Article 2

Purpose
The purpose of this report is to bring to the attention of the science education community a research direction that may have potential for enhancing children's achievement in science. This work suggests to the authors that understanding more about students' academic efficacy beliefs and the role those beliefs may play in science achievement may have important curricular and instructional implications.

This suggestion is offered as a result of a general study that compared the academic efficacy beliefs of seventh and eighth graders from an inner city K-8 school with those from a suburban junior high school. The study was a general one because it collected data relating not only to science but to several other subject areas. In addition, the study obtained data about the context of schooling, which is referred to as the "sociocultural" context of school.

Although this study was not specifically designed to focus on science education, the decision to share pertinent results with the science education research community was made because the attribute(s) the authors are choosing to call "academic efficacy" correlated significantly with teacher assigned grades. It is recognized that grades do not necessarily guarantee literacy, however, it seems illogical to ignore their potential power. Consequently, the directions posed in this paper may be important given the continuing national concern regarding scientific literacy.

Concern for Scientific Literacy
Nearly 50 years ago, a report entitled "The Endless Frontier" was presented to President Franklin Roosevelt that eventually stimulated the formation of the National Science Foundation's efforts in teacher education and curriculum development (Jackson, 1983). Since then, along with their substantial efforts in teacher education, the Foundation has provided over $117,000,000 for more than 50 separate science and mathematics curriculum projects and even after five decades of such support, more than 300 national studies calling for further educational reform were published in the 1980's alone (Champagne, 1989). Citing Arons (1983); Hurd (1985); Fensham (1986); Massey (1989); and Yager (1989), Cannon and Jinks (1992) concluded that "Of all of the concerns of American education, none have been voiced more consistently than the need to establish a scientifically literate general citizenry." They went on to state that, "At the same time, perhaps no concern has been plagued with such an elusive solution."

There is ample evidence to suggest that issues of curriculum and teacher education are important to the goal of scientific literacy. However, one must wonder if the problem can be reduced to just those factors. It seems reasonable to question if scientific literacy will be reached as a result of any more commissions or state or national committees sounding alarms and subsequently authoring additional achievement goals, objectives and standards that ignore underlying student factors such as academic beliefs. It also seems reasonable to question if the key to scientific literacy is to be found in more "new" curriculum projects which treat the same content in different packages. Perhaps goals, objectives, and standards, although important considerations, are not really at the heart of the problem. Additional factors are probably at work and it is submitted that those factors are ones that operate at the individual learner level.

Interest in Affective Variables
By itself, interest in individual learner variables is not new. Science education has a rich literature pertaining to individual factors such as student attitudes, beliefs, motivation, attributions and concept development. Shibeci (1984) reported that over 200 studies dealing with affective variables were published between the years of 1976 and 1983 and even the most cursory contemporary review will reveal that this literature continues to grow. For example, an ERIC search using the descriptors of "student attitudes and science education" will produce nearly 700 citations since Shibeci's review. Restricting the search by adding a parameter such as "achievement" will reduce the number of citations however the list will still number more than 300. This one example (attitude) illustrates that science education's interest in the general field of affective factors continues and that the literature is very large. However, the potential contributions of research dealing with children's efficacy attributes has not drawn the attention of the science education research community to the same extent has as attributes such as student interest, for example (Garaway, 1994; Koballa, 1988). This is said recognizing that the field is huge. Not only are science educators interested in affective research but so are educators in general. To this add the contributions from the discipline of psychology and the task of surveying this literature and sorting out the probable overlaps in terms and concepts becomes daunting, indeed. In fact, this is one of the research directions posed at the conclusion of this paper.

Nonetheless, the premise of this work is that the learner's sense of academic efficacy as it exists and develops within the context of the schooling experience may be an important factor in science achievement and has been largely overlooked. The premise emerges from a general, exploratory study seeking to document the applicability of Albert Bandura's notions of self efficacy to children's sense of overall academic success (Morgan & Jinks, 1994). As a part of that research, data specific to science was obtained and is reported here with the hope that science educators might find it a useful departure for more specific research.

The Promise of Self Efficacy
Social learning theorists suggest that self efficacy can be thought of as a sense of confidence regarding the performance of specific tasks and, consequently, one's sense of self efficacy may influence several aspects of behavior that are important to learning. These aspects can include choice of activities, effort, persistence, learning, and achievement (Bandura, 1977, 1982, 1989; Schunk, 1989a & b; Zimmerman, Bandura, & Martinez-Pons, 1992).

