Abstract
The autistic children (mean DA 4:11, mean CA 10:9), 10 mentally retarded subjects (mean DA 4:6, mean CA 9:1) and 10 normal children (mean CA 4:11) were tested on three visual discrimination reversal problems: Negative-retained partial reversal (NR), positive-retained partial reversal (PR) and standard reversal (R), in which the negative, positive, or both original learning cues were reversed respectively... the non-reversed cues being replaced by new (novel) cues in the NR and PR problems. The results were as follows. The groups were not significantly different in original learning on mean number of trials to criterion. The autistic and retarded children were significantly inferor to normals in overall reversal. The conditions were significantly different in normals. That is, R was more difficult than NR and PR, although these latter two did not differ from each other. The retardates' performance was at the same level for all three conditions. In the autistic group, however, all conditions were same through reversal training but, the analysis of the correct respoeses to the thirteen reversal trials showed that R was more difficult than NR and PR, where the novel cue was positive. The results were interpreted that the autistic and retarded children showed the inability to inhibit a previously acquired habit and the susceptibility to disruption by novel stimuli through reversal learning.
