Optimal motivation in Peruvian high schools: Should learners pursue and teachers promote mastery goals, performance-approach goals or both?

Learning and Individual Differences 55 (2017) 87–96

Contents lists available at ScienceDirect

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

Optimal motivation in Peruvian high schools: Should learners pursue and teachers promote mastery goals, performance-approach goals or both? Lennia Matos a,⁎, Willy Lens b,1, Maarten Vansteenkiste c, Athanasios Mouratidis d a

Pontifical Catholic University of Peru, PUCP, Lima, Peru University of Leuven, Belgium Ghent University, Belgium d Hacettepe University, Ankara, Turkey b c

a r t i c l e

i n f o

Article history: Received 22 March 2014 Received in revised form 26 December 2016 Accepted 25 February 2017 Available online xxxx Keywords: Motivation Achievement goals Learning strategies Academic achievement Mastery/multiple goal perspective

a b s t r a c t Achievement goal theory is an important framework to understand students' achievement goals, motivation, and engagement in academic situations and to study teachers' instructional practices. There has been a debate about whether optimal motivation involves the pursuit of mastery goals only (i.e., mastery goal perspective) or the combined pursuit of mastery and performance-approach goals (i.e., multiple goal perspective; Barron & Harackiewicz, 2001, 2003). In the present correlational research we tested these two goal perspectives in two Peruvian samples of high school students (Sample 1: N = 1505; Sample 2: N = 551) and further examined whether students in classes, in which teachers were perceived to promote mastery goals only or performance-approach goals, would display the most optimal learning pattern. After controlling for learners' performance-avoidance goal pursuit, results provided only slim evidence for the additive goal perspective, as the effects of students' pursuit of mastery goals were more robust and consistent across both samples and outcomes (i.e., learning strategies and math grades). Along similar lines, at the class level, perceived teacher-promoted mastery goals positively predicted deep-level learning strategies, while class-level perceived teacher-promoted performance-avoidance goals related to lower academic achievement. © 2017 Elsevier Inc. All rights reserved.

1. Introduction Achievement goal theory has been extensively used to study students' motivation and achievement (Midgley, Kaplan, & Middleton, 2001; Pajares & Cheong, 2003). According to this theory, the intensity and quality of students' academic engagement is a function of the different purposes, or goals, students endorse when they engage in a specific learning task (Ames, 1992a, b; Anderman & Maehr, 1994; Dweck, 1986; Dweck & Leggett, 1988). Initially, two types of achievement goals were distinguished, that is, mastery goals (i.e., focusing on the development of competence and understanding), and performance goals (i.e., focusing on the demonstration of competence relative to others). Because both types of goals can be framed either as positive outcomes that can be approached or as negative outcomes

⁎ Corresponding author at: Av. Universitaria 1801, San Miguel, Lima 32, Peru. E-mail address: [email protected] (L. Matos). 1 Deceased.

http://dx.doi.org/10.1016/j.lindif.2017.02.003 1041-6080/© 2017 Elsevier Inc. All rights reserved.

that need to be avoided, Elliot and McGregor (2001) developed a 2 × 2 goal framework.2 In achievement goal theory, there has been a controversy regarding which type of achievement goal should be pursued by students and promoted by teachers in their class. Whereas some researchers suggest that students might better exclusively focus on the pursuit of mastery goals (i.e., mastery goal perspective; Midgley et al., 2001), others maintain that the additional pursuit of performance-approach goals is likely to yield incremental learning benefits (i.e., multiple goal perspective; Barron & Harackiewicz, 2001). The present research extends previous work in four significant ways. First, by considering the teacher-promoted achievement goals (i.e., goal structures), we were able to examine possible interactions between personally pursued and contextually promoted goals (Murayama & Elliot, 2009). Contrary to what it may be expected, only few studies

2 Recently, Elliot, Murayama, and Pekrun (2011) proposed a 3 × 2 model of achievement goals and they work around: task (approach/avoidance), self (approach/avoidance), and other (approach/avoidance). However, in this paper, we will focus in mastery (approach) goals and performance-approach and avoidance goals.

88

L. Matos et al. / Learning and Individual Differences 55 (2017) 87–96

have investigated the interplay between personally pursued and contextually promoted goals (Lau & Nie, 2008; Murayama & Elliot, 2009) and this is particularly true for the case of performance-avoidance goals (i.e., avoiding to perform more poorly than others). Therefore – and this is the second point - the present research extends previous work also by considering performance-avoidance goals, both at the personal and at the classroom level. Third, in addressing these questions, we adopted a methodologically more advanced method, that is, multilevel analysis. This approach allowed us to separate the unique contribution of learners' personally endorsed achievement goals at the between-student level and the achievement goals as perceived to prevail in their classroom (i.e., teacher-promoted) at the between-class level. Finally, this research extends previous work by examining different goal perspectives in two independent samples of Peruvian high school students, whereas most previous studies sampled white, middle class North American students (Kaplan, Middleton, Urdan, & Midgley, 2002). In sum, the present study examines the unique and interactive associations of personally endorsed and contextually promoted mastery and performance goals in the prediction of students' use of surface and deep-level learning strategies and achievement in a Mathematics course. Data from two independent samples were collected to test the consistency and replicability of the findings. 1.1. Achievement goals: mastery and performance goals Students pursuing mastery goals are focused on improving their competencies and gaining knowledge and understanding (Covington, 2000; Dweck, 1986; Heyman & Dweck, 1992; Zimmerman, 1994); to do so they use self-referential or task-based standards to evaluate their competence (Elliot, 2005; Senko & Harackiewicz, 2005). In contrast, students pursuing performance goals are focused on comparing favorably to others, either by demonstrating their superiority or by avoiding negative judgments about their competencies relative to others (Covington, 2000; Dweck, 1986; Heyman & Dweck, 1992; Maehr & Midgley, 1996; Zimmerman, 1994). When students adopt performance goals, they make use of normative or interpersonal standards to evaluate their competencies (Elliot, 2005; Senko & Harackiewicz, 2005). Because of the rather inconsistent effects of performance goals, Elliot and Harackiewicz (1996) suggested breaking down the traditional performance goals into two subtypes, that is, performance-approach goals in which the person aims to outperform others and to be among the best in the classroom and performance-avoidance goals in which the person tries to avoid bad judgments and protects oneself from being among the worst. Previous research found this bifurcation useful as it helped clarifying some of the observed inconsistent findings associated with the omnibus construct of performance goals (Harackiewicz, Barron, Tauer, Carter, & Elliot, 2000; Pintrich, 2000a, b, c). Indeed, there exists general agreement regarding the negative effects associated with the pursuit of performance-avoidance goals, which relate to higher test anxiety (Middleton & Midgley, 1997), greater use of self-handicapping strategies (Midgley & Urdan, 2001), and lower grades (Elliot & Church, 1997). Further, there is also general agreement regarding the positive effects of pursuing mastery goals (Kaplan & Maehr, 2007; Linnenbrink & Pintrich, 2002; Pintrich & Schunk, 2002). Mastery goals have been found to relate positively to intrinsic motivation (Elliot & Church, 1997), use of deep-level learning strategies (Meece, Blumenfeld, & Hoyle, 1988), cognitive engagement (Pintrich & Schrauben, 1992), higher levels of self-efficacy (Roeser, Midgley, & Urdan, 1996), and better academic achievement (Botsas & Padeliadu, 2003; Paulick, Watermann, & Nückles, 2013; Van Yperen, Blaga, & Postmes, 2014). These results were observed among both Western and non-western students (e.g., Matos, Lens, & Vansteenkiste, 2007). This line of research has shown an opposite pattern of associations for performance-avoidance goals.