The most prominent self-efficacy theorist is Albert Bandura (Bandura, 1977; 1981; 1982; 1989a & b). Bandura theorizes that individuals develop general anticipation regarding cause and effect based upon experience. He also suggests that individuals develop specific beliefs regarding their own coping abilities within situation- specific constructs. Consequently, if these theories are applied to the study of children's beliefs about learning specific subjects one might predict that children with high self efficacy regarding academic matters (or, as we refer to it, "academic efficacy") would demonstrate greater success. The apparent dynamic is that self efficacy beliefs are "not simply inert predictors of future behavior" but that those with more efficacious beliefs "make things happen" (Bandura, 1989a, p. 731). This makes sense intuitively and it is supported by research (Brookover, et. al., 1978; Chapman, et. al., 1989; Pintrich and DeGroot, 1990, Pintrich, et. al, 1994; Schunck, 1989a; Skinner, 1985; Skinner, et. al., 1988; Stodolsky, Salk, & Glaessner, 1991; Zimmerman, et al, 1992).

The student academic efficacy scale
The Morgan-Jinks Student Efficacy Scale (MJSES) was designed to gain information about student efficacy beliefs that might relate to school success. Although self report measures seeking similar data do exist the majority of these are focused on older students including adults. Gorrell and Capron, 1988 and 1989; Gorrell and Partridge, 1985, Hackett and Campbell, 1987 are a few cases in point. In other cases, self efficacy data are presented with little accompanying information about the measure itself (Andrews & Debus, 1978; Marsh, et. al., 1991). In yet others, self efficacy data is gleaned from a more concrete activity approach (Bandura and Schunk, 1981; Schunk, 1981, 1982, and 1983). However, Pintrich and DeGroot (1990) collected student efficacy data through a self report scale designed for students of elementary school age. However, in that case self-efficacy, which they refer to as a "motivational belief", is inferred from a set of nine items imbedded in the larger "Motivational Strategies for Learning Questionnaire. Those nine items are similar to those of the MJSES although the reading level of their self-efficacy component is fifth grade as determined by Fry's Readability Graph (Fry, 1977)whereas the MJSES has a Fry reading level of early third grade.

Extended development of the MJSES was completed to assure validity and reliability, using a handbook of scale development by DeVellis (1991) for primary guidance. A version of the scale consisting of 53 items divided into three subscales, as indicated by factor analysis was used in this study. The first subscale consisted of items labeled talent items, which were statements designed to obtain information about students' perceptions of their own innate talent or ability. Example items that resulted in this subscale are as follows:
-I am smart.
-I am a good science student.
-My friends ask me for help with their homework.

The second subscale consisted of effort items, which were statements designed to obtain information about students' perceptions of the role of their effort in completing a task. Example items that resulted in this subscale are as follows:
-Most of my classmates work harder than I do.
-I always get good grades when I try hard.
-I work hard in school.

These two groups of items, taken together, seem to the authors consistent with Bandura's self-efficacy construct and would logically appear to have critical influence upon achievement.

The third factor consisted primarily of socio-cultural, or contextual items and appear related to Bandura's construct of outcome expectancy. This factor is mentioned here to give a complete picture of the scale, but is not the focus of this particular study.

A fourth category, task difficulty, consisted of items that attempted to determine if children assign inherent differences in difficulty to particular subjects. For instance, do children believe that science, for example, is more difficult than the study of social studies? Or is reading easier than mathematics? Items pertaining to this category did not load as a result of factor analysis and their item reliabilities were too low to be useful.

All items were designed for a Likert Scale response using a four interval scale of "Really Agree", "Kind of Agree", "Kind of Disagree", and "Really Disagree". The informal nature of the response categories was an attempt to make the choices consistent with children's language patterns and were analogous to the differences of degree adults define with the more traditional language of Likert Scale formats.

The items for the scale were generated by the authors and subjected to content validity evaluation by three separate panels. The first consisted of five university level teacher educators, the second of four middle school teachers, and the third of fifteen public school students representing grades 4-8. In addition to the panels, the items were also reviewed by the principal of the one of the pilot schools and by an individual published in the area of self report research.

