Promoting the will and skill of students at academic risk: an evaluation of an instructional design geared to foster achievement, self-efficacy and motivation.

The main purpose of this study was to investigate the effectiveness of an instructional design in enhancing the academic competence and confidence of students who are at risk for dropping out of school. It is a comparative study aimed at exploring the role instructional practices adopted by schools play in developing and cultivating self-efficacy beliefs. Thirty-seven students enrolled in an Israeli remedial high-school which applied a learner centered structured academic program geared to raise students' academic achievements while providing them with opportunities to gain confidence in their learning abilities were compared to 15 students enrolled in an Israeli conventional remedial high-school. The results indicate that a structured academic program yields significantly higher achievement and self-efficacy scores, and a higher internal motivational orientation. Findings suggest that applying a program geared to foster both academic competence and confidence provides a beneficial synergy to the student. This study supports the contention of social cognitive theory that to increase achievement educational efforts should focus on raising students' self-efficacy through authentic mastery experiences (Bandura, 1977, 1997). Results are further discussed regarding implications for practice and theory.

**********

Perceived academic self-efficacy is defined as personal judgments of one's capabilities to organize and execute courses of action to attain designated types of educational performances (Bandura, 1997; Zimmerman, Bandura & Martinez-Pons, 1992). These judgments influence how students think, motivate themselves and act (Bandura, 1995). Students' belief in their capabilities to master academic activities affects their aspirations, level of interest, in intellectual pursuits and their academic achievements. Furthermore, these beliefs influence emotional states such as stress, anxiety and depression which can intrude on and impair intellectual functioning (Zimmerman, 1995). Students engage in tasks in which they feel competent and confident and avoid those in which they do not. The higher the sense of efficacy, the greater the effort, persistence and resilience students exert in face of adverse situations. As a result of these influences, self-efficacy beliefs are strong determinants of the level of accomplishment that students finally attain (Bandura, 1997; Pajares, 1996). Thus, self-efficacy beliefs play a key role in setting the course of intellectual development and operate as an important contributor to academic success (Bandura, 1995).

An increasing body of evidence provides support for these assumptions. Path analysis of causality indicated that perceived self-efficacy influences students' learning through cognitive as well as motivational mechanisms (Zimmerman, Bandura & Martinez-Pons, 1992). Further findings suggest that students who believe they are capable of performing academic tasks use more cognitive strategies, persist longer and undertake difficult and more challenging assignments than students who doubt their capabilities (Multon, Brown & Lent, 1991). Findings of numerous experimental studies indicated that academic self- efficacy beliefs are correlated with academic performances such as semester and final year grades, in-class seatwork and homework, exams, essays and reports. These findings also suggested that students' perception of their ability to master academic tasks may predict more accurately their motivation and academic achievements than other psychological constructs (see, Pajares, 1996). This implies that self-efficacy plays a mediational or facilitative role in fostering school engagement, that improving self-efficacy may lead to increased use of cognitive strategies and persistence and thereby to higher academic achievements (Schunk, 1989). The practical implication of the said above is that for students to be academically adept they need to have both the "will" and the "skill" (Pintrich & De Groot, 1990). They need to be equipped not only with knowledge and skills but also with efficacy beliefs and intrinsic interest.

Although self-efficacy research has made notable contributions to the understanding of the role these beliefs play in the development of academic mastery, it has done little to inform school practice (Pajares, 1996). For theory to benefit practice researchers need to aid schools in designing and implementing practices that foster self-efficacy. These practices according to social cognitive theory need to focus on a joint effort to raise competence and confidence. This entails designing interventions which allow students to experience success with the performance at hand through authentic mastery experiences (Pajares, 1996).

As stated by Bandura (1977, 1997) performance accomplishments are the most influential source of efficacy formation since they provide the most genuine evidence. Success raises self-efficacy while failure, particularly if it reoccurs, can lower it. Designing and identifying sensible intervention strategies and practical ways to alter self-efficacy beliefs when they are inaccurate and debilitating to students, has been suggested as an important and viable avenue of future research (Pajares, 1996; Bandura, 1997). Hence, this study focuses on students at academic risk who have failed to succeed within the regular school system and as a result lack both the "skill" and the "will" to be academically adept. These students are exposed to an instructional design geared to raise their academic achievements while providing them with opportunities to gain confidence in their learning abilities.

