Taiwanese sixth graders' achievement goals and their motivation strategy use and grades: An examination of the multiple goal perspective

全文

(1)Taiwanese Sixth Graders' Achievement Goals and Their Motivation, Strategy Use, and Grades: An Examination of the Multiple Goal Perspective Shu-Shen Shih National Chengchi University. The Elementary School Journal. Volume 106, Number 1 © 2005 by The University of Chicago. All rights reserved. 0013-5984/2005/10601-0003$05,00. Abstract Using the trichotomous framework of achievement goals, in the present study I investigated the effects of different combinations of achievement goals on Taiwanese sixth graders' motivation, strategy use, and performance. 242 students completed a self-report survey assessing their achievement goal orientations and a range of outcomes including intrinsic motivation, selfhandicapping, and the use of cognitive and metacognitive strategies. I collected students' grades from school records at the end of the semester. Results suggested that mastery goals were positively associated with Taiwanese students' motivation, cognitive engagement, and grades, regardless of their level of performanceapproach orientation. Also, resutts validated the theoretical distinction within performance goals proposed by recently revised goal theory, that is, that performance-approach and performanceavoidance goals may lead to different consequences for motivation, cognition, and achievement. Specifically, performance-approach goals were adaptive in terms of children's use of cognitive strategies and their grades, whereas performance-avoidance goals were related to students' maladaptive motivation, namely, setf-handicapping. Implications for education and future research are discussed. Over the past several decades, achievement goal theory has emerged as the dominant framev^ôrk for studying achievement motivation. As a cognitive approach to motivation, this framevôrk provides a window into the psychological processes through which achievement behavior is created (Midgley et al., 1998; NichoUs, 1984; Pintrich & Schunk, 1996; Weiner, 1990). Achievement goals refer to the purposes or reasons for a person's pursuit in an achievement situation (Dweck & Leggett, 1988; Maehr, 1989). Originally, two qualitatively different.

(2) 40. THE ELEMENTARY SCHOOL JOURNAL. categories of goals were proposed (Ames, 1992; Dweck & Leggett, 1988; Harackiewicz, Barrori, & Elliot, 1998; Maehr & Midgley, 1991; Nicholls, 1984; Wolters, Yu, & Pintrich, 1996). This dichotomous framework assumes that mastery goals (also referred to as learning goals or task involvement) orient individuals to increase their competence or achieve task mastery. In contrast, performance goals (also labeled as ego involvement) focus students on their ability and self-worth. Ability and self-worth are determined by outperforming peers on academic tasks or receiving public recognition for one's superior performance (Ames, 1992; Dweck, 1986; Nicholls, 1984).. & Patashnick, 1989; Nolen, 1988). Moreover, performance goals have more often been correlated with lower intrinsic motivation than have mastery goals (Ames & Archer, 1988; Archer, 1994; Duda & Nicholls, 1992; Harackiewicz, Barron, Carter, Lehto, & Elliot, 1997). In short, mastery goals are more Ukely to foster an adaptive pattern of achievement, whereas performance goals are generally seen as less adaptive in terms of subsequent motivation, strategy use, and performance (Pintrich, 2000a).. Achievement Goals and Patterns of Self-Regulation Much research has suggested that mastery and performance goals may foster different patterns of self-regulation (Ames, 1992; Ames & Archer, 1988; Bandura, 1986; Dweck, 1986; Dweck & Leggett, 1988; Pintrich, 2000b; Pintrich & Schunk, 1996; Urdan, 1997). To master material to be learned, students espousing mastery goals are expected to use cognitive and metacognitive strategies and expend effort. Accordingly, mastery goals are associated with a number of adaptive outcomes, including preference for challenging work (Ames & Archer, 1988; Elliott & Dweck, 1988), persistence in the face of setbacks (Elliott & Dweck, 1988), intrinsic motivation for learning (Butler, 1987; Meece, Blumenfeld, & Hoyle, 1988; Stipek & Kowalski, 1989), and the use of effective learning strategies (Ames & Archer, 1988; Meece et al., 1988; Nolen, 1988). In contrast, an emphasis on performance goals may lead students to view achievement tasks as measuring important aspects of themselves (e.g., ability). Hence, those who adopt performance goals tend to show helpless responses characterized by low ability attributions to explain failure, use of ineffective strategies, and decreases in task involvement when faced with failure (Farrell, 1985; Meece et al., 1988; Nicholls, Cheung, Lauer,. From a Dichotomous to a. Trichotomous Framework of Achievement Goals Although thus far the dichotomous framework of achievement goals has made significant contributions to the achievement motivation literature, some evidence has indicated that strong conclusions about the negative effects of performance goals may be premature (Harackiewicz et al., 1998; Urdan, 1997). Because performance goals orient people toward competence, such goals might promote adaptive achievement behaviors and lead to positive outcomes (Barron & Harackiewicz, 2000; Harackiewicz & Sansone, 1991). To clarify the effects of performance goals and determine the situations in which these goals are advantageous versus harmful, achievement goal theorists have reexamined the performance goal construct (Elliot & Harackiewicz, 1996; Middleton & Midgley, 1997; Skaalvik, 1997). Reconsiderations of the literature have suggested that this construct actually confounds theoretically distinct components, namely, the approach-oriented need to achieve and the avoidance-oriented fear of failure (Harackiewicz, Barron, Pintrich, Elliot, & Thrash, 2002). To distinguish between two types of performance goals, Elliot and Church (1997) proposed an incorporation of Atkinson's (1957) and McClelland's (1951) two achievement motives (i.e., to approach success and to avoid failure) into the original performance-mastery dichotomy. The separation SEPTEMBER 2005.

(3) GOAL PERSPECTrVE. of performance-approach from performance-avoidance goals resulted in the trichotomous achievement goal framework (Elliot & Church, 1997; Elliot & Harackiewicz, 1996). Each of the three goals is supposed to be related to a distinct predictive profile. Mastery goals are expected to lead to a wide range of positive processes and outcomes mentioned previously. Performance-approach goals have been shown to be related to a number of adaptive learning variables such as higher aspiration, absorption during task engagement, and performance attainment (Elliot & Church, 1997; Elliot & Harackiewicz, 1996; Elliot, McGregor, «& Gable, 1999). Individuals who adopt performance-avoidance goals, in contrast, are more likely to experience threatrelated affect while studying, want to escape evaluation, and become anxious prior to as well as during evaluation (Elliot & Church, 1997; ElHot & Harackiewicz, 1996; Skaalvik, 1997).. 41. how one appears to others, these goals may also be positively associated with self-handicapping. To determine the joint effects of performance-approach and performanceavoidance goals on children's reported use of self-handicapping strategies as well as other variables of interest, in the current study I compared combinations of different levels of these two types of goals. I hoped that the comparisons would shed light on differences in students with varied achievement goal profiles.. A Multiple Goal Perspective Because mastery and performance-approach goals are linked to different achievement outcomes, optimal motivation may require both types of goals. Empirical evidence has indicated that pursuing one type of goal does not necessarily exclude pursuit of the other (Ames & Archer, 1988; BouffardBouchard, Boisvert, Vezeau, & Larouche, 1995; Harackiewicz et al., 1997; Middleton & Midgley, 1997). Further, using median Achievement Goats and splits to create groups of college students, Setf-Handicapping Bouffard et al. (1995) found that students in The distinction between the approach and the high-mastery/high-performance group avoidance components of performance had the best learning outcomes (e.g., the goals also allows investigation of the differ- highest motivation, best strategy use, higher ences in their relations to academic self- grades, etc.). Pintrich (2000a) also reported handicapping. Self-handicapping refers to the enhanced effects of adopting both masthe use of strategies such as putting off tery and performance-approach goals on studying until the last moment or fooling junior high school students' achievementaround the night before a test that will serve related behaviors. As for younger students. as ready excuses for potential failure (Cov- Turner, Meyer, Midgley, and Patrick (2003) ington, 1992). Covington's (1992) theory of investigated the relations of combinations of self-worth posits that students' attempts to classroom goal structures to sixth graders' maintain a positive self-image may deter- reported affect (e.g., negative affect followmine their achievement behaviors in schools. ing failure) and achievement behaviors (e.g., Academic self-handicapping is a motiva- positive coping, acadentuc self-regtilation, tional strategy some students use to deflect etc.). They found that, depending on the inothers' perceptions away from lack of abil- structional and motivational contexts within ity should poor performance occur (Midg- the classroom goal structures, a highley & Urdan, 2001; Urdan & Midgley, 2001). mastery/high-performance pattern could Given that self-handicapping is an ex- foster elementary students' self-regulation. ample of avoidant behavior, it is presumed Little is known, however, about how multito be strongly associated with performance- ple goal pursuit operates in cultural conavoidance goals. Nevertheless, because per- texts other than Western societies. By exformance-approach goals are focused on amining the relation between pursuing both.

