Learning and Individual Differences 32 (2014) 1–8

Contents lists available at ScienceDirect

Learning and Individual Differences

j ourna l homepage: www.e lsev ie r .com/ locate / l ind i f

Academic identity, self-efficacy, and self-esteem predict self-determinedmotivation and goals

Meera Komarraju a,⁎, Christopher Dial b

a Southern Illinois University Carbondale, United Statesb Harvard University, United States

⁎ Corresponding author at: College of Liberal ArtsCarbondale, IL 62901-6502, United States. Tel.: +1 618 45

E-mail address: meerak@siu.edu (M. Komarraju).

http://dx.doi.org/10.1016/j.lindif.2014.02.0041041-6080/© 2014 Elsevier Inc. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 4 July 2013Received in revised form 8 January 2014Accepted 21 February 2014Available online xxxx

Keywords:Self-efficacyIdentitySelf-determined motivationGoalsIAT

We conducted the first tests of implicit academic identity (social versus studious) in relation to explicit academicidentity (social versus studious), self-efficacy and self-esteem as predictors of self-determined motivation andgoals (performance and learning). In Study 1, 366 undergraduates completedmeasures of implicit social/studiousidentity and implicit attitude towards social/studious domains along with explicit social/studious identity, aca-demic self-efficacy, and self-determinedmotivation. In Study 2, 128 undergraduates completed implicitmeasuresof social/studious identity, implicit self-esteem, and explicitmeasures of self-esteem, and goals (performance andlearning). Our results offer thefirst evidence of significant correspondence between implicit and explicit academ-ic identity and the unique, incremental contributions of implicitmeasures beyond explicitmeasures in explainingself-determined motivation and goal orientation. We also establish that socially oriented students have loweracademic self-efficacy, lower self-determined motivation, and a preference for performance goals; studiouslyoriented students report higher self-esteem and a preference for learning goals.

© 2014 Elsevier Inc. All rights reserved.

1. Introduction

College years offer a time to forge new identities. Although there aremany dimensions alongwhich students can shape their central identity,we examined two identities that may be particularly salient formotivation and goal striving: 1) ‘scholars’ invested in academic workand 2) ‘socializers’ invested in interpersonal and group socializing. Dothese identities influence students' self-efficacy, self-esteem and moti-vation?Do studentswho identifymore strongly as ‘scholars’have differ-ent goals and motivations than do ‘socializers’? We attempted toanswer these questions by examining college students' implicit andexplicit identities as ‘scholars’ or ‘socializers’ in relation to academicself-efficacy, self-esteem, self-determinedmotivation and goals (perfor-mance and learning).

2. Relevant prior research

2.1. Implicit and explicit academic identity

Implicit identities are typically hidden from conscious awareness,expressed more automatically, and are less susceptible to influencesexerted by impression management or social desirability (Baron &Banaji, 2006). In contrast, explicit identities tend to be within conscious

, Southern Illinois University,3 3543; fax: +1 618 453 3563.

awareness and likely to be expressedmore deliberately. Similarly, implic-it attitudes shaped by preverbal emotional experiences and establishedearly in life tend to be accurate predictors of longer term behavioral pat-terns, whereas explicit attitudes are shaped largely by cognitive contentand cultural norms and are better at predicting specific behaviors in im-mediate contexts (Baron & Banaji, 2006; Rudman, 2004). For instance, ithas been found that implicit responses, asmeasured by the Implicit Asso-ciations Test (IAT), predict natural or spontaneous behavior and explicitself-report ratings predict behavior that is more deliberate or controlled(Asendorpf, Banse, & Mücke, 2002).

Although implicit attitudes are moderately correlated with explicitattitudes (Nosek, Banaji, & Greenwald, 2002), some researchers consid-er them to be more independent and dissociated (Fazio, Sanbonmatsu,Powell, & Kardes, 1986). Research findings regarding the relativestrength of implicit and explicit measures in predicting behavior remaininconsistent and less is known about howandwhen they come togetheror diverge in predicting behavior. Hence, we investigated the relativecontributions of implicit and explicit academic preferences as predictorsof students' self-determined motivation and goal orientation.

2.2. Academic identity and motivation

Research regarding identity-based motivation and the influentialrole of possible selves (Oyserman & James, 2011; Oyserman, 2012;Oyserman & Destin, 2010) suggests that academic identities can bemo-tivating when an aspirational self feels relevant in the academic class-room and when individuals experience environments that match their

2 M. Komarraju, C. Dial / Learning and Individual Differences 32 (2014) 1–8

sense of self. For instance, when individuals believe their identity andactions are congruent they tend to persist longer at a challenging taskbecause they find it to be meaningful and significant (Oyserman &Destin, 2010). On the other hand, if individuals are performing actionsthat they believe are incongruentwith their identity they aremore like-ly to give upwhen they encounter difficulties. Thus, identity congruencehas the potential to influence students' academic choices (such as at-tending class) or behaviors (asking questions in class, or studying for atest). For example, Oyserman and Destin (2010) reported that althoughalmost 90% of students from urban schools in low-income areas seethemselves as being college educated, they do not have an education-dependent identity or a context that evokes such an identity; hence,they are less likely to pursue a college education. Similarly, Devos andCruz Torres (2007) found that Latino students tend to implicitly self-identify as low achievers when they have strong identification withtheir ethnic group and associate it with low-achieving stereotypes. Incontrast, Latino students implicitly identify as high achievers if theyhave strong identification with significant others who are viewed ashigh achievers. Thus, for Latino students, identification with academicendeavors is predicted by a function of their ethnic identity and the ac-ademic achievement of significant others.

