| Abstract: | The generalized effects of self-instructional training on the classroom performance of three “impulsive” preschool children were investigated using a multiple-baseline design across subjects. Measures of child and teacher behavior in the classroom were obtained through direct observations during a daily independent work period. Self-instructional training followed Meichenbaum and Goodman's (1971) approach, except that training materials consisted of naturalistic task worksheets rather than psychometric test items and training sessions were of shorter duration. For all three children, self-instructional training resulted in increased levels of accuracy on worksheets in the classroom that were similar to those used in training. Results related to several supplementary measures were less clear; however, they suggested that rates of on-task behavior may also have improved, and that a mild classroom intervention further strengthened on-task rates and effected consistent work completion for all three children. The findings suggested that generalized increases in accuracy on classroom worksheets were related to the naturalistic format of the self-instructional training sessions. The level of teacher attention was controlled to rule out its effect on changes in child behavior. |