(4) 42. THE ELEMENTARY SCHOOL JOURNAL. mastery and performance-approach goals and Taiwanese elementary children's academic performance, the present study was intended to enhance understanding in this area.. multiple outcomes and test the independent as well as interactive effects of mastery and performance-approach goals on each outcome. Hence, in this study I selected a variety of achievement-relevant variables including motivation (i.e., intrinsic motivation and academic self-handicapping), cognitive engagement (cognitive and metacognitive strategy use), and performance for a more detailed analysis regarding patterns of multiple goal pursuit.. Patterns of Multiple Goal Pursuit The pursuit of multiple goals could facilitate performance in four ways: (1) an additive pattem, (2) an interactive pattem, (3) a specialized goal pattern, and (4) a selective goal pattern (Barron & Harackiewicz, 2000, 2001). In an additive goal pattern, both mastery and performance-approach goals independently exert positive effects on a particular outcome. An interactive goal pattern is characterized by a positive mastery X performance-approach goal interaction effect on a single outcome. These two patterns are different in that in the case of the interactive goal pattern, positive performanceapproach goal effects may depend on a high level of mastery goals (or vice versa). Instead of enhancing a single outcome, in a specialized goal pattern, mastery and performance-approach goals exert independent effects on different outcomes. For instance, mastery goals have an effect on interest, whereas performance-approach goals affect actual performance. The goal patterns mentioned above address the situations in which individuals pursue multiple goals simultaneously. Finally, in the selective goal pattern, people may focus on one goal that is most relevant at a specific time. Achievement goals may be adaptive as a function of the situation in which the goal is pursued. Detection of the selective goal pattern requires a series of analyses yielding main effects of different goals on multiple outcomes at various points in time. Because one purpose of the present study was to investigate the effects of espousing both mastery and performanceapproach goals at the same time within the Taiwanese classroom context, I did not examine the selective goal pattern in this study. To identify any of the three other goal patterns, it is necessary to take into account. The Taiwanese Classroom Context Harackiewicz and Sansone (1991) proposed a matching hypothesis suggesting that the effects of achievement goals might vary with the nature of educational contexts in which goals are pursued. Based on this hypothesis, students who strive to best peers may be optimally motivated in an educational context where competition is salient. The Taiwanese classroom, which is characterized by multidimensional achievement dynamics, thereby constitutes an ideal case for testing the matching hypothesis in terms of the benefits of pursuing multiple goals. In Chinese society, success in competitive national examinations is valued as a key to social mobility and good prospects. To prepare for these competitive examinations, Taiwanese students usually attend cram school at a young age. Within some elementary classrooms, test scores of sixth-grade students are even posted publicly so that children are motivated to attain higher ranks (Grant & Dweck, 2001; Salili, Chiu, & Lai, 2001). Additionally, pupils in Taiwan are influenced by Confucian doctrines such as "Being dihgent in study means devoting one's effort to it for a long time" (Confucius, Zi Zhang chapter). The cultural focus on effort enables Taiwanese students to demonstrate strong willingness to expend effort to achieve academic success (Stigler, Smith, & Mao, 1985). With these cultural backgrounds, Taiwanese students are highly likely to hold both mastery and performance goal orientations at the same time. The competitive atmosphere prevalent SEPTEMBER 2005.

(5) GOAL PERSPECTrVE. in Taiwanese elementary classrooms may help to clarify the question of developmental or age effects raised by Midgley, Kaplan, and Middleton (2001). These researchers indicated that in the United States, as students move up in grade level, the classroom context becomes more performance focused. Accordingly, when examining the individual's achievement goal adoption, the effects of age are often confounded with the effects of the learning environment. It is possible that college or middle-grade students already possess effective learning strategies for attaining their performance-approach goals (Harackiewicz, Barron, Pintrich, et al., 2002). Thus, in response to Midgley et al.'s (2001) call for more research on younger students, in the present study I explored the relations of Taiwanese sixth-grade students' goal orientations to their motivation, cognition, and achievement. I hoped that by studying elementary children in this context, the question about the relevance of a multiple goal perspective for younger students could be answered more clearly.. 43. on median splits of children's achievement goal scores. For the 2 x 2 matrix of performance-approach and performanceavoidance goals, children were also divided into four groups (high performanceapproach/low performance-avoidance, low performance-approach/high performanceavoidance, high/high, and low/low). Such a design allowed the effects of the two combinations (performance-approach goals combined with mastery goals vs. performance-approach goals combined with performance-avoidance goals) on students' patterns of learning to be compared. In doing so, I expected that the more adaptive goal profiles would be identified. Also, the likely varying effects of performanceapproach goals when interacting with mastery versus performance-avoidance goals might be detected. In sum, the following research questions were addressed: (a) What are the relations among Taiwanese children's achievement goals, cognitive and metacognitive strategy use, intrinsic motivation, self-handicapping, and grades? (b) Do children's cognitive and metacogniThe Present Study tive strategy use, intrinsic motivation, selfGiven the multidynamic nature of the Tai- handicapping, and grades differ according wanese educational context, I investigated to the combination of their mastery and the effects of pursuing multiple goals (i.e., performance-approach goal orientations? mastery and performance-approach goals) (c) Do children's cognitive and metacognion Taiwanese children's motivation, cognitive strategy use, intrinsic motivation, selftive engagement, and academic achieve- handicapping, and grades differ according ment. The combined effects of performance- to the combination of their performanceapproach and performance-avoidance goals approach and performance-avoidance goal on the same variables were also explored. orientations? Students can espouse different levels of the two dimensions of performance goals, and differential outcomes may result (Pin- Method Participants trich, 2000a). Accordingly, I created a 2 X 2 matrix of mastery and performance-apThe participants included 242 sixthproach goals, and a 2 X 2 matrix of perfor- grade students from nine classes in three elmance-approach and performance-avoidance ementary schools in the northern part of goals. Taiwan. The 120 girls (50%) and 122 boys Eor the 2 x 2 matrix of mastery and ranged in age from 11 years, 1 month to 12 performance-approach goals, I formed four years, 4 months (M = 11 years, 7 months). groups (high mastery/low performance- The school districts were primarily middle approach, low mastery/high performance- class in terms of socioeconomic status. All approach, high/high, and low/low) based of the participants were Taiwanese..