Themotivating influence of academic identity is also apparent in re-sults showing that first generation college students obtain better gradesafter experiencing a match between their sense of self (independent orinterdependent) and the university's culture (which is more indepen-dent), even after controlling for SAT scores and race (Stephens,Fryberg, Markus, Johnson, & Covarrubias, 2012). Likewise, the negativeconsequences of identity mismatches are evident when women whoimplicitly associate math more strongly with men show a weaker pref-erence for math, less identification with it, less skill, less engagementandmore anxiety inmath-related activities (Nosek & Smyth, 2011). For-tunately, implicit identity appears to bemalleable, as womenwho comeinto contact with a role model with whom they can identify, feel moreconfident and efficacious, show increased class participation, seek helpfrom instructors after class, and reportmore awareness of Science, Tech-nology, Engineering, andMathematics (STEM) careers (Stout, Dasgupta,Hunsinger, & McManus, 2011). Thus, implicit identity measures appearto capture and reflect subtle and complex aspects of the self such as in-fluences from early experiences and messages received from family,teachers, or peers and these seem to have motivating influences.

2.3. Self-efficacy, self-esteem, self-determined motivation, and goals

In addition to the motivational influence of academic identity, re-searchers have also examined academic self-efficacy (academic self-confidence) and self-esteem (self-worth) in relation to academicmotiva-tion assessed as self-determined motivation and goals (performance andmastery). Although there is evidence of a robust relationship betweenself-efficacy and achievement, we need a more thorough understandingof the proximalmechanisms that explainhowandwhy self-efficacy influ-ences themotivation to pursue academic goals (Zimmerman, 2000). Self-determination theory explains how and why individuals may differ intheir academic motivation as it views motivation as a continuum withhigh intrinsicmotivation and amotivation on both ends and extrinsicmo-tivation located in themiddle (Deci & Ryan, 1985; Ryan&Deci, 2000). Ac-cording to this theory, by satisfying three innate needs that includefeeling competent by successfully meeting challenges, feeling autono-mous fromhaving choices, and feeling connected to others in caring rela-tionships, individuals experience a stronger sense of self-determination,‘agency,’ and greater intrinsic motivation (Deci, Vallerand, Pelletier, &Ryan, 1991; Deci & Ryan, 2000). Self-determination theory's basic theo-retical constructs are operationalized by the Academic Motivation Scale,and empirical evidence suggests that students who feel more competent(high academic self-efficacy) tend to experience greater self-determinedmotivation and show greater persistence; in contrast, amotivated stu-dents have the least self-determination and are more apathetic and

disengaged (Vallerand, 2000; Vallerand & Bissonnette, 1992; Vallerandet al., 1992). High academic self-efficacy and implicit beliefs about thechangeable/malleable quality of intelligence are associated with strongermotivation, greater concentration, and deeper processing of study mate-rial (Ommundsen, Haugen, & Lund, 2005). Individuals who internalizeachievement goals display greater self-regulation and self-directed be-haviors that are associated with enhanced learning and performance(Ryan & Deci, 2006). Further, priming individuals with relevant implicitintrinsic and extrinsic cues has been found to influence motivation forperformance behaviors (Levesque, Copeland, & Sutcliffe, 2008).

Regarding self-esteemandmotivation, although there is some confir-mation that a positive academic identity develops from prior successfulacademic experiences, there is an ongoing debate about the reciprocityof this relationship as evidence also exists in support of a positive self-concept leading to successful experiences (Wigfield & Wagner, 2005).Further, good performance in school has been shown to lead to higherself-esteem, whereas self-esteem has shown no impact on academicachievement suggesting that some aspects of this relationship are stillunclear (Baumeister, Campbell, Krueger, & Vohs, 2003). In addition,self-esteem alongwith family support appears to be important for learn-ing and achievement (Sergio, Cuestas, & Fenollar, 2008) and others hintat an indirect relationship between self-esteem and emotional/socialproblems which may lead to attrition (Crocker & Luhtanen, 2003).Thus, the relationship between self-esteem andmotivation remains rid-dled with questions that need to be answered more completely.

Research regarding sense of self and motivation measured via ac-ademic goals suggests that students are motivated by different cueswithin the same environment (Dweck & Leggett, 1988). For instance,individuals who feel competent (high self-efficacy) are driven byperformance–approach goals and those who feel less competentand worry that they will perform poorly are driven by perfor-mance–avoidance goals (Law, Elliot, & Murayama, 2012). Likewise,students pursuing performance goals tend to believe that thosewho have high ability do not need to work hard and having towork hard implies that a student is not very gifted. In contrast, stu-dents pursuing learning goals feel motivated to put forth effort andview this as a path for manifesting their ability (Elliott & Dweck,1988). Consequently, learning more about the complex way inwhich academic identity, academic self-efficacy, and self-esteemare related to self-determined motivation and goals (learning andperformance) would be valuable for educators as they seek to im-prove student motivation.

2.4. Current studies

In Study 1, our aim was to understand the nature and role of stu-dents' academic identity, attitude towards scholarly pursuits, and self-efficacy in predicting self-determined motivation. To achieve this, weperformed the first tests of ‘studious’ versus ‘social’ academic identityusing both implicit (automatic) and explicit (conscious) measures ofidentity and attitude in relation to self-determined motivation. Next,in Study 2, we built on the first study by including implicit and explicitmeasures of self-esteem alongwith implicit and explicit academic iden-tity and also sought a deeper understanding of the relationship betweenacademic identity and academic motivation by using a measure of goalorientation (performance and learning). Common to both studies, weexamined implicit and explicit identity as the predictors andmotivationas the outcome.We assessed implicit preferences and attitudes by usingthe Implicit Associations Test (IAT), a popular and widely used methodformeasuring the strength of implicit associations (Devos & Cruz Torres,2007; Greenwald, Poehlman, Uhlmann, & Banaji, 2009; Nosek et al.,2002). The IAT operates on the assumption that it is effective inaccessing information that is activated automatically or involuntarilyand is less vulnerable to attempts at faking, impression management,or self-deception. Comprehensive reviews evaluating implicit measuresincluding the IAT (Fazio & Olson; 2003; Fiedler, Messner, & Bluemke,