In contrast to the clear-cut findings concerning performance-avoidance and mastery goals, performance-approach goals yielded a more mixed pattern. Although performance-approach goals have been found to relate positively to achievement in college students (Harackiewicz, Barron, Carter, Lehto, & Elliot, 1997; Harackiewicz, Barron, & Elliot, 1998), this was not necessarily the case for younger students (Paulick et al., 2013; Wolters, 2004). Further, performance-approach goals were found to be unrelated to deep-level learning in some studies (Elliot, McGregor, & Gable, 1999; Middleton & Midgley, 1997), but not in others (Al-Emadi, 2001; Pintrich, 2000a, b, c; Wolters, Yu, & Pintrich, 1996). Finally, performance-approach goals have been mostly predictive of superficial learning (Midgley et al., 2001), whereas they were unrelated to enjoyment and intrinsic motivation for learning (see Elliot, 2005). Achievement goal researchers have proposed different explanations for these mixed or inconsistent findings (see also Hulleman, Schrager, Bodmann, & Harackiewicz, 2010). 1.2. Multiple goal perspectives: making sense of inconsistent findings To address the question whether the pursuit of performance-approach goals yields learning benefits and if so under which circumstances, four different types of multiple goal perspectives have been proposed (Barron & Harackiewicz, 2001, 2003). These four multiple goal perspectives were contrasted with the mastery goal perspective, which suggests that the pursuit of mastery goals only promotes optimal learning. First, the additive goal hypothesis suggests that both mastery and performance-approach goals have a positive main effect on the same adaptive outcomes (e.g., Wolters et al., 1996). Second, the specialized goal hypothesis suggests that different types of achievement goals have different effects on different outcomes. While mastery goals would predict one type of outcomes (e.g., intrinsic motivation), performance-approach goals would relate to a different set of outcomes (e.g., achievement; Harackiewicz et al., 1997). Such specialized effects have not been systematically observed. For instance, some scholars have reported evidence for a positive (e.g., Linnenbrink, 2005; Vansteenkiste et al., 2004; Wolters et al., 1996) instead of a null-relation between mastery goals and achievement. In other instances, both performance-approach (e.g., Al-Emadi, 2001; Pintrich, 2000a, b, c) and mastery goals (e.g., Nelson, McInerney, & Craven, 2004) have been found to relate positively to the use of both surface and deep-level learning strategies. These rather inconsistent findings call for additional research. Third, the interactive goal hypothesis proposes that the combined pursuit of both mastery and performance-approach goals yields an additional positive effect on optimal learning that cannot be accounted for by the main effects of both achievement goals (i.e., the additive goal hypothesis). The studies by Bouffard, Boisvert, Vezeau, and Larouche (1995) and Pintrich (2000b) have been cited in this context, as these scholars found that students who simultaneously pursue both types of achievement goals obtained the highest scores on both academic achievement and learning strategies. Finally, the selective goal hypothesis suggests that the effects of achievement goals would depend on the type of goals that are salient in one's learning environment (e.g., match hypothesis). Classroom goal structures refer to the instructional practices, attitudes, values, and messages given by the teacher in the learning context (Urdan, 2007). Goal structures in classrooms would have an effect on students' cognitive engagement and academic achievement (Ames & Archer, 1988; Wolters, 2004), possibly in combination with learners' personally held achievement goals. For example, if the teacher emphasizes the pursuit of performance-approach goals, it would be more adaptive to pursue performance-approach goals, as a ‘match’ is created in this case. The matching hypothesis may explain why Harackiewicz et al. (1997, 1998) found performance-approach goals to predict achievement among college students, that is, because colleges are more competitive. However, in these studies, no measure of promoted achievement goals

L. Matos et al. / Learning and Individual Differences 55 (2017) 87–96

was included, a limitation that was remediated later by Barron and Harackiewicz (2003), Lau and Nie (2008), and Murayama and Elliot (2009). Although this matching hypothesis has been explored with respect to both mastery and performance-approach goals, the question is whether it can be generalized to maladaptive achievement goals (i.e., performance-avoidance goals). Indeed, Lau and Nie (2008) suggested that the maladaptive role of performance-avoidance goals may get exacerbated in a learning environment that is perceived to favor performance goal structures, a possibility that has received little empirical attention. Notably, the perceived contextual goals may not only reinforce the predictive power of learners' individual achievement goals, but may also play a counterbalancing role. For instance, a “dampening pattern” is observed when the perceived classroom context (e.g., performance goal structures) weakens the positive relation between a personal achievement goal (e.g., mastery goals) and a desirable outcome (e.g., interest). Alternatively, a “buffering” pattern occurs when an adaptive environment (e.g., mastery goal structures) weakens the relation between a personal achievement goal (e.g., performance-avoidance goal) and an undesirable outcome (e.g., test anxiety) (Lau & Nie, 2008).

1.3. Present research The broader aim of the present correlational research that involved two independent samples, was to investigate a set of hypotheses that can be derived from the mastery goal and the multiple goal perspective. In doing so, we jointly considered the role of personally endorsed and contextually promoted achievement goals. This was possible because a considerable number of classes (i.e., 61 and 27) were part of each sample. School students filled out the achievement goals measures with respect to Mathematics because it is a core course and it has been an important topic to perform research regarding achievement goals (Murayama & Elliot, 2009). Considering the fact that students need the adequate skills to process information (i.e., acquisition, storage, and usage) as well as strategies that help them to plan, monitor, and modify their cognitions and learning process (Pintrich & De Groot, 1990), we included a measure of both surface (i.e., rehearsal) and deep-level learning strategies (i.e., elaboration, organization, critical thinking, and metacognitive strategies). In addition, students' academic achievement was also included as an outcome. Based on the mastery goal perspective, we predicted that pursuing mastery goals at the personal level and, by extension, promoting these goals within a classroom (as perceived by students) would relate to adaptive learning, as indexed by the use of surface and deep-level learning strategies and better academic achievement. Some authors (e.g. Ames, 1992a, b) however, maintain that mastery goals would relate more to high quality learning involvement, as indexed by the use of deep-level learning strategies. Regarding the multiple goal perspective, we considered the four hypotheses that Barron and Harackiewicz (2001) formulated. According to the additive goal hypothesis, both mastery and performance-approach goals would have a positive contribution to the prediction of learning strategies and academic achievement, whereas the pursuit of performance-avoidance goals would be negatively related to such outcomes. Following the interactive goal hypothesis, the combined presence of both mastery and performance-approach goals would carry an additional benefit in predicting the use of learning strategies and academic achievement, not accounted for by the two main effects. To test this proposition, we examined two particular interactions that seem theoretically sound and practically important. We included the interaction between mastery goals and performance-approach goals (as these two goals share the approach dimension) and the interaction between performance-approach and performance-avoidance goals (as these two goals share the demonstrating competence dimension).

89

Regarding the specialized goal hypothesis, we inferred that mastery goals should relate to the use of deep-level learning strategies, while performance-approach goals should relate to the use of surface level learning strategies. Performance-avoidance goals, in contrast, would negatively predict deep-level learning strategies, while they would positively relate to the use of surface level learning strategies. Although there is inconsistency regarding academic achievement, Harackiewicz and colleagues (i.e. Harackiewicz et al., 1997) have found that performance-approach goals positively predicted academic achievement in college students; alternatively, performance-approach goals may detract from optimal learning as the learning process might be disrupted when one becomes too strongly concerned with evaluations. Finally, the selective goal hypothesis suggests that the effects of achievement goals would depend on the type of goals that are made salient in one's learning environment. One possibility is that a match between the personally endorsed and prevailing goal structure is critical, with students benefitting more from the pursuit of either mastery or performance-approach goals if a corresponding achievement goal is perceived to be highlighted by their teacher. Such an enhancing pattern is, however, unlikely to occur for the combination of personally endorsed and contextually promoted performance-avoidance goals (Lau & Nie, 2008). Alternatively, we also considered the possibility that some prevailing goal structures would weaken (i.e., attenuate) the desirable or protect (i.e., buffer) against the undesirable effects of specific achievement goals (Lau & Nie, 2008). For instance, the positive contribution of personal mastery goals may be undermined if one finds oneself in a performance-avoidance oriented classroom. 2. Method 2.1. Participants Participants in Sample 1 were 1505 high school students from 61 classrooms in 9 schools from Lima (Peru). Students from eighth(N = 538), ninth- (N = 565) and tenth-grade (N = 402) Mathematics classes participated in this research. There were 787 male and 717 female students (1 student failed to report gender). The mean age was 14.55 years (SD = 1.20). The next academic year, we returned to the same 9 schools and this time, we sampled 551 high school students (sample 2) from 27 Mathematics classes that were different from Sample 1. The Sample 2 was composed by 298 male and 253 female students attending eighth (N = 190), ninth (N = 190) and tenth (N = 171) grade. The mean age was 14.52 years (SD = 1.09). The Peruvian educational system (Basic Education) consists of three different stages. The first stage is the Pre-school level for children (up to 5 years old). The second stage is the Primary level, which lasts six years (from 1st to 6th grade). Finally, the Secondary level lasts five years (from grades 7th to 11th). 2.2. Procedure We contacted the principal of each school and we discussed the purpose of our research. All principals gave the approval to apply the questionnaires as they considered it an important source of information and as they included it as part of their school activities during the year. It is not usual that schools perform research projects so it was a good opportunity for them. Then we talked to the Math teachers and we entered the classes that allowed us to apply the questionnaires. All questionnaires were referred to Mathematics and were applied in class sessions and during regular Mathematics class hours in the last quarter of the academic year for both samples. We obtained the students' final grades in Mathematics at the end of the academic year from the teachers. As recommended by Midgley et al. (2000), the purpose of the study was explained to the students and the instructions were read aloud. It was also mentioned that the questionnaire did not constitute a test as if there were good or bad answers and that we were just