(6) 44. THE ELEMENTARY SCHOOL JOURNAL. Procedure The data were collected at the beginning of the year in sixth grade (October). Students were required to fill out a few questionnaires including an achievement goal questionnaire (Elliot & Church, 1997), the subscales of cognitive and metacognitive study strategies, as well as the intrinsic value scale from the Motivated Strategies for Learning Questionnaire (MSLQ; Pintrich & De Groot, 1990; Pintrich, Smith, Garcia, & McKeachie, 1993), and the selfhandicapping questionnaire (Midgley, Arunkiimar, & Urdan, 1996; Urdan, Midgley, & Anderman, 1998) during regular class time. Two trained research assistants were in each class for the data collection. They assured students of the confidentiality of their selfreports and encouraged them to respond to the items as accurately as possible. One research assistant read the items aloud, and the other one walked around the room to check for skipped items and ensure quality responses.. (1997) conducted a principal-components factor analysis with varimax rotation on the 18 items. Three factors were extracted from the analysis. Factor 1 accounted for 33.1% of the total variance and consisted of the six performance-approach goal items. The second factor accounted for 18.2% of the total variance and comprised the six mastery goal items. Factor 3 accounted for 12.0% of the total variance and comprised the six performance-avoidance goal items. Together, the three factors accounted for 63.3% of the total variance. I performed another principal-components factor analysis to examine the validity of the scales. Results showed that the factor solution accounted for 57.43% of the total variance and that 29.20% of the total variance was accounted for by the factor of mastery goal. The factors of performance-approach and performanceavoidance goals accounted for 17.10% and 11.13 % of the total variance, respectively. Cognitive/tnetacognitive study strategies and intrinsic value. I measured children's use of cognitive and metacognitive strategies, as well as their intrinsic interest in course work, by using the adaptation of Pintrich and De Groot's (1990) cognitive/ metacognitive strategies and intrinsic value scales from the Motivated Strategies for Learning Questionnaire. According to Pintrich and his colleagues (Pintrich & De Groot, 1990; Pintrich et al., 1993), the scales of cognitive and metacognitive strategy use were constructed based on a general cognitive model of learning and information processing as well as the results of the factor analysis. Cognitive strategies include students' use of basic (i.e., rehearsal strategies) and complex strategies (i.e., elaboration and organizational strategies) to process information from texts and lectures. Metacognitive strategies involve strategies that help students control and regulate their own cognition such as planning, comprehension monitoring, and regulating. As stated previously, I administered a revised version for use with Taiwanese children (Wu & Cherng, 1992). To test the criterion-related validity. Measures Participants were instructed to respond to all items on five-point Likert scales ranging from 1 (not at all true of me) to 5 (very true of me). Because two of the questionnaires (achievement goal and MSLQ) were originally constructed for assessing college students, I reworded items according to Taiwanese elementary school children's learning experiences. Table 1 displays the description and sample items of all measures used. Achievement goals. Students' goal orientations were assessed by scales adapted from Elliot and Church's (1997) achievement goals questionnaire. This questionnaire is composed of three six-item scales for each of the achievement goals in the trichotomous model. Three scores representing mastery, performance-approach, and performance-avoidance goals for each student were created accordingly. To test the validity of partitioning the goal orientations into several components, Elliot and Church. SEPTEMBER 2005.

(7) GOAL PERSPECTIVE. 45. TABLE 1. Information about Achievement Goals Scales, Cognitive/Metacognitive Strategies Scales, Intrinsic Value Scale, and Self-Handicapping Scale No. of Items. Cronbach's a. Mastery goals. 6. .83. Orientation to develop competence or attain task mastery. Performanceapproach goals. 6. .87. Orientation to demonstrate ability. Performanceavoidance goals. 6. .73. Orientation to avoid the demonstration of incompetence. Cognitive strategy use. 12. .86. Use of elaboration and rehearsal strategies. Metacognitive strategy use. 9. .74. Planning, monitoring comprehension, and regulating cognition. Intrinsic value for learning. 8. .79. Interest in as well as perceived importance of course work. Selfhandicapping. 6. .80. Use of a priori strategies to influence selfpresentation. Scale. What Assessed. of these scales, students' performance on the verbal achievement test was employed as the criterion. Results of the analysis suggested that each scale was correlated with the criterion. Specifically, the scales of cognitive and metacognitive strategy use were positively related to the achievement test (r = .19, p < .01 and r = .21, p < .01, respectively). Children's reported intrinsic value for learning was also positively associated with the criterion (r = .23, p < .01). Such tests supported the validity of these scales. Self-handicapping. Children's use of selfhandicapping strategies was assessed using. Sample Items I want to learn as much as possible from this class. I desire to completely master the material presented in this class. It is important to me to do better than the other students. My goal in this class is to get a better grade than most of the students. I worry about the possibility of getting a bad grade in this class. I often think to myself, "What if I do badly in this class?" When studying, I copy my notes over to help me remember material. I outline the chapters in my book to help me study. When I'm reading I stop once in a while and go over what I have read. I ask myself questions to make sure I know the material I have been studying. I like what I am learning in this class. I prefer class work that is challenging so I can learn new things. Some students purposely don't try hard in school so that if they don't do well they can say it is because they didn't try. How true is this of you? Some students fool around the night before a test so that if they don't do well they can say that is the reason. How true is this of you?. a six-item scale (Urdan et al., 1998). These items were constructed to measure the extent to which students employ a priori strategies to influence self-presentation. Therefore, a phrase such as "they can say that is the reason" or "they can say it is because" was included in each item. Rather than assessing cognitions, this scale measures students' use of active strategies and behaviors. Factor analysis performed by Midgley et al. (1996) yielded a single factor accounting for 59% of the total variance. The principalcomponents factor analysis I conducted also extracted a single factor accounting for.

(8) 46. THE ELEMENTARY SCHOOL JOURNAL. 51.06% of the total variance, with loadings from .63 to .75. Grades. I obtained students' semester grades from student records at the end of the semester as a measure of academic achievement. The grading system in Taiwan for all classrooms assigned points (on a 100point scale) on classroom tasks. Scores in different subjects (e.g., math, language arts) were averaged to represent the overall final semester grades. Before data analysis, collected scores were standardized (converted into r scores) within each classroom to help account for variations in teachers' grading policies.. itive correlation was unexpected in terms of the revised goal theory. Due to the significant correlation between mastery and performance-approach goals, I speculated that the shared variance between the two components of performance goals may have played a role here (Middleton & Midgley, 1997). To investigate the relation between mastery and performance-avoidance goals, I used the partial correlation procedure. After controlling for the effects of performanceapproach goals, I found that mastery goals were unrelated to performance-avoidance goals (r = - .06, p > .01). As to the relations between achievement goals and other variables of interest, both mastery and performance-approach goals were positively related to reported intrinsic motivation, cognitive and metacognitive strategy use, and grades. There were no significant correlations between performanceavoidance goals and these variables. Further, children's self-handicapping was related negatively to intrinsic motivation, effective strategy use, and grades, indicating the negative role of self-handicapping in students' motivation, cognitive engagement, and grades.. Results The data were analyzed according to the following plan. To begin, I computed Pearson correlations to examine relations among key variables (achievement goal orientations, intrinsic motivation, self-handicapping, cognitive and metacognitive strategy use, and grades). Next, I conducted multivariate analyses of covariance (MANCOVAs) to determine the effects of combining two types of goals while controlling for the effects of a third type of goal on students' motivational processes, cognitive engagement, and grades. Finally, because I found gender differences in children's use of selfhandicapping strategies in a preliminary analysis, I conducted a hierarchical multiple-regression analysis to explore multivariate relations among gender, achievement goals, grades, and self-handicapping.. Differences in the Effects of Combining Mastery and Performance-Approach Goals As Table 2 shows, the motivational, cognitive, and achievement variables were correlated with one another and thus were used as dependent variables in the MANCOVA to explore whether students with different Correlation Analyses goal profiles varied on these outcome meaDescriptive statistics and correlation co- sures. To examine the combined effects of efficients are presented in Table 2. Mastery mastery and performance-approach goals, goals were related positively to performance- performance-avoidance goals were included approach goals (r = .52, p < .01). As ex- as a covariate, and for the examination inpected, the two components of perfor- volving the performance-approach/performance goals (i.e., performance-approach mance-avoidance groups, I included masand performance-avoidance goals) were tery goals as a covariate. I used MANCOVAs positively associated with each other (r = in the hope that, by including these covar.54, p < .01). Intriguingly, mastery goals were iates, the combined effects of two types of correlated positively with performance- goals could be determined while the effects avoidance goals (r = .24, p < .01). This pos- of the other type of goal were controlled. To SEPTEMBER 2005.