3M. Komarraju, C. Dial / Learning and Individual Differences 32 (2014) 1–8

2006) acknowledge that implicit measures are useful alternatives to di-rect self-report measures; yet, they also note that the theory underlyingthesemeasures needs to bemore fully developed, that the use of implic-it measures for diagnostic purposes is less compelling, and that a moreconvincing case needs to bemade regarding the automatic and noncon-scious aspect of implicit cognitions. Even though the IAT has provoked ahealthy debate, its psychometric properties have been established byevidence of test–retest consistency (Greenwald & Nosek, 2001), aswell as predictive and convergent validity (Greenwald et al., 2009). Fol-lowing the logic that an individual's associations regarding the self areintertwined within a network that is driven to maintain cognitive con-sistency (Greenwald et al., 2002), we hypothesized that in Study 1:

1. Students who implicitly identify as ‘studious’ rather than ‘social’ aremore likely to implicitly evaluate the related domain of ‘studious’rather than ‘social’, more positively.

2. Students who implicitly identify as ‘studious’ and show greater im-plicit preference for ‘studious’ are likely to report higher levels of ac-ademic self-efficacy.

3. Students who implicitly identify as ‘studious’ and show greater im-plicit preference for ‘studious’ are likely to report higher levels ofself-determined motivation.

Additionally, in Study 2, we hypothesized that:

1. Students who implicitly identify as ‘studious’ rather than ‘social’ aremore likely to prefer learning goals.

2. Students who implicitly identify as ‘social’ rather than ‘studious’ aremore likely to prefer performance goals.

3. Study 1

3.1. Method

3.1.1. ParticipantsRespondents were 366 undergraduates who participated for course

credit. The mean age was 19.38 years, 47% were men, majority wereEuropean American (56%) and African American (31%), 63.4% werefreshmen, 17.2% were sophomores, and 8.5% were juniors and seniors.

3.1.2. Measures

3.1.2.1. Implicit academic identity and attitude. We created two ImplicitAssociation Tests (IAT; Greenwald, McGhee, & Schwartz, 1998) forassessing implicit identity and implicit attitude involving discriminationbetween two target categories: Studious (library, reading, research, tak-ing notes) and Social (fraternity, partying, movies, hanging out). Wetested the association strength between these target categories andtwo sets of attribute categories. For identity, the attribute categorieswere:Me (I, mine, my, self) and Not Me (them, they, theirs, other). Forattitude, the attribute categories were: Good (joy, honest, triumph, rain-bow) and Bad (war, cancer, funeral, poison).

For the IAT identity test, the four categories of words were studious,social,me, and notme; and for the IAT attitude test, theywere studious, so-cial, good, and bad. In condition one, participants categorized together‘studious’ and ‘me’ by pressing one key and categorized together ‘social’with ‘not-me’ by pressing another key. Then, in condition two, partici-pants categorized together ‘social’ and ‘me’ by pressing one key and cat-egorized together ‘studious’ and ‘not-me’ by pressing another key. Thetwo conditionswerepresented in an inter-mixed order andweexaminedthe response latencies in these tasks. For each IAT test, there were sevenblocks of trials including three practice blocks and four blocks for data col-lection that included the critical conditions (Nosek et al., 2002). It was as-sumed that individuals would respond faster when the target categorieswere pairedwith attributes that aremore strongly associated inmemory.For example, an individual who strongly identifies with the academic as-pects of college would respond more quickly when ‘studious + me’

stimuliwere paired thanwhen ‘social+me’ stimuliwere paired. Similar-ly, if participants paired ‘social’with ‘good’, the participant had an implicitattitude that is more positive towards social versus studious.

To obtain an identity IAT score, we calculated the difference in thespeedwithwhich the participant responded to the ‘social+me’ and ‘stu-dious + not me’ blocks versus ‘social + not me’ and ‘studious + me’blocks expressed as an effect size ormodified Cohen's d. Difference scoresin the positive direction were interpreted as stronger implicit identitywith ‘social’ over ‘studious’. To obtain an attitude IAT score, we calculatedthe difference in the speed with which the participant responded to thesocial + good | studious+bad blocks comparedwith social+ bad | stu-dious + good blocks expressed as an effect size or modified Cohen's d(Greenwald, Nosek, & Banaji, 2003). Difference scores in the positive di-rection were interpreted as a stronger implicit preference for ‘social’compared with ‘studious’.

3.1.2.2. Explicit academic identity. Explicit academic identity was mea-sured by averaging across 4 items on a seven-point Likert-type scaleranging from strongly disagree to strongly agree. For the ‘studious’ iden-tity, the four items were, “In college, I ammost myself when I am in thelibrary, reading, working on research, or taking notes in a classroom,”and for the ‘social’ identity, the four items were, “In college, I am mostmyself when I am at a fraternity, partying, watchingmovies, or hangingout.” We created these items to match the words used in the IAT testsfor the superordinate categories of ‘studious’ and ‘social’. The Cronbach'sinternal consistency values for this study were .82 for explicit studiousidentity and .68 for explicit social identity.

3.1.2.3. Academic self-efficacy: the Academic Self-Concept Scale. Academicself-efficacy: the Academic Self-Concept Scale (Reynolds, 1988) is a wellestablished, psychometrically sound measure of academic self-efficacythat consists of 40 items assessing academic self-efficacy on a four-point Likert-type scale ranging from1= strongly disagree to 4= strong-ly agree (sample item, “All in all, I feel I am a capable student”). TheCronbach's internal consistency alpha value for this study was .92.

