Effects of Overtraining on Discrimination Reversal-shift Learning in Children with Moderate and Mild Mental Retardation

The purpose of this study was to investigate effects of overtraining on discrimination reversal-shift learning in children with moderate and mild mental retardation. Subjects were 30 children with moderate mental retardation (mean IQ=40), 30 children with mild mental retardation (mean IQ=59), and 30 children without mental retardation (mean IQ=115). Mean MA was matched to 5-yr level. Each of these groups was subdivided into two conditions: overtraining and no-overtraining. The materials varied on 2 dimensions: size (large and small) and brightness (dark and bright). The relevant dimension was size for all subjects. The overtraining groups received 20 additional trials. The learning criterion on the initial and reversal-shift learning was 10 successive correct responses in a maximum of 60 trials. The main results were as follows: (1) In the group of subjects with moderate mental retardation, the mean number of trials of reversal-shift learning was significantly decreased by the overtraining. The overtraining group had a significantly larger percentage of fast learners (number of trials of initial learning ≧ that of reversal learning) in comparison to the no-overtraining group. Both in the group of subjects with mild mental retardation and in the group with no mental retardation, there was a similar tendency, namely, the number of trials in reversal-shift learning in the overtraining group was fewer than that in the no-overtraining group. (2) Patterns of the percentage of correct responses (every 5 trials) in reversal-shift learning indicated the following results: In the subjects with moderate retardation, no subject belonged to pattern I (attained 100% correct responses with fewer errors) in the no-overtraining group. In the overtraining group, however, 5 subjects belonged to pattern I. This tendency was related to degree of retardation. (3) A higher percentage of the subjects without mental retardation in the overtraining group verbally reported the relevant cue in reversal learning compared to the subjects without mental retardation in the no-overtraining group. On the other hand, no difference was found on this measure between the two groups with mental retardation. From the above results, we concluded that even children with moderate mental retardation could use attentional mediation to resolve discrimination problems. Furthermore, when given overtraining, children with moderate mental retardation could easily extinguish their response to an initially correct cue. For children without mental retardation, overtraining facilitated verbal mediation. But children with either mild or moderate mental retardation could not produce verbal mediation when given non-verbal (attentional) training.