Analysis of the oral discourse of nonhandicapped and emotionally handicapped eight- and nine-year-olds.

Abstract

Language is critical to communication, forming peer relationships, and achieving academic success. Because emotionally handicapped students are described as having difficulty forming peer relationships and achieving academically and because language is critical to these areas, the expressive language characteristics of non-handicapped and emotionally handicapped school age populations must be described and compared. Therefore, the purpose of this research project was to describe and compare the expressive language characteristics of non-handicapped and emotionally handicapped eight- and nine-year-olds, as defined by the mean length of utterance and 13 discourse error categories. This study included 44 subjects, a group of 22 non-handicapped and a group of 22 emotionally handicapped students. Each student was interviewed once by each of two different interviewers. Following the interviews, the language samples (180 communication units for each subject) were transcribed, segmented into communication units, coded, and scored to determine the quality and quantity of discourse errors students made. Multivariate analysis of variance (MANOVA) was used as the primary statistical test of the data to compare group means on all the dependent variables simultaneously. Results indicated that there was no significant difference for the effect of interviewer nor for the interaction effect of interviewer and group. There was, however, a significant difference between the groups, emotionally handicapped and non-handicapped F(11,32) = 3.31, (p < .004) based on the expressive language characteristics considered in this study. While non-handicapped subjects demonstrated a greater mean length utterance, emotionally handicapped subjects committed significantly more errors in the three categories in the area of relation. These categories included poor topic maintenance, inappropriate response, and inappropriate speech style. Using this specific combination of variables, the actual group classification matched the predicated group classification in 86.4% of the cases. The major findings are discussed and implications for education and for future research are also presented.