3.1.2.4. Self-determined motivation: the Academic Motivation Scale. Self-determined motivation: the Academic Motivation Scale (Vallerandet al., 1992) consists of 28 items and is rated on a seven-point Likert-type scale ranging from 1 = strongly disagree to 7 = strongly agree.An index of self-determined motivation is a weighted score that in-cludes the subscales of intrinsic, extrinsic, and amotivation and is com-puted such that a higher score indicates stronger intrinsic or self-determined motivation (Vallerand, 2011, personal communication). Inresponse to a question that asks participants why they attend college,participants rate their degree of agreement with items that representintrinsic motivation (For the pleasure that I experience in broadeningmy knowledge about subjects which appeal to me), extrinsic motiva-tion (In order to have a better salary later on), and amotivation or lackof motivation (Honestly, I don't know; I really feel that I am wastingmy time in school). The internal consistency Cronbach's alpha valuesfor the current study were .90 for intrinsic motivation, .88 for extrinsicmotivation and .84 for amotivation.

3.1.3. ProcedureParticipants were tested individually and responded to two IAT tests

assessing identity and attitude on the computer followedby a paper andpencil survey that included explicitmeasures assessing social and studi-ous identity, academic self-efficacy, and self-determined motivation.

3.2. Results and discussion

3.2.1. Correlation analysesWe found several interesting and meaningful patterns of rela-

tionships. For example, as shown in Table 1, students' implicit iden-tity as ‘social’ rather than ‘studious’ was positively correlated with

Table 1Study 1: Correlation analyses for implicit social identity, implicit attitude ‘social’ is good, academic self-efficacy, explicit studious identity, explicit social identity, and self-determinedmotivation.

Variable Implicit social N studiousidentity

Implicit social N studiousattitude

Academic self-efficacy Explicit studiousidentity

Explicit socialidentity

Self-determinedmotivation

Implicit social N studious identity –

Implicit social N studious attitude .51⁎⁎ –

Academic self-efficacy −.07 −.12⁎ –

Explicit studious identity −.19⁎⁎ −.23⁎⁎ .18⁎ –

Explicit social identity .31⁎⁎ .30⁎⁎ −.03 −.12⁎ –

Self-determined motivation −.17⁎ −.12⁎⁎ .44⁎⁎ .28⁎ .01 –

Mean −.01 −.02 2.87 3.84 5.19SD .32 .35 .39 1.31 1.08

⁎ p b .05.⁎⁎ p b .01.

4 M. Komarraju, C. Dial / Learning and Individual Differences 32 (2014) 1–8

an implicit attitude that favored social pursuits rather than scholarlypursuits (r = .51) and was negatively correlated with explicit self-determined motivation (r = − .17). Further, students who implicitlyevaluated the ‘social’ category more positively than the ‘studious’ cate-gory also reported lower explicit academic self-efficacy (r=− .12) andlower explicit self-determined motivation (r = − .12).

3.2.2. Regression analysesTo test the relative contributions of implicit and explicit measures in

predicting variance in self-determinedmotivation,we used hierarchicalregression analyses (see Table 2).We first entered the explicitmeasuresof identity (studious, social) and explicit academic self-efficacy in step1, and together, these predicted a significant amount (21%) of the vari-ance in self-determined motivation, F (3, 342) = 30.99, p = .00; aca-demic self-efficacy and explicit studious identity were the twosignificant predictors. Next, in step 2, we entered the implicit measuresof identity and attitude and together, these explained an additional 2%of incremental variance in self-determined motivation that was signifi-cant, F (2, 340) = 3.20, p= .04; the implicit identity measure emergedas the significant predictor. The negative beta weight for implicit iden-tity indicates that a stronger identification with the ‘social’ category isassociated with lower self-determined motivation.

3.2.3. Partial mediationTo gain a more complete and thorough understanding of the

complex link between our implicit and explicit measures and self-determined motivation, we examined academic self-efficacy as apotential mediator and found it to partially mediate the followingrelationships: a) between implicit attitude towards ‘studious’ versus‘social’ categories and self-determined motivation, as well as b) the re-lationship between explicit identity as ‘studious’ and self-determinedmotivation, using multiple regression analyses (Baron & Kenny, 1986).

As we had an adequate sample size and normally distributed data,we used the Sobel test ofmediation as it is recommended as having sat-isfactory power and an appropriate type I error rate (Baron & Kenny,

Table 2Study1: Hierarchical regression analyses with all explicit measures, implicit social Nstudious identity, and implicit social N studious attitude as predictors of self-determinedmotivation.

Factor Predictor Beta R2 Change in R2

Self-determined motivationStep 1 Academic self-efficacy .37⁎⁎

Explicit studious identity .19⁎⁎

Explicit social identity .07 .21 .21⁎⁎

Step 2 Implicit social N studious identity −.13⁎

Implicit social N studious attitude .004 .23 .02⁎

⁎ p b .05.⁎⁎ p b .01.

1986; Pituch, Whittaker, & Stapleton, 2005). Specifically, by includingacademic self-efficacy, the relationship between implicit attitude to-wards ‘studious’ versus ‘social’ and self-determined motivation was re-duced from − .12 to − .08, Sobel's test = −2.13, p = .03 (see Fig. 1).Similarly, by including academic self-efficacy, the relationship betweenexplicit identity as ‘studious’ and self-determined motivation was re-duced from .28 to .21, Sobel's test = 3.33, p = .00 (see Fig. 2).