Academic Risk and Recovery

The term students at risk generally refers to several groups of students who have experienced difficulties and/or failure as learners. Our concern is with students who are at high risk for dropping out of school without benefit of diploma or graduation. Research on the causes of dropping out of school has focused on both the individual and the school or institutional level (Rumberger, 1995). At the individual level demographic factors, family background and the school experiences of the dropout population are emphasized (Roderick & Camburn, 1999). The institutional analysis identifies the characteristics and conditions in schools that promote or reduce the number of dropouts. While schools cannot do much about the demographic and social characteristics of their students, they can impact on the students' decision to remain or leave school. Indeed research illustrates that school policies and practices have important implications for the number of student dropouts (Janosz, Le Blanc, Boulerice & Tremblay, 2000; Rumberger, 1995).

An analysis of high achieving schools that enroll a high proportion of students at risk found that these schools set high academic standards with a firm belief in students' capabilities to fulfill them. They implement mastery-oriented instruction that enables students to exercise control over their academic performances. In particular, such schools refrain from excusing students' poor academic performance on the grounds of low inherent ability or adverse family backgrounds that supposedly render students unmodifiable (Mortimore, 1995).

A review of different alternative programs for students at risk indicates that schools can actively help students overcome their failings (Wehlage et. al., 1989). Two major models have been devised to enhance learning by disadvantaged youth who do not benefit much from conventional instruction (Comer, 1988; Levin, 1996). Both models encompass self- governance and parental involvement but differ in some important respects. The first (Comer, 1988) relies heavily on a positive interpersonal climate to promote educational growth, whereas the second (Levin, 1996) makes greater use of curriculum changes and a variety of proven instructional strategies. A positive school climate will not necessarily cultivate high-level competencies unless it is supported by a stimulating curriculum and informative ongoing monitoring of students' academic progress. Schools can maximize intellectual benefits by combining interpersonal relationships conducive to learning with instructional practices that guide it (Bandura, 1997). This study suggests one such program.

The current study examines the efficacy of a structured academic program in enhancing the academic achievements and confidence of students who are at risk for dropping out of school. These students failed to succeed within the regular school system and as a result lack both the "skill" and "will" to be academically adept. It is a comparative study situated in a natural school setting. As such it explores the role that instructional practices adopted by schools play in developing and cultivating self-efficacy beliefs. Most comparative studies done to date examine differential effects of instruction within a very specific academic domain such as math or writing skills on self-efficacy. An ERIC and PsyINFO search to the end of 2001 yielded no comparative work on self-efficacy within schools. To the best of our knowledge, therefore, this study constitutes the first attempt at an empirical examination of the effectiveness of a school-wide instructional program in promoting perceived academic self-efficacy beliefs while fostering academic competencies.

Method

The participants of this study were 52 high school students characterized as potential school dropouts, who lack the "will" and "skill" to be academically adept. All students had dropped out of the regular school setup and were enrolled at alternative high schools geared for under-achieving students at high academic risk. The schools were randomly selected from a pool of alternative schools in a specific urban school district. The schools served a similar student body, deriving largely from families with a low socioeconomic status. Students age ranged from 14 to 17 (M = 15.95 SD = 1.47). Thirty students were enrolled in classes at an eighth grade level (age M = 15.83 SD = 1.49) and twenty-two students were enrolled in classes at a ninth grade level (age M = 16.11 SD = 1.46). Placement in the different levels was determined according to their acquired academic skills. The amount of hours devoted to academic studies was similar at both grade levels.

The experimental group consisted of 37 students enrolled in an alternative school that adopted a structured academic program, which entailed organizational and instructional modifications of the existing practices within the school. The control group was comprised of 15 students enrolled at an alternative school that used conventional remedial instruction and curricula.

The composition of the groups with respect to ethnic origin was similar. Chi square analyses did not find significant differences between groups in the distribution of the demographic variables of gender _2 (1, N = 52) = 2.75, p >.05, social-economical status_2 (1, N = 51) = 1.49, p >.05, and age_2 (1, N = 52) = 1.44, p >.05. To further ensure that the groups selected were comparable a structured interview was administered to both school principals. An analysis of the responses to the structured interview verified that the schools participating in the study serve a similar student body and expect students to learn similar contents in the realm of language and math skills at the different grade levels. Moreover, the analysis corroborated previous findings that the scholastic performance of students enrolled in 8th and 9th grade level courses in alternative schools for students at high academic risk is equivalent to that of students in the third and fourth grade of elementary school (Cohen-Navot & Givon, 1998). The schools were found to provide the same amount of academic and vocational instruction and were equally equipped with libraries, laboratories and computers. The school faculty all had similar professional training and years of experience.

