Achievement goals, study strategies, and achievement

Learning and Individual Differences 24 (2013) 1–10

Contents lists available at SciVerse ScienceDirect

Learning and Individual Differences journal homepage: www.elsevier.com/locate/lindif

Achievement goals, study strategies, and achievement: A test of the “learning agenda” framework Corwin Senko ⁎, Hidetoshi Hama, Kimberly Belmonte Psychology Department, State University of New York-New Paltz, New Paltz, NY 12561, US

a r t i c l e

i n f o

Article history: Received 2 March 2012 Received in revised form 14 October 2012 Accepted 22 November 2012 Keywords: Achievement goals Interest Learning strategies

a b s t r a c t Two classroom studies tested whether mastery-approach goals and performance-approach goals nudge students to pursue different learning agendas. Each showed that mastery-approach goals promote an interest-based studying approach in which students allocate study time disproportionately to personally interesting material over duller material. Study 2 showed that this approach can jeopardize their academic achievement. Conversely, performance-approach goals promote a vigilant approach in which students seek cues about how to succeed and allocate study time toward material they believe is most important to their instructors. Study 1 showed that this approach encourages flexibility in how deeply they study material, and Study 2 showed that it facilitates achievement for students who are accurate in their beliefs about which material is instructionally important. These findings counter the assumption that performanceapproach goals trigger a rigid reliance on superficial learning. They also therefore contribute to the broader discussion about when and why achievement goals affect achievement. © 2012 Elsevier Inc. All rights reserved.

1. Introduction Achievement goal theory has been a prominent theory of motivation in the past three decades. Much of its research highlights the effects of mastery versus performance goals. Students pursuing mastery goals strive to learn and develop their topic competence, while those pursuing performance goals strive to demonstrate their competence by outperforming peers (Dweck, 1986; Nicholls, 1984). These two broad classes of goals further divide along approachavoidance dimensions, thus allowing the possibility of a masteryavoidance goal to avoid a decline in skill or a misunderstanding of concepts, and a performance-avoidance goal to avoid doing worse than others (Elliot & McGregor, 2001). Generally, goal theory posits that the two avoidance goals produce negative outcomes, that performance-approach goals produce a mix of mildly negative and mildly positive outcomes, and that masteryapproach goals produce entirely positive outcomes. The research supports much of this premise (for extensive reviews, see Baranik, Stanley, Bynum, & Lance, 2010; Hulleman, Schrager, Bodmann, & Harackiewicz, 2010; Moller & Elliot, 2006; Payne, Youngcourt, & Beaubien, 2007; Senko, Hulleman, & Harackiewicz, 2011). Performance-avoidance (PAV) and mastery-avoidance (MAV) goals are regularly linked to anxiety, disinterest, disorganized study strategies, and poor achievement, for example. Performance-approach (PAP) goals, as theorized, do predict a few negative outcomes such ⁎ Corresponding author. Tel.: +1 845 257 3602; fax: +1 845 257 3474. E-mail address: [email protected] (C. Senko). 1041-6080/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.lindif.2012.11.003

as moderate levels of anxiety and poor sportsmanship, but also some positive outcomes such as high effort, persistence, and feelings of pride. Finally, mastery-approach (MAP) goals are routinely linked to high interest, effective self-regulation, cooperativeness, and various other benefits. Thus, a vast body of research supports the theory across many educational outcomes. There is, however, one noteworthy exception: academic achievement. A recent meta-analysis of approximately 100 studies (Hulleman et al., 2010) shows that PAP goals predict achievement (e.g., exam grades) more robustly than do MAP goals. 1 This finding contravenes the long-standing assumption that MAP goals match or surpass all other achievement goals in producing any desirable educational outcome. Various explanations for it are now emerging. This paper explores two of them. As this research question focuses squarely on the two approach goals, so too do this paper and the present studies. 1.1. Goals, learning strategies, and achievement One common explanation (Brophy, 2005), dubbed here the “depth of learning” explanation, is rooted to the learning strategies triggered by achievement goals. PAP goals often promote a surface learning strategy emphasizing rote learning and memorization, while MAP goals promote a deep learning strategy emphasizing elaboration and 1 Hulleman et al.'s (2010) meta-analysis revealed that MAP goals predict high achievement when the goal measure includes challenge-seeking or interest elements. When stripped of those confounding elements and defined purely in terms of learning and task mastery, however, MAP goals are unrelated to achievement.

2

C. Senko et al. / Learning and Individual Differences 24 (2013) 1–10

evaluation of course concepts (Moller & Elliot, 2006). According to the depth of learning explanation, PAP goals promote achievement because the surface learning strategies they trigger are wellmatched to the level of topic knowledge typically assessed by instructors. Were instructors to demand deeper topic knowledge, however, we might see that MAP goals are the stronger predictor of achievement because they trigger deeper engagement. This explanation rests on two assumptions. One is that each goal promotes a single learning strategy – a surface one for PAP goals, a deep one for MAP goals – in a largely fixed manner across different learning contexts. The other is that surface learning strategies can account for the PAP goal's relationships with achievement. In a recent review, however, Senko et al. (2011) found little empirical support for either assumption. 2 In contrast to the first, although MAP goals and PAP goals do regularly promote deep learning or surface learning strategies, respectively, they each also sometimes promote both of the learning strategies (e.g., Diseth, 2011; Koopman, Den Brok, Beijaard, & Teune, 2011; Liem et al., 2008; Vrugt & Oort, 2008). 3 In contrast to the second assumption, surface learning strategies typically fail to aid achievement or mediate PAP goal effects. Senko et al. (2011) therefore proposed an alternative explanation for goal effects on academic achievement. Their premise, dubbed here the “learning agenda” explanation, is rooted to the different standards that the two goals use to define success and failure. Attaining PAP goals requires outperforming peers, usually on teacher-set criteria (Elliot, 2005; Nicholls, 1984). Consequently, success demands they pursue their teacher's learning agenda—that is, to maintain vigilance for cues (e.g., teacher hints, study guides) about the topic knowledge which the teacher values and is most likely to assess, and then study that material faithfully. By contrast, attaining MAP goals requires satisfying either task-based standards or, more often than not, students' own subjective standards (i.e., feeling one has learned). These standards are inherently more flexible than those for PAP goals (Dweck, 1986), thus allowing students greater freedom to pursue their own learning agenda—that is, to chase their interests in the material. This, of course, is desirable, for it may lead to deeper mastery of the material they explore (Hidi & Renninger, 2006), but it can also jeopardize students' achievement if it leads them to neglect any course material that they consider uninteresting but that their teacher values and assesses. This may help explain why a deep learning strategy, so often linked to high achievement in the study strategies literature (Credé & Kuncel, 2008), has had uneven effects in the achievement goal literature: though it sometimes has aided mastery-focused students' achievement (e.g., Fenollar et al., 2007; Grant & Dweck, 2003; Liem et al., 2008), it often has not (e.g., Diseth, 2011; Elliot, McGregor, & Gable, 1999; Harackiewicz et al., 2000; Koopman et al., 2011; Senko & Miles, 2008; Vrugt & Oort, 2008). It may be that those 2 Senko et al. (2011) also explored whether MAP goals promote achievement on tasks that demand deeper topic knowledge. They identified a few classroom studies and several laboratory experiments that attempt to test this. In both settings, the findings were evenly mixed, with some studies showing MAP goal gains on deep knowledge tasks and others showing none. 3 Surface learning has been conceptualized in two ways. In the Student Approach to Learning tradition (Biggs, 1985; Entwistle, Tait, & McCune, 2000), it is conceptualized as maladaptive and is defined with elements of laziness and confusion (e.g., “Much of what I'm studying makes little sense: it's like unrelated bits and pieces.”). Such measures correlate positively with work-avoidance goals (described in Section 1.2) and either negatively or not at all with MAP and PAP goals (e.g., Diseth, 2011; Fenollar, Roman, & Cuestas, 2007). By contrast, in the Self-Regulated Learning tradition (Pintrich, 2004), surface learning is conceptualized more positively and is defined strictly in terms of memorization, without connoting confusion or laziness (e.g., “I memorize key words to remind me of important concepts in this class.”). Such measures correlate negatively with work-avoidance goals, positively with performance goals, and either positively (e.g., Wolters, 1998) or not at all with MAP goals (e.g., Harackiewicz et al., 2000). This second type is more common in achievement goal research and is integral to the depth of learning framework. We therefore adopt this conceptualization throughout the paper. Also, due to the mixed pattern of positive versus null correlations between MAP goals and this type of surface learning (Senko et al., 2011), we made no a priori hypotheses about this relationship in the present studies.