90

L. Matos et al. / Learning and Individual Differences 55 (2017) 87–96

interested in their real thoughts and experiences regarding the aspects we were asking about. An item example was given to help students understand how to answer Likert scale items. Students made questions if some issues required clarification and these were addressed during the application. We emphasized that their participation was voluntary and that they could withdraw from the study at any moment. Instead of writing down their names in the questionnaires, the researcher wrote a code to each questionnaire. This code was only known by the researcher and it was explained that the only purpose of this was to match the code with students' grades. No student denied participation. 2.3. Measures The instruments used for this research were the Patterns of Adaptive Learning Scales (PALS, 2000; Midgley et al., 2000) and the Motivated Strategies for Learning Questionnaire (MSLQ, Pintrich, Smith, Garcia, & McKeachie, 1991). Following translation and back-translation procedures, we translated the scales from English to Spanish and then we asked from a Spanish speaking expert to back-translate the scales. An expert researcher working with adolescents carefully read the translated items and suggested the rephrasing of some of them. Both, the PALS and the MSLQ are self-report instruments with a Likert-type scale that ranged from 1 (not at all true) to 5 (very true). 2.3.1. Students' achievement goals (PALS, 2000; Midgley et al., 2000) The mastery goal subscale measures students' purpose of developing competence and skills as well as knowledge and understanding (e.g., “In this Math class, it is important to me that I improve my skills this year”; 5 items). The performance-approach goal subscale assesses students' purpose of comparing favorably to others, of demonstrating their competence and superiority, and of outperforming others (e.g., “In this Math class, one of my goals is to show others that I'm good at my class work”; 5 items). The performance-avoidance goals subscale assesses students' purpose to avoid demonstration of incompetence (e.g., “One of my goals in this Math class is to avoid looking like I have trouble doing the work”; 4 items). Confirmatory Factor Analysis (CFA, LISREL 8.50, Jöreskog & Sörbom, 2001; Hu & Bentler, 1999) indicated that a three-factor model (mastery, performance-approach and performance-avoidance) yielded a good fit, both in Sample 1, χ2 (101, N = 1344) = 380.41, p b 0.001; RMSEA = 0.045; SRMR = 0.039, and Sample 2, χ2 (74, N = 527) = 160.84, p b 0.001; RMSEA = 0.047; SRMR = 0.039. Alpha coefficients for students' mastery, performance-approach and performance-avoidance ranged between 0.68 and 0.89 across samples, with an average alpha = 0.80. 2.3.2. Perceived goal structures Based on the PALS, we assessed the extent to which students perceived their Math teachers to stress the importance of engaging in academic activities to gain competence and understanding (i.e., mastery goal structures; e.g., “In this Math class, my teacher gives us time to really explore and understand new ideas”; 5 items), to demonstrate superiority compared to others (i.e., performance-approach goal structures; e.g., “In this Math class, my teacher lets us know which students get the highest scores on a test”; 3 items) and to avoid any demonstration of incompetence relative to others (i.e., performance-avoidance goal structures; e.g., “My Math teacher says that showing others that we are not bad at class work should be our goal”, 4 items). Confirmatory Factor Analysis (CFA, LISREL 8.50, Jöreskog & Sörbom, 2001; Hu & Bentler, 1999) indicated that a three-factor model (mastery, performance-approach and performance-avoidance) yielded a good fit, both in Sample 1: χ2 (51, N = 1387) = 123.01, p b 0.001; RMSEA = 0.032; SRMR = 0.032 and Sample 2: χ2 (51, N = 538) = 165.43, p b 0.001; RMSEA = 0.065; SRMR = 0.062. The internal consistency coefficients for students' perceptions of teachers' mastery, performance-approach, and performance-avoidance goal structures ranged between 0.61 and 0.79 across samples, with an average alpha = 0.70. To properly

use this scale as an index of the classroom environment, we aggregate students' responses that belonged to the same classroom. 2.3.3. Learning strategies (MSLQ, 1991, Pintrich et al., 1991) To measure surface level learning strategies, we considered rehearsal strategies which refer to practices used in basic memory activities, such as, reciting items to be learned (e.g., “When I study for this Math class, I practice saying the material to myself over and over”; 4 items). To assess deep-level learning strategies, we obtained the average scores of elaboration strategies (e.g., “When studying for this Math class, I try to relate the material to what I already know”; 6 items); organization strategies (e.g., “When I study for this Math course, I go over my class notes and make an outline of important concepts”; 4 items); critical thinking (e.g., “I try to play around with ideas of my own related to what I am learning in this Math course”; 5 items); and metacognitive strategies (e.g., “I ask myself questions to make sure I understand the material I have been studying in this Math course”; 10 items). Confirmatory Factor Analysis (CFA, LISREL 8.50, Jöreskog & Sörbom, 2001; Hu & Bentler, 1999) indicated that a two factor model (with rehearsal items loading on the surface learning strategies latent factor and the remaining items – that is, elaboration, organization, critical thinking, and metacognitive strategies loading on the deep-level learning strategies latent factor) yielded a good fit, both in Sample 1, χ2 (376, N = 1296) = 2391.75, p b 0.001; RMSEA = 0.064; SRMR = 0.044 and Sample 2, χ2 (376, N = 524) = 1935.27, p b 0.001; RMSEA = 0.089; SRMR = 0.061. Alpha coefficients for surface and deep-level learning strategies were 0.66 and 0.92 in Sample 1 and 0.69 and 0.93 in Sample 2, respectively. Two items from the original metacognitive strategies scale were dropped due to low corrected item-total correlation. Hence, 10 of the 12 original items were used in both samples. 2.3.4. Academic achievement We asked for the students' final grade of Mathematics according to the Peruvian school grade system that goes from zero to twenty (0− 20). We collected the scores that each teacher had given to each student at the end of the academic year (for both samples). 2.4. Plan of analyses Before testing our main hypotheses, we first inspected our measures through descriptive statistics, and bivariate correlations. We then examined the variance lying at the student level and classroom level so as to test whether a multilevel model was necessary. To test our hypotheses, we examined in three separate multilevel models to what extent surface, deep-level learning strategies, and grades could be explained by student-level and classroom-level predictors. At the student-level we included students' gender, students' personally endorsed achievement goals (i.e., mastery goals, performance-approach goals, and performance-avoidance goals) as well as the interaction between mastery and performance-approach goals and the interaction between performance-approach and performance-avoidance goals (after centering all the first-order predictors – see Cohen, Cohen, West, & Aiken, 2003). At the classroom level, aggregate scores of mastery, performanceapproach, and performance-avoidance goal structures were entered as contextual predictors of the three outcomes. These aggregated scores were created based on the scores of students belonging to the same class. These three classroom-level predictors were also tested as moderators of the slopes (i.e., the relations) between gender, the three personally endorsed achievement goals and the two two-way interactions (i.e., mastery by performance-approach goals interaction and performanceapproach by performance-avoidance goals interaction) on the one hand and the three outcomes (i.e., surface and deep-level learning strategies and grades) on the other hand. Given the multiple tests that we employed (i.e., 6 predictors at the student level × 3 predictors at the classroom level × 3 outcomes), we considered as noteworthy only those cross-level interactions which would be statistically significant

L. Matos et al. / Learning and Individual Differences 55 (2017) 87–96

at least at the alpha level of 0.01 and which would show some consistency (i.e., emerge as statistically significant in two out of the three studied outcomes). In line with the recommendations provided by Enders and Tofighi (2007), we group-mean centered all the variables at the student-level except gender, which was uncentered (−1 = males; 1 = females) so that the intercept to represent the average score for the whole sample. At the classroom-level the three variables were grand-mean centered. As concerns the slopes (i.e., the relations between student-level predictors and the outcomes), only those which were found to significantly vary from classroom to classroom were modeled as random. Fixing the slopes that were not statistically significant was opted for computation ease as the model contained six different slopes, (i.e., student-level predictors). 3. Results 3.1. Descriptive analyses and correlations Descriptive statistics (means and standard deviations) as well as bivariate correlations between the studied variables in Samples 1 and 2 are presented in Table 1. As it can be seen, there were significant point biserial correlations between gender and the other variables for both samples, meaning that gender will have to be controlled for in further analyses. Students' mastery, performanceapproach and performance-avoidance goals positively correlated to each other in both samples as well as to students' perceptions of mastery, performance-approach and performance-avoidance goal structures. Regarding, surface and deep-level learning strategies they were strongly, and positively, correlated to each other and to students' achievement goals and perceived goal structures in both samples (although their correlations to deep-level as compared to surface level learning strategies were higher in all cases). In both samples, academic achievement correlated positively with students' mastery goals and negatively with students' performance-approach goals and performance-avoidance goals. Also, in both samples academic achievement correlated negatively with perceived performance-avoidance goal structures. Only in Sample 2, academic achievement was negatively correlated with perceived performance-approach goal structures. In Sample 1, academic achievement correlated negatively with surface level learning strategies.

