When people encounter problems in translating their goals into action (e.g., failing to get started, becoming distracted, or falling into bad habits), they may strategically call on automatic processes in an attempt to secure goal attain-ment. This can be achieved by plans in the form of imple-mentation intentions that link anticipated critical situations to goal-directed responses ("Whenever situation x arises, I will initiate the goal-directed response y!"). Implementa-tion intentions delegate the control of goal-directed re-sponses to anticipated situational cues, which (when actu-ally encountered) elicit these responses automatically. A program of research demonstrates that implementation intentions further the attainment of goals, and it reveals the underlying processes.

We studied how specific motivational processes are related to the salience of mastery and performance goals in actual classroom settings. One hundred seventy-six students attending a junior high/high school for academically advanced students were randomly selected from one of their classes and responded to a questionnaire on their perceptions of the classroom goal orientation, use of effective learning strategies, task choices, attitudes, and causal attributions. Students who perceived an emphasis on mastery goals in the classroom reported using more effective strategies, preferred challenging tasks, had a more positive attitude toward the class, and had a stronger belief that success follows from one's effort. Students who perceived performance goals as salient tended to focus on their ability, evaluating their ability negatively and attributing failure to lack of ability. The pattern and strength of the findings suggest that the classroom goal orientation may facilitate the maintenance of adaptive motivation patterns when mastery goals are salient and are adopted by students. Recent research on achievement motivation has focused on identifying different types of goal orientations among students, the motivational processes that are associated with these different goals, and the conditions that elicit them. These goal orientations have been contrasted as task involved versus ego involved (Maehr, 1983;Maehr&Nicholls, 1980;Nicholls, 1979, 1984; see also deCharms, 1968, 1976), as learning

The purpose of this article is to examine the contribution made by the self-efficacy component of Bandura's (1986) social cognitive theory to the study of self-regulation and motivation in academic settings. The difference between self-efficacy beliefs and other expectancy constructs is first explained, fol lowed by a brief overview of problems in self-efficacy research. Findings on the relationship between self-efficacy, motivation constructs, and academic performances are then summarized. These findings demonstrate that particu larized measures of self-efficacy that correspond to the criterial tasks with which they are compared surpass global measures in the explanation and prediction of related outcomes. The conceptual difference between the defi nition and use of expectancy beliefs in social cognitive theory and in expect ancy value and self-concept theory is then clarified. Last, strategies to guide future research are offered. In Social Foundations of Thought and Action, Albert Bandura (1986) wrote that individuals possess a self system that enables them to exercise a measure of

The theoretical distinction between goal intentions ("I intend to achieve-c") and implementation intentions ("I intend to perform goal-directed behavior y when I encounter situation z " ; P. M. Gollwitzer, 1993) is explored by assessing the completion rate of various goal projects. In correla-tional Study 1, difficult goal intentions were completed about 3 times more often when participants had furnished them with implementation intentions. In experimental Study 2, all participants were assigned the same difficult goal intention, and half were instructed to form implementation intentions. The beneficial effects of implementation intentions paralleled diose of Study 1. In experimental Study 3, implementation intentions were observed to facilitate the immediate initiation of goal-directed action when the intended opportunity was encountered. Implementation intentions are inter-preted to be powerful self-regulatory tools for overcoming the typical obstacles associated with the initiation of goal-directed actions.

Modeling Modeling is not merely a process of behavioral mimicry. Highly functional patterns of behavior, which constitute the proven skills and established customs of a culture, may be adopted in essentially the same form as they are exemplified. There is little leeway for 25 improvisation on how to drive automobiles or to perform arithmetic operations. However, in many activities, subskills must be improvised to suit varying circumstances. Modeling influences can convey rules for generative and innovative behavior as well. This higher-level learning is achieved through abstract modeling. Rule-governed behavior differs in specific content and other details but it contains the same underlying rule. For example, the modeled statements, "The dog is being petted," and "the window was opened" refer to different things but the linguistic rule-- the passive form--is the same. In abstract modeling, observers extract the rule embodied in the specific behavior exhibited by others. Once they lear...