students sometimes are not deeply studying the “instructionally important” material. Senko and Miles (2008) provided initial evidence that masteryfocused students pursue their own learning agenda. In their study, MAP goals predicted a deep learning strategy, as usual, but also an interest-based studying approach characterized by allocating more study time toward the personally interesting material, even to the neglect of duller material. This latter approach in turn predicted low achievement in the course, thus demonstrating the risk of pursuing one's own agenda. The flip side of the learning agenda model – that PAP goals direct students toward their teacher's learning agenda – has not yet been tested. 1.2. Current research questions The present studies provide this test. They are guided by two basic research questions. First, do PAP goals arouse a vigilant approach marked by seeking and following cues about how to succeed? Second, is this vigilance beneficial? With regard to the first question, Senko and Miles (2008) defined vigilance as (a) calculating the importance of each course topic, for example, by actively seeking cues from the instructor, and (b) then allocating their study efforts accordingly.4 No existing measure fully captures this construct, but a few studies, viewed together, nevertheless hint strongly at high vigilance among students pursuing PAP goals. In particular, those students ask teachers questions aimed at identifying how to study (Shell & Husman, 2008), rely heavily on teacher guidelines when studying (Vermetten, Lodewijks, & Vermunt, 2001), and consider clarity about how to succeed one of the most desirable qualities in a teacher (Senko, Belmonte, & Yakhkind, 2012). With this foundation, we developed a measure of vigilance and, in the two current studies, tested whether it is promoted by PAP goals but not MAP goals. The second research question is even more vital. If PAP goals do promote a vigilant approach, does it pay dividends? There are two competing and equally plausible perspectives. On the one hand, vigilance could be a lazy, maladaptive strategy that leads students to rely excessively on teachers, self-regulate ineffectively, experience crippling confusion when encountering challenge, and perform poorly. On the other hand, it could be an effortful, adaptive strategy that leads students to shrewdly adopt their teachers' learning objectives, self-regulate effectively by monitoring their progress and adjusting their learning strategies to meet those objectives, and perform well as a result. No studies have tested this issue, yet a few provide suggestive evidence for the adaptive characterization. Poortvliet, Janssen, Van Yperen, and Van de Vliert (2007) found that participants assigned a PAP goal, but not those assigned a MAP goal, accurately discerned the quality and usefulness of suggestions given by a partner in a cooperative task. Likewise, trait-like competitiveness, an antecedent to PAP goals (Harackiewicz et al., 1997), correlates with an “achieving” learning style, which is characterized by high awareness of instructor demands and effective time management (Entwistle, 1988) and is conducive to academic achievement (Biggs, 1985). The current studies provide the first direct test of whether the vigilance spurred by PAP goals is truly adaptive. We rely on two general tactics to this end. One is to examine vigilance effects on a key aspect of self-regulation. Alongside planning and monitoring of one's progress, one hallmark of effective self-regulation is adjusting strategies to match task demands (Cantwell & Moore, 1996; Winne, 1995). Many studies (e.g., Wolters, 1998) have documented that MAP goals strongly, and PAP goals weakly, promote planning and monitoring, but no study, to our knowledge, has yet tested goal relationships with strategy 4 The word vigilant means to be alert or watchful for cues in the situation. Sometimes it is also used specifically to refer to alertness to threats, but we use it here in the neutral, more traditional sense,


3

(Harackiewicz et al., 1997; Wolters, 2003). We, too, included WAV goals in Study 1 because they provide a useful contrast for PAP goals. In particular, PAP goals should predict high vigilance but WAV and MAP goals should not; conversely, WAV goals should predict syllabus-boundness (Hypothesis 7) but PAP and MAP goals should not. For predictive validity, we tested if vigilance leads students to be flexible in their learning strategies, strategically adjusting them to the demands of the instructor and course, in which case PAP goals should predict high study flexibility by virtue of sparking a vigilant approach (i.e., an indirect effect; Hypothesis 8). We had no a priori hypothesis about whether MAP goals also promote study flexibility, though it is plausible insofar as they promote effective self-regulation more generally.

adjustment. If the vigilance aroused by PAP goals is truly a positive form of regulation, then it should nudge students to strategically alter the learning strategy as dictated by the demands of the course. Those students might, for example, rely on a surface learning strategy where possible but astutely adopt a deeper strategy when needed. If so, this would suggest that performance-focused students, due to their heightened vigilance, use a more strategic and flexible learning strategy than sometimes assumed by goal theorists (e.g., Kozlowski et al., 2001). Study 1 will test this possibility by assessing students' tendency to adjust their strategies to course demands. Another way to examine if vigilance is truly beneficial is to test its effect on achievement. The learning agenda framework assumes that vigilance aids performance-focused students' achievement—provided their vigilance leads to accurate appraisals of which course topics are most likely to be featured on the exam, that is. Study 2 will test this assumption.

2.1. Method 2.1.1. Participants We emailed all students at an American public university (N=6500) with a solicitation to voluntarily complete an online survey about their school motivation and study strategies. Of the 256 who replied, 163 students (85% female; 80% Caucasian; M age=23.4 years) completed the survey at mid-semester. This sample matched campus norms for all demographics except it included disproportionately more females.