(9) GOAL PERSPECTIVE. 47. TABLE 2. Descriptive Statistics and Correlations for Study Variables {N = 242). 1. 2. 3. 4. 5. 6. 7. 8.. Mastery Performance-approach Performance-avoidance Intrinsic motivation Self-handicapping Cognitive strategies Metacognitive strategies Grade. M. SD. 1. 2. 3.69 3.21 2.94 3.50 1.55 3.28 3.23 87.61. .90 1.06 .88 .73 .62 .77 .68 11.50. .52** .24** .70** -.19** .68*» .62*» .28**. .54*" .32*" -.07 .41*" .30*" .21*"'. 3. .12 .15* .14*. .05 -.06. 4. 5. 6. 7. 8. -.21** .68** .59** .28**. - .25** - .25** - .33**. .78** .31**. .27**. .... * p < .05. **p< .01.. test for interactions, I used the mediansplits method on the three achievement goal measures to form three low/high categorical variables. These variables then served as independent variables in the MANCOVAs. I used median splits because this method allows the use of MANCOVA with multiple dependent variables. In contrast, regression analyses can use only one dependent variable at a time (Pintrich, 2000a). Because tbe primary purpose of this study was to examine effects of different goal profiles on a number of achievement-related outcomes, the median-splits procedure was appropriate. Table 3 displays the observed means, adjusted means, and tbe standard deviations of the dependent variables according to the different goal profiles. Before I computed MANCOVA, I performed a preliminary ANOVA to compare students at the three elementary schools on each of the eight variables in Table 2. Using the Bonferroni method to correct for inflated probability levels associated witb significance when conducting multiple tests (familywise a = .05), I found no significant difference among students at the three schools. Additionally, t tests were performed to determine gender differences in the same variables. I used the Bonferroni method when making comparisons. Results of t tests suggested that boys (M = 1.67) differed significantly from girls (M = 1.43) in self-handicapping, t(240) = 3.03, p < .01. However, preliminary MANCOVAs yielded no significant main effects of gender or in-. teracting effects for gender and achievement goals on tbe dependent variables. Consequently, I did not include gender as an independent factor in subsequent analyses. Two assumptions for the MANCOVA had been examined before the analysis was performed. Because cell sizes for the independent variables were unequal, I conducted Box's M test first to check for homogeneity of variance-covariance matrices. Tbe result of this test was not significant (F = 1.48, p > .05), indicating the confirmation of this assumption. Moreover, the test for homogeneity of regression also yielded insignificant results. Hence, using a common regression coefficient to adjust for the covariate in all groups was appropriate (Tabachnick & Fidell, 1996). When I controlled for performanceavoidance goals, tbe MANCOVA revealed significant main effects for mastery goals, Wilks's X = .71, f (5, 233) = 18.79, p < .001. The effect size was calculated by means of an eta-squared value {xf), which represented the proportion of variance of the dependent variables tbat was attributable to the effect (Tabacbnick & Fidell, 1996). The eta-squared value for the effects of mastery goals was .29. The univariate analyses of tbe main effects of mastery goals were significant for intrinsic motivation, f(l, 237) = 79.15, p < .001, ri^ = .25; self-handicapping, Fil, 237) = 4.51, p < .05, ri^ = .02; cognitive strategies, f(l, 237) = 61.67, p < .001, ri^ = .21; metacognitive strategies, F(l, 237) = 50.58, p < .001, ri^ = .18; and grades..

(10) S to. ;§5 ^ CO vo CO o ^ m tN ^ CO csTfininoi—jOt-^f-ô CO CO r-^ .~~* CO CO CO CO vO ^. Q. tn. COCOrHrHCOCOCOCO. CO"*. Q to. \.s.o T3 C. Q. b CD CO. Q to. 'S. S. CO. in. 1 CC. I o. 0-i ^. I. X. c o. Q to. •n. ns. <. in. ^. in. in in. CO. I. D. COCOr-l—tCOCOCOCOt. I T3. c. if. Q to. oi. OVOvOOCOOvi—tinvo oqoOTt-^voiNioin O I CO CO T-^ T—I CO CO CO CO. a. s o. •.oa (U. Cl.. a.. s. '2. •3. c. S. I i. Cog. c. 8 2 U.

(11) GOAL PERSPEGTTVE. f(l, 237) = 7.17, p < .01, Ti^ = .03. A post hoc Scbeffe analysis sbowed that, in terms of motivational processes, students in both bigh mastery/bigb performance-approach (adjusted M = 3.89) and bigh mastery/low performance-approach (adjusted M = 3.91) groups reported significantly bigber intrinsic motivation tban did students in the low mastery/bigb performance-approach (adjusted M = 3.24) and low mastery/low performance-approacb (adjusted M = 3.03) groups. In addition, students in tbe bigh mastery/high performance-approacb group (adjusted M = 1.43) and bigb mastery/ low performance-approach group (adjusted M = 1.45) reported engaging in less selfhandicapping tban did students in tbe low mastery/low performance-approach group (adjusted M = 1.74). In regard to tbe cognitive components, students bigh in mastery orientation (i.e., tbose in botb bigb mastery/bigb performance-approacb and higb mastery/low performance-approacb groups) reported significantly greater use of cognitive (adjusted Ms = 3.71 and 3.64, respectively) as well as metacognitive strategies (adjusted Ms = 3.56 and 3.52, respectively) than did students in tbe low mastery/bigb performanceapproacb group (adjusted M = 3.15 for cognitive strategies and 3.00 for metacognitive strategies) and low mastery/low performance-approacb group (adjusted M = 2.75 for cognitive strategies and 2.86 for metacognitive strategies). As for academic acbievement, children in tbe bigb mastery/ bigb performance-approach (adjusted M = 90.52) and higb mastery/low performanceapproacb (adjusted M = 90.34) groups got bigber grades tban did children in the low mastery/low performance-approach group (adjusted M = 83.20). Wbereas tbe main effects for performance-approach goals did not reacb significance at tbe multivariate level, the univariate tests were significant for cognitive strategies, f(l, 237) = 5.18, p < .05, TI^ = .02, and grades, f(l, 237) = 4.34, p < .05, ri^ = .02. Cbildren wbo scored bigher on the. 49. performance-approach orientation reported more use of cognitive strategies and bad better grades tban did those wbo were less performance-approacb oriented. Moreover, the univariate test showed significant effects for tbe covariate of performanceavoidance goals on self-bandicapping, f (1, 237) = 6.08, p < .03, ri^ = .03. Students who obtained bigher scores on performanceavoidance orientation were more likely to use self-bandicapping strategies tban were cbildren low in performance-avoidance orientation. Also, tbere were significant interaction effects for the mastery and performance-approacb goals on students' grades at the univariate level, f (1, 237) = 4.87, p < .05, Ti^ = .02. Tbe test of simple main effects sbowed a significant difference between students bigb in performance-approach orientation (adjusted M = 89.79) and students who were less performance-approacb oriented (adjusted M = 82.68) within tbe lowmastery group. Differences in tbe Combined Effects of Performance-Approacb and Performance-Avoidance Goals Tbis set of analyses concerned bow students' academic functioning and performance varied with tbe combinations of different levels of performance-approach and performance-avoidance goals. To examine the pure effects of both components of performance goals, I used MANCOVA with mastery goal as a covariate. I checked tbe assumptions regarding bomogeneity of variance-covariance matrices as well as homogeneity of regression before proceeding witb tbe analyses. Results of tbe tests were not significant, indicating tbat tbese assumptions were met. Tbe MANCOVA sbowed significant effects for the covariate of mastery goals, Wilks's X = .71, f (5, 233) = 19.30, p < .001, yf = .29. Tbe univariate tests indicated significant effects on intrinsic motivation, f (1, 237) = 80.81, p < .001, ^^ = .25; selfbandicapping, f(l, 237) = 4.84, p < .05, ri^ = .02; cognitive strategies, f(l, 237) =.