As mentioned above all students were placed in the different grade levels based upon their acquired skills as assessed by common tests of math and language skills administered by both schools. In order to verify that students of both groups were not only of comparable academic skill but also of comparable intelligence the Standard Progressive Matrices (SPM) was administered to all students. The findings confirmed that the difference between the groups was insignificant F(1,50) = .59, p >.05. The general mean performance on the SPM (M = 34.98 SD = 5.66) revealed that the mean intelligence of students in this study was within the below average range of intellectual functioning.

Schools

The schools participating in this study were remedial institutions aimed at students aged 12-18 who had not been able to adjust to the regular school system and were at high academic risk. The mandate of these schools was to keep the students within the educational system and to enable them to acquire the scholastic, social and work skills required to function in society. To this end, these schools offered a combined program of vocational training in workshops, academic studies, and a variety of other enrichment activities. Upon completing the program the school granted their graduates certificates reflective of their years of schooling.

The experimental school within this study had adopted a learner centered structured academic program developed and adapted for remedial schools by the Youth Advancement Service of the Ministry of Education (1993, 1996). Applying the program entailed organizational and instructional modifications of the schools' existing practices. The structured academic program mainly addresses three levels of study from 8th to 10th grade. However, the number of hours and learning domains at the 10th grade level is significantly different to those at the other grade levels. Hence the study focuses solely on students in 8th and 9th grade level classes.

Learner Centered Structured Academic Program

The structured academic program is an instructional design developed to cater for students who have not been able to adjust to the regular school setup. The program is based on the rationale that learning difficulties are mainly a result of obstacles rather than lack of ability. The major goal of the program is to decrease dropout rate by bringing a significant improvement in students' learning achievements, academic confidence and motivation. Operational practices of the program include individual educational programs (IEPs) that allow for student choice and flexible scheduling within the boundaries of uniform achievement goals in addition to teacher ongoing training and teamwork. It is a modular academic program that includes detailed syllabi of the academic skills and knowledge acquired in the eighth and ninth grades of high school. Students enter the program according to their level of acquired skills. Once they complete the requirements of their current grade they receive a formal and recognized diploma that grants them entrance in to the next grade. The structured and modular nature of the academic program gives students an opportunity to reintegrate into the regular school system.

The learner centered structured academic program is adapted to the needs and interests of each learner and is based on a close relationship between teacher and student. The program is individualized and non-competitive and it allows students to advance at their own pace without having to adhere to dictated norms and content.

Based upon uniform and structured assessments teachers detect students' strengths and weaknesses and construct individualized academic programs. The academic program is developed in relation to detailed syllabi that are available for every subject at every level. The syllabi are meant to provide a guide rather than a prescription. They include recommended sequences of contents and skills in addition to a varied and heterogeneous list of textbooks and computerized learning materials. This list contains a short description of the contents and skills covered in each textbook and/or computer program. The syllabi and assessments create objective criteria and a uniform set of expectations common to all students and teachers. To allow for student choice the selection of instructional strategies and learning materials is left to the joint decision of the instructor and student. Different means are therefore used to the same ends.

Evaluation of student progress is done monthly. The evaluation is based upon short quizzes, performance tasks and projects that are gathered in a student portfolio. The teacher shares the evaluation with the student and together they plan learning goals, topics and materials for the coming month. Instruction is divided in to small units. Examinations and assessments are administered frequently. This allows students to experience success and develop feelings of competence. At the end of the program students are expected to master the contents and skills on the syllabi and pass external exams which are common to all students. Upon completion of the required coursework at each level, students receive formal certificates that are commensurate to their level of achievement.

The joint collaboration and professional development of the school faculty is organized by key figures within the school. The school principal allocates resources and allows for flexible scheduling to provide time for teacher collaboration. The program manager, a contact person who supervises the implementation of the program within the schools, organizes workshops for the teachers and trains program coordinators. The workshops aimed at changing teachers' educational methods and attitudes are crucial to the success of the program and deal with a variety of issues inherent to the program. The program coordinator, who is a school faculty member, organizes and supervises faculty teams in developing learning materials adapted to students' diverse needs. The learning materials become part of a library of resources for the purpose of developing program syllabi. The faculty teams also share their teaching experiences with each other, cooperatively formulate academic expectations and discuss effective learning strategies and teaching methods. This evokes teachers thinking about learning while learning about teaching.