The purpose of this study was to propose and test a motivational model of high school dropout. The model posits that teachers, parents, and the school administration's behaviors toward students influence students' perceptions of competence and autonomy. The less autonomy supportive the social agents' behaviors are, the less positive are students' perceptions of competence and autonomy. In turn, the less positive students* perceptions are, the lower their levels of self-determined school motivation are. Finally, low levels of self-determined motivation lead students to develop intentions to drop out of high school, which are later implemented, leading to actual dropout behavior. This model was tested with high school students (N = 4,537) by means of a prospective design. Results from analyses of variance and a structural equation modeling analysis (with L1SREL) were found to support the model for all participants and for each gender separately. High school dropout represents an important problem that affects thousands of students each year. Roughly one third of all students will drop out of high school without having received their high school diplomas, both in Canada (Canada Manpower and Immigration, 1990) and in the United States A survey of research on high school dropout reveals that one factor in a student's decision to drop out of school may be motivation (see Preparation of this article was facilitated through a grant and a doctoral fellowship from the Social Sciences and Humanities Research Council of Canada and grants from le Fonds pour la Formation des Chercheurs et l'Aide a la Recherche and the University du Quebec a Montreal. We would like to thank Ed Deci, Andrew Elliott, Marc Blais, Stephane Perreault, and Luc Pelletier for their feedback on an earlier version of this article and Pierre Provencher for his help in collecting the data. Correspondence concerning this article should be addressed to Robert J. Vallerand, Laboratoire de Recherche sur le Comportement Social, Departement de Psychologie, University du Quebec a Montreal, C.P. 8888, Station "Centre-Ville," Montreal, Quebec, Canada H3C 3P8. Electronic mail may be sent via the Internet to vallerand.robertj@uqam.ca. a motivational model of high school dropout on the basis of theory and research in the field of intrinsic and extrinsic motivation. We believe that such an undertaking can yield benefits on three counts. First, it can provide a real-life test of current theory and research on intrinsic and extrinsic motivation. Second, because the proposed model deals with an unfolding sequence, it may serve to integrate existing knowledge on intrinsic and extrinsic motivation, especially as pertains to their antecedents and consequences. Finally, such a model, if proven valid, should provide a better understanding of the process involved in dropping out of high school, thereby leading to potential insights concerning future interventions with this population. Below, we present the motivational model with the supportive evidence. A Motivational Model of High School Dropout The motivational model is depicted in Motivation Toward School One motivational approach that has been found useful in education posits that behavior can be intrinsically or extrinsically motivated Much research supports the validity of the concepts of intrinsic and extrinsic motivation and amotivation in education. For instance, results from several studies have supported the validity of different subscales that assess the concepts of intrinsic and extrinsic motivation and amotivation. Thus, the various subscales have been found to distinguish themselves clearly in factor analyses and to display adequate levels of reliability (Ryan & Connell, 1989, Study 1; On the Social Determinants of School Motivation Research reveals that the social context in education can have an important influence on motivation (see Much research supports this line of reasoning. For instance, controlling behavior from the teacher It is important to emphasize that the motivational model posits that social agents do not influence students' motivation directly. Rather, in line with self-determination theory, social agents' effects are hypothesized to be mediated by students' perceptions of competence and autonomy. Some evidence for this proposi-MOTIVATIONAL MODEL OF HIGH SCHOOL DROPOUT 1163 tion exists. For instance, research has shown that the impact of feedback from a supervisor Dropping Out of High School as a Motivational Consequence Considerable research reveals that motivation can lead to important outcomes. Although most studies have focused on the effects of intrinsic motivation (see , for review), more recent research based on the tenets of selfdetermination theory has dealt with the whole spectrum of motivations. The various forms of motivation are posited to differ in their inherent levels of self-determination. Listed from low to high levels of self-determination, these motivations are amotivation, external regulation, introjection, identification, and intrinsic motivation. Because self-determination has been hypothesized to be associated with enhanced psychological functioning In addition, much experimental (laboratory) research reveals that individuals who are induced to become externally regulated persist much less than those who are intrinsically motivated (see Deci & Ryan, 1985, for a review). This is especially likely to be the case when subsequent engagement in the activity is not compulsory. We feel that the parallel with high school dropout is striking. Although certain students do drop out of high school when they are only 14 or 15 years old, it appears that the majority of students pursue their schooling at least until they are legally obliged lo do so (i.e., until the age of 16, in the Province of Quebec). During that period, they form intentions of either persisting or dropping out of school. A large number of the students who have developed an interest in certain subjects or who desire a career that requires a college degree intend to pursue their schooling. For a substantial