2. Study 1 Study 1 examines if MAP goals predict a deep learning strategy (Hypothesis 1), PAP goals predict a surface learning strategy (Hypothesis 2), and WAV goals discourage deep and surface learning strategies (Hypotheses 3 and 4). This allows a context of replication in which to then test the learning agenda framework: in particular, if MAP goals predict an interest-based studying approach (Hypothesis 5) and PAP goals predict a vigilant approach (Hypothesis 6). To this end, we developed a measure of vigilance and tested its discriminant and predictive validity in multiple ways. For discriminant validity, we compared vigilance to syllabusboundness (Entwistle et al., 2000). Each captures adherence to cues from the teacher, and so they may be moderately correlated. But the similarities end there. Whereas vigilance is conceptualized as an effortful and adaptive form of regulation, syllabus-boundness is a rigid and lazy learning approach (sample item: “I tended to read very little beyond what was actually required to pass”) linked to ineffective self-regulation and poor performance (Entwistle et al., 2000). We also measured students' pursuit of work-avoidance goals. These goals are not technically achievement goals because, unlike MAP or PAP goals, they do not represent striving for competence (Elliot, 2005). They instead represent striving to satisfy task demands with minimal effort. Nonetheless, many studies have tested work-avoidance (WAV) goals alongside achievement goals, and they routinely link WAV goals to procrastination, lack of self-regulation, low effort, and poor achievement

2.1.2. Measures and procedure The survey was hosted on a secure website. Participants were instructed to select any one of their current courses and to complete the survey with that class in mind. All survey measures used 1 (strongly disagree) to 5 (strongly agree) scales. Table 1 lists their internal reliabilities. The survey comprised three parts. The first assessed students' achievement goals. MAP Goals (3 items; e.g., “I have been striving to understand the content as thoroughly as possible”) and PAP Goals (3 items; e.g., “My aim has been to perform well relative to other students”) were assessed with the Achievement Goal QuestionnaireRevised (AGQ-R; Elliot & Murayama, 2008). WAV Goals (3 items; e.g., “My aim in this class has been to get a passing grade with as little studying as possible”) were assessed with Shell and Husman's (2008) measure, slightly modified so that each item's opening stem used the goal-based language of the AGQ-R. The second part of the survey assessed participants' exam preparation strategies. These included the rehearsal (i.e., Surface Learning; 4 items; e.g., “I memorize key words to remind me of important concepts in this class”) and elaboration (i.e., Deep Learning 6 items;

Table 1 Zero-order correlations among all measures (Study 1). 1 1. 2. 3. 4.

MAP goal PAP goal WAV goal Surface learning 5. Deep learning 6. Interestbased studying 7. Vigilance 8. Syllabusboundness 9. Study flexibility M SD α

2

3

4

5

6

7

8

−.03 −.14

– .46

–

.37

.13

.35

.15

3.85 0.72 .79

3.08 1.11 .80

3.97 0.71 .90

3.34 0.81 .66

– .07 −.48 .17

– .02 .14

– −.27

.35

.09

−.30

.40

.20

.16

−.01

−.01

.32

.04 −.31

.25 .07

−.03 .42

.51 .17

.43 −.13

.26

−.10

−.14

.32

4.24 0.75 .76

3.81 0.97 .84

2.20 1.04 .82

3.59 0.86 .70

9

– – –

– 3.49 0.81 .89

Note: N = 163. MAP = mastery-approach. PAP = performance-approach. WAV = work-avoidant. Descriptives for both goals reflect untransformed goals, while correlations reflect normalized goal effects. Correlation coefficients >.16 are significant, p b .05. α = internal reliability for multi-item measures.

4


“I try to relate ideas in this subject to those in other courses whenever possible”) measures from the MSLQ (Garcia Duncan & McKeachie, 2005), as well as Senko and Miles's (2008) Interest-Based Studying measure (5 items; e.g., “I rarely get to the dry material because I spent so much time on the interesting material”). It also included a new measure of Vigilance capturing students' attempts to calculate the importance of different course topics for the exam and to allocate studying efforts to those topics (10 items; see Appendix A for full measure). The calculation and allocation aspects are united in a sequential link: students make these calculations in order to decide how to allocate their studying resources; likewise, to allocate resources requires first making calculations that guide the allocation decision. Additionally, as a point of comparison for the vigilance measure, the survey assessed Syllabus-boundness (4 items; e.g., “I concentrate on learning just those bits of information I had to know to pass”; Entwistle et al., 2000). Finally, for the third part of the survey, we developed a measure of Study Flexibility intended to capture if students tailor their studying approaches to match the demands of their courses (e.g., “My study technique has changed from class to class, based on how deeply the teacher wanted us to delve into the course material”; see Appendix A for full measure). This measure is modeled after Cantwell and Moore's (1996) “adaptive” flexibility measure, which assesses whether students adjust their approach primarily to satisfy their desire for challenge, intellectual curiosity, or enjoyment (e.g., “I often find the most interesting part of an assignment is in discovering new ways of tying my material together, and this often leads me to change the way I go about completing the task”). Concerned that those underlying motives may be more germane to some students than others, we constrained our measure to assess study flexibility strictly as a tool to meet the different demands of various courses and assignments, without any reference to challenge-seeking, enjoyment, or other motives. 2.2. Results and discussion 2.2.1. Construct validation The vigilance and study flexibility measures are both new. We examined the construct validity of each in two ways. One was to assess the structural integrity of each measure through confirmatory factor analyses (CFA) with maximum likelihood estimation, using AMOS 16.0 (Arbuckle, 2007). We used the following statistical criteria to evaluate model fit: a normed chi-square value (χ 2/df) below 3.0, a Comparative Fit Index (CFI) value above .95, a root-mean-square error of approximation (RMSEA) under .06, and a standardized root mean square (SRMR) under .05 together indicate a strong fit of the data with the model; a CFI above .90, a RMSEA under .08, and a SRMR under .08 indicate an adequate fit (Kline, 2005). In addition, for convergent and discriminant validity tests, we examined the bivariate correlations between the vigilance measure and other key measures (see Table 1). 2.2.1.1. Vigilance. As noted earlier (Section 2.1.2), we conceptualize vigilance as a single higher-order construct. A CFA of the 10 vigilance items showed that a single factor solution fit the data well, χ 2/df = 2.06, CFI = .95, RMSEA = .08, SRMR = .05. A two factor model distinguishing the five calculation items from the five allocation items fit the data similarly, χ 2/df = 2.12, CFI = .95, RMSEA = .08, SRMR = .05. We retained the single-factor solution for parsimony (Kline, 2005) and on theoretical grounds. This measure had high internal reliability (α = .90). The null correlation (r = − .05) between vigilance and interestbased studying shows that the two are unique instead of polar opposites. Additionally, although vigilance and syllabus-boundness are correlated (r = .46), they have diverging relationships with other key variables. In particular, vigilance correlated positively, while syllabus-boundness correlated negatively, with the deep learning

strategy (rs = .43 vs. − .13). As well, vigilance correlated positively with PAP goals (r = .25) but not WAV goals (r = − .03), while the opposite was true for syllabus-boundness (rs=.07 and .42, respectively). Thus, despite their shared external focus, vigilance and syllabusboundness are distinct from one another. 2.2.1.2. Study flexibility. A CFA showed that the expected single factor model for study flexibility fit the data well, χ 2/df = 1.45, CFI = .98, RMSEA = .05, SRMR = .04. The measure also showed high internal reliability (α = .89). Thus, the study flexibility measure was structurally sound. 2.2.2. Hypothesis testing We tested our key hypotheses next. Given the medium sample size and our desire to explore all possible effects involving the two new measures, we chose multiple regression over structural equation modeling for both studies (Kline, 2005). The MAP and PAP goals were negatively skewed and therefore were square-root transformed for the analyses (Cohen, Cohen, Aiken, & West, 2002). All interactions between goal terms were non-significant in preliminary analyses and therefore were omitted from the final analyses reported here. Table 1 provides the means and correlations for all variables. In this and the next study, all indirect effects (B) are bootstrap estimates (based on 5000 trials) tested with bias-corrected 95% confidence intervals (Hayes, 2012). Significant indirect effects are indicated when the confidence interval (CI) range excludes zero. 2.2.2.1. Effects on learning strategies. First, we sought to establish a context of replication by linking goals to surface learning and deep learning strategies. Table 2 provides the results. MAP goals predicted the use of the deep learning strategy (β = .26) and PAP goals predicted the use of the surface learning strategy (β = .15), thus supporting Hypotheses 1 and 2. WAV goals instead predicted low use of deep learning (β = − .18) and surface learning (β = − .25) strategies alike, thus supporting Hypotheses 3 and 4. We then tested the learning agenda framework. As shown in Table 2, in support of Hypotheses 5 and 6, MAP goals predicted use of an interest-based studying approach (β = .16) but not a vigilant approach, while PAP goals predicted use of a vigilant approach (β = .27) but not an interest-based studying approach. The WAV goal was unrelated to interest-based studying and vigilance. Then, in an effort to see whether the PAP goal effect on vigilance represents an adaptive effect, we also examined the goal relationships with syllabus-boundness, a maladaptive form of regulation. Table 2 shows that WAV goals, in accord with Hypothesis 7, predicted syllabus-boundness (β = .35). By contrast, PAP goals were unrelated, and MAP goals were marginally negatively related (β = − .15), to it. 2.2.2.2. Effects on study flexibility. The diverging effects on vigilance and syllabus-boundness illustrate that vigilance may reflect an adaptive form of regulation. To further test this possibility, we examined whether PAP goals, by virtue of prompting a vigilant approach, also encourage

