variance-covariance was met [F(6, 2507.55)= .28, p = .95]. Therefore, Wilks' Lambda test statistic was used to interpret the MANOVA results. No significant administration differences were found [Wilks' Lambda= .99, (3, 95)= .34, p = .80, r2 = .01]. A one- way MANOVA was also undertaken to examine administration differences for body dissatisfaction, drive for thinness, and perfectionism scores for Time 2. The Box's test was not significant thus, homogeneity of variance-covariance was met [F(6, 1792.57) = 1.11, p = .35]. Therefore, Wilks' Lambda test statistic was used to interpret the MANOVA results. No significant administration differences were found [Wilks' Lambda =.95, (3, 66) = 1.22, p = .31, r2 = .05]. Mean and standard deviation scores for the direct and mail administration are presented in Table 4-4. Table 4-4. Mean (M) and Standard Deviation (SD) Scores for Gymnasts Who Completed the Study Via Mail and Direct Administration. Time 1 Time 2 Direct Administration Mail-Based Direct Administration Mail-Based Study Measures n = 14 n = 85 n = 12 n = 61 M (SD) M (SD) M (SD) M (SD) Eating Disorder Inventory-2 Subscales Body Dissatisfaction 9.62 (6.60) 8.22 (6.27) 11.18 (8.21) 7.75 (6.80) Drive for Thinness 5.31 (5.42) 5.35 (5.73) 5.45 (5.05) 5.03 (5.84) Perfectionism 6.69 (3.73) 7.01 (4.16) 6.36 (3.75) 7.15 (4.57) Social Physique Anxiety Scale 25.71 (6.37) 25.01 (7.94) 29.25 (6.03) 23.77 (8.39) Comparison of Time Differences The primary purpose of this study was to examine if disordered eating correlates differed for gymnasts between preseason and competitive season. A repeated measures ANOVA indicated no significant time effect for social physique anxiety [F(1,65) = 1.95,