Conventional Academic Program

The conventional academic program offered to the control group is similar in content but differs in structure from the learner-centered program. On the basis of assessment procedures which depict students' strengths and weaknesses teachers devise an IEP for each student. The IEPs include instructional goals only, they do not relate to learning materials or the sequence of curricula. In addition assessment procedures are used to search for commonalties between students with the purpose of placing similar students in the same class. This results in homogenous classes allowing teachers to select a textbook that is suitable to the particular class. This differs from the assessment methods described above, which emphasize the pinpointing of weaknesses and strengths for each individual student as a means for devising an individualized learning program based upon the offered syllabi. The sequence of instruction and use of learning materials is usually determined according to that suggested in the selected textbooks. Instruction is generally applied by using the lecture method in addition to individualized instruction. Individualized instruction is provided when students work on worksheets that are usually common to all students. Students are evaluated by teacher devised internal tests. Upon completion of their studies students receive certificates reflective of their years of schooling. Ongoing training of teachers is organized by the school district to all teachers within the district. Faculty teams of the school meet periodically to discuss student progress.

Administration of Measures

Permission was granted by the school district to conduct the study in the present form. The Raven Standard Progressive Matrices was administered to students participating in 8th and 9th grade classes prior to initiation of the study. Achievement tests, a standardized reading test, measures of self-efficacy and motivation scales were administered to all students towards the end of the school year.

Measures

The Internal consistency reliability of all measures was calculated by Cronbach's alpha. The achievement tests were determined by the students' level of study and reflected the topics and skills taught within each subject. Independent raters, certified teachers in the specific academic domains, agreed on the appropriateness of the questions on the different tests. The total score of all achievement tests was expressed in percentage terms.

Math skill tests. A 25-item test was used to assess students' mathematics skills. The test was based on traditional evaluation procedures involving multiple choice items and open-ended computation problems. The eighth grade test ([alpha] = .83) was divided into six sections and covered the following topics: reading and writing of numbers; computation of natural numbers; simple fractions; decimal fractions; measurements; and number relations. The ninth grade test ([alpha] = .90) was divided into four sections and covered the following topics: percentages; rational numbers; algebra; and geometry.

Language skill tests. These tests assessed the student' s reading comprehension, verbal knowledge and functional writing skills. The eighth grade language skills test ([alpha] = .82) consisted of 22 items which examined five sections. Explicit and implicit reading comprehension skills; the students ability to use a dictionary; conjunctions and prepositions; and basic functional writing skills. The ninth grade language skills test ([alpha] = .78) consisted of 16 questions divided into three major sections. Explicit and implicit reading comprehension skills; application of verbal knowledge such as word meaning, conjunctions and prepositions; and writing skills.

Standardized Reading Test. (Ortar, 1987). The Ortar Reading Test ([alpha] = .80) consists of 68 items followed by several-multiple choice questions. The first 21 questions relate to decoding competency, the next 47 questions tap reading comprehension. The 10 reading comprehension passages are mainly of expository text and vary in their length from 35 to 135 words. The content of the passages represents the content found in most public schools (e.g., social sciences, natural sciences and the arts). This test is widely used in Israel to diagnose students' level of reading comprehension and identify those who perform below their grade level.

Academic Self-Efficacy Scale. Academic self-efficacy was assessed in math ([alpha] = .96) and language skills ([alpha] = .94). The six items of each scale are from the Patterns of Learning Survey (PALS: Midgley et al., 1996). Each scale refers to students' judgments of their capability to complete their schoolwork successfully in that specific domain. The items were scored on a five-point scale with high scores reflecting a high academic self-efficacy and low scores reflecting a low self-efficacy.

Intrinsic versus Extrinsic Orientation Scale. (Hatter, 1981) This instrument is composed of 30 items that relate to motivation for classroom learning in five dimensions along an intrinsic to extrinsic continuum. The first three relate to intrinsic or extrinsic motivational orientation: (a) preference for challenge versus preference for easy work, (b) curiosity versus working to please the teacher and/or to get good grades (c) independent mastery (liking to figure things out on one's own) versus dependence on teacher. The remaining two subscales relate to cognitive-informational orientation: (d) independent judgement about what to do in the classroom versus reliance on teacher's opinion, and (e) internal criteria for evaluation of success or failure versus dependence on external criteria (i.e., grades, approval). The items were scored on a four-point scale with high scores reflecting an intrinsic motivation and low scores reflecting an extrinsic motivation.