portion of students, however, self-determined motivation toward school has become so low (i.e., they have developed low levels of intrinsic motivation and identified regulation but high levels of amotivation and external regulation) that they eventually develop intentions to drop out. The critical point comes when the obligation to remain in school is terminated, that is, when students are 16 years old and have to decide either to stay in or to drop out of school. It is postulated that it is at this point that students act in line with their intentions. Students who have developed intentions to quit school will do so; the others will remain in school. Research in the attitude literature reveals that intentions represent a key predictor of behavior (see Ajzen& Fishbein, 1980). We believe that this relationship applies to the dropout situation as well. Not only is the motivational model theoretically sound, it is also in line with the high school dropout literature. First, with respect to the social context, dropout students, relative to persistent students, report that they participate much less in the decision-making process at school, that they are told to improve more often, and that they are disciplined much more Although several studies from both the motivation and dropout literatures support the different parts of the motivational model, only one, to the best of our knowledge, has focused on the school dropout issue from an intrinsic-extrinsic motivation perspective. In that study, In sum, the motivational analysis presented above and the 1164 VALLERAND, FORTIER, AND GUAY results of the Vallerand and Bissonnette (1992) study suggest that motivation comes into play in the decision to drop out of high school. However, the Vallerand and Bissonnette study did not provide a complete assessment of the motivational model and did not deal with dropping out from high school as such. The purpose of the present study, therefore, was to provide a more thorough test of the motivational model of high school dropout using a prospective design and structural equation modeling. Overall, we believe that the present study should allow us to better understand the psychological processes involved in dropping out of high school as well as provide a test of intrinsicextrinsic motivation theory and research, which underlie the motivational model of high school dropout. Method Participants Participants were 4,537 9th-and lOth-grade French-Canadian students (2,280 boys and 2,245 girls; 12 did not indicate their gender). Participants had a mean age of 14.97 years and came from seven Montreal public high schools. Questionnaire The questionnaire was made up of five parts. In the first part, participants completed three scales that assessed perceptions of different social agents' (parents, teachers, and the school administration) autonomy support in the school domain. 1 Each scale consisted of three items. The Parental Autonomy Support Scale (e.g., "My parents provide me with lots of opportunity to make personal decisions concerning my school activities"), the Teacher Autonomy Support Scale (e.g., "I feel that my teachers pressure me to do what they want"; this scale used reverse scoring), and finally, the School Administration Autonomy Support Scale (e.g., "The school administration generally consults students before introducing new school policies'') had Cronbach alphas of .54, .56, and .65, respectively. 2 Participants rated items on a 7-point Likert scale, ranging from not at all in agreement (I) to completely in agreement (7). These scales were adapted from the Perceived Interpersonal Style Scale In the second part of the questionnaire, participants completed two scales that assessed educational motivational mediators , namely, perceived school competence and perceived autonomy at school. The Perceived School Competence Scale measured students" perceptions of competence in the academic domain. This scale consisted of three items (e.g., "I consider myself to be a good student") and had a Cronbach alpha of .61. Adapted from the Perceived Competence in Life Domains Scale The third part of the questionnaire was the French version of the Academic Motivation Scale (AMS; In the fourth part of the questionnaire, students were asked to complete two items that measured their future schooling intentions ("I often consider dropping out of school" and "I intend to drop out of school"). Responses to this behavioral intention scale were rated on a 7-point Likert scale, ranging from not at all in agreement (1) to completely in 1 One might suggest that it would have been preferable to assess teachers', parents', and school administrators' behaviors instead of students' perceptions. However, cognitive evaluation theory insists that it is not the behavior of others per se that influences one's motivation but rather one's perceptions of such behavior. In addition, research (e.g., Smith, Srnoll, & Curtis, 1979) reveals that children's perceptions of adults in authority positions are quite accurate and in fact more accurate than adults' reports of their own behaviors. We therefore feel that our strategy was appropriate. 2 That some of the scales used in this study yielded alphas in the .50 and .60 range may be regarded as problematic by some researchers. We feel that there is no need for concern on this issue, however, because internal consistency that is based on the alpha coefficient may not be an adequate reliability estimate for a scale made up of a small number of items. Indeed, as noted by Cronbach (1951}, given a small number of items, low alphas can underestimate scale item intercorrelations that are the basis for internal consistency. For instance, given the same average item intercorrelations, the three-item teacher autonomy-support scale, which yielded an alpha coefficient of .56 in this study, would produce an alpha of .81 if there were eight items. With short scales such as the ones used in this study, the adequacy of the underlying measurement model is generally more indicative of the quality of contruct measurement than