91

3.1.1. Primary analyses Inspection of the unconditional (i.e., no-predictor) model revealed that across both samples and outcomes, the amount of variance lying at the student level exceeded, as expected, the variance at the between-class level; yet the variance at the latter level was still meaningful (see Table 2). Given the considerable variance lying at the classroom-level, we proceeded to examine our main hypotheses through the full models described above. The results of the full model for the three outcomes for Samples 1 and 2 are shown in, respectively, Tables 3 and 4. 3.1.2. Student-level predictors At the student-level and after controlling for gender, both mastery and performance-approach goals positively related to the three studied outcomes in Sample 1 (however one of the effects of performance-approach goals was significant at the 0.05 level), while performanceavoidance goals related negatively to achievement. For Sample 2, mastery goals were the only consistent and positive predictor of all the three outcomes (at 0.01 level), while performance-approach goals were related positively to surface level strategies only (however it was at the 0.05 level). These findings provide evidence to the mastery goal hypothesis and regarding the multiple goal hypotheses it gives some evidence to the additive hypothesis. Moreover, out of the possible twelve interactions (i.e., the 2 two-way interactions × 2 samples × 3 outcomes), two of them only (i.e., 17%) were statistically significant (at the 0.05 level). Specifically, in Sample 2, mastery and performance-approach goals interacted in the prediction of deep-level learning strategies. A test of simple slopes revealed that the positive relation between mastery goals and deep-level learning strategies was stronger among students who favored less performance-approach goals (γ20 [at −1 SD in performance-approach goals] = 0.43, SE = 0.04, z = 11.44, p b 0.01) than among students who favored more performance-approach goals (γ20 [at +1 SD in performance-approach goals] = 0.28, SE = 0.05, z = 5.53, p b 0.01). Also, in Sample 2 the two-way interaction between performanceapproach goals and performance-avoidance goals in the prediction of grades was found to be statistically significant. Although this second interaction implies that students who pursue both performance-approach and performance-avoidance goals might have somewhat higher grades, a test of simple slopes indicated that the relation between performanceapproach goals and grades was statistically non-significant irrespective of whether students endorsed performance-avoidance goals weakly

Table 1 Means (M) and standard deviations (SD) and bivariate correlations of the measured variables for Sample 1 (above diagonal) and for Sample 2 (below diagonal). Mathematics

1

2

3

4

5

6

7

8

9

10

M

SD

1. Gendera Students' goals 2. Mastery 3. Pap goals 4. Pav goals Perceived goal structures 5. Mastery goal structures 6. Pap goal structures 7. Pav goal structures Outcome variables 8. Surface level LS 9. Deep level LS 10. Math grades

–

0.06⁎

−0.11⁎⁎

0.00

0.04

−0.01

−0.01

0.02

−0.00

0.05

–

–

0.04 −0.02 0.00

– 0.32⁎⁎⁎ 0.27⁎⁎⁎

0.31⁎⁎⁎ – 0.82⁎⁎⁎

0.17⁎⁎⁎ 0.63⁎⁎⁎ –

0.44⁎⁎⁎ 0.23⁎⁎⁎ 0.13⁎⁎⁎

0.07⁎⁎ 0.19⁎⁎⁎ 0.21⁎⁎⁎

0.19⁎⁎⁎ 0.50⁎⁎⁎ 0.52⁎⁎⁎

0.34⁎⁎⁎ 0.37⁎⁎⁎ 0.28⁎⁎⁎

0.46⁎⁎⁎ 0.44⁎⁎⁎ 0.31⁎⁎⁎

0.09⁎⁎ −0.07⁎ −0.17⁎⁎⁎

4.35 3.40 3.22

0.58 0.86 0.76

0.05 −0.00 −0.05

0.41⁎⁎⁎ 0.12⁎⁎ 0.17⁎⁎⁎

0.13⁎⁎ 0.25⁎⁎⁎ 0.54⁎⁎⁎

0.06 0.29⁎⁎⁎ 0.50⁎⁎⁎

– 0.07⁎ 0.10⁎⁎

0.08⁎⁎ – 0.41⁎⁎⁎

0.25⁎⁎⁎ 0.28⁎⁎⁎ –

0.32⁎⁎⁎ 0.14⁎⁎⁎ 0.37⁎⁎⁎

0.44⁎⁎⁎ 0.19⁎⁎⁎ 0.44⁎⁎⁎

0.03 −0.03 −0.22⁎⁎⁎

3.91 3.15 3.03

0.79 0.99 0.88

0.08⁎ 0.04 0.09⁎

0.37⁎⁎⁎ 0.50⁎⁎⁎ 0.19⁎

0.36⁎⁎⁎ 0.37⁎⁎⁎ −0.09⁎

0.29⁎⁎⁎ 0.30⁎⁎⁎ −0.09⁎

0.27⁎⁎⁎ 0.38⁎⁎⁎ 0.05

0.18⁎⁎⁎ 0.24⁎⁎⁎ −0.10⁎

0.29⁎⁎⁎ 0.37⁎⁎⁎ −0.21⁎⁎⁎

– 0.73⁎⁎⁎ −0.00

0.82⁎⁎⁎ – 0.05

−0.08⁎ −0.04 –

3.33 3.41 13.07

0.84 0.67 2.54

–

3.95 0.70

2.99 0.87

2.87 0.84

3.65 0.73

2.87 0.98

2.63 0.77

3.14 0.74

3.19 0.61

12.91 2.46

M SD

a Correlations with gender was dummy-coded as 0 = males; 1 = females and it was point biserial; LS = learning strategies; Pap = performance-approach; Pav = performanceavoidance. ⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.001.

92

L. Matos et al. / Learning and Individual Differences 55 (2017) 87–96

Table 2 Percentages of variance situated at the between-class and between-student level for the outcomes in both samples.

Variance lying at the Student level Classroom level Variance explained at the Student level Classroom level

Surface level learning

Deep-level Learning

Math grades

Sample 1

Sample 2

Sample 1

Sample 2

Sample 1

Sample 2

84.5% 15.5%

85.3% 14.7%

76.1% 23.9%

81.5% 18.5%

85.9% 14.1%

83.6% 16.4%

25.2% 81.2%

17.8% 73.5%

42.3% 80.5%

24.5% 81.1%

23.9% 64.5%

8.7% 31.5%

(γ30 [at −1 SD in performance-avoidance goals] = −0.11, SE = 0.26, z = − 0.41, p = 0.68), moderately (γ30 [at mean levels of in performance-avoidance goals] = 0.14, SE = 0.23, z = 0.60, p = 0.55), or strongly (γ30 [at + 1 SD in performance-avoidance goals] = 0.38, SE = 0.23, z = 1.68, p = 0.09). 3.1.3. Classroom-level predictors With respect to the classroom level, mastery goal structures as a classroom characteristic, positively predicted deep-level strategies in both samples (at Sample 1 at 0.01 level; at Sample 2 at 0.05 level), while they also positively predicted surface level learning strategies in Sample 1 (at 0.05 level). These findings suggest that students who belonged to classrooms which were perceived as being high in mastery goal structures reported using more learning strategies (especially deep-level). Interestingly, class-level performance-avoidance goal structures as a classroom feature positively predicted in both samples both surface and, rather unexpectedly, deep-level, learning strategies (but see supplementary analyses below). On the other hand, however, performance-avoidance goal structures as a classroom feature negatively predicted grades in both samples. The explained variance at the classroom level is presented in Table 2.

Table 3 Surface and deep-level learning strategies and math grades as a Function of students' gender and achievement goals and class-level goal structures for Sample 1. Surface-level strategiesa

Deep-level strategiesa

Math gradesb

Fixed effects

B

SE

B

B

Intercept Student-level predictors Gender Mastery goals (M) Performance-approach goals (Pap) Performance-avoidance goals (Pav) M × Pap goals Pap × Pav goals Classroom-level predictors Mastery (M) goals structures Pap goals structures Pav goals structures Random effects Intercept u0j M goals slopes Pap goals slopes Pav goals slopes Pap × Pav goals slopes Level 1rij

3.30

(0.03) 3.37

(0.02) 13.20

(0.11)

0.01 0.32⁎⁎ 0.15⁎⁎

(0.02) −0.02 (0.02) 0.12 (0.04) 0.39⁎⁎ (0.03) 0.78⁎⁎ (0.03) 0.16⁎⁎ (0.03) 0.34⁎

(0.08) (0.14) (0.14)

0.07

(0.05) 0.02

SE

SE

(0.03) −0.48⁎⁎ (0.14)

−0.02 (0.04) −0.01 (0.03) 0.08 0.01 (0.03) 0.03 (0.03) 0.31

(0.14) (0.15)

0.16⁎

(0.07) 0.23⁎⁎

(0.06) 0.10

(0.30)

0.09 0.51⁎⁎

(0.07) 0.08 (0.06) 0.48⁎⁎

(0.06) 0.54 (0.31) (0.04) −1.39⁎⁎ (0.26)

0.02⁎⁎ – – 0.02⁎

0.02⁎⁎ 0.02⁎ –

0.32⁎⁎ – 0.14⁎⁎

0.52

0.24

0.25⁎⁎ 4.84

⁎ p b 0.05. ⁎⁎ p b 0.01. a Based on N = 1505 students (belonging to 61 math classes). b Based on N = 859 students (belonging to 43 classes).

