Social Education 57(3), 1993, pp. 119
1993 National Council for the Social Studies
Fouts found no statistically significant relationship between social studies teachers who had coaching responsibilities and those who did not on eight of the nine scales. Noncoaching teachers had higher CES scores on the involvement, affiliation, and teacher support scales, but they scored lower than the coaches on the organization, rule clarity, and teacher control scales. These differences, however, were not statistically significant. On scales measuring openness to system change and innovation, the teachers who coach scored significantly lower than the noncoaches. Fouts concluded that the poorer performance of coaches on the change and innovation scales suggests that social studies teachers who coach have less desirable classroom environments.
We attempted to follow Fouts's procedures in selecting participants for the study. A random sample of fifty-one classrooms with a total of 1,046 students was selected for this study. Included were thirty-five high school and sixteen junior high/middle school social studies classes, selected from six high schools and three middle schools in eight suburbs of a large, midwestern, metropolitan area. The fifty-one classrooms consisted of one class each for forty-seven teachers and two classes each for two teachers. (The Fouts sample included forty-seven secondary classrooms which included twenty-seven high schools, twenty junior high/middle schools, and 1,180 students selected from four high schools and three junior high/middle schools from a suburban area of a large, West Coast, metropolitan area.) We excluded, as did Fouts, honors classes for advanced students and basic skills classes for remedial students. Of the fifty-one classes, twenty-three were taught by teachers coaching one or more sports per year. Twenty-eight of the classes were taught by teachers with no coaching assignments. The teachers' average age was forty-three years and the average number of years of teaching experience was nineteen. Eleven teachers were female and thirty-eight were male.
To assess classroom environment we used the CES used by Fouts. The CES is based on the theory that perception of reality is important in influencing behavior. Evidence is available that students' perceptions of classroom environments, as measured by the CES, are accurate predictors of both achievement and attitude. 2
The CES contains ninety items on four major environmental dimensions: relationship, personal development, system maintenance, and system change. Each dimension has one or more scales. Possible scores for each scale range from zero to ten, and we used the mean scores from all students in the classroom to develop a class profile.
Students in the replication study completed the CES in group settings. The teachers administered the survey, under the direction of a school contact person who had been trained by the research investigator during two months of the second half of the 1990 school year. All parties-teachers, students, and contact people-were made aware of the confidentiality of the data collected. (In the Fouts study, outside research workers administered the CES. Teachers were not informed about the nature of the research-they were simply asked to give the CES to one of their regular classes and return the survey to their contact person.) That contact person verified that the class fit the study design and that the information gathered was similar to the previous study. Teacher data including age, sex, years of teaching experience, and extra-duty coaching responsibilities were provided by school districts, principals, school districts' administrative offices, or by the nonteaching contact person in that school. The coaching teachers were classified as either head coaches or assistant coaches.
Both studies calculated classroom means, standard deviations, and effect sizes on the nine scales of the CES for both studies and used analysis of variance tests to check for significant differences. Fouts notes the usefulness of reporting effect sizes when conducting exploratory research. He explains that an effect size of .2 can be interpreted as small, .5 as medium, and .8 as large (Welkowitz, Ewen, and Cohen 1982). The other measure used in this research is an interpretation of effect size. Fouts uses the effect size and the normal curve to calculate the percentage of one group that is below or above the average of the second group.
We compared our results with Fouts's to see whether the two studies showed similar patterns in classroom environments created by coaching and noncoaching social studies teachers.
In the original study, Fouts reported no statistically significant differences between the two groups of classrooms and teachers concerning the variables of class size, teacher age, and years of teaching experience.
The replication study does show some differences among the teachers. The coaches in the replication study were significantly younger and had taught fewer years than those in the original study.
Using the classroom as the unit of analysis, table 1 shows the CES mean scale scores, standard deviations, and effect sizes for coaching and noncoaching teachers in both the original and replication studies. Table 1 also shows the percentages of noncoaching classes that were above the average coaching teachers' classroom scores for both studies.
Analysis of the replication data shows no statistically significant differences between the classes of coaching and noncoaching teachers on seven scales of the nine-scale environmental dimensions of the CES. Statistically significant differences concerning the relationship and system change dimensions of the CES occurred between the groups on the teacher support (F = 14.16, p.00) and innovation scales (F = 6.71, p. 01) of this instrument with coaching teachers scoring higher than noncoaching teachers.
It is also useful to compare the results of the original study to the replication study. Statistical analyses of the personal development and system maintenance dimensions indicate that no statistically significant differences appeared among the groups' mean scores on these five scales in either study. The analysis of the relationship dimension shows no differences among the involvement or affiliation scales, but indicates a significant difference in teacher support between the coaching and noncoaching social studies teachers in the replication study. The final difference is the system change dimension of the replication study. The innovation scale in this dimension not only is different for coaching and noncoaching social studies teachers but is almost a direct reversal of that reported by Fouts.
Our results differed from Fouts's in two ways. First, the classrooms with coaching teachers in the replication study had higher scores than those of noncoaching teachers on the three scales of involvement, affiliation, and teacher support. Although the differences were not statistically significant in the involvement and affiliation scales, they were significant in the teacher support scale-with effect sizes of -.51 (medium), -.09 (probably no difference), and -1.47 (large), respectively. These effect sizes suggest that on these three scales only 31 percent, 46 percent, and 7 percent of the noncoaching teachers scored above the average coaching teachers' score. This is a reversal of the results of Fouts's study.
