Criterion-related versus non-criterion-related prompt training with severely mentally handicapped children.

Thirty-three severely mentally handicapped children were involved in a study to evaluate the effectiveness of three training programmes in teaching visual discriminations. Three matched sets of subjects each received both a size and an intensity discrimination with either: (1) size prompting; (2) intensity prompting; or (3) no prompting (trial-and-error training). The prompted groups therefore received both criterion-related (CR) and non-criterion-related (NCR) training. Order of presentation of the discriminations was counterbalanced across subjects. There was a significant main effect of training group with both size and intensity prompting groups making significantly fewer errors than the trial-and-error group (P less than 0.05 in both cases), but not differing significantly from each other. Significant linear trends for increasing errors across CR, NCR and trial-and-error training programmes were found for both the size discrimination (P less than 0.01), and intensity discrimination (P less than 0.05). More children acquired the discriminations when trained with a CR prompt than with a NCR prompt or trial-and-error procedure, and all four children who failed to acquire discriminations with NCR prompts subsequently acquired them when trained with a CR programme. The results are discussed in relation to cue discriminability and theoretical explanations involving overshadowing are considered.