| 520 |
|
|
|a Until recently no time-efficient teacher rating scale of social skills has been available. With the development of the Teacher Ratings of Social Skills (TROSS; Gresham & Elliott, 1984) this gap has possibly been filled. Preliminary research on the scale has shown it to possess adequate reliability and validity (Clark, Gresham, & Elliott, 1985). Little is known yet, however, regarding the possible utility of the scale in discriminating among school-aged children of differing handicapping conditions. The present study examined the discriminative efficiency of the TROSS with a sample of 250 school-age children from four different groups: (a) Behavior Disordered, (b) Learning Disabled, (c) Mildly Mentally Retarded/Educationally Handicapped, and, (d) Nonhandicapped ( normals ). Rater, ratee and sex biases in the ratings by teachers, concurrent validity, and reliability were also examined. Additionally, the scale was factor analyzed and compared to previous findings (Clark, Gresham, & Elliott, 1985). Results indicate that the TROSS, particularly in view of its short length, does discriminate between handicapped and nonhandicapped students at a reasonably high level. The scale does not, however, adequately discriminate among exceptionality groups (BD, LD, MR/EH). No rater, ratee or sex biases were found. Correlations of the TROSS, the WISC-R, and Woodcock-Johnson (concurrent validity) were low, indicating little relationship among the measures. Cronbach's Alpha indicates that the TROSS is a highly reliable instrument, particularly on the Frequency Dimension. The factor structures of the present and previous research are essentially equivalent. The present research indicates that the Tross is a highly reliable instrument which is relatively free of bias and is capable of discriminating between handicapped and nonhandicapped students. Based on these results, the TROSS appears to be an instrument which can confidently be used as a screening instrument in a social skills assessment package. The scale could also be included as part of the general referral/screening package used to refer students for possible special education placement. The Importance Dimension should prove useful in target behavior selection for students receiving social skills training or special education services
|