internal consistency MOTIVATIONAL MODEL OF HIGH SCHOOL DROPOUT 1165 agreement (7). There was a correlation of .63 between these two items. Finally, in the fifth and last part of the questionnaire, participants were asked to indicate their age, student identification number, gender, and date of birth. Procedure In October, during the fall semester, students were asked to complete the questionnaire described above in class. The questionnaire was administered by a trained experimenter according to standardized instructions. The experimenter explained that the purpose of the questionnaire was to learn about the feelings and behaviors of high school students. Students were told that additional information would be gathered later on, and so it was important that they write their student identification numbers on the questionnaire. The experimenter also explained the types of questions that students would be asked to answer and provided examples. It was clearly stated that confidentiality of their answers would prevail at all times. Following these instructions, the experimenter answered questions, and students completed die questionnaire individually. Following completion of the questionnaire, students were thanked for their cooperation. A year later, we contacted the Quebec Ministry of Education to establish a list of students who did not re-enroll in any high school in the Province of Quebec. Once we had this initial list, we contacted the seven schools individually to determine which of the students were true dropouts, that is, those who had hot simply moved to another province or died. Through these procedures, we identified a total of 282 dropout cases. The number of dropout students recorded amounts to a 6% dropout rate. There were 161 boys and 121 girls in the dropout sample, leaving 4,255 participants (2,119 boys and 2,124 girls, and 12 of unspecified gender) in the "persistent" group. Thus, 57% of all dropouts were boys, whereas only 43% were girls. This difference is significant, Results Motivation Toward School and Behavioral Intentions A 2 (type of student: dropout vs. persistent) X 2 (gender) x 7 (type of school motivation) analysis of variance (ANOVA) with repeated measures on the motivation variable was performed on the data. 3 This approach involved taking the scores on the seven motivation subscales and incorporating them into the Type of Student X Gender design as a repeated measure (BMDP; Dixon, Results revealed a significant main effect for school motivation, F(3.20, 14000) = 895.27, p < .0001. Newman-Keuls post hoc analyses revealed that the seven motivational subscales were significantly different from one another. The most important forms of motivation for participants in this sample were, in decreasing order, identified regulation, external regulation, introjected regulation, intrinsic knowledge, intrinsic accomplishment, intrinsic stimulation, and amotivation. Results also revealed a significant main effect for type of student, F( 1, 4521) = 50.75, p < .0001. Results indicated that overall, persistent students scored higher on the motivation subscales than dropout students. Similarly, results revealed a significant main effect for gender, F( 1,4521) -23.01, p < .0001, where overall, female students scored higher on the motivation subscales than male students. Of greater interest, however, results also revealed a significant Type of Student x Motivation interaction, F(3.20, 14000) = 45.51, p < .0001. Simple effect analyses indicated that six of the seven motivational subscales yielded significant differences (p < .0001) between the two types of students. Results showed that dropout students were significantly less intrinsically motivated to accomplish, to know, and to experience stimulation and were less identified and introjected toward education than persistent students. However, dropout students displayed significantly more amotivation than persistent students. Finally, no differences were found between the two types on external regulation. The means and standard deviations of the motivational subscales as a function of type of student appear in With respect to behavioral intentions, 4 results revealed a main effect for type of student, F(l, 4521) = 213.06, p < .0001. Students who eventually dropped out had greater intentions to drop out of school early in the school year than did those who persisted. A gender main effect was also obtained, F(l, 4521) = 11.58, p < .001. Male students reported higher dropout intentions than female students. The interaction was not significant Perceived School Competence and Autonomy We also conducted two 2 (type of student: dropout students vs. persistent students) x 2 (gender) ANOV^s on students' 3 We also performed a 2 (type of student: dropout vs. persistent) x 2 (gender) x 7 (type of school motivation) AN0Y\ with repeated measures on the motivation variables using a random sample of 282 persistent students and the 282 dropout students. The results were remarkably similar to those obtained with the whole sample. 4 Because the measure of intention was not normally distributed, we performed a logarithmic transformation on this variable. The analyses were conducted with this transformed variable. 5 We also conducted a regression analysis to predict behavioral intentions from the motivation scales. Results revealed that four predictors were significant (p < .01): amotivation (/3 = .50), identification (0 = -.11), intrinsic motivation toward stimulation (/? = -.07), and intrinsic motivation toward accomplishment {0 = -.06). These four predictors accounted for 38% of the variance in dropout intentions. perceptions of competence and autonomy. 