Table 4 Surface and deep-level learning strategies and math grades as a function of students' gender and achievement goals and class-level goal structures for Sample 2.a Surface learning

Deep level strategies

Math grades

B

B

Fixed effects

B

SE

Intercept Student-level predictors Gender Mastery goals (M) Performance-approach goals (Pap) Performance-avoidance goals (Pav) M × Pap goals Pap × Pav goals Classroom-level predictors Mastery (M) goals structures Pap goals structures Pav goals structures Random effects Intercept u0j Pav goals slopes M × Pap goals slopes Level 1rij

3.11

(0.04) 3.17

(0.03) 13.00

SE

(0.19)

0.06 0.34⁎⁎ 0.16⁎

(0.03) 0.02 (0.05) 0.35⁎⁎ (0.07) 0.08

(0.02) 0.19 (0.04) 0.96⁎⁎ (0.05) 0.14

(0.10) (0.18) (0.23)

0.00

(0.07) 0.01

(0.06) −0.31

(0.23)

−0.01 (0.08) −0.11⁎ (0.04) 0.02 −0.02 (0.04) 0.00 (0.03) 0.28⁎

SE

(0.20) (0.13)

0.18

(0.11) 0.20⁎

(0.09) 0.31

0.03 0.56⁎⁎

(0.11) 0.07 (0.13) 0.50⁎⁎

(0.10) 0.26 0.55 (0.10) −1.82⁎⁎ 0.62

0.02⁎⁎ – 0.06⁎⁎

0.01⁎⁎ 0.02⁎ – 0.24

0.40

0.50

0.67⁎⁎ – – 4.79

⁎ p b 0.05. ⁎⁎ p b 0.01. a Based on N = 551 students (belonging to 27 math classrooms).

3.1.4. Cross-level interactions Concerning the cross-level interactions between the six studentlevel predictors (i.e., gender, the three personally endorsed achievement goals, and the two-way interactions between mastery and performance-approach goals and between performance-approach and performance-avoidance goals) and the three class-level predictors (i.e., class-level mastery, performance-approach, and performanceavoidance goal structures), only five cross-level interactions for Sample 1, and three cross-level interactions for Sample 2 (out of 54 possible interactions for each sample;14.81%) were statistically significant at the 0.05 level (not shown in Tables 3 and 4); among them, only three were statistically significant at the 0.01 level (5.55%). The absence of systematic cross-level interactions between personally endorsed and context-promoted achievement goals failed to provide support to the selective goal hypothesis. The only consistent finding concerned the cross-level interaction between personally endorsed performance-approach goals and class-level performance-avoidance goal structures. Specifically, perceived teacherpromoted performance-avoidance goals moderated the positive relation between performance-approach goals and surface (γ33 = 0.22, SE = 0.07, p b 0.01) and deep-level learning strategies (γ33 = 0.17, SE = 0.06, p b 0.01). A test of simple slopes for surface and deep-level learning strategies revealed that the relation between performance-approach goals and surface as well as deep-level learning strategies was positive among students belonging to classrooms which were characterized as being high (i.e., +1 SD) in performance-avoidance goal structures (respectively, γ30 [at +1 SD in performance-avoidance goal structures] = 0.34, SE = 0.08, z = 4.48, p b 0.01 and γ30 [at +1 SD in performanceavoidance goal structures] = 0.30, SE = 0.07, z = 4.57, p b 0.01) or average (respectively, γ30 [at mean levels of in performance-avoidance goal structures] = 0.15, SE = 0.03, z = 4.45, p b 0.01 and γ30 [at mean levels of in performance-avoidance goal structures] = 0.16, SE = 0.03, z = 5.60, p b 0.01); in contrast the relation between performance-approach goals and surface or deep-level learning strategies was nonsignificant among students belonging to classroom which were low (i.e., −1 SD) in performance-avoidance goal structures (respectively, γ30 [at −1 SD in performance-avoidance goal structures] = −0.05, SE = 0.06, z =

L. Matos et al. / Learning and Individual Differences 55 (2017) 87–96

−0.80, p = 0.42 and γ30 [at −1 SD in performance-avoidance goal structures] = 0.01, SE = 0.05, z = 0.19, p = 0.85). In sum, these interactions suggest that students who pursued performance-approach goals tended to report even higher surface and deep-level learning strategies if they belonged to classrooms where performance-avoidance goals were collectively perceived to be salient. These findings seem at odds with our expectations and as described below supplementary analyses revealed that both cross-level interactions became nonsignificant once we controlled for surface level learning strategies (in the model of deep-level learning strategies) and deeplevel learning strategies (in the model of surface level learning strategies). 3.1.5. Supplementary analyses Because (a) both performance-approach and mastery goals yielded similar positive relations to surface level learning strategies in both samples and to deep-level learning strategies in Sample 1 (providing thus some support to the additive goal hypothesis), and because (b) performance-avoidance goal structures as a classroom predictor positively predicted, rather unexpectedly, both surface and deep-level learning strategies, and because (c) the two learning strategies were fairly strongly correlated, we further examined the contribution of the three achievement goals and performance-avoidance goal structures in the prediction of surface level learning strategies after controlling for deep-level learning strategies at both levels and in the prediction of deep-level learning strategies, after controlling (again, at both levels) for surface level learning strategies. In essence, we ran exactly the same models as before but by adding deep-level learning strategies as a student-level and classroom-level covariate in the model of surface-level learning strategies and the latter as a student-level and classroomlevel covariate in the model of deep-level learning strategies. Interestingly, when surface level strategies were entered as a student-level and class-level covariate in the model of deep-level learning strategies in Sample 1, both mastery goals (γ20 = 0.22, SE = 0.02, p b 0.01) and performance-approach goals (γ30 = 0.07, SE = 0.02, p b 0.01) remained statistically significant predictors of deep-level learning strategies, whereas performance-avoidance goals continued remaining nonsignificant (γ20 = − 0.01, SE = 0.01, p = 0.59). These findings were in line with the additive goal hypothesis. Mastery as well as performance-avoidance goal structures that have been previously found to positively predict deep-level learning strategies became statistically nonsignificant class-level predictors of deep-level learning strategies (respectively, γ01 = 0.00, SE = 0.00, p = 0.33 and γ01 = 0.00, SE = 0.00, p = 0.97). In the same sample, when deep-level learning strategies were entered as a covariate in the model of surface level strategies, mastery goals emerged as the sole negative (rather than positive) predictor of surface level learning strategies both at the student (γ20 = − 0.10, SE = 0.03, p b 0.01) and the classroom level (γ01 = −0.08, SE = 0.03, p b 0.01); neither performance-approach goals, nor performance-avoidance goals were statistically significant (respectively, γ30 = −0.02, SE = 0.02, p = 0.46 and γ40 = 0.04, SE = 0.03, p = 0.14); these findings were mainly in line with the mastery (and not the multiple) goal perspective. Also, performance-avoidance goal structures were nonsignificant class-level predictors (γ20 = 0.03, SE = 0.04, p = 0.44). Notably also, the two previously found cross-level interactions between personally endorsed performance-approach goals and class-level performance-avoidance goal structures became nonsignificant. These results were further against the selective goal hypothesis. In Sample 2, none of the predictors (either student-level or classlevel) remained statistically significant in the prediction of surface level learning strategies when deep-level learning strategies were taken into account. In contrast, when surface level learning strategies were controlled for in the model of deep-level learning strategies the only significant predictors that emerged were mastery goals (γ20 = 0.19, SE = 0.03, p b 0.01) and their interaction with performance-