A second pattern of responses suggests another reversal. For the personal development dimension scales, the classrooms with noncoaching teachers in the replication study had a higher mean score in task orientation and a lower mean score in competition than the classrooms of coaching teachers. The effect sizes were .59 (medium) and .14 (probably no difference) respectively, showing that on these two scales, 72 percent and 44 percent of the noncoaching teachers' classrooms scored above the average coaching teachers' classrooms. These scores are a reversal of the original results, where the noncoaches scored highest on the task orientation scale of the CES personal development dimension. The high score of the replication classes seems to indicate greater effectiveness on the part of the noncoaching teachers in keeping students on task in their classrooms.
The reversed pattern of responses in the replication study is further suggested by the mean scores on the CES system maintenance dimension with its three scales on order and organization, rule clarity, and teacher control. For these three scales, the classrooms with coaching teachers had mean scores that were consistently lower than or equal to the mean scores of the classrooms of the noncoaching teachers. The effect sizes were .07 (probably no difference), -.01 (probably no difference), and .24 (small) respectively, suggesting that on these three scales 57 percent, 50 percent, and 59 percent of the noncoaching teachers scored above the average of the coaching teachers. Although no statistically significant difference between coaching and noncoaching teachers appeared on this dimension, the pattern suggests a reversal of results reported in the original study. Fouts found that the classes of noncoaching teachers had much lower scores in this dimension than did the coaches' classrooms.
Scores from the system change dimension and the innovation scale suggest that coaching teachers are more innovative in their teaching practices than their noncoaching counterparts. This is shown by the significantly higher mean scores for the coaching social studies teachers and by the relatively large effect sizes. From the effect size -.68 (medium-high), it is calculated that only 25 percent of the classes taught by noncoaching teachers scored above the average of classes taught by coaches. This result again contrasts with Fouts's results, which indicated that coaches lacked innovation in their classrooms.
Finally, the results of both the replication study and the original study seem to indicate that all social studies teachers-coaches and noncoaches alike-are not very innovative in their teaching styles. In both studies, the lowest overall scores were found in this dimension of the CES.
A comparison of the results of these two studies raises doubt about the legitimacy of blaming coaching teachers for poor teaching performance or lack of innovation. The lack of consistent results suggests that athletic coaching by itself may not be a factor.
Why do results differ between the original and the replication study? The simplest explanation is that in spite of our best efforts at replication, the teachers in the two studies were significantly different and not comparable. For example, the replication study included several younger and less experienced teachers who were coaches. If increased age and experience reduce innovation, as measured in these studies, then perhaps the replication study results reflect age and experience differences. The original study did not report age and years of teaching as variables, so it is difficult to be certain.
A second way to look at the two sets of results is to conclude that social studies teachers in general seem unable to get students excited about social studies. The notion is made plausible by a profile of a typical social studies teacher working in our schools today. This profile, based on survey responses of over ten thousand teachers nationwide, suggests that
social studies teachers assume considerable responsibilities in their schools, they appear to be the most alienated. Social studies teachers spend more total hours engaged in school activities, social studies teachers teach larger percentages of less capable students, have larger class loads, spend more time supervising study halls, filling out forms, counseling students, and coaching athletics than teachers in most other academic area subjects. Social studies teachers are isolated from professional contact with colleagues. They have little time for staff development, rarely visit other classrooms to observe and discuss teaching. (Rutter 1986, 253, 255)
Perhaps social studies teachers work in an environment that does not reward effective classroom planning and implementation. The problem may be the extraordinary expectations placed on social studies teachers in the workplace.
The large variation between the studies in innovation scale scores might be explained by the intensity of individual teacher teaching and coaching environment. Pressure on coaches varies greatly from community to community. Some teacher/coaches may be more able than others to balance their coaching and teaching duties effectively. The type and number of coaching assignments, along with the degree of interest teachers have in their coaching assignment, may be responsible for the differences in the data found on this scale.
The results of the replication study suggest that teaching social studies and coaching athletics is not necessarily a combination that is detrimental to social studies programs. More replications in other parts of the country would help us to improve our understanding of this issue. The high teacher support scores evident in the replication may indicate some undiscovered positive effects from coaching and teaching social studies that will emerge in future studies.
1Rudolph H. Moos and Edison J. Trickett, Classroom Environment Scale: Manual and Form R (Palo Alto, Calif.: Consulting Psychologists Press, 1974).2See, for example, Chavez 1984; Fouts 1989; Murray 1938; and Herbert J. Walberg and Geneva D. Haertel, "Validity and Use of Educational Environment Assessments," Studies in Educational Evaluation 6 (1980): 225-38.
Bruckerhoff, Charles E. Between Classes: Faculty Life at Truman High. New York: Teachers College Press, 1991.Chavez, Rudolfo C. "The Use of High-Inference Measures to Study Classroom Climates: A Review." Review of Educational Research 54 (1984): 237-61.Fouts, Jeffrey T. "Coaching Athletics and the Social Studies Classroom." Social Education 53 (February 1989): 117-19, 123.Murray, H. A. Explorations and Personality. New York: Oxford University Press, 1938.Rutter, Robert A. "Profile of the Profession." Social Education 50 (April/May 1986): 382-87.Welkowitz, Joan, Robert B. Ewen, and Jacob Cohen. Introductory Statistics for the Behavioral Sciences. 3d ed. New York: Harcourt Brace Jovanovich, 1982.
Alan Van Deraa is a history teacher at Greenfield High School in Greenfield, Wisconsin, and Mark C. Schug is Professor of Education at the University of Wisconsin-Milwaukee.