6 The results of the Perceived School Competence Scale revealed the presence of a type of students main effect, F(1, 4521) = 118.52, p < .0001. Results indicated that dropout students perceived themselves as significantly less competent in school activities than persistent students. A significant gender main effect was also revealed, F(l, 4521) = 6.42. p < .01. Female students perceived themselves as more academically competent than male students. Results of the Perceived School Autonomy Scale also revealed a significant type of student main effect, F(l, 4521) -27.89, p < .0001. Results indicated that dropout students reported feeling significantly less autonomous at school than persistent students. Finally, the gender main effect was also significant, F(l, 4521) = 7.13, p < .01. Results showed that female students felt more autonomous at school than male students. The interactions for these two analyses were not significant (Fs < 1). The means and standard deviations of the Perceived School Competence and Autonomy Scales as a function of type of student appear in Social Agents' Autonomy Support The ANOVAs 7 with the Parental Autonomy Support Scale revealed the presence of a type of student main effect, F(l, 4521) = 51.12, p < .0001. Results indicated that dropout students perceived their parents as significantly less autonomy supportive than persistent students. The gender main effect for this variable, F( 1, 4521) = 1.36, p = .24, was found to be nonsignificant. The results with the School Administration Autonomy Support Scale also revealed a significant type of student main effect, F(l, 4521) -30.34, p < .0001, in which dropouts perceived the school administration as significantly less autonomy supportive than did persistent students. A gender main effect was also revealed, F(l, 4521) -11.27, p < .001. Results indicated that female students perceived the school administration as more autonomy supportive than did male students. Finally, with respect to the Teacher Autonomy Support Scale, a significant gender main effect was revealed, F(l, 4521) = 50.40, p < .0001. Results showed that female students perceived their teachers as significantly more autonomy supportive than did male students. Similarly, the type of student main effect for this variable, F(l, 4521) = 7.33,p < .01, was also significant; dropout students perceived their teachers as being less autonomy supportive than did persistent students. All interactions were not significant (Fs < 1, except for the parental measure, p = .03). The means and standard deviations of the social agents' autonomy support scales as a function of type of student and gender also appear in The Motivational Model of High School Dropout We tested the proposed model (see The latent constructs of parental autonomy support, teacher autonomy support, school administration autonomy support, school competence, and school autonomy were measured by three items each (see Questionnaire section). In light of the already high number of variables in the model, we decided to reduce the number of latent variables assessing motivation to one. This variable, self-determined school motivation, was measured by four separate composite scales reflecting self-deter- 6 We also performed 2 (type of student: dropout vs. persistent) X 2 (gender) ANOVAs on the perceptions of competence and autonomy variables using a random sample of 282 persistent students and the 282 dropout students. The results were very similar to those obtained with the overall sample. 1 We also performed 2 (type of student: dropout vs. persistent) X 2 (gender) ANOVAs on the parental, teacher, and school direction autonomy-support variables using a random sample of 282 persistent students and the 282 dropout students. Once again, the results were almost identical to those obtained with the whole sample. Note. Scores ranged from 1 to 7 and are based on four items for each of the seven motivational subscales. Means differ significantly at p value listed. D Means for this scale are based on logarithmic transformation of scores. mined motivation indexes. A self-determined motivation index MOTIVATIONAL MODEL OF HIGH SCHOOL DROPOUT 1167 Two manifest indicators were used to create the behavioral intentions latent variable (see Questionnaire section). Actual dropout behavior was assessed through a dichotomous variable that reflected enrollment status the following fall semester (0 = re-enrolled; 1 = dropped out). The variance-covariance matrix of the 22 observed variables was used as the database for the analysis. The variance-covariance matrix of the observed variables (which also include the means) is shown in the Appendix. The model was statistically tested using L1SREL VII Structural and measurement coefficients from the completely standardized solution under maximum likelihood are displayed in The less autonomy supportive the parents (/3 = .51) and teachers (P -.35) were, the less competent the students felt. Similarly, the less autonomy supportive the parents (j3 = .41), the teachers (/? = .22), and the school administration (ft = .28) were, the less autonomous the students felt at school. In turn, the less competent {(3 = .32) and autonomous (/5 = .65) students felt, the less self-determined their school motivation. Low levels of self-determined motivation (/? --.67) led to intentions to drop out of high school, which were later implemented (/? = .24). In sum, results from the structural equation modeling analysis strongly supported the motivational model of high school dropout. We also tested the model separately for boys and girls. Results The purpose of this study was to propose a model that integrates existing knowledge on the determinants and consequences of intrinsic and extrinsic motivation and to test it with the real-life social problem of dropping out of high school. This model posits that teachers', parents', and the school administration's autonomy-supportive behaviors toward students influence their perceptions of competence and autonomy. The less autonomy supportive (or the more controlling) the social agents' behaviors are, the less positive are students' perceptions of competence and autonomy. In turn, the less positive students' perceptions are, the lower their levels of self-determined motivation are. Finally, low levels of self-determined motivation are expected to lead students to develop intentions to drop out of high school, which are later acted out. The present results provide strong support for the