93

approach goals (γ01 = − 0.09, SE = 0.03, p = 0.015) (which implied that the relation between mastery goals and deep-level learning strategies was attenuated when students endorsed performance-approach goals). In essence, the findings for Sample 2 favored more the mastery goal perspective, while the statistically significant interaction implied that the concurrent pursuit of mastery and performance-approach goals (reflecting the multiple goal perspective) relates to less than more use of deep-level learning strategies. Taken together, these supplementary analyses revealed a more adaptive profile for mastery goals than performance-approach goals as the former positively predicted deep-level learning strategies in both samples while they negatively predicted surface-level learning strategies in Sample 1; in contrast, performance-approach goals (a) was weaker predictor than mastery goals of deep-level learning strategies (in Sample 1) while (b) they were found to attenuate the positive association between mastery goals and deep-level learning strategies (in Sample 2). 4. Discussion In this research we aimed to gain further insight with respect to the debated question in Achievement Goal Theory regarding what can be considered as optimal goal pursuit and motivation. To do so, we extended the previous work by (a) testing tenets of achievement goal theory in an understudied cultural sample (i.e., Peru); (b) in light of the paucity of studies looking at performance-avoidance goals at the classroom level, whether contextually promoted performance-avoidance goals would come with similar costs at the contextual level as they do at the personal level, (c) examining whether learners' personally pursued goals and the goals as perceived to be prevailing in their classrooms would in conjunction predict outcomes not accounted for by the main effects of both (d) by adopting a more rigorous methodological approach, that is, multilevel analyses, which allowed us to break down the variance situated at the between-student and between-class level. Based on the mastery goal perspective, one should expect that only the pursuit and, by extension, the promotion by the teacher of mastery goals (as perceived by students) would be associated with optimal learning; based on the multiple goal perspective however, one should anticipate that the additional presence and, by extension, the additional promotion by the teacher of performance-approach (as perceived by students) goals would yield an incremental positive contribution, at least with respect to some outcomes, or under some circumstances. To test the multiple goal perspective, we considered the four different hypotheses proposed by Barron and Harackiewicz (2001): the additive hypothesis in which both goals would be adaptive for the same outcome, the specialized hypothesis in which each type of goal would predict a different type of outcome, the interactive hypothesis in which endorsing both types of achievement goals simultaneously would be more adaptive than the combined pursuit of each of them (as implied within the additive goal hypothesis), and the selective hypothesis in which a match between students' personal achievement goals and the achievement goals promoted in the academic context would be most beneficial for students' academic thriving. These hypotheses were complemented by the buffering and counterbalancing hypotheses stated by Lau and Nie (2008). 4.1. Differences between students in pursued achievement goals The results concerning students' personal achievement goals obtained in our research with Peruvian high school students, mainly favored the mastery goal perspective and, to lesser extent, the additive goal perspective, although a number of side remarks must be made. At first, students' mastery goals were found to be a consistently positive predictor of the use of both types of learning strategies (i.e., deep and surface) across both samples. However, when partialing out the shared variance between deep and surface level learning strategies in the

94

L. Matos et al. / Learning and Individual Differences 55 (2017) 87–96

supplementary analyses as to obtain a “purer” indicator of employed learning strategies, mastery goals were still a significant positive predictor of deep-level learning strategies in both samples, yet, they became a significant negative predictor of surface level learning strategies in Sample 1 (but not in Sample 2). It is important to comment that surface level learning strategies (rehearsal, as measured in this research) influence the attention and encoding processes but they do not imply to help students build internal connections between new information and prior knowledge (Pintrich et al., 1991). Therefore, it is not surprising that mastery goals were negative predictors of this outcome and similar findings have been reported in previous work in both Western (e.g., Diseth, 2011) and non-Western samples (e.g., Ng, 2015). Interestingly, mastery goals even consistently predicted academic achievement across samples. Also this significant contribution is in line with some previous studies. Although Meece, Anderman, and Anderman (2006) commented that several researchers failed to demonstrate a positive relation between mastery goals and academic achievement (i.e., Elliot & McGregor, 2001; Skaalvik, 1997), several other studies provided evidence for such a positive relation (e.g., Ames, 1992a, b, Wolters et al., 1996; Botsas & Padeliadu, 2003; Linnenbrink, 2005; Matos et al., 2007, Paulick et al., 2013). Second, performance-approach goals in Sample 1 positively predicted the three outcomes (as mastery goals did). However, these associations were not consistent across both samples: in Sample 2 performance-approach goals only positively predicted surface level learning strategies. Besides, the statistically significant associations that were found were less strong than the respective ones found for mastery goals, suggesting that the pursuit of performance-approach goals is perhaps less conclusive. In any case, more research is needed to test whether the pursuit of performance-approach goals do provide additional benefits at least among high school students. Regarding performance-avoidance goals, in Sample 1, it was a negative predictor of grades, but this result was not replicated in Sample 2. Regarding the interactive hypothesis in which the combination of different achievement goals could lead to positive outcomes, only a handful of interactions emerged, although none of them was systematically replicated across samples and outcomes. First, in Sample 2, the interaction between mastery and performance-approach goals was significant in the prediction of deep-level learning strategies such that the positive contribution of mastery goals in the prediction of deeplevel learning was attenuated for those holding strong performance-approach goals. It should be noted however that this interaction was not significant in Sample 1. Second, an unexpected significant interaction emerged in Sample 2 with performance-approach and performanceavoidance goals in conjunction predicting grades. Yet, follow-up analyses indicated that this analysis should be interpreted with caution. Specifically, although there was a significant trend for the association between performance-approach goals and achievement to shift as a function of the level of held performance-avoidance goals, when considered at each level, the relation between performance-approach goals and grades appeared non-significant. Hence, from a statistical viewpoint, the relation between performance-approach goals and grades was similar across different levels of pursued performance-avoidance goals. Moreover, the trend of the interaction was in opposition to what could be expected. Given these considerations, we call for further research is needed around this topic. In fact, overall, only a small percentage (i.e., 17%) of the studied interactions across samples was found significant. 4.2. Differences due to perceived goal structures Regarding the contextually promoted achievement goals (i.e., goal structures), our results indicate that a substantial percentage of variance in the perceived class-room goal structures can be situated at the classroom level (see Table 2). The breakdown of the variance at the student and class-level, one of the more novel aspects of the study, is not only

methodologically stronger, but also reasonable in light of the fact that the prevailing achievement goal structures were truly assessed at the classroom level. That is, students indicated which achievement goals were salient for the class as a whole rather than for them personally. Specifically, when the teacher is perceived as emphasizing masteryoriented instructional practices, students report making more use of surface (Sample 1) and deep-level learning strategies (both samples). Interestingly though, our more conservative test, which involved partialling out the shared variance between both learning strategies, indicated that class-level mastery goal structures predicted negatively (rather than positively) surface level learning strategies (yet, only in Sample 1). This suggests that some of observed correlates between mastery goals and surface level learning in past work (e.g., Liem, Lau, & Nie, 2008) may have been rather spurious. That is, because deep-level and surface level learning may go hand in hand, being reflective of the fact that some learners make use of a broad variety of learning strategies, the impression is created that mastery goal pursuit is conducive to surface level learning. Yet, when controlling for the shared variance between both, it seems that mastery goal pursuit rather uniquely predicts deep-level learning. Regarding perceptions of performance-approach goal structures, no significant effects were found in any of the samples. Thus, while the personal pursuit of performance-approach goals relates to some benefits, that is not the case when these goals are promoted at the contextual level. Similar discrepancies have been reported by other researchers such as Wolters (2004). For instance, he found that performance-approach goal structures positively predicted the use of cognitive and metacognitive learning strategies but these effects disappeared after entering personal performance-approach goals. However, performanceapproach goal structure predicted less-adaptive outcomes (less persistence and more procrastination) even after entering performance-approach goals into equation. More studies are required. Regarding performance-avoidance goal structures, our findings showed that students in classes in which they perceived their teacher to promote these goals reported both, more surface and deep-level learning strategies, while being negatively related to achievement. Although performance-avoidance goal structures have been hardly studied in past work, both the promotion and personal pursuit of performance-avoidance goals seems to come with less desirable outcomes. Indeed, past work (Liem et al., 2008) has shown that performance avoidance goals positively predicted surface level learning strategies. Yet, the fact that performance-avoidance goal structure positively predicted deep-level learning strategies was an unexpected finding. In an attempt to unravel this issue, we tested whether (class-level) performance-avoidance goal structures would remain significant predictors of deep level learning strategies, once we controlled for the shared variance between the use of deep and surface level learning strategies. Much as this approach provided a clearer picture for the perceived promotion of mastery goals, this was also the case for performance-avoidance goals. That is, performance-avoidance goal structures were no longer positive predictors of the use of deep-level learning strategies, suggesting that the initial positive contribution was equally spurious, that is, due to the elevated use of superficial learning strategies. In addition, performance-avoidance goal structures negatively predicted grades in both samples, a finding that resonates with past work showing similar effects for personal pursued of performance-avoidance goals (e.g. Zusho, Pintrich, & Cortina, 2005). Although the personal pursuit of achievement goals and their contextual promotion have received substantial attention in the past, the possibility of both working in conjunction has received far less empirical attention (as already said by Murayama & Elliot, 2009). Herein, we studied the cross-level interactions between personally pursued and contextually promoted goals at the class-level, as maintained within the selective goal hypothesis. Yet, we failed to provide consistent and systematic evidence across samples and outcomes for a particular combination of personally endorsed and contextually promoted