Table 2 Regression analyses of achievement goal relationships with study strategies (Study 1).

MAP Goal PAP Goal WAV Goal R2

Deep Surface Learning learning

Interest-based studying

.26⁎⁎ .04 .08 .15⁎ −.18⁎ −.25⁎⁎ .15⁎⁎⁎ .10⁎⁎

.16⁎ .13 .07 .07⁎

Vigilance .01 .27⁎⁎ −.03 .07⁎

Notes: Values for goals are standardized regression coefficients (β). † p = .07. ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.

Syllabusboundness −.15† .07 .35⁎⁎⁎ .20⁎⁎⁎

C. Senko et al. / Learning and Individual Differences 24 (2013) 1–10 Table 3 Hierarchical regression analysis of goal and study strategy relationships with study flexibility (Study 1). Study flexibility Step 1: MAP goal PAP goal WAV goal ΔR2

.26⁎⁎ −.10 −.01 .08⁎

Step 2: Surface learning Deep learning Interest-based studying Vigilance Syllabus-boundness ΔR2

.13 .16 .08 .24⁎ .07 .19⁎⁎⁎

Notes: Values for goals and strategies are standardized regression coefficients (β). ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.

flexibility in students' studying. Specifically, we regressed study flexibility onto a hierarchical model comprising the three goals at Step 1 and the five different learning strategies at Step 2. This allowed us to first assess the goal effects directly and then to assess if vigilance provided the PAP goal the expected indirect pathway. Table 3 provides the results. Step 1 shows that students pursuing MAP goals reported being flexible in their study approach (β = .26). WAV and PAP goals were not directly related to study flexibility. However, in accord with Hypothesis 8, Step 2 of the model showed that students who were vigilant in their studying approach also reported being flexible in how they study (β = .24), and the indirect effect of PAP goals through vigilance was significant, B = .36, 95% CI [.06, 1.04]. Thus, MAP and PAP goals both predicted strategic flexibility in studying tactics: MAP goals directly, and PAP goals indirectly by virtue of triggering a vigilant approach.

2.3. Summary The learning agenda framework (Senko & Miles, 2008) posits that MAP goals spark not only a deep learning strategy (Payne et al., 2007), but also an interest-based studying approach characterized by prioritizing interesting over dull material. Study 1 supports this premise: MAP goals predicted the usage of deep learning and interest-based studying approaches alike. This framework also posits that PAP goals spark a vigilant approach in which students seek cues about how to succeed and follow those cues carefully, even if this requires strategically adjusting what and how they study. Study 1 provided the first test of this proposition and generally supported it. PAP goals predicted a vigilant approach, which in turn promoted flexibility in one's studying approach. This pattern, coupled with the fact that PAP goals were unrelated to syllabus-boundness, suggests that PAP goals foster a fairly positive regulatory profile. This contrasts with the view that these goals create rigidly shallow learners who are unable to self-regulate effectively (e.g., Kozlowski et al., 2001). Those negative qualities perhaps better characterize avoidance-oriented goals such as WAV goals, which in Study 1 promoted syllabusboundness but discouraged use of the surface and deep learning strategies. Finally, MAP goals also directly predicted study flexibility. Students pursuing these goals reported adjusting their studying approach to meet the demands of the course, the teacher's style, and type of test. This effect corresponds well with prior research showing that MAP goals promote active planning and monitoring efforts (e.g., Wolters, 1998).

5

3. Study 2 Study 2 provides a stronger test of the learning agenda framework. It does so in three ways. First, we measured goals and learning strategies early in the semester to rule out the possibility of students' class performance biasing their self-reports. Second, we measured utility value (i.e., finding personal meaningfulness in the material), an established catalyst for the development of students' individual interest in domains (Hidi & Renninger, 2006; Hulleman, Durik, Schweigert, & Harackiewiz, 2008). Senko and Miles (2008) found utility value to be a mediator of the MAP goal effect on interestbased studying and deep learning strategies. Including it in Study 2 allowed a stronger replication of their study. We expected MAP goals, but not PAP goals, to predict high utility value (Hypothesis 1). We also expected MAP goals to predict both a deep learning strategy (Hypothesis 2) and an interest-based studying approach (Hypothesis 3), as in Study 1, and that utility value would mediate these two effects (Hypotheses 4 and 5). We also expected to replicate Study 1 by finding that PAP goals predict both a surface learning strategy (Hypotheses 6) and a vigilant approach (Hypothesis 7). Third, we tested the impact that the goals and learning strategies have on students' exam performance. Based on prior research (Hulleman et al., 2010), we expected PAP goals to predict high grades (Hypothesis 8) but did not expect this of MAP goals. We also tested if the learning strategies affected academic achievement. In achievement goal studies, academic achievement tends to have a mix of null or negative links with surface learning strategies, and a mix of null or positive links with deep learning strategies (see Section 1.1 above). Accordingly, we made no firm hypotheses about either strategy promoting achievement. We did, however, expect the interest-based studying approach sparked by MAP goals to facilitate high achievement (Hypothesis 9), as in Senko and Miles (2008). More importantly, we tested if the vigilance triggered by PAP goals translates into high achievement. We reasoned that it would, provided that it is accompanied by an accurate appraisal of how likely the different course topics are to feature on the exam (Hypothesis 10). If accurate, the student will have studied the “right” material. If inaccurate, as might occur due to faulty judgments on the student's part or vagueness on the instructor's part, then vigilance is unlikely to translate into achievement gains. This last hypothesis, therefore, assumes an interaction effect (i.e., moderated mediation) between students' vigilance and the accuracy of their “instructional importance” judgments (Broekkamp et al., 2002). 3.1. Method 3.1.1. Participants Participants were 157 undergraduate students (64% female; 76% Caucasian; M age = 20.9 years) at an American public university. They were recruited from three courses, one in Psychology (n = 59), another in Chemistry (n = 44), and another in Business (n = 54). Students were offered a $5 incentive plus entry into a lottery to win one of several $25 gift certificates to local retailers. 3.1.2. Measures and procedure Students completed an online survey hosted on a secure website. It was accessible only within 36 hours before their class's first exam, thereby ensuring that most would have already begun exam preparations (see Broekkamp et al., 2002). Indeed, students reported having already studied for nearly 4 hours (M = 235 minutes), over halfway through their total anticipated study time (M = 445 minutes). The survey comprised the MAP goal and PAP goal measures first, followed by the surface learning, deep learning, interest-based studying, and vigilance measures. Each of these measures was identical to those from Study 1, though this time using the present progressive verb tense. In addition, we added a 3-item measure of utility value (“I find the content in this course personally meaningful,” “I can apply what

