Participants were from 14 to 16 years of age with their mean and standard division being 14.61 and 0.69, respectively. In the social support variable, the mean for normal students was 96.46 and for the students with learning disabilities their mean was 88.53; in self-injurious behaviors, the mean for normal students was calculated as 16.46 and for students with learning disabilities was 21.66; in physical aggression, the mean for normal students was calculated to be 22.46 whereas for students with learning disabilities, it was 28.43; in verbal aggression, the mean for normal student equaled 20.93 and for students with learning disabilities it was 27.26; and in IQ, the mean for normal students was 105.40 and for students with learning disabilities, it was 104.73.
Before using parametric tests, for considering the main assumptions, Box and Levene’s test was utilized and the homogeneity assumption of variance/covariance was regarded. Significance levels of all tests permitted application of the MANOVA test. The result of Wilks’ lambda indicated that the effect of groups on social support, aggression and self-injurious behaviors was meaningful (Wilks’ lambda = 0.656, F = 7.20, P < 0.001); this result denoted that there was a significant relationship amongst social support, aggression and self-injurious behaviors in students with and without learning disabilities.
As it can be observed in Table 1, the MANOVA result showed there was a significant difference between normal students and students with learning disabilities; in other words, although students with learning disabilities did not get higher scores in social support, they showed higher scores in self-injurious behaviors and aggression.
Table 1.
Results of MANOVA Test to Compare the Variables in Normal Students and Students with Learning Disabilities
Variable | df | MS | F | P | ES |
---|
Social support | 1 | 944.06 | 8.93 | 0.004 | 0.133 |
Self-injurious behaviors | 1 | 405.60 | 23.71 | 0.001 | 0.290 |
Physical aggression | 1 | 534.01 | 8.20 | 0.006 | 0.124 |
Verbal aggression | 1 | 601.66 | 8.99 | 0.004 | 0.134 |
Abbreviations: df, degree of Freedom; ES, Effect size; MS, mean square; P, P value.
To determine the effects of each under-studied variable on the variance of self-injurious behaviors, we entered social support, physical aggression and verbal aggression into the regression model as predictor variables. Moreover, self-injurious behaviors were assigned as the criterion variable. As Table 2 shows, the observed F was meaningful and 14% of the variance of self-injurious behaviors can be explained by social support, physical aggression and verbal aggression.
Table 2.
Coefficients of Predicting Self-injurious Behaviors Based on Social Support, Physical Aggression and Verbal Aggression
Predicting | Non-Standardized Coefficients | | | | | |
---|
Variables | B | SE | β | t | F | R2 | P |
---|
Constant | 22.487 | 7.350 | - | 3.059 | 4.29 | 0.144 | 0.003 |
Social Support | -0.086 | 0.064 | -0.193 | -1.334 | 8 | | 0.188 |
Physical aggression | 0.177 | 0.098 | 0.311 | 1.817 | | | 0.075 |
Verbal aggression | 0.001 | 0.104 | 0.001 | 0.008 | | | 0.994 |
Abbreviations: B, unstandardized partial regression coefficient; β, Standardized partial regression coefficient; t, t tests to assess the significance of individual B coefficients; SE, standard error.
LEAVE A COMMENT HERE: