivation: a cluster analysis. Journal of Sport and

Exercise Psychology, 22, 256–272.Hussong, A. M. (2000). Perceived peer context and adolescent adjustment. Journal of Research on Adolescence, 10,

391–415.Kaplan, A. (2004). Achievement goals and intergroup relations. In P. R. Pintrich & M. L. Maehr (Eds.), Advances

in research on motivation and achievement, Vol. 13: Motivating students, improving schools: The legacy of CarolMidgley (pp. 97–136). United Kingdom: Elsevier.

Kaplan, A., & Flum, H. (2005). Achievement goal orientations and identity formation. Paper presented at theannual meeting of the American Psychological Association, Washington DC, USA.

Kaplan, A., Gheen, M., & Midgley, C. (2002). Classroom goal structure and student disruptive behavior. BritishJournal of Educational Psychology., 72, 191–211.

Kaplan, A., & Maehr, M. L. (1999). Achievement goals and student well-being. Contemporary Educational Psy-chology, 24, 330–358.

Kaplan, A., & Maehr, M. (2002). Adolescents’ achievement goals: situating motivation in socio-cultural contexts.In T. Urdan & F. Pajaers (Eds.), Academic motivation of adolescents (Vol. 2). Adolescence and education (pp.125–167). Greenwich, CT: Information Age.

Kaplan, A., & Midgley, C. (1997). The eVect of achievement goals: does level of perceived academic competencemake a diVerence? Contemporary Educational Psychology, 22, 415–435.

Kaplan, A., Middleton, M. J., Urdan, T., & Midgley, C. (2002). Achievement goals and goal structures. In C.Midgley (Ed.), Goals, goal structures, and patterns of adaptive learning (pp. 21–53). Mahwah, NJ: Erlbaum.

Kegan, R. (1982). The evolving self: Problem and process in human development. Harvard University press.Kindermann, T. A. (1993). Natural peer groups as contexts for individual development: the case of children’s

motivation in school. Developmental Psychology, 29, 970–977.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.Levy-Tossman, I., Kaplan, A., & Assor, A. (2005). Academic goal orientations and young adolescents’ preferences

for peers as cooperation partners: a longitudinal study. Manuscript in preparation.Maccoby, E. E. (1990). Gender and relationships: a developmental account. American Psychologist, 45, 755–775.Markstrom-Adams, C., Sabino, V., Turner, B., & Berman, R. (1994). Adolescent Ego Resiliency: the Eriksonian

Measure of Ego Strengths. Paper presented in The Thirteen Biennial Meeting of the International Society forthe Study of Behavioral Developmental, Amsterdam.

Meece, J. L., & Holt, K. (1993). A pattern analysis of students’ achievement goals. Journal of Educational Psychol-ogy, 85, 582–590.

Middleton, M. J., & Midgley, C. (1997). Avoiding the demonstration of lack of ability: an under-explored aspectof goal theory. Journal of Educational Psychology, 89, 710–718.

Midgley, C., Kaplan, A., & Middleton, M. (2001). Performance-approach goals: good for what, for whom, underwhat circumstances, and at what cost? Journal of Educational Psychology, 93, 77–86.

Midgley, C., Maehr, M. L., Hruda, L. Z., Anderman, E., Anderman, L., Freeman, K. E., et al. (2000). Patterns ofadaptive learning survey (PALS). University of Michigan.

Nicholls, J. G. (1984). Achievement motivation: conceptions of ability, subjective experience, task choice, and per-formance. Psychological Review, 91, 328–346.

Nicholls, J. G. (1989). The competitive ethos and democratic education. Cambridge, MA: Harvard University Press.Nicholls, G. J. (1992). Students as educational theorists. In D. Schunk & J. Meece (Eds.), Student perceptions in the

classroom (pp. 267–286). Hillsdale, NJ: Lawrence Erlbaum.Nicholls, J. G., Cheung, P. C., Lauer, J., & Patashnick, M. (1989). Individual diVerences in academic motivation:

perceived ability, goals, beliefs, and values. Learning and Individual DiVerences, 1, 63–84.Noack, P., Kretterk, C., & Walper, S. (2001). Peer relations of adolescents from nuclear and separated families.

Journal of Adolescence, 24, 535–548.

I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252 251

Orlofsky, J. L. (1993). Intimacy status: theory and research. In J. E. Marcia, A. S. Waterman, D. R. Matteson, S. L.Archer, & J. L. Orlofsky (Eds.), Ego-identity: A handbook for psychosocial research. New York: Springer-Verlag.

Parker, J. G., & Asher, S. R. (1987). Peer relations and later personal adjustment: are low-accepted children atrisk? Psychological Bulletin, 102, 357–389.

Parker, J. G., & Asher, S. T. (1993). Friendship and friendship quality in middle childhood: links with peer groupacceptance and feelings of loneliness and social dissatisfaction. Developmental Psychology, 29, 611–621.