(12) 50. THE ELEMENTARY SCHOOL JOURNAL. 63.94, p < .001, ri^ = .21; metacognitive strategies, F(l, 237) = 51.59, p < .001, TI^ = .18; and grades, HI, 237) = 7.96, p < .01, r|2 = .03. Students who scored higher on mastery orientation reported higher intrinsic motivation, greater use of cognitive and metacognitive strategies, less self-handicapping, and had higher grades than did students low in mastery orientation. Although there were no main effects of the two components of performance goals at the multivariate level, results of the univariate tests demonstrated significant effects of performance-avoidance goals on self-handicapping, f (1,237) = 5.52, p < .02, ri^ = .02. Significant interacting effects for the two types of performance goals on cognitive strategies were also found at the univariate level, f (1, 237) = 6.36, p < .02, rj^ = .03. Results of the test of simple main effects revealed that, within the low performanceavoidance group, children espousing performance-approach goals reported significantly greater use of cognitive strategies (adjusted M = 3.30) than did children low in performance-approach orientation (adjusted M = 3.14). In spite of these findings concerning the effects of performance goals, results of this set of analyses suggested that, all in all, students' achievement-relevant processes as well as their grades varied primarily according to their level of mastery goal orientation rather than different combinations of performance-approach and performance-avoidance goals. Table 4 summarizes the results of the MANCOVAs.. formance-avoidance goals to the model in the second step. Mastery goals negatively predicted handicapping (|3 = — .21, p < .01) when other predictors were controlled for in the regression model. Performanceavoidance goals positively predicted children's use of self-handicapping strategies (P = .19, p < .05), whereas performanceapproach goals were not a significant predictor. In the final step, performance alone was entered and accounted for an additional 7% of the variance in handicapping. Students' grades negatively predicted handicapping (P = -.28, p < .001). In the final regression model, gender, mastery goals, and grades were significant predictors of handicapping. There were no significant interactions by gender. Table 5 shows the standardized regression coefficients and R^ values associated with each step.. Regression Analysis Because girls reported engaging in less self-handicapping than did boys, it seemed informative to explore the multivariate relations among gender, achievement goal orientations, grades, and self-handicapping. To do this, I performed a hierarchical multiple-regression analysis. In the first step of the analysis, gender was entered and emerged as a significant predictor of selfhandicapping (P = .19, p < .01). I added mastery, performance-approach, and per-. Discussion The present study explores a more differentiated view of the role of achievement goals in Taiwanese children's motivation, strategy use, and academic performance. This study represents an initial investigation to compare the effects of performanceapproach goals combined with mastery versus performance-avoidance goals in order to obtain a clearer picture of the differential pathways that different goal profiles may lead students to follow in the classroom. Overall, my findings best support the adaptiveness of mastery-based goals, a theory that has been maintained by a traditional achievement goal perspective developed in the Western cultures (Barron & Harackiewicz, 2001). Mastery goals are positively associated with Taiwanese children's motivation, cognitive engagement, and academic performance, regardless of their level of performance-approach orientation. Additionally, some results of the present research also fit the recently revised goal theory, for they provide empirical evidence to validate the theoretical distinction within performance goals (Harackiewicz, Barron, Pintrich, et al., 2002; Wolters, 2004). PerformanceSEPTEMBER 2005.

(13) GOAL PERSPECTIVE. 51. TABLE 4, Summary of the Results of MANCOVAs {N = 242). Covariate (performanceavoidance goals) Mastery goals Performance-approach goals Mastery X performanceapproach Covariate (mastery goals) Performance-approach goals Performanceavoidance goals Performance-approach X performanceavoid. Multivariate Analyses (F Values). Motivation. Handicapping. Cognitive. Metacognitive. Grade. 1,37 18,79***. ,00 79,15***. 6.08* 4.51*. ,35 61,67***. .39 50,58***. 2.06 7.18**. Univariate Analyses (F Values). 1,84. 1,03. 1,69. 5,18*. ,92. 4.34*. 1,75. 1,80. 1,53. 3,46. .28. 4.87*. 51.59***. 7,96**. 19,30***. 80,81***. 4,84*. 63,94***. 1,34. .47. 1,24. 3,11. .37. 3,66. 1.27. ,04. 5,52*. .07. .14. 1,88. 1,45. 2,30. ,31. 6.36*. 4.64. .04. * p < .05. **p< .01. ***p< .001.. TABLE 5. Hierarchical Regression Analysis Examining Predictors of Self-Handicapping (N = 242). Predictor Variables Step 1: Gender Step 2: Mastery goals Performance-approach goals Performance-avoidance goals Step 3: Grades Change in R^ Total adjusted R^. Step 1. Step 2. Step 3. P. P. P. self-handicapping. I discuss the results in terms of components of academic functioning examined in the study and their relation to achievement goals.. Intrinsic Motivation When the combined effects of mastery -.21** -,15* and performance-approach goals on Taiwanese children's intrinsic motivation were -.05 ,01 examined, students high in mastery orien,19* .13 tation reported higher intrinsic interest than did their counterparts low in mastery ori-.28*** entation, regardless of their level of perfor,06*** .07*** .04 ,10 .17 mance-approach goals. This finding replicates a host of previous studies revealing NOTE.—Gender coded 0 = girl, 1 = boy. positive main effects of mastery goals on in* p < .05. **p < .01, terest outcomes (e.g.. Archer, 1994; Barron ***p< .001. & Harackiewicz, 2001; Church, Elliot, & Gable, 2001; Elliot & Church, 1997; Harackiewapproach and performance-avoidance goals icz, Barron, Tauer, Carter, & Elhot, 2000; may lead to different consequences for mo- Harackiewicz, Barron, Tauer, & Elliot, 2002; tivation, cognition, and achievement. Spe- Pintrich & Garcia, 1991), Mastery goals decifically, performance-approach goals are crease social comparisons and enable stuadaptive in terms of children's use of cog- dents to use self-set standards. Positive feelnitive strategies and grades, whereas per- ings associated with making progress are formance-avoidance goals are related to likely to heighten the degree to which stustudents' maladaptive motivation, namely. dents value being competent and their in.19**. .21**. ,20**.

(14) 52. THE ELEMENTARY SCHOOL JOURNAL. volvement in tasks. In turn, students are more likely to discover the enjoyable aspects of learning tasks, and intrinsic interest is thereby promoted. In regard to effects of different goal profiles on intrinsic motivation, children who espoused goals of outperforming others and mastering nev^ skills or knowledge at the same time appeared to follow the similarly adaptive trajectory of students who pursued mastery goals only. In other words, a performance-approach goal is not necessarily deleterious to a student's intrinsic interest as long as it is coupled with a mastery orientation. Nevertheless, when the adoption of performance-approach goals is not accompanied by a focus on mastery orientation, intrinsic motivation is likely to be reduced (Ames, 1992; Dweck & Leggett, 1988). Students in the low mastery/high performance-approach group reported significantly lower intrinsic interest than did students in the high mastery/high performance-approach group. These findings are consistent with research showing that performance-approach goals seem unrelated to interest outcomes (e.g.. Church et al., 2001; Elliot & Church, 1997; Harackiewicz et al., 1997, 2000). The matching hypothesis (Harackiewicz, Barron, Pintrich, et al., 2002) speculating that performance-approach goals may become more adaptive when pursued in classroom settings that afford a good match is not supported in my study with respect to intrinsic motivation.. in performance-avoidance orientation reported greater use of handicapping strategies than did children low in performanceavoidance orientation. Further, performanceapproach goals had no significant effects on self-handicapping when the avoidance component of the performance goal was taken into consideration in the regression model. As other researchers have shown (Elliot & Harackiewicz, 1996; Middleton & Midgley, 1997; Ryan, Gheen, & Midgley, 1998; Skaalvik, 1997), I found that avoidance motives are associated with a range of negative outcomes. The fear of failure, in which performance-avoidance goals are grounded, motivates students to engage in handicapping for self-protection. However, performance-approach goals are at least not maladaptive under these circumstances (Harackiewicz et al., 1998; Harackiewicz, Barron, Pintrich, et al., 2002; Urdan, 1997). Another noteworthy finding concerns the role of mastery goals in students' engagement in handicapping. Results of both MANCOVA and regression analysis reveal significant effects for this type of goal. In the analysis comparing groups of different goal profiles, students who were more mastery oriented reported less use of handicapping strategies than did students low in mastery orientation. Further, when additional controls such as other types of goals and students' grades were included in the regression model, the mastery goal remained a fairly robust negative predictor of selfhandicapping. Because mastery goals focus students on understanding and mastery of the task, this orientation reduces the importance of self-perceptions and self-consciousness in achievement situations. Thus, students holding this type of goal may be less likely to engage in handicapping to protect self-image (Maehr & Kaplan, 2000).. Self-Handicapping The current study is the first to compare the effects of different combinations of performance-approach and performanceavoidance goals on children's reported handicapping. Clearly, when avoidance behaviors such as self-handicapping are considered, the performance-avoidance goal rather than performance-approach goal emerges as the relevant variable (Midgley & Urdan, 2001; Urdan, 2004). Results of MANCOVA indicate that, regardless of performance-approach goals, children high. Cognitive Strategy Use On the basis of a general cognitive model of learning and information processing (Pintrich et al., 1993), children's cognitive engagement in the current study is repSEPTEMBER 2005.

