The consistency of a class of coercive child behaviors across school settings for individual subjects.

This study addressed the problem of the intraindividual consistency of a class of coercive child behaviors across two school settings. Observational data collected on eight coercive behaviors of 53 boys in classrooms and on the playground. Previous Guttman scale analyses had shown that these eight coercive behaviors constituted a response class. The rates of the eight coercive responses in the response class were rank-ordered separately for each of the 53 subjects in each of the two settings. The intraindividual rank-order correlations were all positive, and 37 were statistically reliable. This number of significant correlations was further shown to be significantly greater than chance. It was concluded that there is a substantial amount of cross-situational consistency if intraresponse class analyses are done for individuals.