motivational model of high school dropout. First, results revealed that dropout students had lower levels of intrinsic motivation, identification, and introjection, but higher levels of amotivation, toward school activities than persistent students. Second, as expected, dropout students perceived themselves as being less competent and autonomous at school activities. Third, in line with the motivational model, dropout students perceived their teachers, parents, and the school administration as being less supportive of autonomy than persistent students. Finally, results from structural equation modeling provided support for all proposed relations among the model variables except for the link between the school administration's autonomy-supportive behavior and students' perceived school competence. The present results have important implications for intrinsic-extrinsic motivation theory and research, gender differences, and the issue of high school dropout. We discuss each of these issues in turn. Implications for Intrinsic-Extrinsic Motivation Theory and Research The findings from this study have a number of implications for intrinsic-extrinsic motivation theory and research. A first implication is that motivation, and more specifically, self-determined motivation (or the lack of it), leads to important reallife outcomes, such as dropping out of high school. This is in agreement with recent motivation research, which has shown that self-determined motivation has a host of positive affective (positive affect, psychological adjustment, satisfaction, etc.) and cognitive (concentration, learning, etc.) consequences (see Deci&Ryan, 1985 Vallerand, in press). The present results add to this literature by showing that motivation can also help predict behavioral consequences. These findings are in line with recent research of Vallerand and Bissonnette (1992), which also showed that motivation assessed early in the academic semester can predict future academic behavior, and research of Williams, Grow, Freedman, Ryan, and Deci (1996), which showed that over a 23-month period, self-determined motivation positively predicted attendance at weight loss program meetings. Thus, the present findings reinforce the notion that motivation is a powerful force that leads to action and support self-determination theory's position on the role of self-determination in human behavior. In certain instances, we believe, the impact of motivation on behavior is not direct but is mediated by behavioral intentions. Such a position is in line with abundant research in the attitude literature that reveals that intentions mediate the impact of attitudes on behavior (see We believe that considering behavioral intentions in such situations should lead to at least three advantages. First, incorporating behavioral intentions in the motivational model may better reflect the actual process through which people come to implement behavior. Often, motivation does not lead directly to behavior, especially if the latter occurs months later. However, motivation is still important, as it plants the seed (the intentions) that will eventually grow into behavior. Second, using behavioral intentions should allow a better prediction of behavior. To achieve this goal, however, there should be correspondence between intentions and behavior with respect to elements of action, R The measurement and structural coefficients from the models run separately with male and female students can be obtained from Robert J. Vallerand. 9 To further assess the validity of the model, we tested three alternative models. Alternative Model 1 involved the unmediated direct effects of the exogenous variables (parental, teacher, and school administration autonomy support) on dropout behavior. Alternative Model 2 involved the direct effects of the exogenous variables (as in Alternative Model 1) plus the direct effects of perceived competence and perceived autonomy on dropout behavior. The final model included all of these effects and also incorporated the direct effect of self-determined motivation on dropout behavior. Results with the overall sample provided support for the basic model. First, none of the alternative models yielded an improvement in fit over that of the motivational model (GFI = .96, AGFI = .94, for all three alternative models). Second, of all six direct effects, only one was worth noting, namely that of perceived competence ((3 = -.14 in Alternative Model 2; /? = -.16 in Alternative Model 3). However, it should be noted that this direct effect was smaller than that of intentions (fi = .22) and much smaller than that of competence on self-determined motivation (0 = .32), thereby supporting the basic hypothesis that the impact of perceived competence on dropout behavior is mediated by motivation and behavioral intentions. Nevertheless, in light of these findings, we tested a final alternative model in which only the direct effect of perceived competence was added to the motivational model. Results revealed that this model had the same fit as that of the motivational model (GFI = .96, AGFI = .94). The beta linking perceived competence and dropout behavior was only -.09. The same basic findings were obtained in separate analyses for males (GFI = .95, AGFI = .94; beta for perceived competence = -.12) and females (GFI = .95, AGFI = .93; beta for competence = -.06). Overall, the results of these analyses provide additional support for the validity of the motivational model. 