L. Matos et al. / Learning and Individual Differences 55 (2017) 87–96

achievement goals. Given that only few interactions emerged, we argue that it is premature, if not misleading to draw any firm conclusions regarding the selective (person by context) hypothesis. Further research is simply required. This has also been the case in previous studies addressing interactions as they did not show clear results for different outcomes, different samples in the same studies or across studies (e.g., Lau & Nie, 2008; Wolters, 2004). The present research included two different student samples from Peru, which seems to be underrepresented in the international literature. Given that the findings replicate prior ones conducted in Western cultures and in non-western ones (e.g. Asian, Lau & Nie, 2008), we provide further evidence for the adaptive nature of mastery goals across different cultures such as the Peruvian context. In Peru, as in many countries, the educational system posits much importance in grades. For instance, a way to enter to the university is by getting the highest scores in an entry test (each university has its own test). Every school student can apply to the university he/she wants regarding of his/her school performance and so, some university exams are massive (for instance, they may have more than 60,000 applicants for less than 5000 spots). If a student is among the ones having good grades at school, he/she might be invited to enter to a university through a different exam (in private universities). Further research is needed in samples of different ages (not only high school or primary but also university students and even the ones that finished the school but are not yet at the university). 4.3. Practical implications So, what are the implications at the practical level? Students' mastery goals were found to consistently positively predict the use of both types of learning strategies (especially deep-level learning strategies under the conservative test) and academic achievement, so it would be important that teachers encourage their students pursuing mastery goals. This goes more in line with a mastery goal perspective. Moreover, we found no evidence against pursuing performance-approach goals; based on our findings we cannot recommend to the teachers either favoring or discouraging this particular type of goals. Regarding pursuing and promoting performance-avoidance goals the present evidence suggests that it is not recommended. We tried to address all these issues by considering the nested structure of our data (as students were nested into classes) so multilevel analyses were performed. 4.4. Limitations One of the limitations of our research could be its correlational nature, especially for the fact that some authors (e.g., Nolen & Haladyna, 1990) have commented that personal goals and perceived classroom goals are highly correlated. We perfectly understand that it is very important to consider objective measures of the classroom goal structures; Ames (1992a) however, stressed the importance of considering what students perceive from their learning contexts, because at the end, it could be that the students' perceptions of the classroom environment are the ones that shape their own motivation and behavior. Despite this, there is still the need to include more (i.e. experimental) research regarding what is really happening at the classroom level (Kaplan et al., 2002). Moreover, insight can be gained from research that includes a combination of survey data of teachers and students as well as class observations in order to get a better understanding of what teachers actually do, what they actually report about themselves, and finally how students perceive their teachers (Anderman & Midgley, 2002). It is noteworthy to comment that our results may be due to the type of the measures we used to assess achievement goals. Based on more recent theoretical evolutions, (which however are not necessarily shared by all achievement goal researchers), as well as on a meta-analysis (Hulleman et al., 2010), the more recent measures (e.g. Elliot & McGregor, 2001; Elliot et al., 2011), clearly separate the “aims” from

95

the “reasons” components of the achievement goals. However, in the PALS, which was used in the present research, both aspects are not separated and each of them could account for the observed correlates. It would be very useful and clarifying to examine the unique and interactive contribution of the achievement aims as such and its underlying reasons (see Urdan & Mestas, 2006; Vansteenkiste, Lens, Elliot, Soenens, & Mouratidis, 2014). However, the PALS is still used by several researchers (e.g., Lau & Nie, 2008; Shim & Finch, 2014). We are aware also that some of the subscales used in this research have reliabilities lower than 0.70 (higher than 0.60) and that there is room for further scale development and improvement. Also, the number of higher-order units (i.e., classrooms) in Sample 2 was rather marginal. Future research would benefit to include bigger samples and a wide set of outcomes that, according to the literature, relate differently to the achievement goals (i.e., interest, help-seeking, achievement). 5. Conclusion If we consider all our results, they are more in line with a Mastery goal perspective, which favors the pursuit and promotion of mastery goals, regardless of which other achievement goals students endorse or which achievement goals are perceived to prevail in the classroom. This does not mean that holding performance-approach goals is maladaptive; on the contrary, the observed unique effects were positive, yet more limited. Future research can continue to address the unique and potentially interactive contribution between personally endorsed achievement goals and those being salient in the classroom, making use of different measures, designs (e.g., experimental), samples (e.g., younger children) and outcomes (e.g., effort-expenditure, selfhandicapping). References Al-Emadi, A. A. (2001). The relationships among achievement, goal orientation, and study strategies. Social Behavior and Personality, 29(8), 823–832. Ames, C. (1992a). Classrooms: Goals, structures, and student motivation. Journal of Educational Psychology, 84, 261–271. http://dx.doi.org/10.1037/0022-0663.84.3.261. Ames, C. (1992b). Achievement goals and the classroom motivational climate. In D. H. Schunk, & J. L. Meece (Eds.), Student perceptions in the classrooms (pp. 327–348). New Jersey: Lawrence Erlbaum. Ames, C., & Archer, J. (1988). Achievement goals in the classroom: Students' learning strategies and motivation processes. Journal of Educational Psychology, 80, 260–267. http://dx.doi.org/10.1037//0022-0663.80.3.260. Anderman, E. M., & Maehr, M. L. (1994). Motivation and schooling in the middle grades. Review of Educational Research, 64, 287–309. http://dx.doi.org/10.3102/ 00346543064002287. Anderman, E. M., & Midgley, C. (2002). Methods for studying goals, goal structures, and patterns of adaptive learning. In C. Midgley (Ed.), Goals, goal structures, and patterns of adaptive learning (pp. 1–20). New Jersey: Lawrence Erlbaum. Barron, K. E., & Harackiewicz, J. M. (2001). Achievement goals and optimal motivation: Testing multiple goal models. Journal of Personality and Social Psychology, 80, 706–722. http://dx.doi.org/10.1037/0022-3514.80.5.706. Barron, K. E., & Harackiewicz, J. M. (2003). Revisiting the benefits of performance-approach goals in the college classroom: Exploring the role of goals in advanced college courses. International Journal of Educational Research, 39, 357–374. http://dx.doi.org/ 10.1016/j.ijer.2004.06.004. Botsas, G., & Padeliadu, S. (2003). Goal orientation and reading comprehension strategy use among students with and without reading difficulties. International Journal of Educational Research, 39, 477–495. http://dx.doi.org/10.1016/j.ijer.2004.06.010. Bouffard, T., Boisvert, J., Vezeau, C., & Larouche, C. (1995). The impact of goal orientation on self-regulation and performance among college students. British Journal of Educational Psychology, 65, 317–329. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis in the behavioral sciences (3rd ed.). Mahwah, NJ: Erlbaum. Covington, M. V. (2000). Goal theory, motivation and school achievement: An integrative review. Annual Review of Psychology, 51, 171–200. http://dx.doi.org/10.1146/annurev. psych.51.1.171. Diseth, Å. (2011). Self-efficacy, goal orientations and learning strategies as mediators between preceding and subsequent academic achievement. Learning and Individual Differences, 21, 191–195. http://dx.doi.org/10.1016/j.lindif.2011.01.003. Dweck, C. S. (1986). Motivational processes affecting learning. American Psychologist, 41, 1040–1048. http://dx.doi.org/10.1037//0003-066x.41.10.1040. Dweck, C. S., & Leggett, E. (1988). A social-cognitive approach to motivation and personality. Psychological Review, 95, 256–273. http://dx.doi.org/10.1037//0033-295x.95.2. 256.