(15) GOAL PERSPECTIVE. resented by tbeir use of cognitive and metacognitive strategies. As expected, Taiwanese children bigb in mastery orientation reported more frequent use of cognitive as well as metacognitive strategies than did cbildren low^ in mastery orientation. The "strategic or self-regulated learning" bypotbesis proposed by Linnenbrink and Pintricb (2000) suggests bow a particular type of achievement goal may affect cognition. According to tbis hypothesis, because mastery goals lead students to concentrate on mastering and understanding the task, this type of goal should result in effective cognitive processing as well as more self-regulated learning. Similar to other studies (e.g., Bouffard et al., 1995; Pintrich, 2000b; Pintricb & Schrauben, 1992; Wolters, 2004), results of the current research apparently sustain this bjôthesis. As to the effects of performance-approach goals, the univariate tests were significant for cognitive strategies. Children high in performance-approach orientation reported greater use of cognitive strategies than did children low in performanceapproach orientation. Additionally, I found significant interaction effects for performance-approach and performance-avoidance goals on cogrutive strategies at the univariate level. Students low in performanceavoidance orientation who were more performance-approacb oriented used cognitive strategies more frequently in comparison to cbildren wbo were less performance-approach oriented. It appears that a high performance-approach orientation without an accompanying avoidance motive is associated wîth more use of cognitive strategies. As suggested earlier, the cognitive strategies assessed in the current study are supposed to facilitate effective processing of information. A focus on doing better tban others is likely to orient students toward the use of these strategies, so tbat strategic processing of information from texts and lectures migbt result in better academic performance. These findings to some extent support the revised goal theory, which pos-. 53. its that performance-approach goals may promote students' cognitive engagement in learning (Harackiewicz, Barron, Pintrich, et al., 2002; Wolters, 2004). Academic Performance The majority of researchers have not found a direct link between mastery goals and academic performance (Barron & Harackievi'icz, 2001; Elliot & McGregor, 2001; Elliot et al., 1999; Harackiewicz et al., 1997, 2000; McWbaw & Abrami, 2001; Pintrich, 2000a; Skaalvik, 1997; Wolters, 2004), in spite of the ties with students' cognitive engagement. Nonetheless, in the present study mastery goals are positively related to children's grades. Regardless of their level of performance-approach goal, students high in mastery goal orientation had higher grades in comparison with other students. Because an association between mastery goals and achievement has not been found in most previous studies (Harackiewicz, Barron, Tauer, et al., 2002), future research using tbe method of structural equation modeling may illuminate my findings. In terms of the effects of performanceapproach goals, results of the present study are in line with established findings (Harackiewicz, Barron, Pintrich, et al., 2002; Wolters, 2004). Performance-approach goals have positive effects on children's achievement. Evidently, in the current study, mastery and performance-approach goals individually are positively related to students' grades. These findings are somewhat different from tbe now-familiar conclusion tbat only performance-approacb goals are beneficial to an individual's performance (Wolters, 2004). In addition, I obtained significant interacting effects for mastery and performance-approach goals on students' grades. Children low in mastery goal orientation who were more performance-approach oriented performed better than did their counterparts who were less performance-approach oriented. On the basis of these findings, both tbe additive and interactive goal patterns Barron and Harackiew-.

(16) 54. THE ELEMENTARY SCHOOL JOURNAL. icz (2000, 2001) described can be identified in regard to the effects of pursuing multiple goals on Taiwanese students' academic performance. A thorough investigation of the multiple goal patterns shown in the present study also reveals an additive goal pattern for Taiwanese children's use of cognitive strategies. Again, mastery and performance-approach goals are both positively associated with this outcome. Together, these identified patterns illustrate the advantage of multiple goal pursuit relative to single mastery goal pursuit (Harackiewicz, Barron, Pintrich, et al., 2002).. classrooms with a high-mastery/high-performance goal structure may be sensitive to protecting self-worth and vulnerable to avoidance behaviors, such as handicapping, because of the strong achievement messages (Covington, 1992). Consistent with findings of the current study in respect to the negative relation of the mastery goal to self-handicapping, previous research (Turner et al., 2002, 2003) indicated that teachers' provision of mastery-oriented motivational support in classrooms with an emphasis on demonstrating ability may reduce the incidence of avoidance behaviors. Specifically, teachers may assure students of their increasing intellectual competence by responding to their progress and accomplishments. It is also crucial that the teacher explicitly convey to students that asking questions and making mistakes are a natural part of learning. If teachers can address students' fears of being incompetent through their instructional practices, students will be more likely to perceive the classroom environment as affirming. As a consequence, they may be less likely to feel their self-worth threatened, so that selfhandicapping may be reduced (Billings & Moos, 1982; Covington, 1992). In summary, in terms of implications for classroom practices, my findings parallel previous studies (Patrick, Anderman, Ryan, Edelin, & Midgley, 2001; Turner et al., 2002, 2003) investigating the effects of teacher discourse on students' achievement outcomes and use of avoidance strategies through the lens of goal theory. Although an emphasis on both the development of ability (mastery goals) and the demonstration of ability (performance-approacb goals) may lead to tbe benefits of multiple goal pursuit relative to single mastery goal pursuit, such multiple goal structures could bring forth the motivational hazards associated with avoidance bebaviors. To address this problem, teachers create an environment in which students are free to take risks, make mistakes, and try again on their way to success without. Implications for Classroom Practice Results of my study corroborate the optimal effects of pursuing multiple goals on children's academic functioning and achievement. It appears that classroom practices emphasizing both understanding (mastery goals) and high normative achievement (performance-approach goals) should support students' cognitive engagement and academic performance. In contrast, this study points to the role of performanceavoidance goals in students' use of handicapping strategies. Whereas an emphasis on both mastery and performance goals at the classroom level may be positively related to adaptive achievement-relevant processes and outcomes (Midgley et al., 2001), students may engage in avoidance behaviors in a context characterized by the same goal structures when self-worth becomes an issue (Turner et al., 2003). Hence, it is important to consider the instructional practices that may mitigate circumstances that put students' self-worth on the line in classrooms perceived to be bigb in both mastery and performance-approach goals. By examining the relations between classroom discourse and sixth graders' reported achievement behaviors in two high-mastery/ high-performance classrooms. Turner et al. (2003) found that teacher discourse indeed made a difference between classrooms with similar goal patterns. Clearly, students in. SEPTEMBER 2005.