6


we are learning in this course to real life,” and “The topics we are covering in this class are important to my career plans”) to test if it mediates the MAP goal's relationships with interest-based studying and deep learning strategies. As shown in Table 2, each measure was sufficiently reliable. We also created a measure of instructional importance accuracy using Broekkamp et al.'s (2002) method. Without consulting their course notes or the textbook, students considered 20 different topics randomly selected from a textbook chapter relevant to their upcoming exam. They rated how likely each topic was to feature on the exam. Instructors did the same after the exam. This allowed us to create an intra-class coefficient for each student that assessed how well their importance judgments aligned with their instructor's (i.e., inter-rater reliability). Higher scores indicate greater accuracy. We expected this measure to moderate any effect of vigilance on exam grades, such that vigilance predicts grades only if students are successful in identifying the relative importance of each course topic. Given the survey's timing, we assumed students would be familiar with the course chapter listed. We checked this by having them rate their familiarity with each topic, from 1 (“not at all”) to 7 (“extremely”), prior to judging its instructional importance. Their reports confirmed our assumption: students indicated being “not at all” familiar with a topic in only 8% of the ratings; their average ratings indicate moderate to strong familiarity with the set of topics (M = 4.83, SD = .92). These familiarity ratings were uncorrelated with instructional important judgments, and will not be discussed further. Instructors provided participants' exam scores on a 0–100% scale. Because the Chemistry course had a lower average and wider range in scores than the other courses, we standardized each student's exam score relative to the course average before aggregating grade data.

number of classrooms to test multilevel effects, we, in accord with best practices (Raudenbush & Bryk, 2002), aggregated the data from the three courses and used two dummy codes as covariates to account for any differences between them. Table 4 provides the means and correlations for all variables. 3.2.1. Effects on utility value We regressed utility value onto a model comprising the MAP and PAP goals, both of which were first square-root transformed due to negative skew, plus, as control variables, the two class dummy codes. Table 5 shows that, in support of Hypothesis 1, students pursuing MAP goals reported high utility value (β=.20). Thus, utility value could serve as a mediator of MAP goal effects on the learning strategies. The two dummy codes were also significant predictors (βs=.27 and−.20), indicating that one class was more likely than the others' to report high utility value. 3.2.2. Effects on learning strategies To test the hypothesized effects on learning strategies, the regression model was expanded to include utility value at Step 2, thereby allowing us to examine if utility value mediates any MAP goal effects observed at Step 1. Table 5 provides the results. The results support most of the hypotheses concerning MAP goals. Matching Hypothesis 2, MAP goals predicted high usage of a deep learning strategy (β = .20), but in contrast with Hypothesis 3, these goals did not directly predict interest-based studying (β = .03). Based on prior research (Senko & Miles, 2008), however, we expected utility value to provide an indirect pathway for MAP goals to these two learning strategies. The results demonstrate this. Utility value predicted high usage of the deep learning strategy (β = .20) and interest-based studying (β = .21) alike, and, matching Hypotheses 4 and 5, utility value provided a significant indirect pathway for MAP goals to the deep learning (B = .25, 95% CI [.05, .60]) and interestbased studying (B = .26, 95% CI [.03, .74]) strategies. The results also support each hypothesis for PAP goals. Matching Hypotheses 6 and 7, PAP goals predicted high usage of a surface learning strategy (β = .26) and vigilant approach to studying (β = .22). Unexpectedly, PAP goals also predicted high usage of a deep learning strategy (β = .16).

3.2. Results and discussion We tested the hypotheses with three sets of regression analyses. Each regression model excluded the interaction between these two goals, time already spent studying for the exam prior to participation, and familiarity with each chapter topic, all of which were nonsignificant predictors in preliminary analyses. Also, lacking sufficient Table 4 Zero-order correlations among all measures (Study 2). 1 1. MAP goal 2. PAP goal 3. Surface learning 4. Deep learning 5. Interestbased studying 6. Vigilance 7. Utility value 8. Instructional Importance accuracy 9. Exam performance 10. Prior time studying M SD α

2

3

4

5

6

7

8

9

10

– .33 .27

– .25

.31

.11

.35

.05

.00

−.02

.35

.13 .26

.28 −.02

.40 .26

.08 .41

−.03 .20

– −.04

.05

.13

−.01

−.18

−.12

.11

−.12

.28

.21

.02

.15

−.14

.09

.14

−.05

.26

.08

.06

.03

.03

.07

.12

−.01

.07

–

5.73

5.48

5.38

5.21

4.00

5.10

5.10

.06

.12

233.95

1.13 .84

1.15 .87

1.00 .73

1.19 .76

1.19 .75

1.14 .88

1.27 .83

.15 –

0.92 –

185.49 –

– – –

– –

–

Note: N = 157. Descriptives for both goals reflect untransformed goals, while correlations reflect normalized goal effects. Correlation coefficients≥.16 are significant, pb .05. α=internal reliability for multi-item measures.


7

Table 5 Regression analyses of achievement goal relationships with utility value and study strategies (Study 2). Utility value ΔR2 Step 1: MAP Goal PAP Goal Class code 1 Class code 2 Step 2: Utility Value

Deep learning β

.22⁎⁎⁎

ΔR2

–

ΔR2

β

.27⁎⁎⁎ .20⁎ .01 .27⁎⁎ −.20⁎⁎

Interest-based studying

.20⁎⁎ .16⁎ .33⁎⁎⁎ −.17⁎

.03⁎⁎

.20⁎⁎

–

β

Vigilance

ΔR2

ΔR2

β

.19⁎⁎⁎

.01

04⁎

Surface learning

.03 .02 .13 .03 .21⁎

β

.13⁎⁎⁎ .14 .26⁎⁎ .11 −.25⁎⁎

.01

.07 .22⁎⁎ −.27⁎⁎ −.17⁎ .00

.13

−.02

Notes: Class codes 1 and 2 refer to the two dummy codes used to represent the three courses in the sample. Values for goals and classes are standardized regression coefficients (β). ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.