96

L. Matos et al. / Learning and Individual Differences 55 (2017) 87–96

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: Guilford Press. Elliot, A. J., & Church, M. A. (1997). A hierarchical model of approach and avoidance achievement motivation. Journal of Personality and Social Psychology, 72, 218–232. http://dx.doi.org/10.1037/0022-3514.72.1.218. Elliot, A. J., & Harackiewicz, J. M. (1996). Approach and avoidance achievement goals and intrinsic motivation: A mediational analysis. Journal of Personality and Social Psychology, 70, 461–475. http://dx.doi.org/10.1037/0022-3514.70.3.461. Elliot, A. J., & McGregor, H. A. (2001). A 2 × 2 achievement goal framework. Journal of Personality and Social Psychology, 80, 501–519. http://dx.doi.org/10.1037/00223514.80.3.501. 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. http://dx.doi.org/10.1037/0022-0663.91.3.549. Elliot, A. J., Murayama, K., & Pekrun, R. (2011). A 3 × 2 achievement goal model. Journal of Educational Psychology, 103, 632–648. http://dx.doi.org/10.1037/a0023952. Enders, C. K., & Tofighi, D. (2007). Centering predictor variables in cross-sectional multilevel models: A new look at an old issue. Psychological Methods, 12, 121–138. http:// dx.doi.org/10.1037/1082-989x.12.2.121. 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. http://dx.doi.org/10.1037//0022-0663.92.2.316. Harackiewicz, J. M., Barron, K. E., & Elliot, A. J. (1998). Rethinking achievement goals: When are they adaptive for college students and why? Educational Psychologist, 33, 1–21. http://dx.doi.org/10.1207/s15326985ep3301_1. 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. http://dx.doi.org/ 10.1037//0022–0663,92.2.316. Heyman, G. D., & Dweck, C. S. (1992). Achievement goals and intrinsic motivation: Their relation and their role in adaptive motivation. Motivation and Emotion, 16, 231–247. http://dx.doi.org/10.1111/j.1467-8624.1998.tb06197.x. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. http://dx.doi.org/10.1080/10705519909540118. 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. http://dx.doi.org/10.1037/a0018947. Jöreskog, K., & Sörbom, D. (2001). Lisrel 8: Structural equation modeling with the Simplis command language. Hillsdale, NJ: Laurence Erlbaum. Kaplan, A., & Maehr, M. (2007). The contributions and prospects of goal orientation theory. Educational Psychology Review, 19, 141–184. http://dx.doi.org/10.1007/s10648006-9012-5. Kaplan, A., Middleton, M. J., Urdan, T., & Midgley, C. (2002). Achievement goals and goal structures. In C. Midgley (Ed.), Goals, goal structures, and patterns of adaptive learning (pp. 21–53). New Jersey: Lawrence Erlbaum. Lau, S., & Nie, Y. (2008). Interplay between personal goals and classroom goal structures in predicting student outcomes: A multilevel analysis of person-context interactions. Journal of Educational Psychology, 100, 15–29. http://dx.doi.org/10.1037/0022-0663. 100.1.15. 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. http:// dx.doi.org/10.1016/j.cedpsych.2007.08.001. Linnenbrink, E. (2005). The dilemma of performance-approach goals: The use of multiple goal contexts to promote students' motivation and learning. Journal of Educational Psychology, 97, 197–213. http://dx.doi.org/10.1037/0022-0663.97.2.197. Linnenbrink, E. A., & Pintrich, P. R. (2002). Motivation as an enabler for academic success. School Psychology Review, 31, 313–327. Maehr, M. L., & Midgley, C. (1996). Transforming school cultures. Boulder: Westview Press. Matos, L., Lens, W., & Vansteenkiste, M. (2007). Achievement goals, learning strategies and language academic achievement among Peruvian high school students. Psychologica Belgica, 47, 51–70. Meece, J. L., Anderman, E. M., & Anderman, L. H. (2006). Classroom goal structure, students' motivation, and academic achievement. Annual Review of Psychology, 57, 487–503. http://dx.doi.org/10.1146/annurev.psych.56.091103.070258. Meece, J. L., Blumenfeld, P. C., & Hoyle, R. H. (1988). Students' goal orientations and cognitive engagement in classroom activities. Journal of Educational Psychology, 80, 514–523. http://dx.doi.org/10.1037//0022-0663.80.4.514. Middleton, M., & Midgley, C. (1997). Avoiding the demonstration of lack of ability: An underexplored aspect of goal theory. Journal of Educational Psychology, 89, 710–718. http://dx.doi.org/10.1037/0022-0663.89.4.710. 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, 103, 77–86. http://dx.doi.org/10.1037/0022-0663.93.1.77. Midgley, C., Maehr, M. L., Hruda, L. Z., Anderman, E., Anderman, L., Freeman, K. E., et al. (2000). Manual for the patterns of adaptive learning scales (PALS). Ann Arbor, Michigan: University of Michigan. Midgley, C., & Urdan, T. (2001). Academic self-handicapping and achievement goals: A further examination. Contemporary Educational Psychology, 26, 61–75. http://dx.doi. org/10.1006/ceps.2000.1041.

Murayama, K., & Elliot, A. J. (2009). The joint influence of personal achievement goals and classroom goal structures on achievement-relevant outcomes. Journal of Educational Psychology, 101, 432–447. http://dx.doi.org/10.1037/a0014221. Nelson, G. F., McInerney, D. M., & Craven, R. G. (2004). Similarities and differences in motivation and processes of learning between Papua New Guinea and Australian school students. Conference presented in the third international SELF research conference. Selfconcept, motivation and identity: Where to from here?. Berlin, Germany: Max Planck Institute. Ng, C. (2015). Learners' goal profiles and their learning patterns over an academic year. The International Review of Research in Open and Distributed Learning, 16(3), 86–109. Nolen, S., & Haladyna, T. M. (1990). Personal and environmental influences on students' beliefs about effective study strategies. Contemporary Educational Psychology, 15, 116–130. http://dx.doi.org/10.1016/0361-476X(90)90011-0. Pajares, F., & Cheong, Y. F. (2003). Achievement goal orientation in writing: A developmental perspective. International Journal of Educational Research, 39, 437–455. http://dx.doi.org/10.1016/j.ijer.2004.06.008. Paulick, I., Watermann, R., & Nückles, M. (2013). Achievement goals and school achievement: The transition to different school tracks in secondary school. Contemporary Educational Psychology, 38, 75–86. http://dx.doi.org/10.1016/j.cedpsych.2012.10.003. Pintrich, P. R. (2000a). An achievement goal theory perspective on issues in motivation terminology, theory, and research. Contemporary Educational Psychology, 25, 92–104. http://dx.doi.org/10.1006/ceps.1999.1017. Pintrich, P. R. (2000b). The role of goal orientation in self-regulated learning. In M. Boekaerts, P. R. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 451–502). San Diego: Academic Press. Pintrich, P. R. (2000c). Multiple goals, multiple pathways: The role of goal orientation in learning and achievement. Journal of Educational Psychology, 92, 544–555. http://dx. doi.org/10.1037//0022-0663.92.3.544. Pintrich, P. R., & De Groot, E. V. (1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology, 82, 33–40. http://dx.doi.org/10.1037/0022-0663.82.1.33. Pintrich, P. R., & Schrauben, B. (1992). Students' motivational beliefs and their cognitive engagement in classroom academic tasks. In D. H. Schunk, & J. L. Meece (Eds.), Student perceptions in the classrooms (pp. 149–183). New Jersey: Lawrence Erlbaum. Pintrich, P. R., & Schunk, D. H. (2002). Motivation in education: Theory, research and application (2nd ed.). Upper Saddle River, NJ: Prentice Hall. Pintrich, P. R., Smith, D. A. F., Garcia, T., & McKeachie, W. J. (1991). A manual for the use of motivated strategies for learning questionnaire (MSLQ). Ann Arbor, Michigan: University of Michigan. Roeser, R. W., Midgley, C., & Urdan, T. C. (1996). Perceptions of the school psychological environment and early adolescents' psychological and behavioral functioning in school: The mediating role of goals and belonging. Journal of Educational Psychology, 88, 408–422. http://dx.doi.org/10.1037/0022-0663.88.3.408. Senko, C. M., & Harackiewicz, J. M. (2005). Achievement goals, performance and task interest: Why perceived difficulty matters. Personality and Social Psychology Bulletin, 31, 1739–1753. http://dx.doi.org/10.1177/0146167205281128. Shim, S. S., & Finch, W. H. (2014). Academic and social achievement goals and early adolescents' adjustment: A latent class approach. Learning and Individual Differences, 30, 98–105. http://dx.doi.org/10.1016/j.lindif.2013.10.015 (Inc. All rights reserved). Skaalvik, E. M. (1997). Self-enhancing and self-defeating ego orientation: Relations with task and avoidance orientation, achievement, self-perceptions, and anxiety. Journal of Educational Psychology, 89, 71–81. http://dx.doi.org/10.1037/0022-0663.89.1.71. Urdan, T. (2007). What are classroom goal structures and how much do they matter? Paper presented at the American Educational Research Association Conference. Chicago: USA. Urdan, T., & Mestas, M. (2006). The goals behind performance goals. Journal of Educational Psychology, 98, 354. http://dx.doi.org/10.1037/0022-0663.98.2.354. Van Yperen, N. W., Blaga, M., & Postmes, T. (2014). A meta-analysis of self-reported achievement goals and nonself-report performance across three achievement domains (work, sports, and education). PloS One, 9. http://dx.doi.org/10.1371/journal. pone.0093594. Vansteenkiste, M., Lens, W., Elliot, A. J., Soenens, B., & Mouratidis, A. (2014). Moving the achievement goal approach one step forward: Toward a systematic examination of the autonomous and controlled reasons underlying achievement goals. Educational Psychologist, 49, 153–174. http://dx.doi.org/10.1080/00461520.2014.928598. Vansteenkiste, M., Simons, J., Lens, W., Soenens, B., Matos, L., & Lacante, M. (2004). Less is sometimes more: Goal-content matters. Journal of Educational Psychology, 96, 755–764. http://dx.doi.org/10.1037/0022-0663.96.4.755. Wolters, C. A. (2004). Advancing achievement goal theory: Using goal structures and goal orientations to predict students' motivation, cognition, and achievement. Journal of Educational Psychology, 96, 236–250. http://dx.doi.org/10.1037/0022-0663.96.2.236. Wolters, C. A., Yu, S. L., & Pintrich, P. R. (1996). The relation between goal orientation and students' motivational beliefs and self-regulated learning. Learning and Individual Differences, 8, 211–238. http://dx.doi.org/10.1016/s1041-6080(96)90015-1. Zimmerman, B. J. (1994). Dimensions of academic self-regulation: A conceptual framework for education. In D. Schunk, & B. Zimmerman (Eds.), Self-regulation of learning and performance (pp. 3–21). New Jersey: Lawrence Erlbaum. Zusho, A., Pintrich, P. R., & Cortina, K. S. (2005). Motives, goals, and adaptive patterns of performance in Asian American and Anglo American students. Learning and Individual Differences, 15, 141–158. http://dx.doi.org/10.1016/j.lindif.2004.11.003.