The teachers and teacher educators were asked to categorize the content of each item into one of the original four influence categories- talent, effort, task difficulty or context based on written definitions and examples. They were also asked to rate their confidence on a scale of 1 (not sure) to 5 (very sure) with those decisions. Ambiguous items were either rewritten or eliminated. Items that may have been categorized consistently but in which the judges' confidence about that categorization was low were also rewritten or eliminated. This process resulted in a group of 53 items plus four items requesting grade performance.

In addition to the adult panels, the children were divided in two panels, one consisting of the fourth and fifth graders and the other consisting of the sixth, seventh and eighth graders. The children were given the scale and led through a "think aloud" exercise which was intended to determine if the items were readable, clear in content and within the children's frame of school experience. All of the items were readable and few required adjustment for clarity. The students also expressed comfort with the choices provided on the Likert and demonstrated ability to detect differences among the choices.

The public school principal reviewed the items for appropriate content and readability and concurred with the earlier reviews regarding these issues. The self report research expert also concluded that the items were not ambiguous.

The MJSES makes use of self-report grades as a variable. Subjects were asked what their last grades were in reading, mathematics, science, and social studies, as reported on a progress report or report card. These items were the last four items on the scale. Although having actual academic performance information would be clearly preferable to self-reported data the attendant difficulties in acquiring such data are equally clear. However Dornbusch (1987) and his colleagues made use of self-report achievement data with adolescents and were able to demonstrate .76 correlation between such data and actual grades. Furthermore they demonstrated that the correlation did not decline with poorer grades. Although far from perfect from several perspectives including the fact that Dornbusch worked with high school students where this study deals with junior high youngsters, it was felt that self- reported grade data was sufficiently accurate to warrant including such items at the end of the scale. In circumstances where actual data can be ethically obtained those should be used.

Subjects
Subjects were students in two separate school districts. The first school was part of a district located in a major midwestern urban setting. The K-8 building consisted entirely of an African-American population, was 100% low-income (as defined by criteria used to determine student participants in a federally sponsored free lunch program), and experienced a 53% mobility rate during a typical academic year. Forty-two percent of the students scored at, or above the mean on nationally normed achievement measure with about 70% of students scoring in the middle stanine range. The second district was located in a midwestern suburban setting with a population of approximately 90,000. This participating school housed grades six through eight with 19% of the students considered low-income, using the same criteria as the urban school. Eighty-eight percent of the students were Caucasian, with the remaining 12% consisting of African-American, Hispanic, Asian, and Native American students. This school experienced a mobility rate of 15% during a typical academic year. A total of about 570 responses were obtained from these two schools.

Data Collection Procedures
Students were administered the scale in groups anywhere from 20 in number to 75. In some cases the researchers administered the scale themselves and in others, classroom teachers were trained to administer the scale. In all cases, the same procedures were followed.

After giving informed consent, students were led through example questions to ensure understanding about the purpose and procedures to complete the scale. For example, a sample question such as "I work hard in school" was given to students orally, with discussion following about the various ways a person might respond to it on the Likert scale, emphasizing the individual nature of the answering process. There was additional discussion and practice in actually indicating answers on the computer-scannable sheets provided to students. Students were asked to read carefully and answer honestly. In addition, they were told that while they were to consider the statements and their answers carefully, they were not to labor over any question, so that they could complete the scale quickly. Students were reminded that their answers were anonymous. Generally, students completed the scale within twenty minutes. The reliability coefficient as a result of using Cronbach's Alpha was .78 on the subscale containing the talent items, .66 on the subscale containing the effort items, and .82 on the overall scale.

Findings
The first analysis was to determine if any significant differences between the two schools existed on any of the subscales or on the overall scale. T-tests revealed no differences between the two schools. Consequently, for purposes of this report, all data have been aggregated.

Although a considerable amount of data has been generated from the administration of the scale, this report deals specifically with information concerning science. Subjects' performance in science (as indicated by their self-report science grade)was examined as it related to self-efficacy (talent and effort items) and the overall scale. Table 1 summarizes these correlations.

Table 1

Correlations Between Science Grade and Talent Items, Effort Items,

and Overall Scale

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Items                           Correlations    Significance Level

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Students (n = 570)



Talent Items                         .53            .000



Effort Items                         .35            .000



Overall Scale                        .40            .000

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Conclusions and Departures for Future Research
The purpose of this report was to bring to the attention of the science education community a research direction that our studies suggest could yield useful insights into children's achievement in science. The results reported here demonstrate positive correlations between students' science grades and their sense of academic efficacy as measured by the MJSES. It is recognized that a student's grades in science classes may not guarantee scientific literacy however their power cannot be simply dismissed. Intuitively it must be accepted that grades influence college admissions, scholarship awards, academic choice and persistence as well as career choice.