In order to reduce the number of variables and simplify data presentation only the high-order factors motivational orientation (MTO) and cognitive informational orientation (CIO) were selected to be reported in this study. Coefficient alpha reliability estimates were .91 for the MTO scale and .84 for the CIO scale. Findings in support of the predictive, discriminant, and construct validity can be found in Harter (1981).

Results

The main objective of the present study was to investigate the efficacy of a learner centered structured academic program designed to promote competence and confidence when it was compared to traditional methods of remedial instruction. Data on four dependent variables were collected: achievement scores in language arts, math skills and on a standardized reading measure in addition to measures of academic self-efficacy and motivational orientation.

Achievement Measures

To test the significance of group differences in achievement scores on math and language skill tests a one-way MANOVA was conducted separately for 8th (n=30) and 9th (n=22) grade students across the two study groups. The analyses revealed significant differences among 8th grade students F(2,27) = 43.92, p < .001, and 9th grade students F(2,19) = 44.64, p < .001. The means, standard deviations and univariate ANOVA of both variables for each grade are presented in Table 1.

As can be seen in Table 1 differences were found among groups on all measures across grades. The experimental group differed significantly from the control group on all achievement measures, with the experimental group obtaining higher scores.

Standardized Reading Measure

An ANOVA was performed using a 2 (control group vs. experimental group) X 2 (8th grade vs. 9th grade) design in order to examine whether achievement scores on the standardized reading measure varied across groups and grades. The analysis revealed a significant difference in the mean scores on the standardized measure of reading comprehension between groups F(1,48) = 45.61, p < .001, with the experimental group obtaining higher scores (M = 52.32; SD = 4.29) than the control group (M = 41.87; SD = 6.59). However, the analysis did not reveal differences in the mean scores of the different grades F(1,48) = 1.12, p > .05, or a significant interaction effect of Group X Grade on the standardized comprehension test, F(1,48) = 1.43, p > .05.

Measures of Self-efficacy

Data on self-efficacy in the realms of language arts and math were collected. Possible scores ranged from 1 to 5 with high scores reflecting high self-efficacy and low scores reflecting low self-efficacy. To test the significance of group differences in self-efficacy a 2 X 2 (Group X Grade) MANOVA was conducted. The results revealed significant differences between the two study groups F(2,47) = 37.99, p <.001. The means, standard deviations and univariate ANOVA of both variables are presented in Table 2.

As can be seen in Table 2 the experimental group differed significantly from the control group and obtained higher mean scores on both measures of self-efficacy.

The MANOVA did not reveal significant differences between grades, F(2,47) = .19, p > .05, or a significant interaction of Group X Grade F(2,47) = .06, p > .05.

To test the relationship between self-efficacy and achievement scores, Pearson correlations were calculated. Significant positive correlations were found between achievement scores in language arts and measures of self-efficacy in language arts and math, r= .42,p < .05; r= .59,p < .01. The achievement scores in math also correlated positively and significantly with self-efficacy measures in language arts and math, r = .56, p < .01; r = .62, p < .01.

Multivariate ANCOVA analyses were carried out with corresponding self-efficacy measures (in the domains of language arts and math) as covariates to determine whether self-efficacy accounts for the variance found between groups on achievement scores. The MANCOVA analyses revealed significant differences between groups but with smaller F scores than those found in analyses which did not use the self-efficacy measures as covariates (see Table 1). Results of the MANCOVA performed on eighth grade students yielded significant effects for method of instruction in language arts, F(1,26) = 32.48, p < .001 and in math F(1,26) = 25.68, p < .001. And for ninth grade students the MANCOVA yielded significant effects for method of instruction in language arts, F(1, 18) = 5.24, p <.001 and in math F(1, 18) = 18.53, p < .001.

Effect sizes were calculated to portray the contribution of self-efficacy to the differences found between groups on achievement scores. Results of the analyses performed on 8th grade students indicated that self-efficacy accounts for 13% of the variance between groups in language arts, and 21% of the variance between groups in math. And for 9th grade students the analyses indicated that self-efficacy accounts for 24% of the variance between groups in language arts, and 27% of the variance between groups in math.