3.2.4. DiscussionThe findings from Study 1 constitute the first test of implicit academic

identity as ‘studious’ and ‘social’ and also demonstrate the incrementalcontributions of implicitmeasures beyond explicitmeasures in explainingself-determined motivation. We have established that implicit measuresfaithfully capture orientations towards scholarly and social pursuits andalso predict students' explicit academic self-efficacy andmotivational ori-entation. Thus, students who implicitly subscribe to a ‘social’ identity aremore likely to implicitly endorse a positive evaluation of the social do-main. This provides verification for the logic that thoughts, feelings, andattitudes maintain consistency within a domain's network of implicitand explicit associations. Our findings also suggest that it is importantto assess both implicit and explicit identities as each provides unique in-formation about different aspects of the self.

Further, implicit social identity and explicit social identity were pos-itively correlated as were implicit social identity and attitude, and im-plicit social identity was a significant predictor of self-determinedmotivation. For example, studentswhohad a strong implicit social iden-tity also reported a stronger explicit social identity. Similarly studentswho implicitly identified as being more ‘social’ than ‘studious’ also val-ued social pursuits more than scholarly activities. Further, althoughthe associations were weak tomoderate in strength, the results suggestthat studentswhohad amore implicit social orientationwith greater in-vestments in their interpersonal and group activities also reported lessself-determined motivation (i.e. they were less intrinsically motivated)indicating that they were more likely to respond to external rewardsand punishments from the environment. Students implicitly evaluatingsocial pursuits more positively than scholarly pursuits reported loweracademic self-efficacy suggesting that they were less confident in theacademic domain. In contrast, students who implicitly and explicitlyidentified as being more ‘studious’ were also more likely to be self-assured in their academic competency and reported being more self-determined or intrinsically interested in schoolwork and motivated bya love of learning.

Educators might find it interesting and useful to know that eventhough students' explicit identification as being studious and feelinggreater academic self-efficacy predicted a significant amount of vari-ance in self-determination, implicit identity predicted a unique amountof incremental variance, even though it was by a small amount. Practi-cally speaking, it may be helpful to consider the value of recognizingthe potential importance of implicit cognitions in the academic context.

Academic Self-Efficacy

Implicit Attitude (‘Social’ Is Good) Self-DeterminedMotivation

.44-.12

-.12 (reduced to -.08 with mediator)

Fig. 1. Implicit attitude (‘social’ is good) and self-determined motivation partially mediated by academic self-efficacy.

5M. Komarraju, C. Dial / Learning and Individual Differences 32 (2014) 1–8

Thus, these findings are unique to this study and highlight the impor-tance of recognizing how students' automatically activated implicit cog-nitions might influence their feelings of competence and motivation.These results support previous work attesting to the unique contribu-tion of implicitmeasures as in the instance of implicit conscientiousnesspredicting incremental variance in academic achievement beyond thatexplained by explicitmeasures (Vianello, Robusto, & Anselmi, 2010). Al-though a majority of work in this area has examined explicit identityand attitudes, our results signal theneed to build on thiswork by furtherexploring implicit identity and attitudes.

Further, our understanding of the relationship among implicit iden-tity, attitudes and self-determined motivation is enhanced by resultssupporting themediating role of academic self-efficacy. Being confidentin one's ability to employ study strategies and learn effectively appearsto be a critical ingredient for feeling motivated and energized at school.This implies that students who implicitly value social pursuits overscholarly activities feel less self-motivated partly because they lackself-efficacy or confidence and have self-doubts about their skill andability to succeed at school. This raises the possibility that helping stu-dents develop an implicit scholarly orientation and strengtheningtheir self-efficacymight help increase their level of self-determinedmo-tivation. College administrators, academic advisors, and instructorscould provide resources and strategies for building students' self-efficacy such as opportunities for experiencing success, encouragement,positive rolemodels, and learning how tomanage performance anxiety.

Academ

Explicit ‘Studious’ Identity

.18

.28 (reduc

Fig. 2. Explicit ‘studious’ identity and self-determined mo

4. Study 2

4.1. Method

4.1.1. ParticipantsRespondents included 128 undergraduate students who participat-

ed for course credit. The mean age was 19 years, 56% were men, 57%were European American, 25% were African American, and 14% wereother; 72% were freshmen, 13% were sophomores, 10% were juniorsand 5% were seniors.

4.1.2. Measures

4.1.2.1. Implicit academic identity and implicit self-esteem. The two implic-itmeasures included one IAT assessing identity that involved discrimina-tion between two target categories: Studious (library, reading, research,taking notes) and Social (fraternity, partying, movies, hanging out). Wetested the strength of association between these target categories andtwo attribute categories: Me (I, mine, my, self) and Not Me (them, they,theirs, other). A second IAT, assessing attitude or self-esteem, involveddiscrimination between two target categories:Good (win, peace, heaven,cheer, amazing, love), and Bad (enemy, pain, stink, vomit, fear, wrong).We tested the strength of association between these target categoriesand same two attribute categories: (Me and Not Me).

ic Self-Efficacy

Self-DeterminedMotivation

.44

ed to .21 with mediator)

tivation partially mediated by academic self-efficacy.

6 M. Komarraju, C. Dial / Learning and Individual Differences 32 (2014) 1–8

The first IAT identity test was the same as in Study 1, and the twocritical conditions involved four categories of words: Studious, Social,Me, and Not me. For the second IAT test which assessed self-evaluation or self-esteem, the four categories were Me, Not me, Good,and Bad. As in Study 1, the conditions were presented in an inter-mixed order and we analyzed participants' response latencies to theIAT tasks based on the assumption that responses would be fasterwhen the target categories were paired with attributes that had stron-ger associations (Lane, Banaji, Nosek, & Greenwald, 2007). For example,an individual who had a strong identificationwith the academic aspectsof college would respond more quickly when Studious + Me stimuliwere paired than when Studious + Not Me stimuli were paired. Simi-larly, a student with higher self-esteem would respond faster whenGood + Me stimuli were paired than when Good + Not Me stimuliwere paired.