Since the literature suggests that one's sense of efficacy is learned, as opposed to being a deeper psychological construct, it would seem reasonable to explore curriculum designs that would focus on enhancing students' sense of efficacy in science achievement. On the other hand, correlation and causality are not the same thing. At this point it is not known if higher grades result in higher efficacy or if higher efficacy results in higher grades or if both are the products of some factor(s) that remain, as yet, unrevealed. Regardless, additional research that seeks to pinpoint cause and effect should determine if interventions designed to increase students' science efficacy, with the expectation that higher science achievement would follow, are justified. Research also needs to be conducted that might clarify the place of efficacy within the broader realm of attribution theory. We also need to understand more about the influence of other potentially contributing factors such as locus of control, gender, age, and teacher's efficacy.

If it does turn out that science achievement is likely to be enhanced by increasing students' self efficacy beliefs, the individual items of the MJSES talent/effort subscale might serve as the illustrative "I" statements for such efforts. For example, such items as "I work hard in school", "My teacher thinks I am smart", and "I am a good science student" could be viewed as learner outcomes which might serve as the basis for developing the mind set that probably is one of the building blocks of scientific literacy. Here, again, additional research is needed. Although we have shown that such beliefs are related to grade performance we do not know if there are other efficacy belief indicators that might be even more strongly related.

One of the more intriguing findings in this study, although tangential, was the considerable similarity that existed among the efficacy beliefs of the two groups of students. Although the national concern for the quality and future of the urban school may be at an all time high these results suggest that regardless of the differences in environment, culture, and previous achievement the two groups of students scored similarly on both the talent and effort subscales, as well as the overall scale. Demographically and environmentally these schools are very different. Yet this fact did not make a difference in how students responded to the scale. This may suggest that subject specific efficacy instrumentation could be developed with less concern for inherent cultural and environmental bias than might be the case with achievement assessments. Certainly more research is needed before this speculation can be accepted as fact.

In conclusion, this report has been written with the intent of drawing the attention of the science education research community to the possibility that student efficacy beliefs may be a contributing factor to science achievement and, ultimately, science literacy. Current work suggests that there may be considerable potential here, however, more research is needed before the potential of such beliefs can be fully evaluated.

     Andrews, G. R. & R. L. Debus. (1978). Persistence and the causal perception of failure: Modifying cognitive attributions. Journal of Educational Psychology, 70 (2), 154-166.

     Arons, A. B. (1983). Achieving wider scientific literacy. Daedalus, 112, 91-122.

     Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84 (2), 191-215.

     Bandura, A. (1981). Self-referent thought: A developmental analysis of self-efficacy. In J. H. Flavel & L. Ross (Eds.), Social Cognitive Development: Frontiers and Possible Futures, pp. 200-239. Cambridge, England: Cambridge University Press.

     Bandura, A. & Schunk, D. H. (1981). Cultivating competence, self-efficacy, and intrinsic interest through proximal self-motivation. Journal of Personality and Social Psychology, 41, 586-598.

     Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122-147.

     Bandura, A. (1989a). Regulation of cognitive processes through perceived self-efficacy. Developmental Psychology, 25, 729-735.

     Bandura, A. (1989b). Human agency in social cognitive theory. American Psychologist, 44, 1175-1184.

     Brookover, W., Schweitzer, J., Schneider, J., Beady, C., Flood, P., & Wisenbaker, J. (1978). Elementary school social climate and school achievement. American Educational Research Journal, 15 (2), 301-318.

     Cannon, J. R. & J. L. Jinks. (1992). A cultural literacy approach to assessing general scientific literacy. School Science and Mathematics, 92 (4), 196-200.

     Champagne, A. (1989). Defining scientific literacy. Educational Leadership, 112, 85-86.

     Chapman, M., Skinner, E., & Baltes, P. (1989). Interpreting correlations between children's cognitive performance: Control, agency, or means-ends beliefs? Developmental Psychology, 246-253.

     Dornbusch, S., Ritter, P., Leiderman, P., Roberts, D., & Fraleigh, M. (1987). The relation of parenting style to adolescent school performance. Child Development, 58, 1244 - 1257.