Internal Versus External Motivational Measures

Data on two measures of the Internal versus External Motivational Scale were collected. Possible scores ranged from 1 to 4 with high scores reflecting an intrinsic motivation and low scores reflecting an extrinsic motivation. A 2 X 2 (Group X Grade) MANOVA was conducted in order to examine whether motivational orientation varied across groups and grades. The results revealed significant differences between the two study groups F(2,47) = 6.69, p <.001 on the motivational orientation measure only. Means, standard deviations and the univariate F values obtained from the MANOVA for the two groups are presented in Table 3.

As can be seen in Table 3 the experimental group obtained higher mean scores than the control group on both motivational measures, reflecting an internal motivational orientation. However, as mentioned above, the difference between groups was found to be significant on the motivational orientation scale only.

The MANOVA did not reveal significant differences between grades, F(2,47) = 1.17, p > .05, or a significant interaction of Group X Grade F(2,47) = 2.94, p > .05.

To determine whether the differences between groups revealed on achievement scores account for the differences found between groups on motivational measures, MANCOVA analyses were carried out with both motivational measures (MTO and CIO) as covariates. The MANCOVA analyses revealed significant differences between groups similar to those found without using the motivational measures as covariates, even though the F scores tended to be slightly smaller. Results of the MANCOVA performed on eighth grade students yielded, F(4,72) = 12.63, p < .001. And for ninth grade students the MANCOVA yielded F(4,52) = 20.06, p < .001. These findings indicate that the motivational orientation scales do not impact the differences found between group achievement scores.

To test the relationship between self-efficacy and motivational orientation measures, Pearson correlations were calculated. Significant positive correlations were found between measures of self-efficacy in language arts and math and the motivational orientation (MTO) scale only, r = .35, p < .01; r= .34, p < .01.

Discussion

During the past decade, self-efficacy beliefs have attracted increasing attention in educational research. Experimental studies have verified the contention that self-efficacy plays a key role in the development of academic mastery. However, the movement from theory to practice has been slow. Little has been done to determine whether and how educational institutions may make use of self-efficacy beliefs to benefit students. There is a need to acquaint schools with practical ways to foster and alter self-efficacy beliefs especially when they are inaccurate and debilitating to students. Moreover, the role that schools play in developing and fostering self-efficacy beliefs should be investigated. This study attempts to do this by examining the effectiveness of an applied school-wide instructional design aimed at raising both the academic competence and confidence of students at risk.

The two principal hypotheses of this study is that students at risk who are exposed to a structured academic program will show greater achievements and higher self-efficacy than students who are exposed to conventional methods of instruction. The study focuses on math and language arts, and the results strongly support these hypotheses.

In general potential dropouts tend to score very low on basic math and language skill tests (Cohen- Navot & Givon, 1998; Rumberger, 1987). And, indeed this study finds that students at risk who participate in conventional remedial instruction programs score a failing grade (between 22% to 54%) on achievement tests. However, this study also finds that students who participate in a structured academic program score an average 70% on achievement tests, a score which is equal to a score of all high school students. Moreover, these findings are further corroborated using a standardized measure of reading comprehension. While students belonging to the experimental group (structured program) score an average 70% on the standardized reading comprehension measure which is equivalent to an 8th grade level of reading comprehension. The students from the control group (conventional program) score a failing grade of 50% which is equivalent to a 4th grade level of reading comprehension. These results suggest that a structured academic program may yield significantly better achievement results than conventional instruction. In particular, the results demonstrate the effectiveness of the structured academic program in engaging students who had experienced academic failure in the past. These findings can have important policy implications.

Self-efficacy measures revealed a similar pattern to that of achievement tests. Students exposed to the structured academic program displayed significantly higher perceived self-efficacy in language arts and math than students exposed to the conventional program. This coincides with previous findings of an experimental paradigm in which low achieving students were exposed to an academic program that was supplemented with different instructional treatments to examine their effect on students' self-efficacy. These studies demonstrated that exposing students to components of the structured academic program enhances students' academic confidence and competence (Schunk, 1985; Bandura, 1997). For example, it was found that teaching students to set their own proximal goals, giving frequent feedback to students on performance, and providing a clear standard against which students can gauge their progress all positively affect academic self-efficacy and achievement. These studies, however, examined each factor separately in a particular academic domain and as such had very specific aims and limited applications. In contrast, introducing a multi component instructional design applied school wide, that is geared to raise academic competence and confidence of students at risk may have broad educational implications. Results of this study suggest that learner centered academic structured programs are a viable form of school intervention for students at academic risk whose self-efficacy beliefs seem low and debilitating.