Finally, Table 5 shows various differences in utility value and learning strategies due to the three different classes, as represented by the two dummy codes. Together, they show that students in one class reported higher vigilance than those in the other two classes, that students in another class reported higher utility value and deep learning strategy usage than did those in the other two classes, and that students in one class reported less surface learning strategy usage than did students in the other two classes. All of these are to be expected, of course, given the different teaching styles and course content between the classes. From our vantage, what matters more is that the hypothesized goal effects still emerge when controlling for these course differences. 3.2.3. Effects on exam performance Last, we examined the links that achievement goals and their corresponding learning strategies have with exam performance. We

Table 6 Hierarchical regression analyses of effects on exam performance (Study 2). Exam performance Step 1:

ΔR2

.21⁎⁎ .18⁎⁎ .12 −.01 .11⁎⁎⁎

ΔR2

.06 .00

MAP goal PAP goal Class code #1 Class code #3

Step 2: Utility value

Step 3: Surface learning Deep learning Interest-based studying Vigilance Instructional importance accuracy Vigilance × importance accuracy ΔR2

−.19⁎⁎ .10 −.22⁎⁎⁎ .16⁎ −.05 .17⁎⁎ .10⁎⁎

Notes: Values for goals and learning strategies are standardized regression coefficients (β). ⁎ p = .06. ⁎⁎ p b .05. ⁎⁎⁎ p b .01. ⁎⁎⁎⁎ p b .001.

used a hierarchical model that mirrored those used above, with the goals and covariates entered at Step 1, utility value entered at Step 2, and the four learning strategies, Instructional Importance Accuracy, and the Vigilance × Instructional Importance interaction term all entered at Step 3. Table 6 provides the results. Matching Hypothesis 8, students pursuing PAP goals earned high grades on the exam (β = .18). So, unexpectedly, did those pursuing MAP goals (β = .21). No other terms were significant at Step 1. Nor was utility value significant at Step 2. At Step 3, however, several of the learning strategies predicted exam performance. Matching Hypothesis 10, interest-based studying predicted poor exam performance (β = − .22). This directly replicates Senko and Miles (2008). This effect also provided a small but significant negative indirect pathway for MAP goals to influence exam performance through the utility value → interest-based studying mediation sequence, B = − .04, 95% CI [− .15, − .01]. Vigilance also predicted marginally greater exam performance (β = .16), but this was qualified by the expected vigilance × instructional importance interaction effect (β = .17). Matching Hypothesis 10, simple slope analyses (Cohen et al., 2002) show that vigilance predicted high performance when participants were relatively accurate in their appraisals of the instructional importance of course topics (β = .33, p b .05), and, furthermore, vigilance provided a positive indirect pathway for those accurate students pursuing PAP goals, B = .29, 95% CI [.04, .77]. Vigilance, however, had no effect on exam performance when students' appraisals were inaccurate (β = −.02, p >.05). Finally, deep learning strategies were unrelated to achievement, while surface learning strategies predicted low exam performance (β = − .19). Surface learning strategies failed to provide a negative indirect effect for PAP goals to achievement, however, B = − .14, 95% CI [− .39, .00]. 3.3. Summary Overall, these findings replicate Study 1 and support the learning agenda framework. MAP goals again predicted use of a deep learning strategy, and although they did not directly predict interest-based studying, they did promote high utility value. Utility value, in turn, predicted use of a deep learning strategy as well as an interest-based studying strategy, thereby providing an indirect pathway for MAP goals to each strategy. Deep learning and interest-based studying strategies had null and negative effects on exam performance, respectively, thus mitigating the unexpected direct benefit of MAP goals on exam performance. PAP goals also directly predicted high achievement, as expected, and this effect was partly mediated by their impact on vigilance: students pursuing PAP goals adopted a vigilant approach

8


characterized by seeking and following cues about how to succeed, and this aided exam performance for those who were accurate in their appraisals about which course topics were instructionally important. PAP goals also predicted the surface learning strategy, which, as in prior research (see Senko et al., 2011), failed to parlay into exam success.

Hulleman et al., 2010). Thus, in Study 2, the two goals were similarly beneficial to achievement. When considering the learning strategy effects as well, however, we see that the interest-based studying approach weakened the MAP goal's benefit, while the vigilant approach enhanced the PAP goal's benefit. 4.1. Limitations and new directions

4. General discussion In contrast to early theorizing, PAP goals have been linked to high achievement more reliably than have MAP goals (Hulleman et al., 2010). This unexpected finding has triggered a lively discussion about when and why achievement goals aid achievement (Brophy, 2005; Midgley, Kaplan, & Middleton, 2001; Senko et al., 2011). The current studies contribute to this discussion. The predominant perspective has been the “depth of learning” framework (e.g., Brophy, 2005). It contends that academic achievement depends upon the fit between how deeply students learn, as ordained by their goals, and how deeply their knowledge is assessed by teachers. In particular, perhaps PAP goals often aid achievement simply because the surface-oriented learning strategies they spark is rewarded by instructors demanding only a superficial comprehension of course material. Were instructors to demand deeper comprehension, MAP goals, by sparking a deeper learning strategy, might aid achievement. As noted earlier (Section 1.1), the evidence for this framework is weak, especially its assumption that surface learning strategies explain why PAP goals facilitate achievement. The learning agenda framework tested here provides an alternative explanation for these goal effects. It, too, asserts that success depends on the match between students' learning and the knowledge tested by instructors, but instead of focusing on how deeply students study, it focuses on what they study. In particular, in accord with research on instructional importance (Broekkamp et al., 2002), this perspective assumes that students are most likely to perform well if they study the material valued and tested by instructors. PAP goals, due to their fixed and external standards for defining success, demand that students maintain a vigilant approach and be flexible and responsive to the task demands set forth by instructors. This approach, if done well, should aid achievement. By contrast, MAP goals, due to their internal and often subjective standards for defining success, may lead students to prioritize developing their interest. This approach, though noble in many ways, can at times hamper achievement if it leads students away from any material assessed by instructors. In sum, the two goals may spur students to pursue different learning agendas, with distinct consequences to their academic achievement. Our findings support both parts of this learning agenda framework. First, replicating Senko and Miles (2008), each study showed that MAP goals spark not only a deep learning strategy but also an interest-based studying approach that prioritizes studying personally interesting topics over the duller yet potentially important ones. This approach may interfere with those students' achievement, as shown in Study 2. Each study also showed that PAP goals spark not only a surface learning strategy but also a vigilant approach characterized by actively seeking out cues about the teacher's learning agenda and then allocating their study efforts accordingly. Furthermore, Study 1 showed that this vigilant approach promotes strategic flexibility in how students approach their learning, and Study 2 showed that this vigilance, when successful, may aid achievement. Finally, although our interest is in the learning strategies linking goals to academic achievement, it is worth noting that the PAP and MAP goals also each directly predicted achievement in Study 2. This was expected of PAP goals. But the MAP goal's effect was somewhat surprising because, though other studies have found it as well (e.g., Grant & Dweck, 2003; Wolters, 1998), it has been inconsistent and less reliable than the PAP goal's effect on achievement (see