     Fensham, P. J. (1986). Science for all. Educational Leadership, 44, 18-23.

     Fry, E. (1977). Fry's readability graph: Clarifications, validity, and extensions to level 17. Journal of Reading, 21, 249.

     Garaway, G. B. (1994). Language, culture and attitude in mathematics and science learning: A review of the literature. Journal of Research and Development in Education, 27(3), 53-61.

     Gorrell, J.and Capron, E. W. (1988). Effects of instructional type and feedback on prospective teachers' self-efficacy beliefs. Journal of Experimental Education, 56, 120-123.

     Gorrell, J and Capron, E. W. (1989). Cognitive modeling effects on preservice teachers with low and moderate success expectations. Journal of Experimental Education, 57, 231-244.

     Gorrell, J. and Partridge, M. E. (1985). Effects of effort attributions on college students' self-efficacy judgements, persistence, and essay writing. College Student Journal, 19, 227-231.

     Hurd, P. D. (1985). Science education for a new age: The reform movement.
NASSP Bulletin, 9, 83-92.

     Jackson, P. W. (1983). The reform of science education: A cautionary tale. Phi Delta Kappan, 67, 353-358.

     Koballa, T. R. (1988). Attitude and related concepts in science education. Science Education, 72(2), 115-126.

     Massey, W. (1988). Making science accessible. Liberal Education, 74, 16-19.

     Morgan, V. L. and J. L. Jinks. (1994). Self-efficacy and achievement: A comparison of children' beliefs from urban, suburban and rural schools. In J. H. Divine & R. S. Tompkins (Eds.) Interdisciplinary Studies. Proceedings of the 17th National Conference of The Society of Educators and Scholars. Evansville, Indiana. The University of Southern Indiana Press. pp 216-224

     Pintrich, P. R. And E. V. DeGroot. (1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology, 82 (1), 33-40.

     Pintrich, P. R., R. W. Roeser, and E. A. M. DeGroot. (1994). Classroom and individual differences in early adolescents' motivation and self-regulated learning. Journal of Early Adolescence, 14 (2), 139-161.

     Schibeci, R. A. (1984). Attitudes to science: An update. Studies in Science Education, 11, 26-59.

     Schunk, D. H. (1981). Modeling and attributional effects on children's achievement: A self-efficacy analysis. Journal of Educational Psychology, 73, 93-105.

     Schunk, D. H. (1982). Effects of effort attributional feedback on children's perceived self-efficacy and achievement. Journal of Educational Psychology, 74, 548-556.

     Schunk, D. H. (1989a). Self-efficacy and achievement behaviors. Educational Psychology Review, 1, 173-208.

     Schunk, D. H. (1989b). Social cognitive theory and self- regulated learning. In B. J. Zimmerman & D. H. Schunk (Eds.), Self- regulated learning and academic achievement: Theory, research, and practice. New York: Springer-Verlag.

     Skinner, E. (1985). Action, control judgments, and the structure of control experience. Psychological Review, 92, (1), pp. 39 - 58.

     Skinner, E., Chapman, M., & Baltes, P. (1988). Control, means- ends, and agency beliefs: A new conceptualization and its measurement during childhood. Journal of Personality and Social Psychology, 54, (1), 117 - 133.

     Stodolsky, S. S., Salk, S., & Glaessner, B. (1991). Student views about learning math and social studies. American Educational Research Journal, 28, 89-116.

     Yager, R. (1989). Approaching scientific literacy goals in general science education. Journal of College Science Teaching, 58, 273-275.

     Zimmerman, B. J., Bandura, A., & Martininez-Pons, M. (1992). Self-motivation for academic attainment: The role of self-efficacy beliefs and personal goal setting. American Educational Research Journal, 29, 663-676.

The Authors...

Assistant Professor Vicky Morgan is a junior high/middle school educator in the Department of Curriculum and Instruction at Illinois State University at Illinois State University in Normal, IL. She received her Ph.D. from the University of Nebraska-Lincoln in Educational Psychology and held a position at Nebraska Wesleyan University before coming to Illinois State. Her current research interests include student and teacher efficacy, teacher education in professional development schools, and adolescent education in schools that implement middle school philosophy.

To post a comment about this article to the EJSE discussion list, click here.

To get to the top of this page, click here.

To get back to the current issue of the EJSE, click here.

To get back to the EJSE's Archive page, click here.