The significant positive correlation found between measures of self-efficacy and achievement scores further supports the effectiveness of the structured academic program. However, it does not indicate whether enhancing achievement fosters self- efficacy, or visa-versa. This is an important issue as it may suggest that the efficacy of the intervention is partially due to the joint focus of the structured academic program to raise competence and confidence.

According to social cognitive theory perceived self-efficacy affects performance accomplishments both directly and indirectly (Zimmerman, Bandura & Martinez-Pons, 1992; Bandura, 1995). Indeed, causality tests consistently show that self-efficacy beliefs contribute significantly to students attainments. In practical terms this means that fostering self-efficacy increases achievement scores over and beyond that gained through teaching academic skills. The findings of this study revealed that self-efficacy beliefs accounted approximately for a third of the variance found between groups on achievement scores. This indicates that the significant difference found between groups on achievement scores is partially attributed to the self-efficacy component inherent to the structured academic program. Thus, applying school practices that foster both competence and the necessary accompanying confidence appears to provide a beneficial synergy to the student at-risk. Furthermore, this finding supports the social cognitive model of academic achievement which contends that to increase achievement educational efforts should also focus on raising students feelings of confidence (Pajares, 1996).

An additional finding of this paper is that students at risk display greater intrinsic motivation on the scale that measures students' preferences for challenge, their curiosity and their independence in mastering skills (MTO) than students at risk exposed to conventional methods of instruction. On the second scale that measures the extent to which students are able to make autonomous classroom decisions and judge whether they are doing well (CIO) no significant differences were found between the groups. This fits with other studies which find that shifts in one dimension of intrinsic orientation does not in general entail a similar shift in the other dimension (Harter, 1981).

It is important to note, however, that directions of the orientation patterns found in this study were markedly different from those discovered for mainstream students of a similar age group. Previous research has established that as students move up the grades they shift toward a lower interest in the studies as ends in themselves (greater extrinsic orientation in the first scale) while becoming more self sufficient in these studies (greater intrinsic orientation in the second scale) (Hatter, 1981). The students considered in this study displayed the opposite orientations. They displayed high intrinsic interest in their studies and a low intrinsic motivation to judge their schoolwork. Potential school dropouts are typically students who have failed to engage in their studies. For such students, therefore, effective learning is a new experience that is bound to challenge and stimulate them. In addition, students at-risk tend to be students who have not had much experience or skill at mastering the school system and as a result may lack an understanding of the system and its requirements.

Measures of self-efficacy in language arts and math correlated positively with the motivational orientation (MTO) scale only. Self-efficacy theory claims that students with high confidence in given domains approach difficult tasks as challenges to be mastered rather than threats to be avoided. These efficacious beliefs foster in turn intrinsic interest (Bandura, 1995). In contrast, theory and research does not relate self-efficacy to autonomous classroom decisions and judgments. This may explain the discrepancies found in this study between the measures of the MTO and the CIO scales. Findings of the motivational scale provide additional support that a learner centered structured academic program geared to foster both academic competence and confidence is an instructional design with much promise for students at risk for dropping out of school.

This study supports social cognitive theory which contends that a joint effort to raise confidence and competence is accomplished through successful performance of authentic mastery experiences (Bandura, 1977). Nonetheless, findings of this study are limited to potential school dropouts enrolled in alternative schools. Research suggests that when integrative high schools adopt reforms aimed at providing an infrastructure for school improvement, low achieving students benefit the most (Janosz, et. al., 2000). Even though this suggests similar findings may be extrapolated to integrated settings with other kinds of low achieving students, further research is greatly needed to generalize the efficacy of structured academic programs to other settings.

Overall, this study has demonstrated the educational benefits of a learner centered structured academic program for students who are potential dropouts. Furthermore, in congruence with social-cognitive theory it suggests that applying an instructional design that is geared to foster both academic competence and confidence provides a beneficial synergy to the student at-risk. Moreover, it suggests that instructional practices adopted by schools may play a major role in fostering self- efficacy. These findings thus suggest that prevailing methods of school design should be altered, at least in the case of students at risk.

