Impulsivity and academic performance in learning and behavior disordered children

Within this last decade a considerable literature has been developed dealing with the investigation of impulsivity in school-age children. Evidence of the detrimental effects of an impulsive response tempo has prompted numerous attempts to modify such a disposition. While the general effectiveness of a variety of approaches has been demonstrated, the generalizability of reflective training, using match-to-sample figural tasks, to the performance of academic tasks has received little attention. A further question concerns the relative effectiveness of curriculum- based training materials as compared to conventional figural tasks as means of promoting reflective responding.

The present study was designed to address these concerns. Specifically, the purpose was to compare the effects of two reflective training conditions and a control condition on measures of response latency and accuracy and on arithmetic computation and reading comprehension proficiency.

Subjects for the investigation were attending a private residential school for learning disabled and emotionally disturbed children. The entire population of the school was administered the Matching Familiar Figures Test (MFF) and the 30 most impulsive students were identified according to dual latency an1 accuracy criteria. The subjects were then randomly assigned to one of three experimental conditions and administered pretests of arithmetic computation, reading comprehension, and latency and accuracy on selected reading and arithmetic training items.

Those assigned to a Materials Training (MT) condition were trained to use a reflective scanning strategy on items analogous to the MFF and were reinforced for accurate responses. Subjects under a Skills Training (ST) condition were instructed and reinforced in an identical manner but the training items were match-to-sample arithmetic problems and sentences. Those under a control (C) condition were exposed to ST materials but received neither instructions nor reinforcers.

Subjects were seen individually for ten minutes per day over a ten-day period. MFF, training item, and proficiency posttests were administered immediately following training.

Analyses of covariance indicated. significant treatment effects for reading latency (p < .01) and for MFF and reading error (p < .01). Arithmetic error approached significance (.05 < p < .10). Post hoc analyses for reading found significantly longer mean latencies and fewer errors under the ST condition than under either MT or C conditions. The difference between MT and ST conditions was not significant for MFF error although both were superior to the control condition. No significant differences obtained for arithmetic computation or reading comprehension.

Some of the results were at variance with those of previous investigations and the possible causes for these findings and the lack of improvement on the achievement measures were explored. Finally, an overall superiority for ST over MT training was noted in the latency and error data. It was suggested that ST materials, as employed in this study, would be more appropriate in future efforts to modify impulsivity, particularly where improvements in academic performance are sought.