(17) GOAL PERSPECTIVE. cor\cerns about their self-worth (Patrick et al., 2001). Limitations and Future Research Although the results of the present study provide insights into the cross-cultural validity of the achievement goal theory, a few limitations need to be addressed in future research. First, there is a need to explore more fully the interaction between the student's personal goal orientation and the classroom goal structure. Fven if the general Taiwanese educational context is predominantly performance oriented, some variations in classroom orientations likely exist. Depending on the extent to which personal and classroom goals are congruent, the moderator effect may be positive or negative (Harackiewicz et al., 1997, 1998; Newman, 1998). Such explorations could offer valuable insights into the effects of classroom practices on children's adaptive learning behaviors. Second, other personal characteristics may mediate relations between goals and outcomes. For example, Molden and Dweck (2000) contended that recent expansions of goal theory tended to overlook critical issues regarding the meaning of achievement goals. In hght of this need, they proposed a framework through which meaning of goals can moderate the selection and implementation of achievement goals. This framework posits that individuals can hold different beliefs about the personal qualities being evaluated in achievement situations. Some students believe that performance tasks measure their fundamental and fixed intelligence, whereas others regard these tasks as examinations of more specific, acquirable skills. By considering the meaning system in which achievement goals are embedded, researchers may determine when a certain type of goal will be beneficial versus detrimental. Future research should examine how individuals' systems of meaning may alter the way they perceive and pursue goals. The dynamics of achievement motivation cannot be fully captured without. 55. taking meaning into account (Molden & Dweck, 2000). Third, my findings suggest the possibility that students can pursue multiple pathways to achieve similar outcomes. In other words, they may employ different types of strategies to regulate their cognition, motivation, and affect to attain their goals (Pintrich, 2000b). It is important that future research explore the ways in which different goal profiles lead to the development of self-regulation. In this case, researchers can use path analysis to clarify the underlying processes. As Harackiewicz, Barron, Pintrich, et al. (2002) pointed out, despite a challenging research agenda, the multiple goal perspective offers the potential for a more profound grasp of the complex interactions among motivation, cognition, and achievement.. Note This study was supported by grant no. NSC 92-2413-H-004-007 from the National Science Council, Taiwan, Special thanks go to ChiungHui Tzeng and Ya-Ru Tsai for their assistance with this project.. References Ames, C. (1992), Classrooms: Goals, structures, and student motivation. Journal of Educational Psychology, 84, 261-271, Ames, C, & Archer, J, (1988), Achievement goals in the classroom: Students' learning strategies and motivation processes. Journal of Educational Psychology, 80, 260-267, Archer, J, (1994), Achievement goals as a measure of motivation in university students. Contemporary Educational Psychology, 19,430446, Atkinson, J, W, (1957), Motivational determinants of risk-taking behavior. Psychological Review, 6, 359-372, Bandura, A, (1986), Social foundations of thought and action: A social cognitive theoiy. Englewood Cliffs, NJ: Prenttce-Hall,.

(18) 56. THE ELEMENTARY SCHOOL JOURNAL. Barron, K. E., & Harackiewicz, J. M. (2000). ysis. Journal of Educational Psychology, 76, Achievement goats and optimal motivation: 628-644. A multipte goats approach, tn C. Sansone & EUiott, E. S., & Dweck, C. S. (1988). Goats: An J. M. Harackiewicz (Eds.), Intrinsic and extrinapproach to motivation and achievement. sic motivation: The search for optimal motivation Journal of Personality and Social Psychology, 54, and performance (pp. 229-254). San Diego, 5-12. CA: Academic Press. Farrett, E. (1985). The role of motivational processes Barron, K. E., & Harackiewicz, J. M (2001). in the transfer of learning. UnpubUstied docAchievement goats and optimat motivation: torat dissertation. Harvard University, CamTesting muttipte goat modets. journal of Perbridge, MA. sonality and Social Psychology, 80, 706-722. Grant, H., & Dweck, C. S. (2001). Cross-cutturat Bittings, A. G., & Moos, R. H. (1982). Sociat supresponse to faiture: Considering outcome atport and functioning among community and tributions with different goats. In F. Satiti, C. ctinicat groups: A panet modet. Journal of BeChiu, & Y. Hong (Eds.), Student motivation: havioral Medicine, 5, 295-312. The culture and context of learning (pp. 203219). New York: Ktuwer Academic. Bouffard-Bouchard, T., Boisvert, ]., Vezeau, C, & Larouche, C. (1995). The impact of goat ori- Harackiewicz, J. M., Barron, K. E., Carter, S. M., entation on self-regutation and performance Lehto, A. T, & Ettiot, A. J. (1997). Predictors among cottege students. British journal of and consequences of achievement goats in Educational Psychology, 65, 317-329. the cottege ctassroom: Maintaining interest and making the grade. Journal of Personality Butter, R. (1987). Task-invotving and ego-invotvand Social Psychology, 73,1284-1295. ing properties of evatuation: Effects of different feedback conditions on motivationat per- Harackiewicz, J. M., Barron, K. E., & Ettiot, A. J. ceptions, interest, and performance. Journal (1998). Rethinking achievement goats: When of Educational Psychology, 58,1-14. are they adaptive for cottege students and why? Educational Psychologist, 33,1-21. Church, M. A., Ettiot, A. J., & Gabte, S. L. (2001). Perceptions of ctassroom environment, Harackiewicz, J. M., Barron, K. E., Pintrich, P. R., achievement goats, and achievement outEttiot, A. J., & Thrash, T M. (2002). Revision comes. Journal of Educational Psychology, 93, of achievement goat theory: Necessary and 43-54. ittuminating. Journal of Educational PsycholCovington, M. V. (1992). Making the grade: A self- ogy, 94, 638-645. worth perspective on motivation and school re-Harackiewicz, J. M., Barron, K. E., Tauer, J. M., form. New York: Cambridge University Carter, S. M., & Ettiot, A. J. (2000). Short-term Press. and tong-term consequences of actuevement goats in cottege: Predicting continued interDuda, J. L., & Nichotts, J. G. (1992). Dimensions est and performance over time. Journal of of achievement motivation in schootwork and sport. Journal of Educational Psychology, Educational Psychology, 92, 316-330. 84, 290-299. Haractdewicz, J. M., Barron, K. E., Tauer, J. M., & Ettiot, A. J. (2002). Predicting success in tDweck, C. S. (1986). Motivationat processes afcottege: A tongitudinal study of achievement fecting teaming. American Psychologist, 41, goats and ability measures as predictors of 1040-1048. interest and performance from frestvman Dweck, C. S., & Leggett, E. L. (1988). A sociatyear through graduation. Journal of Educacognitive approach to motivation and pertional Psychology, 94, 562-575. sonatity. Psychological Review, 95, 256-273. Ettiot, A. J., & Church, M. A. (1997). A hierarchi- Harackiewicz, J. M., cfe Sansone, C. (1991). Goats cat model of approach and avoidance and intrinsic motivation: You can get there achievement motivation. Journal of Personalfrom here. In M. L. Maehr & P R. Pintrich ity and Social Psychology, 72, 218-232. (Eds.), Advances in motivation and achievement (Vot. 7, pp. 21^9). Greenwich, CT: JAI. Ettiot, A. J., & Harackiewicz, J. M. (1996). Approach and avoidance achievement goats Linnenbrink, E. A., & Pintrich, P. R. (2000). Muland intrinsic motivation: A mediationat analtiple pathways to learning and achievement: ysis. Journal of Personality and Social Psychol- The rote of goat orientation in fostering ogy, 76, 549-563. adaptive motivation, affect, and cognition. In C. Sansone & J. M. Harackiewicz (Eds.), InEttiot, A. J., & McGregor, H. A. (2001). A 2 X 2 trinsic and extrinsic motivation: The search for actiievement goat framework. Journal of Peroptimal motivation and performance (pp. 195sonality and Social Psychology, 80, 501-519. 227). San Diego, CA: Academic Press. Eltiot, A. J., McGregor, H. A., & Gabte, S. L. (1999). Achievement goats, study strategies, Maehr, M. L. (1989). Thoughts about motivation. and exam performance: A mediationat anatIn C. Ames & R. Ames (Eds.), Research on moSEPTEMBER2005.