Pastor, D. A., Barron, K. E., Miller, B. J., & Davis, S. L. (2004). College students’ achievement goal orientation pro-Wles. Paper presented at the annual meeting of the American Educational Research Association, San Diego, CA.

Patrick, H., Anderman, L. H., & Ryan, A. M. (2002). Social motivation and the classroom social environment.In Goals, goal structures, and patterns of adaptive learning (pp. 85–108). Mahwah, NJ: Erlbaum.

Paul, E. L., & White, K. M. (1990). The development of intimate relationships in late adolescence. Adolescence, 25,375–400.

Pedhazur, E. J. (1982). Multiple regression in behavioral research: Explanation and prediction (2nd ed.). FortWorth, TX: Harcourt Brace Jovanovich.

Pintrich, P. R. (1994). Continuities and discontinuities: future directions for research in educational psychology.Educational Psychologist, 29, 137–148.

Pintrich, P. R. (2000a). Multiple goals, multiple pathways: the role of goal orientation in learning and achieve-ment. Journal of Educational Psychology, 92, 544–555.

Pintrich, P. R. (2000b). The role of goal orientation in self-regulated learning. In M. Boekaerts, P. R. Pintrich, &M. Zeidner (Eds.), Handbook of self-regulation (pp. 451–502). San Diego, CA: Academic Press.

Pintrich, P. R. (2003). A motivational science perspective on the role of student motivation in learning and teach-ing contexts. Journal of Educational Psychology, 95, 667–686.

Prager, K. J. (1995). The psychology of intimacy. New York: The Guilford Press.Reis, H. T., & Shaver, P. (1988). Intimacy as an interpersonal process. In S. Duck (Ed.), Handbook of personal rela-

tionships (pp. 367–389). New York: John Willey.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.Ryan, A. M., Hicks, L. H., & Midgley, C. (1997). Social goals, academic goals, and avoiding seeking help in the

classroom. Journal of Early Adolescence, 17, 152–171.Savin-Williams, R., & Brendt, T. (1990). Friendship and peer relations. In S. Feldman & G. Elliott (Eds.), At the

threshold: The developing adolescent (pp. 277–307). Cambridge, MA: Harvard University Press.Schunk, D. H. (1991). Self-eYcacy and academic motivation. Educational Psychologist, 26, 207–231.Schunk, D. H., & Miller, S. D. (2002). Self-eYcacy and adolescents’ motivation. In F. Pajares & T. Urdan (Eds.),

Academic motivation of adolescents (pp. 29–52). USA: Information Age Publishing.Shulman, S., Laursen, B., Kalman, Z., & Karpovsky, S. (1997). Adolescent intimacy revisited. Journal of Youth and

Adolescence, 26, 597–617.Skaalvik, E. M. (1997). Self-enhancing and self-defeating ego orientation: relations with task and avoidance orien-

tation, achievement, self-perceptions, and anxiety. Journal of Educational Psychology, 89, 1–11.Steele, C. M. (1988). The psychology of self-aYrmation: sustaining the integrity of the self. In L. Berkowitz (Ed.),

Advances in experimental social psychology (Vol. 21, pp. 261–302). New York: Academic Press.Sullivan, H. S. (1953). The interpersonal theory of psychiatry. New York: Norton.Tesser, A. (2001). Self-esteem. In A. Tesser & N. Schwarz (Eds.), Blackwell handbook of social psychology: Intrain-

dividual processes (pp. 479–498). Malden, MA: Blackwell.Urdan, T. (1997a). Achievement goal theory: past results, future directions. In M. L. Maehr & P. R. Pintrich

(Eds.), Advances in motivation and achievement (Vol. 10, pp. 99–141). Greenwich, CT: JAI Press.Urdan, T. C. (1997b). Examining the relations among early adolescent students’ goals and friends’ orientation

toward eVort and achievement in school. Contemporary Educational Psychology, 22, 165–191.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.Vandel, D. L., & Hembree, S. E. (1994). Peer social status and friendship: independent contributiors to children’s

social and academic adjustment. Merrill-Palmer Quarterly, 40, 461–477.Weimer, B. L., Kerns, K. A., & Oldenburg, C. M. (2004). Adolescents’ interactions with a best friend: associations

with attachment style. Journal of Experimental Child Psychology, 88, 102–120.White, K. M., Speisman, J. C., Costos, D., Jackson, D., & Batris, S. (1989). Intimacy interview and scoring manual.

Boston University.

252 I. Levy-Tossman et al. / Contemporary Educational Psychology 32 (2007) 231–252

White, K. M., Speisman, J. C., Jackson, D., Batris, S., & Costos, D. (1986). Intimacy maturity and its correlates inyoung married couples. Journal of Personality and Social Psychology, 50, 152–162.

Youniss, J., & Smollar, J. (1985). Adolescent relations with mothers, fathers, and friends. Chicago, IL: University ofChicago Press.

Zarabatany, L., Conley, R., & Pepper, S. (2004). Personality and gender diVerences in friendship needs andexperiences in preadolescence and young adulthood. International Journal of Behavioral Development, 28,299–310.