Performance Differences and School Size Focused upon in the first research question was the effect of school size (i.e., small, medium, or large) upon the percentage of boys taking AP or IB exams and the percentage of girls taking AP or IB exams. A statistically significant difference was present for all groups of students, Wilks’ Λ = .95, p < .001, n2 = .02. The effect size for this statistically significant difference was small (Cohen, 1988). Statistically significant differences were present in the univariate follow-up analysis of variance procedures for the percentage of boys taking AP or IB exams, F(2, 900) = 19.02, p < .001, n2 = .04, small effect size, and for the percentage of girls taking AP or IB exams, F(2, 900) = 21.19, p < .001, n2 = .05, small effect size (Cohen, 1988). Thus, both the percentage of boys taking AP or IB exams and the percentage of girls taking AP or IB exams differed as a function of school size. Scheffe` post hoc procedures revealed statistically significant differences were present between students enrolled in small and large schools and between students enrolled in medium and larger schools, but no difference was noted between students enrolled in small and medium schools. The percentage of boys taking AP or IB exams was higher in large schools (i.e., 22.92%) than the percentage of boys taking AP or IB exams was in either medium schools (i.e., 18.20%) or small schools (17.95%). Similar results were present for the percentage of girls taking AP or IB tests. Larger schools had a higher percentage of girls taking either AP or IB exams (i.e., 28.04%) than medium schools or small schools (i.e., from 23.44% to 22.64%). Readers are referred to Table 1 for the descriptive statistics for the percentages of boys and girls taking AP or IB tests as a function of school size. Addressed in the second research question was the effect of school size upon the percentage of boys scoring above the criterion on AP or IB exams and on the percentage of girls scoring above the criterion on AP or IB exams. The MANOVA indicated a statistically significant difference, Wilks’ Λ = .84, p < .001, n2 = .08. The effect size for this statistically significant difference was moderate (Cohen, 1988). Statistically significant differences were present in the univariate follow-up analysis of variance procedures for the percentage of boys scoring above the criterion on AP or IB exams, F(2, 898) = 72.13, p < .001, n2 = .14, large effect size and for the percentage of girls scoring above the criterion on AP or IB exams, F(2, 898) = 84.76, p < .001, n2 = .16, large effect size (Cohen, 1988). Thus, both the percentage of boys scoring above the criterion on AP or IB exams and the percentage of girls scoring above the criterion on AP or IB exams differed as a function of school size. Scheffe` post hoc procedures revealed statistically significant differences were present for all groups. The percentage of boys scoring above the criterion increased as the school size increased (i.e., from 9.78% to 20.49%). Similar results were present for the percentage of girls scoring above the criterion. Small schools had the lowest percentage of girls meeting the criterion (i.e., 12.37%), followed by medium schools (i.e., 20.50%), and larger schools (i.e., 25.69%). Readers are referred to Table 2 for the descriptive statistics for the percentages of boys and girls scoring above the criterion on AP or IB tests as a function of school size. December 2012 / ACEF 10

ACEF Journal Vol 3 Issue 1 December 2012

To see the actual publication please follow the link above