We anticipate several useful new directions. Two concern methodology. The current studies, like most, relied on students' selfreports of their learning strategies. Given the potential inaccuracy of their reports (Winne, 2010), it is vital for future research to test the learning agenda framework and depth of learning framework with laboratory or internet procedures that simulate the studying process and provide behavioral indicators of the different learning strategies. They might, for example, trace students' actual memorization and elaboration efforts, as well as how they much they attend to contextual cues about which material to study closely or how much study time they allocate to topics of varying levels of interest. Such methods would allow a robust test of either theoretical framework. Along the same lines, our survey method prohibits strong causal interpretations of our findings. For example, in Study 1, the goals, learning strategies, and study flexibility were measured in sequence on the same survey. Our analyses retained this sequential logic by testing if goals promote strategy usage and if strategy usage promotes study flexibility. We believe this is the most sensible sequence based on the learning agenda framework as well as other models of motivation and learning strategies (e.g., Pintrich, 2004; Wolters, 1998). Nevertheless, there could be other ways to model these relationships. Similarly, although prior research shows that goal effects remain even when controlling for student ability, confidence, or personality (Senko et al., 2011; Steinmayr, Bipp, & Spinath, 2011), we cannot fully rule out their influence in the current findings. Thus, it will be useful in future research to test the causal dynamics, perhaps with experimental methods. It may also be useful to compare different types of PAP goals. These goals have historically been defined in two different ways (see Elliot, 2005). One is to focus on outperforming others, the other to appear talented. The two definitions can overlap insofar as outperforming peers is an effective way to appear talented. But they need not overlap. One can strive to outperform others simply because of pride and satisfaction; likewise, one need not outperform others to appear talented. Thus, theoretically at least, the two types of PAP goals are independent (Senko et al., 2011). They also appear to have different effects, according to Hulleman et al. (2010), whose metaanalysis showed that competition-based PAP goals promote a small positive effect on achievement, as in the present research, whereas appearance-based PAP goals promote a small negative effect. This pattern parallels recent work by Vansteenkiste, Mouratidis, and Lens (2010), who found stronger PAP goal benefits when the goals are pursued for autonomous reasons (e.g., enjoying the challenge of outperforming others, or finding normative success more personally worthwhile) than for controlling reasons (e.g., attaining rewards or meeting others' expectations of you). We suspect that appearancebased goals are often pursued for controlling reasons (e.g., to please a parent or teacher), whereas normative-based goals are pursued for a more diverse set of intrinsic and extrinsic reasons. If so, then we may find a less adaptive self-regulation profile for appearancebased PAP goals than for the competition-based PAP goals used in the current studies. Though our focus here has been on PAP goals and vigilance, the findings for MAP goals are noteworthy too. Replicating Senko and Miles (2008), both studies showed that MAP goals promote an interest-based studying approach characterized by prioritizing the interesting material. This effect was direct in Study 1 and indirect via utility value in Study 2. The latter also showed that interest-


based studying can sometimes jeopardize students' achievement. Of course, interest-based studying need not always undermine achievement. It may even benefit achievement if the topics students favor and prioritize also happen to be the ones that teachers consider most important; in such cases, mastery-focused students' personal learning agenda would also be their teacher's learning agenda. Along similar lines, it will be important to consider the role that course characteristics may have in these effects. The current studies were conducted in introductory and mid-level college courses. One wonders what effects would emerge in seminars and other advanced courses, which presumably require deeper topic comprehension than in the lower-level courses. The depth of learning framework and leaning agenda framework make different predictions. The depth of learning framework posits that MAP goals will aid achievement because they promote the deep learning strategy that is likely to pay dividends in those courses. It further assumes that PAP goals will trigger only superficial learning strategies that will likely disadvantage students in advanced courses. The learning agenda framework, by contrast, assumes that, for any type of class, success requires a match between the student's learning agenda and the teacher's. Thus, PAP goals, due to the vigilance and strategic flexibility they õencourage, should foster success in any type of class so long as teachers clearly articulate their learning agendas and provide sufficient clues about how to succeed (i.e., so students can make accurate “instructional importance” appraisals of course content). Lacking that, students pursuing these goals are unlikely to reap much benefit for their effort, regardless of whether the class is advanced or introductory. MAP goals, too, should enable success in any type of class so long as students' learning agenda matches their teacher's. This could in principle occur in introductory courses if the teacher inspires students to find interesting those topics that the teacher values and tests. Yet it is more likely to occur in seminars and other advanced courses in which students are provided sufficient autonomy to chase their interests. In such classes, their deep learning strategy will likely be rewarded with high achievement. Thus, both frameworks allow for the possibility that, in advanced courses, MAP goals might facilitate success and PAP goals might not. But their explanations differ greatly. Studies are needed to test these perspectives. 4.2. Conclusion and applied considerations The current findings further support the contention that MAP and PAP goal relationships with academic achievement are due in part to the learning agendas sparked by these goals. In particular, MAP goals trigger numerous educational benefits, but one of their potential risks is that, by nudging students to chase their own interests in the material, they may sometimes jeopardize students' achievement. By contrast, PAP goals trigger fewer overall benefits than MAP goals, but one of their vital benefits is that, by nudging students to strategically pursue the instructionally important material, they may sometimes facilitate students' achievement. These findings, when placed within the broader literature, support a “multiple goals perspective” that considers how the two goals can be beneficial in unique ways (Harackiewicz et al., 1997). This multiple goals perspective, like the traditional “mastery goal perspective” (Midgley et al., 2001), recognizes the many benefits of MAP goals and encourages teachers to try to cultivate a masteryoriented climate. It also, however, recognize that PAP goals can work quite well for students who naturally pursue these goals, and therefore advises teachers to avoid actively discouraging those students from pursuing PAP goals. We share this viewpoint and advocate that teachers clearly articulate their own learning agenda and also seek ways to align it with their students' own learning agendas, ideally in a way that demands deep topic knowledge for all students. We believe that this approach can serve students with either goal, enabling them to learn deeply and perform well.

9

Appendix A Vigilance Measure (Studies 1 and 2) Calculation Items 1. For each course topic, I tried to calculate how likely it was to be tested on the exam. 2. I tried to find out what information would be on the exam by paying attention to cues given by the lecturer. 3. The first thing I did was try to determine which topics were likely to be tested on the exam. 4. I tried to figure out what the professor thought was important because it gave me clues about which topics were tested on the exam. 5. I tried to understand the teacher's expectations so that I could be prepared for the test. Allocation Items 6. I focused more on material my teachers thought was important. 7. When doing an assignment or class project, I kept in mind what the instructor was likely to be looking for. 8. I tried to prioritize the topics that are most likely to be on the exam, regardless of how interesting I found them. 9. I made sure to allocate more study time to the topics that I believe the instructor considered most important. 10. I tried to figure out which material is most important for the exam so that I could spend most of my study time learning it. Flexible Studying Measure (Study 1 only) 1. I tried to change the way I study in order to fit the instructor's teaching style. 2. I tried to adjust my study strategies to match the demands of the teacher. 3. I tried to change the way I study in order to fit the course requirements. 4. I often found I changed from the way I prefer to study, in order to meet particular topic requirements. 5. My study technique changed from class to class, based on how deeply the teacher wanted us to delve into the course material. 6. I used different preparation strategies for multiple choice tests than I used for essay tests. 7. When I studied I tried to tailor my approach to the type of exam I was taking. 8. Even if I had a preferred study method, I adjusted my study approach to the format of the exam. 9. My studying approach depended a lot on the nature of the exam. References Arbuckle, J. L. (2007). Amos (Version 16.0) [Computer Program]. Chicago: SPSS. Baranik, L. E., Stanley, L. J., Bynum, B. H., & Lance, C. E. (2010). Examining the construct validity of mastery-avoidance achievement goals: A meta-analysis. Human Performance, 23, 265–282. Biggs, J. B. (1985). The role of metalearning in study processes. The British Journal of Educational Psychology, 55, 185–212. Broekkamp, J., van Hout-Wolters, B., Rijlaarsdam, G., & van den Bergh, H. (2002). Importance in instructional text: Teachers' and students' perceptions of task demands. Journal of Educational Psychology, 94, 260–271. Brophy, J. (2005). Goal theorists should move on from performance goals. Educational Psychologist, 40, 167–176. Cantwell, R. H., & Moore, P. J. (1996). The development of measures of individual differences in self-regulatory control and their relationship to academic performance. Contemporary Educational Psychology, 21, 500–517. Cohen, J., Cohen, P., Aiken, S. G., & West, L. S. (2002). Applied multiple regression/correlation analysis for the behavioral sciences (3rd Edition). Mahwah, NJ: Lawrence Erlbaum. Credé, M., & Kuncel, N. R. (2008). Study habits, skills, and attitudes: The third pillar supporting collegiate academic performance. Perspectives on Psychological Science, 3, 425–453.