4.1.2.2. Explicit academic identity. This was assessed by asking partici-pants to select one item from a set of seven that ranged from, I am afar more social than studious person to I am a farmore studious than so-cial person.

4.1.2.3. Explicit self-esteem. The Self-Esteem Scale (Rosenberg, 1965)consisted of 10 items and used a five-point, Likert-type rating scaleranging from 1 = strong disagree to 5 = strongly agree (sample item,“I feel that I have a number of good qualities”); for this study, the inter-nal consistency Cronbach's alpha value was .85.

4.1.2.4. Explicit goal orientation. Drawing from Dweck and Leggett's(1988) scale, we used a single item to measure each goal orientation;performance goals (Although I hate to admit it, I sometimeswould rath-er do well in a class than learn a lot) and learning goals (It's muchmoreimportant for me to learn things in my classes than it is to get the bestgrades), and each of these was rated on a six-point, Likert-type scaleranging from1= strongly disagree to 6 = strongly agree.

4.1.3. ProcedureParticipants were tested individually; after responding to two IAT

tests on the computer, each participant completed a paper and pencilsurvey including explicit self-report measures of identity, self-esteem,learning and performance goal orientations.

4.2. Results and discussion

4.2.1. Correlation analysesStudentswho implicitly identifiedwith ‘social’ rather than ‘scholarly’

pursuits also expressed a stronger explicit identity as ‘social’ (r= .33), astronger preference for performance goals (r= .29), andweaker prefer-ence for learning goals (r=− .23) (see Table 3). Similarly, studentswitha stronger explicit social identity were more likely to pursue perfor-mance goals (r = .29). Regarding self-esteem, students with strongerimplicit self-esteem showed no preference for performance goals and

Table 3Study 2: Correlation analyses between the implicit and explicit measures of identity and self-e

Variable Implicit social N studiousidentity

Implicit self-esteemme + good

Implicit social N studious identity –

Implicit self-esteem me + good −.13 –

Explicit social N studious identity .33⁎⁎ −.16Explicit self-esteem −.03 .07Performance goal .29⁎⁎ −.07Learning goal −.23⁎⁎ .25⁎⁎

Mean .02 −.39SD .30 .34

⁎ p b .05.⁎⁎ p b .01.

a stronger preference for learning goals (r= .25); and students who re-ported a stronger explicit self-esteem were less likely to pursue perfor-mance goals (r = − .18).

4.2.2. Regression analysesTo test whether implicit measures of identity and self-esteem pre-

dicted incremental variance in performance goals, beyond explicit mea-sures of identity and self-esteem, we used hierarchical regressionanalyses and first entered the explicit measures of identity (studious, so-cial) and explicit self-esteem in step 1, and these predicted a significantamount (10%) of the variance in performance goals, F (2, 119) = 6.86,p = .00 (see Table 4). Next, in step 2, we entered the implicit measuresof identity (studious, social) and self-esteem, and these explained anoth-er 5% of incremental variance in performance goals thatwas significant, F(2, 117) = 3.04, p = .05. These analyses indicated that students whoexpressed amore social rather than scholarly identity, both at the explic-it and at the implicit level, preferred performance goals.

Next, we used hierarchical regression analyses to test whether im-plicit measures predicted incremental variance in learning goals, be-yond explicit measures. First, we entered the explicit measures ofidentity (studious, social) and explicit self-esteem in step 1, and theseexplained variance (3%) in learning goals and this was not statisticallysignificant, F (2, 119) = 1.64, p = .19 (see Table 5). Next, in step 2,we entered the implicit measures of identity and self-esteem andthese explained an additional 10% of incremental variance in learninggoals that was significant, F (2, 117)= 7.19, p= .00. These results indi-cate that only the implicit measures succeeded in detecting that stu-dents with a more social identity show a weaker preference forlearning goals and thosewith stronger implicit self-esteemprefer learn-ing goals.

4.2.3. DiscussionResults from Study 2 illuminated the relationship between explicit

and implicit identity, as well as implicit and explicit self-esteem, withregard to goal orientation and established the sensitivity of implicitmeasures in assessing a preference for learning goal orientation.Although the correlationswereweak tomoderate, studentswho explic-itly and implicitly identified as beingmore social (rather than studious)clearly preferredperformance goals. In contrast, studentswho implicitlyidentified as being more studious (rather than social) and alsoexpressed higher implicit self-esteem preferred learning goals. Thus,students who show a preference for social pursuits such as spendingtime with peers and friends are more likely to focus on performancegoals. However, students who implicitly think of themselves as beingmore studious andhold themselves in high regard seem to bemotivatedto maximize their learning. Their focus appears to be onmastery of ma-terial, rather than achieving the highest grade.

It is interesting that the results for performance goals were support-ed by both explicit and implicit measures, but for learning goals, onlythe implicit measures were significant predictors. In addition, for bothperformance and learning goals, implicit measures predicted significant

steem, performance goals and learning goals.

Explicit social N studiousidentity

Explicit self-esteem Performancegoals

Learninggoals

–

−.09 –

.29⁎⁎ −.18⁎ –

−.07 .14 −.30⁎⁎ –

.07 4.06 4.15 3.021.38 .62 1.24 1.08

Table 4Study 2: Hierarchical regression analyses with explicit social N studious identity, self-esteem and implicit social N studious identity, self-esteem as predictors of performancegoal orientation.