1170 VALLERAND, FORTIER, AND GUAY target, situation, and time 10 By assessing behavior and intentions at the same level of correspondence, future researchers should be able to show an intimate relation between these two constructs. Finally, the integration of behavioral intentions in our motivational model allows us to make use of knowledge acquired from the attitude-behavior literature in order to better predict and understand behavior. For instance, attitude research has revealed the existence of moderators of the link between attitude and intentions or behavior. An individual's direct experience with a task A second implication of the present findings for motivation theory and research is that there is an important parallel between the present findings and those from the experimental (laboratory) research on intrinsic motivation. Such research typically induces a loss of intrinsic motivation through a manipulated independent variable (e.g., a controlling reward, such as money) and leaves the participant alone in the room with the opportunity to engage further in the activity. Results generally reveal that rewarded participants persist much less on the activity during the subsequent free-choice period than nonrewarded participants. The present findings reveal that, like the rewarded participants from laboratory experiments, dropout students find the task (school) less interesting and eventually disengage from it. However, the present findings have also shown that it is not only a lack of intrinsic motivation that is at fault but also a loss of identified regulation (or purposeful extrinsic motivation) and a concomitant increase in amotivation. In light of the present findings, researchers should assess whether these other motivational states (identified regulation and amotivation) play a causal role in the lack of persistence observed in laboratory settings. It should be noted that the results of the present study that pertained to introjected regulation were somewhat surprising. Persistent students displayed higher levels of introjection than dropout students. Considering that introjection is a non-selfdetermined form of motivation, one would have expected dropout students to score higher than persistent students-and even more so in light of findings in other life contexts, such as politics A third implication of interest deals with the role of the social context as a source of influence on motivation. Three points must be made in this respect. First, as shown by the present results, the impact of social context on motivation is mediated by the individual's perceptions of competence and autonomy. Thus, others' behaviors will influence our motivation only in cases in which they affect our perceptions of competence and autonomy. Several theories, such as cognitive evaluation theory , self-efficacy theory Much less research has focused on the link between perceptions of autonomy and motivation. A second, related point deals with the relative impact of social-context-engendered perceptions of competence and autonomy on motivation. On one hand, theorists such as Bandura (1986) suggested that perceptions of autonomy are not especially useful because only perceptions of competence (or selfefficacy) are important predictors of motivation. On the other 10 Another potential explanation for the facl that the link between intention and behavior was not stronger in this study is that there was very little variation in behavior, as only 6% of the students dropped out. This lack of variance in behavior may have reduced the intentionbehavior relationship. Tb further test this hypothesis, we conducted structural equation modeling analyses (with LISREL) on the motivational model with equal numbers of dropout students {« = 282) and persistent students (a random sample of 282 students). Results replicated those obtained with the overall sample (GFI = .92, AGFI = .90), and all links were very similar. However, the beta for the intention-behavior path moved from .24 to .43, thereby providing some support for our hypothesis. MOTIVATIONAL MODEL OF HIGH SCHOOL DROPOUT 1171 hand, theorists such as n Future research is needed to better understand the relative impact of perceived competence and autonomy on motivation. A third and final point with respect to the social context refers to the strategy used to assess the influence of social agents on motivation. Much of the relevant intrinsic-extrinsic motivation research in education has focused on the impact of either teachers (e.g., It should also be noted that in the present study, we focused on autonomy-supportive behavior from important social agents. However, Deci, Ryan, and then: colleagues Gender Differences in Motivation and Behavior The present findings also revealed that girls displayed a more self-determined motivational profile than boys. This is in line with research that has explored gender differences in intrinsic and extrinsic motivation in various life contexts. Such research generally reveals that women display a more self-determined motivational profile than men in a diversity of life activities, such as sports, leisure, interpersonal relationships, and education (see The present findings, which show that girls display a more self-determined motivational profile than boys, are intriguing because they seem to run counter to past findings that women display higher levels of learned helplessness than men (see 1172 VALLERAND, FORTIER, AND GUAY life contexts (see In light of the important consequences for school persistence (and other outcomes) that these gender differences in motivation seem to engender, research on the determinants of these motivation differences is definitely called for. We feel that the motivational model includes important determinants of these gender differences. For instance, the social context at school may be responsible to some extent for these gender differences. Results from this study revealed that boys perceived their teachers as being less supportive of autonomy than girls did. Past research has also revealed that teachers do not act in the same way toward boys and girls in the classroom: Boys receive much more criticism from teachers (e.g., On the High School Dropout Issue Results from the present study have at least two important implications for the high school dropout issue. First, motivation, and specifically self-determined motivation, is a key variable for one to consider when attempting to predict high school dropout. The present results showed that four types of motivation were found to distinguish dropout students from persistent students: lower levels of intrinsic, identified, and introjected regulation but higher levels of amotivation. These results are in direct agreement with those of These findings are also in line with research on high school dropout that has shown dropout students to display more alienation (a construct akin to that of amotivation) but fewer positive attitudes and less liking (constructs related to intrinsic motivation) toward school than persistent students (e.g., Aslone & McLanahan, 1991;BeardenetaL, 1989; A second implication is that the social context plays a fundamental role in the dropout process. Much theorizing and research has focused almost exclusively on the personal determinants of dropping out of high school {see Finally, results revealed that parents' influence on motivation (through their impact on their children's perceptions of competence and autonomy) proved significantly more important than that of teachers and school administrators. 