FACTORIAL VALIDITY OF THE DEVEREUX BEHAVIOR RATING SCALE-SCHOOL FORM

The construct validity of the Devereux Behavior Rating Scale-School Fo rm was examined, primarily through use of factor analysis. The subject s in this study were 384 children ages 5-18 who were either receiving special education services as seriously emotionally disturbed (SED; n = 171) or receiving no special education services (non-SED; n = 213). Parents and/or teachers rated each student on the Devereux Behavior Ra ting Scale-School Form. Item-level factor analyses were conducted for both the 5-12 age version of the scale and the 13-18 age version. Thes e analyses were originally conducted separately by the SED sample and the non-SED sample. Because of differences in these factor structures, the final analyses were conducted with the SED and non-SED samples co mbined but with the effects of group placement partialed out. These an alyses did not replicate the four Devereux subscales. For both age gro ups, a three-factor solution was found to best represent the data. Ite ms did group together by original subscale to some extent, but there w as a substantial amount of item-crossover. Internal consistency reliab ilities of the subscale scores and obtained factor scores were calcula ted and were consistently high. Mean differences between the SED and n on-SED students on the subscale scores and factor scores were also cal culated. SED students were rated as having a significantly higher leve l of problems on all subscale and factor scores. These results are dis cussed and implications related to the use of this scale are highlight ed.