Factor Predictor Beta R2 Change in R2

Performance goalStep 1 Explicit social N studious .26⁎⁎ .10 .10⁎⁎

Explicit self-esteem −.16Step 2 Implicit social N studious .23⁎ .15 .05⁎

Implicit self-esteem .03

⁎ p b .05.⁎⁎ p b .01.

7M. Komarraju, C. Dial / Learning and Individual Differences 32 (2014) 1–8

incremental variance beyond explicit measures indicating the greatersensitivity of implicit measures in detecting identity-motivation inter-relationships. It appears that implicit measures capture an aspect ofstudents' academic identity that is unique and not reflected in explicitself-report measures. This greater sensitivity of implicit measures isparticularly true when it comes to learning goals as both implicit aca-demic identity (more studious than social) and implicit self-esteem sug-gest a stronger association with the motivation to learn and masterchallenging material. These results suggest that although studentsmay be reluctant to explicitly state their preference for learning goals,those who implicitly adhere to a scholarly identity and have higherself-esteem prefer mastery/learning goals.

Educators could address these implicit orientations by emphasizing,encouraging, and rewarding adherence to learning goals so that stu-dents feel more comfortable in acknowledging and pursuing learninggoals. Perhaps this might also help make sense of the negative directionof the correlation between explicit self-esteem and performance goalswhich suggests that students who consciously hold themselves inhigh regard show less preference for focusing on good grades only. Al-though the direction of causality between identity and motivation can-not be ascertained in these cross-sectional data, this is the first evidencethat college studentswith a social versus scholarly identity have a stron-ger preference for performance goals, students with stronger implicitself-esteem and an implicit scholarly identity prefer learning goals,and implicit measures capture unique aspects of academic identity.

5. Limitations, future directions, and conclusions

Taken together, the results of our studies establish the importance ofexamining both implicit and explicit academic identities as well as aca-demic self-efficacy in predicting college students' self-determinedmoti-vation and goal orientation. However, we also acknowledge somelimitations. Because we utilized a cross-sectional design, we cannotdraw conclusions of causality; as our samples included a majority offreshmen, we are cautious in generalizing our findings to upperclassmen; and as we used some single-item measures in Study 2, werecommend lengthier measures in future work. Further, although themajority of the research in this area is correlational in nature andmakes causal inferences inappropriate, future researchers could con-duct longitudinal studies that follow college cohorts starting with

Table 5Study 2: Hierarchical regression analyses with explicit social N studious identity, self-esteem and implicit social N studious identity, self-esteem as predictors of learning goalorientation.

Factor Predictor Beta R2 Change in R2

Learning goalStep 1 Explicit social N studious −.08 .03 .03

Explicit self-esteem .13Step 2 Implicit social N studious −.22⁎ .13 .10⁎⁎

Implicit self-esteem .24⁎⁎

⁎ p b .05.⁎⁎ p b .01.

baseline measures at college entry and track students' actual perfor-mance across the college years. Findings from such studies are likelyto help identify critical predictors of academic outcomes and demon-strate the relative strength of implicit and explicit identity in shapinglife goals, both scholarly and social. Such research could vastly improveour efforts to understand and improve studentmotivation and goal pur-suit. Instructors, educators, academic advisors, and college administra-tors could apply these findings in strengthening students' academicidentity, self-efficacy, and self-esteem, so that they remain motivatedand focused on goal attainment during their college years.

Acknowledgment

The authors would like to acknowledge the contributions of Profes-sorMahzarin Banaji and her Social Cognition Research Lab, Departmentof Psychology, Harvard University, in conducting our research studiesand preparing our manuscript.

References

Asendorpf, J., Banse, R., & Mücke, D. (2002). Double dissociation between implicit and ex-plicit personality self-concept: The case of shy behavior. Journal of Personality andSocial Psychology, 83, 380–393.

Baron, A. S., & Banaji, M. R. (2006). The development of implicit attitudes: Evidence ofrace evaluations from age 6 and 10 and adulthood. Psychological Science, 17, 53–58.

Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in socialpsychological research: Conceptual, strategic, and statistical considerations. Journal ofPersonality and Social Psychology, 51, 1173–1182.

Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003). Does high self-esteemcause better performance, interpersonal success, happiness, or healthier lifestyles?Psychological Science in the Public Interest, 4, 1–44.

Crocker, J., & Luhtanen, R. K. (2003). Level of self-esteem and contingencies of self-worth:Unique effects on academic, social, and financial problems in college students.Personality and Social Psychology Bulletin, 29, 701–712.

Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human be-havior. New York: Plenum Press.

Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs andthe self-determination of behavior. Psychological Inquiry, 11, 227–268.

Deci, E. L., Vallerand, R. J., Pelletier, L. G., & Ryan, R. M. (1991). Motivation and education:The self-determination perspective. Educational Psychologist, 26, 325–347.

Devos, T., & Cruz Torres, J. A. (2007). Implicit identification with academic achievementamong Latino college students: The role of ethnic identity and significant others.Basic and Applied Social Psychology, 29, 293–310.

Dweck, C. S., & Leggett, E. L. (1988). A social-cognitive approach to motivation and per-sonality. Psychological Review, 95, 256–273.

Elliott, E. S., & Dweck, C. S. (1988). Goal: An approach to motivation and achievement.Journal of Personality and Social Psychology, 54, 5–12.

Fazio, R. H., & Olson, M. A. (2003). Implicit measures in social cognition research: Theirmeaning and use. Annual review of psychology, 54(1), 297–327.

Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardes, F. R. (1986). On the automaticactivation of attitudes. Journal of Personality and Social Psychology, 50, 229–238.

Fiedler, K., Messner, C., & Bluemke, M. (2006). Unresolved problems with the “I”, the “A”,and the “T”: A logical and psychometric critique of the Implicit Association Test (IAT).European Review of Social Psychology, 17, 74–147.