10


Diseth, A. (2011). Self-efficacy, goal orientations and learning strategies as mediators between preceding and subsequent academic achievement. Learning and Individual Differences, 21, 191–195. Dweck, C. S. (1986). Motivational processes affecting learning. The American Psychologist, 41, 1040–1048. Elliot, A. J. (2005). A conceptual history of the achievement goal construct. In A. J. Elliot, & C. S. Dweck (Eds.), Handbook of competence and motivation (pp. 52–72). New York, NY: Guilford Publications. Elliot, A. J., McGregor, H. A., & Gable, S. (1999). Achievement goals, study strategies, and exam performance: A mediational analysis. Journal of Educational Psychology, 91, 549–563. Elliot, A. J., & McGregor, H. A. (2001). A 2 × 2 achievement goal framework. Journal of Personality and Social Psychology, 80, 501–519. Elliot, A. J., & Murayama, K. (2008). On the measurement of achievement goal: Critique, illustration, and application. Journal of Educational Psychology, 100, 613–628. Entwistle, N. (1988). Motivational factors in students' approaches to learning. In R. Schmeck (Ed.), Learning strategies and learning styles: Perspectives on individual differences (pp. 21–51). New York: Plenum Press. Entwistle, N., Tait, H., & McCune, V. (2000). Patterns of response to an approaches to studying inventory across contrasting groups and contexts. European Journal of Psychology of Education, 1, 33–48. Fenollar, P., Roman, S., & Cuestas, P. J. (2007). University students' academic performance: An integrative conceptual framework and empirical analysis. The British Journal of Educational Psychology, 77, 873–891. Garcia Duncan, T., & McKeachie, W. J. (2005). The making of the motivated strategies for learning questionnaire. Educational Psychologist, 40, 117–128. Grant, H., & Dweck, C. S. (2003). Clarifying achievement goals and their impact. Journal of Personality and Social Psychology, 85, 541–553. Harackiewicz, J. M., Barron, K. E., Carter, S. M., Lehto, A. T., & Elliot, A. J. (1997). Predictors and consequences of achievement goals in the college classroom: Maintaining interest and making the grade. Journal of Personality and Social Psychology, 73, 1284–1295. Harackiewicz, J. M., Barron, K. E., Tauer, J. M., Carter, S. M., & Elliot, A. J. (2000). Short-term and long-term consequences of achievement goals: Predicting interest and performance over time. Journal of Educational Psychology, 92, 316–330. Hayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable mediation, moderation, and conditional process modeling [White paper]. Retrieved from. http://www.afhayes.com/public/process2012.pdf Hidi, S., & Renninger, K. A. (2006). The four-phase model of interest development. Educational Psychologist, 41, 111–127. Hulleman, C. S., Durik, A. M., Schweigert, S. A., & Harackiewicz, J. M. (2008). Task values, achievement goals, and interest: An integrative analysis. Journal of Educational Psychology, 100, 398–416. Hulleman, C. S., Schrager, S. M., Bodmann, S. M., & Harackiewicz, J. M. (2010). A meta-analytic review of achievement goal measures: Different labels for the same constructs or different constructs with similar labels? Psychological Bulletin, 136, 422–449. Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd edition). New York: The Guilford Press. Koopman, M., Den Brok, P., Beijaard, D., & Teune, P. (2011). Learning processes of students in pre-vocational secondary education: Relations between goal orientations, information processing strategies and development of conceptual knowledge. Learning and Individual Differences, 21, 426–431. Kozlowski, S. W. J., Gully, S. M., Brown, K. G., Salas, E., Smith, E. M., & Nason, E. R. (2001). Effects of training goals and goal orientation traits on multidimensional training

outcomes and performance adaptability. Organizational Behavior and Human Decision Processes, 85, 1–31. Liem, A. D., Lau, S., & Nie, Y. (2008). The role of self-efficacy, task value, and achievement goals in predicting learning strategies, task disengagement, peer relationship, and achievement outcome. Contemporary Educational Psychology, 33, 486–512. Midgley, C., Kaplan, A., & Middleton, M. (2001). Performance-approach goals: Good for what, for whom, under what circumstances, and at what cost? Journal of Educational Psychology, 93, 77–86. Moller, A. C., & Elliot, A. J. (2006). The 2×2 achievement goal framework: An overview of empirical research. In A. Mittel (Ed.), Focus on educational psychology (pp. 307–326). NY: Nova Science Publishers, Inc. Nicholls, J. G. (1984). Achievement motivation: Conceptions of ability, subjective experience, task choice, and performance. Psychological Review, 91, 328–346. Payne, S. C., Youngcourt, S. S., & Beaubien, J. M. (2007). A meta-analytic examination of the goal orientation nomological net. The Journal of Applied Psychology, 92, 128–150. Pintrich, P. R. (2004). A conceptual framework for assessing motivation and self-regulated learning in college students. Educational Psychology Review, 16, 385–407. Poortvliet, P. M., Janssen, O., Van Yperen, N. W., & Van de Vliert, E. (2007). Achievement goals and interpersonal behavior: How mastery and performance goals shape information exchange. Personality and Social Psychology Bulletin, 33, 1435–1447. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods. Thousand Oaks, CA: Sage Publications, Inc. Senko, C., Belmonte, K., & Yakhkind, A. (2012). How students' achievement goals shape their beliefs about effective teaching: A “Build-A-Professor” study. The British Journal of Educational Psychology, 82, 420–435. Senko, C., Hulleman, C. S., & Harackiewicz, J. M. (2011). Achievement goal theory at the crossroads: Old controversies, current challenges, and new directions. Educational Psychologist, 46, 26–47. Senko, C., & Miles, K. M. (2008). Pursuing their own learning agenda: How masteryoriented students jeopardize their class performance. Contemporary Educational Psychology, 33, 561–583. Shell, D. F., & Husman, J. (2008). Control, motivation, and strategic self-regulation in the college classroom: A multidimensional phenomenon. Journal of Educational Psychology, 100, 443–459. Steinmayr, R., Bipp, T., & Spinath, B. (2011). Goal orientations predict academic performance beyond intelligence and personality. Learning and Individual Differences, 21, 196–200. Vansteenkiste, M., Mouratidis, A., & Lens, W. (2010). Detaching reasons from aims: Fair play and well-being in soccer as a function of pursuing performance-approach goals for autonomous or controlling reasons. Journal of Sport and Exercise Psychology, 32, 217–242. Vermetten, Y. J., Lodewijks, H. G., & Vermunt, J. D. (2001). The role of personality traits and goal orientations in strategy use. Contemporary Educational Psychology, 26, 149–170. Vrugt, A., & Oort, F. J. (2008). Metacognition, achievement goals, study strategies and academic achievement: pathways to achievement. Metacognition and Learning, 3, 123–146. Winne, P. H. (1995). Inherent details in self-regulated learning. Educational Psychologist, 30, 173–187. Winne, P. H. (2010). Improving measurements of self-regulated learning. Educational Psychologist, 45, 267–276. Wolters, C. A. (1998). Self-regulated learning and college students' regulation of motivation. Journal of Educational Psychology, 90, 224–235. Wolters, C. A. (2003). Understanding procrastination from a self-regulated learning perspective. Journal of Educational Psychology, 95, 179–187.