本研究は精薄児と健常児を対象に,同一次元内で操作された手がかり類似性が弁別逆転学習に及ぼす影響を精神発達との関連をふまえて比較検討することを目的とした。このため,実験は手がかりの類似度が低いと考えられる円-三角形条件と,高いと考えられる円-楕円条件の2条件を設定し,形次元のみを適切次元として実施した。 その主な結果は以下の通りであった。 1. 低MA (IQ)段階(MA5・6∼7歳)において,原学習所要試行数は円-三角形条件群に比べ円-楕円条件群が多かった。これは精薄児,健常児群ともに言えた。また後学習についてみると,精薄児群は円-三角形条件群,円-楕円条件群とも健常児群に比べて多くの試行数を要した。 2. 高MA (IQ)段階(MA8∼10歳)をみると,円-三角形条件群では,精薄児群,健常児群ともに,原,後両学習の所要試行数が低MA (IQ)段階に比べて減少した。円-楕円条件群においても,精薄児群の所要試行数は低MA(IQ)段階より少なかった。しかし,この傾向は健常児ほど顕著ではなかった。 3. 原学習と後学習の所要試行数の分布から弁別逆転移行の難易をみると,低MA(IQ)段階の精薄児群は両条件群とも健常児群に比べて逆転移行が困難であった。高MA(IQ)段階になると,精薄児群は健常児群ほどではないが,逆転移行が容易になった。 4. 手がかり名を適切に言語報告した被験児の割合について比較すると,低MA(IQ)段階では,精薄児群,健常児群とも両条件群において50∼60%であり,顕著な条件差,被験児差はみられなかった。高MA(IQ)段階の円-三角形条件群では,両被験児群とも低MA(IQ)段階に比べて割合が増加した。これに対し,円-楕円条件群における精薄児群の増加は,健常児群のそれに比べて少なかった。 以上から,精薄児,健常児とも媒介型反応様式に至らない段階では手がかり類似性の影響を受け,弁別逆転学習が困難になること,また精薄児はMA水準が高くなると媒介型反応様式が可能になるが,依然として手がかり類似性のような知覚的要因の影響を受けやすいことが示唆された。

Quite a number of studies on the discrimination reversal-shift learning in mentally retarded children found that it was difficult for them to grasp the conception of the relevant dimension. This study was intended to investigate the influence of relative cue similarity, changed within one dimension, on the discrimination reversal-shift learning in mentally retarded and normal children in relation to Ss' MA level. Ss were 60 mentally retarded children without remarkable speech disorders and 60 normal ones. Their MAs (IQs) were 6:3-10:9 (41-87) for the mentally retarded children and 5:8-10:9 (103-143) for the normal ones. Retarded and normal Ss were divided into two MA (IQ) levels, respectively. In discrimination reversal-shift learning, the group of Ss were furthermore divided into two condition groups. They were circle-triangle condition group (C-T Group) whose degree of the relative cue similarity might be low and circle-ellipse one (C-E Group) whose degree might be high. The relevant dimension of the present experiment was only "form". After discrimination reversal-shift learning, all Ss were given dimensional preference test. The main results were as follows: 1. In C-T Group of Iow-MA (IQ) level, mentally retarded children learned reversal shift significantly slower than normal ones. In C-T Group of high-MA (IQ) level, however, retarded children learned as fast as normal ones. 2. In C-E Group of low-MA (IQ) level, both mentally retarded and normal children required many trials of preshift learning. The number of trials of postshift learning in mentally retarded children was more than that of normal children. The number of trials of pre- and postshift in retarded children in high-MA (IQ) level was fewer than in low-MA (IQ) level. 3. It was confirmed that there was no difference in the proportion of appropriate verbal reports between two condition groups, in both retarded and normal children of low-MA (IQ) level. This tendency of retarded children was not changed in high-MA (IQ) level, especially in C-E Group. From the above results, we were found that discrimination reversal-shift learning of normal children done through verbal mediation was not influenced by the relative cue similarity. And it may be suggested that mentally retarded children were easily influenced by relative cue similarity even if their discrimination learning mechanism was dimensional by using verbal mediator.