Greenwald, A. G., Banaji, M. R., Rudman, L. A., Farnham, S. D., Nosek, B. A., & Mellot, D. S.(2002). A unified theory of implicit attitudes, stereotypes, self-esteem, andselfconcept. Psychological Review, 109, 3–25.

Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differ-ences in implicit cognition: The Implicit Association Test. Journal of Personality andSocial Psychology, 74, 1464–1480.

Greenwald, A. G., & Nosek, B. A. (2001). Health of the Implicit Association Test at age 3.Journal of Experimental Psychology, 48, 85–93.

Greenwald, A. G., Nosek, B. A., & Banaji, M. R. (2003). Understanding and using the Implic-it Association Test: I. An improved scoring algorithm. Journal of Personality and SocialPsychology, 85, 197–216.

Greenwald, A. G., Poehlman, T. A., Uhlmann, E. L., & Banaji, M. R. (2009). Understandingand using the Implicit Association Test: III. Meta-analysis of predictive validity.Journal of Personality and Social Psychology, 97, 17–41.

Lane, K. A., Banaji, M. R., Nosek, B. A., & Greenwald, A. G. (2007). Understanding and usingthe Implicit Association Test: IV. What we know (so far) about the method. In B.Wittenbrink, & N. Schwarz (Eds.), Implicit measures of attitudes: Procedures and con-troversies (pp. 59–102). New York: Guilford Press.

Law, W., Elliot, A. J., & Murayama, K. (2012). Perceived competence moderates the rela-tion between performance–approach and performance–avoidance goals. Journal ofEducational Psychology, 104, 806–819.

Levesque, C., Copeland, K. J., & Sutcliffe, R. A. (2008). Conscious and nonconscious process-es: Implications for self-determination theory. Canadian Psychology, 49, 218–224.

Nosek, B. A., Banaji, M. R., & Greenwald, A. G. (2002). Math = male, me = female, there-fore math ≠ me. Journal of Personality and Social Psychology, 83, 44–59.

8 M. Komarraju, C. Dial / Learning and Individual Differences 32 (2014) 1–8

Nosek, B. A., & Smyth, F. L. (2011). Implicit social cognitions predict sex differences inmath engagement and achievement. American Educational Research Journal, 48,1125–1156.

Ommundsen, Y., Haugen, R., & Lund, T. (2005). Academic self-concept, implicit theories ofability, and self-regulation strategies. Scandinavian Journal of Educational Research, 49,461–474.

Oyserman, D. (2012). Not just any path: Implications of identity-basedmotivation for dis-parities in school outcomes. Economics of Education Review, 33, 179–190.

Oyserman, D., & Destin, M. (2010). Identity-based motivation: Implications for interven-tion. The Counseling Psychologist, 38, 1001–1043.

Oyserman, D., & James, L. (2011). Possible identities. Handbook of identity theory and re-search (pp. 117–145). New York: Springer.

Pituch, K. A., Whittaker, T. A., & Stapleton, L. M. (2005). A comparison of methods to testfor mediation in multisite experiments.Multivariate Behavioral Research, 40(1), 1–23.

Reynolds, W. M. (1988). Measurement of academic self-concept in college students.Journal of Personality Assessment, 52, 223–240.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton Uni-versity Press.

Rudman, L. A. (2004). Sources of implicit attitudes. Current Directions in PsychologicalScience, 13, 79–82.

Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsicmotivation, social development, and well-being. American Psychologist, 55, 68–78.

Ryan, R. M., & Deci, E. L. (2006). Self-regulation and the problem of human autonomy:Does psychology need choice, self-determination, and will? Journal of Personality,74, 1557–1585.

Sergio, R., Cuestas, P. J., & Fenollar, P. (2008). An examination of the interrelationships be-tween self‐esteem, others' expectations, family support, learning approaches and ac-ademic achievement, Studies in Higher Education, 33, 127–138.

Stephens, N. M., Fryberg, S. A., Markus, H. R., Johnson, C. S., & Covarrubias, R. (2012). Un-seen disadvantage: How American universities' focus on independence underminesthe academic performance of first-generation college students. Journal of Personalityand Social Psychology, 102, 1178–1197.

Stout, J. G., Dasgupta, N., Hunsinger, M., & McManus, M. A. (2011). STEMing the tide:Using ingroup experts to inoculate women's self-concept in science, technology, en-gineering, and mathematics (STEM). Journal of Personality and Social Psychology, 100,255–270.

Vallerand, R. J. (2000). Deci and Ryan's self-determination theory: A view from the hier-archical model of intrinsic and extrinsic motivation. Psychological Inquiry, 11,312–318.

Vallerand, R. J. (2011). Personal Communication.Vallerand, R. J., & Bissonnette, R. (1992). Intrinsic, extrinsic, and amotivational

styles as predictors of behavior: A prospective study. Journal of Personality,60, 599–620.

Vallerand, R. J., Pelletier, L. G., Blais, M. R., Briere, N. M., Senecal, C., & Vallieres, E. F. (1992).The Academic Motivation Scale: A measure of intrinsic, extrinsic, and amotivation ineducation. Educational and Psychological Measurement, 52, 1003–1017.

Vianello, M., Robusto, E., & Anselmi, P. (2010). Implicit conscientiousness predicts aca-demic performance. Personality and Individual Differences, 48, 452–457.

Wigfield, A., & Wagner, A. L. (2005). Competence, motivation, and identity developmentduring adolescence. In A. J. Elliot, & C. S. Dweck (Eds.), Handbook of competence andmotivation (pp. 222–229). New York, N. Y. The Guilford Press.

Zimmerman, B. J. (2000). Self-efficacy: An essential motive to learn. ContemporaryEducational Psychology, 25, 82–91.