12 These findings, especially with respect to teachers, may have resulted from two methodological aspects of this study. First, the teacher autonomy-support measure was taken in October, early in the school year. It is thus possible that students had not yet formed clear impressions of their teachers. Second, the measure that involved teachers asked students to assess their teachers "in general." 12 Using the same strategy as that discussed in footnote 11, we compared the paths involving parents, teachers, and the school administration, on one hand, and perceptions of competence and autonomy, on Lhe other. Here again, all differences in chi-square values from the equality models relative to the motivational model were significant, average &X 2 ( 1, AT = 4,537) -33.39, p < .05. These results revealed that the paths involving parental autonomy support and perceptions of competence and autonomy are statistically different from those involving teachers and the school administration. MOTIVATIONAL MODEL OF HIGH SCHOOL DROPOUT 1173 This task of "averaging out" perceptions of their teachers may have proven difficult for students. Still, we believe that our finding that parents exert a stronger influence than teachers and school administrators on their children's perceptions of competence and autonomy is valid and important. One is reminded that teachers usually teach students for a year and that most students have few direct interactions with the school administration. Conversely, by the time their children reach the age of 15, most parents will have spent close to 10 years engaging in school-related activities with them. The greater impact of the parents is therefore understandable. This finding implies that if we are to do something about the high school dropout problem, we must involve the parents and not simply focus on teachers or the school administration (see Identified and introjected forms of political intcmalization: Extending self-determination theory. Journal of Personality and Social Psychology. 70, 1025-1036

The causal role of students ' self-efficacy beliefs and academic goals in self-motivated academic attainment was studied using path analysis pro-cedures. Parental goal setting and students ' self-efficacy and personal goals at the beginning of the semester served as predictors of students' final course grades in social studies. In addition, their grades in a prior course in social studies were included in the analyses. A path model of four self-motivation variables and prior grades predicted students ' final grades in social studies, R =.56. Students ' beliefs in their efficacy for self-regulated learning affected their perceived self-efficacy for academic achievement, which in turn influenced the academic goals they set for themselves and their final academic achievement. Students ' prior grades were predictive of their parents ' grade goals for them, which in turn were linked to the grade goals students set for themselves. These findings were interpreted in terms of the social cognitive theory of academic self-motivation. BARRY J. ZIMMERMAN is a professor of educational psychology in the Graduate

Abstract not found

During the past two decades, self-efficacy has emerged as a highly effective predictor of students ’ motivation and learning. As a performance-based measure of perceived capability, self-efficacy differs conceptually and psychometrically from related motivational constructs, such as outcome expectations, self-concept, or locus of control. Researchers have succeeded in verifying its discriminant validity as well as convergent validity in predicting common motivational outcomes, such as students’ activity choices, effort, persistence, and emotional reactions. Self-efficacy beliefs have been found to be sensitive to subtle changes in students ’ performance context, to interact with self-regulated learning processes, and to mediate students’ academic achievement. © 2000 Academic Press Educators have long recognized that students ’ beliefs about their academic capabilities play an essential role in their motivation to achieve, but selfconceptions regarding academic performance initially proved difficult to measure in a scientifically valid way. Initial efforts to study students ’ selfbeliefs gave little attention to the role of environmental influences, such as

Fantasy realization theory states that when people contrast their fantasies about a desired future with reflections on present reality, a necessity to act is induced that leads to the activation and use of relevant expectations. Strong goal commitment arises in light of favorable expectations, and weak goal commitment arises in light of unfavorable expectations. To the contrary, when people only fantasize about a desired future or only reflect on present reality, expectancy-independent moderate goal commitment emerges. Four experiments pertaining to various life domains supported these hypotheses. Strength of goal commitment was assessed in cognitive (e.g., making plans), affective (e.g., felt attachment), and behavioral terms (e.g., effort expenditure, quality of performance). Implications for theories on goal setting and goal striving are discussed.