Factor structure replication and bias investigation of the teacher rating of social skills

The factor congruence and an analysis of potential bias of the Teacher Ratings of Social Skills (TROSS) were the focus of this study. Preliminary research on the TROSS has shown it to possess adequate reliability and validity. A sample of 250 mainstreamed school-age children from four different groups that were behavior-disordered, learning-disabled mildly mentally retarded/educationally handicapped, and nonhandicapped was used to examine (a) rater, ratee, and sex biases in TROSS ratings by teachers, (b) concurrent validity and reliability, and (c) factor congruence with a previous investigation of nonhandicapped children. The results indicate that the TROSS discriminated between mainstreamed handicapped and nonhandicapped students at a reasonably high level. No rater, ratee, or sex biases were found. Coefficient alphas indicate that the TROSS is a highly reliable instrument. The factor structures of the present and previous research were essentially equivalent. In view of these results, the TROSS appears to be an instrument that can confidently be used as a screening instrument in a social skills assessment package.