Table 1
Means, Standard Deviations and Univariate F Tests of
Achievement Measures for Experimental and Control
Groups by Grade

                Experimental       Control

Subject           M      SD       M      SD         F

Grade 8
Language Arts   72.88   12.26   35.76    5.29   59.61 ***
Math            68.38   12.58   22.32   14.52   67.21 ***

Grade 9
Language Arts   69.99    9.26   53.54    8.27   17.29 ***
Math            70.08   11.38   27.49    6.45   93.57 ***

1 df of the 8th grade (1,28), df of the 9th grade (1,20)

*** p < .001

Table 2
Means, Standard Deviations and Univariate F Tests of
Achievement Measures for Experimental and Control Groups

                  Experimental       Control

Self-Efficacy      M       SD       M       SD      F (1, 48)

Language Arts     4.24     .50     3.20     .41     42.26 ***
Math              4.29     .56     2.91     .46     60.14 ***

*** p < .001

Table 3
Means, Standard Deviations and Univariate F Tests of
Motivational Orientation Measures for Experimental and
Control Groups

             Experimental       Control

Measures      M       SD       M       SD      F (1,48)

MTO          2.95     .49     2.43     .41     13.65 ***
CIO          2.29     .67     2.24     .43       .15 ***

*** p < .001

References

Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215.

Bandura, A. (1995). Exercise of personal and collective efficacy in changing societies. In A. Bandura (Ed.), Self-efficacy in changing societies (pp. 1-45). Cambridge University Press.

Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W. H. Freeman and Company.

Cohen-Navot, M., & Givon, S. (1998). Evaluation of the Miftanim final research report. Jerusalem: The Center for Youth and Children.

Comer, J. P. (1988). Educating poor minority children. Scientific American, 259, 42-48.

Harter, S. (1981). A new self-report scale of intrinsic versus extrinsic orientation in the classroom: Motivational and informational components. Developmental Psychology, 17(3), 300-312.

Janosz, M., Le Blanc, M., Boulerice, B., & Tremblay, R.E. (2000). Predicting different types of school dropouts: A typological approach with two longitudinal samples. Journal of Educational Psychology, 92(1), 171-190.

Levin, H. M. (1996). Accelerated schools after eight years. In R. Glaser and L. Schauble (Eds.), Innovations in learning: New environments in education (pp.329-352). Mahway, N.J.: Lawrence Erlbaum.

Midgley, C., Maehr, M., Hicks, L., Roeser, R., Urdan, T., Anderson, E., & Kaplan, A. (1996). Patterns of Adaptive Learning Survey (PALS). Ann Arbor, MI: Center for Leadership and Learning.

Ministry of Education (1993). The HILA Program. The Youth Advancement Service and Ort. Israel: Jerusalem.

Ministry of Education (1996). The HILA Program and the Miftanim Alternative Schools. The Youth Advancement Service: Tel-Aviv.

Mortimore, P. (1995). The positive effects of schooling. In M. Rutter (Ed.), Psychosocial disturbances in young people (pp. 333-363). New York: Cambridge University.

Multon, K., Brown, S., & Lent, R. (1991). Relation of self-efficacy beliefs to academic outcomes: A meta-analytic investigation. Journal of Counseling Psychology, 38, 30-38.

Ortar, G. (1987). Lets Read: A reading Comprehension Test for Schools. Achiever, Jerusalem.

Pajares, F. (1996). Self beliefs in academic settings. Review of Educational Research, 66, 543-578.

Pintrich, P. R., & De Groot, E. V. (1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology, 82, 33-40.

Roderick, M., & Camburn, E. (1999). Risk and recovery from course failure in the early years of high school. American Educational Research Journal, 36(2), 303-343.

Rumberger, R. W. (1995). Dropping out of middle school: A multilevel analysis of students and schools. American Educational Research Journal, 32 (3), 583-625.

Schunk, D. H. (1985). Participation in goal setting: Effects on self-efficacy and skills of learning-disabled children. Journal of Special Education, 19, 307-317.

Wehlage, G. G., Rutter, R. A., Smith, G. A., Lesko, N., & Fernandez, R. R. (1989). Reducing the risk: Schools as communities of support. New-York: The Falmer Press.

Zimmerman, B. J. (1995). Self-efficacy and educational development. In A. Bandura (Ed.), Self-efficacy in changing societies (pp. 46-68). Cambridge University Press.

Zimmerman, B. J., Bandura, A., & Martinez-Pons, M. (1992). Self-motivation for academic attainment: The role of self-efficacy beliefs and personal goal-setting. American Educational Research Journal, 23, 614-628.

Miriam Alfassi, Bar-Ilan University, School of Education, Ramat-Gan.

Correspondence concerning this article should be addressed to Miriam Alfasse, Bar-Ilan University, School of Education, Ramt-Gan 52900, Isreal. Email: [e-mail]