Factorial Validity of the Devereux Behavior Rating Scale-School Form

• Published in: Journal of psychoeducational assessment 1996-12, Vol.14 (4), p.334-348
• Main Authors: Gimpel, Gretchen A., Nagle, Richard J.
• Format: Article
• Language: English
• ISSN: 0734-2829; 1557-5144
• DOI: 10.1177/073428299601400402

The construct validity of the Devereux Behavior Rating Scale-School Form was examined, primarily through use of factor analysis. The subjects 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 Rating 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. These 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 combined but with the effects of group placement partialed out. These analyses did not replicate the four Devereux subscales. For both age groups, a three-factor solution was found to best represent the data. Items did group together by original subscale to some extent, but there was a substantial amount of item-crossover. Internal consistency reliabilities of the subscale scores and obtained factor scores were calculated and were consistently high. Mean differences between the SED and nonSED students on the subscale scores and factor scores were also calculated. SED students were rated as having a significantly higher level of problems on all subscale and factor scores. These results are discussed and implications related to the use of this scale are highlighted.