(19) GOAL PERSPECTIVE. 57. tivation in education: Goals and cognitions (Vot. Nicholts, J. G. (1984). Achievement motivation: 3, pp. 299-315). New York: Academic Press. Conceptions of abitity, subjective experience, Maehr, M. L., & Kaplan, A. (2000, Aprit). It might task choice, and performance. Psychological be all about self: Self consciousness as an orga- Review, 91, 328-346. nizing scheme for integrating understanding Nichotls, J. G., Cheung, P C, Lauer, J., & Pafrom self-determination theory and achievement tashnick, M. (1989). Individual differences in goal theory. Paper presented at the annuat academic motivation: Perceived ability, meeting of the American Educational Regoats, betiefs, and values. Learning and Indisearch Association, New Orleans. vidual Differences, 1, 63-84. Maehr, M. L., & Midgtey, C. (1991). Enhancing Nolen, S. (1988). Reasons for studying: Motivastudent motivation: A schootwide approach. tional orientations and study strategies. CogEducational Psychologist, 26, 399-427. nition and Instruction, 5, 269-287. McCtettand, D. C. (1951). Measuring motivation Patrick, H., Anderman, L. H., Ryan, A. M., Edein fantasy: The achievement motive, tn H. lin, K. C, & Midgtey, C. (2001). Teachers' Guetzkow (Ed.), Groups, leadership, and men communication of goat orientations in four (pp. 191-205). Pittsburgh: Carnegie Press. fifth-grade ctassrooms. Elementary School McWhaw, K., & Abrami, P. (2001). Student goat Journal, 102, 35-58. orientation and interest: Effects on students' Pintrich, P. R. (2000a). Muttipte goats, muttipte use of setf-regutated teaming strategies. Conpathways: The rote of goat orientation in temporary Educational Psychology, 26,311-329. learning and achievement. Journal of EducaMeece, J. L., Btumenfeld, P. C, & Hoyle, R. H. tional Psychology, 92, 544-555. (1988). Students' goat orientations and cog- Pintrich, P. R. (2000b). The role of goat orientanitive engagement in ctassroom activities. tion in setf-regutated teaming. In M. BoeJournal of Educational Psychology, 80,514-523. kaerts, P R. Pintrich, & M. Zeidner (Eds.), The Middleton, M. J., & Midgley, C. (1997). Avoiding handbook of self-regulation (pp. 451-502). San the demonstration of lack of ability: An unexDiego, CA: Academic Press. plored aspect of goal theory. Journal of Edu- Pintrich, P. R., & De Groot, E. V. (1990). Motivacational Psychology, 89, 710-718. tionat and setf-regutated teaming compoMidgtey, C, Arunkumar, R., & Urdan, T (1996). nents of ctassroom academic performance. "If I don't do weU tomorrow, there's a reaJournal of Educational Psychology, 82, 33-40. son": Predictors of adotescents' use of aca- Pintrich, P R., & Garcia, T. (1991). Student goat demic setf-handicapping strategies. Journal of orientation and self-regulation in the coltege Educational Psychology, 88, 423-434. classroom. In M. L. Maehr & P. R. Pintrich Midgtey, C, Kaplan, A., & Middteton, M. (2001). (Eds.), Advances in motivation and achievement Performance-approach goats: Good for what, (Vol. 7, pp. 371-402). Greenwich, CT JAI. for whom, under what circumstances, and at Pintrich, P R., & Schrauben, B. (1992). Students' what cost? Journal of Educational Psychology, motivational beliefs and their cognitive en93, 77-86. gagement in ctassroom academic tasks. In D. Midgtey, C, Kaplan, A., Middteton, M., Maehr, Schunk & J. Meece (Eds.), Student perceptions M., Urdan, T., Anderman, L., Anderman, E., in the classroom (pp. 149-183). Hitlsdale, NJ: & Roeser, R. (1998). The devetopment and Ertbaum. validation of scates assessing students' Pintrich, P. R., & Schunk, D. H. (1996). Motivation achievement goal orientations. Contemporary in education: Theory, research, and applications. Educational Psychology, 23,113-131. Engtewood Ctiffs, NJ: Merritl. Midgley, C, & Urdan, T. (2001). Academic self- Pintrich, P R., Smith, D. A., Garcia, T, & Mchandicapping and achievement goals: A furKeachie, W. (1993). Retiabitity and predictive ther examination. Contemporary Educational validity of the motivated strategies for learnPsychology, 26, 61-75. ing questionnaire (MSLQ). Educational and Psychological Measurement, 53, 801-813. Molden, D. C, & Dweck, C. S. (2000). Meaning and motivation. In C. Sansone & J. M. Har- Ryan, A. M., Gheen, M. H., & Midgley, C. (1998). ackiewicz (Eds.), Intrinsic and extrinsic motiWhy do some students avoid asking for vation: The search for optimal motivation and hetp? An examination of the interplay performance (pp. 131-159). San Diego, CA: among students' academic efficacy, teachers' Academic Press. sociat-emotionat rote, and the ctassroom goat structure. Journal of Educational Psychology, Newman, R. (1998). Students' help-seeking dur90, 528-535. ing problem solving: Influences of personal and contextual goals. Journal of Educational Satiti, F., Chiu, C, & Lai, S. (2001). The influence Psychology, 90, 644-658. of culture and context on students' motiva-.

(20) 58. THE ELEMENTARY SGHOOL JOURNAL. tional orientation and performance. In F. Salment in school. Contemporary Educational Psyili, C. Chiu, & Y. Hong (Eds.), Student motichology, 22,165-191. vation: The culture and context of learning (pp. Urdan, T (2004). Predictors of academic self221-247). New York: Kluwer Academic. handicapping and achievement: Examining Skaalvik, E. M. (1997). Self-enhancing and setfachievement goats, ctassroom goat strucdefeating ego orientation: Relations with tures, and cutture. Journal of Educational Psytask and avoidance orientation, achievechology, 96, 251-264. ment, self-perception, and anxiety. Journal of Urdan, T, & Midgley, C. (2001). Academic selfEducational Psychology, 89, 71-81. handicapping: What we know, what more there is to learn. Educational Psychology ReStigler, J. W., Smith, S., & Mao, L. (1985). The selfview, 13,115-138. perception of competence by Chinese children. Child Development, 56,1259-1270. Urdan, T, Midgley, C, & Anderman, E. (1998). Stipek, D. J., & Kowatski, P S. (1989). Leamed The role of classroom goal structure in stuhelplessness in task-orienting versus perfordents' use of setf-handicapping strategies. mance-orienting testing conditions. Journal of American Educational Research Journal, 35, Educational Psychology, 81, 384-391. 101-122. Tabachnick, B. G., & Fidell, L. S. (1996). Using Weiner, B. (1990). History of motivational remultivariate statistics. New York: Harpersearch in education. Journal of Educational CoUins Coltege Publishers. Psychology, 82, 616-622. Turner, J. C, Meyer, D. K., Midgley, C, & Patrick, Wolters, C. A. (2004). Advancing achievement H. (2003). Teacher discourse and sixth gradgoal theory: Using goat structures and goat ers' reported affect and achievement behavorientations to predict students' motivation, iors in two high-mastery/high-performance cognition, and achievement. Journal of Edumathematics classrooms. Elementary School cational Psychology, 96, 236-250. Journal, 103, 357-382. Wolters, C. A., Yu, S. L., & Pintrich, P R. (1996). Turner, J. C, Midgley, C, Meyer, D. K., Gheen, The relation between goal orientation and M., Anderman, E. M., & Kang, Y. (2002). The students' motivational beliefs and self-reguctassroom environment and students' related learning. Learning and Individual Differports of avoidance strategies in mathematics: ences, 6, 211-238. A multimethod study. Journal of Educational Wu, J., & Cherng, B. (1992). Motivated strategies Psychology, 94, 88-106. for learning questionnaire (MSLQ): A reUrdan, T C. (1997). Examining the relations vised version for use with Chinese elemenamong earty adotescent students' goats and tary and junior high schoot students. Psychofriends' orientation toward effort and achievelogical Testing, 39, 59-78.. SEPTEMBER 2005.

(21)

(22)