幼児における過剰訓練逆転効果と報酬量

概要

論文の詳細を見る
An incentive model (Theios & Blosser, 1965a, 1965b) of overtraining reversal effect (ORE) in animal discrimination learning predicts that there is an interaction between degree of training and magnitude of reward in such way that ORE occurs more often with increasing reward magnitude. The purpose of the present experiment was to test this prediction using kindergarten children. A 2×3 factorial design was used, with degree of original training (criterion and overtraining) and reward magnitude (0, 1, and 3 marbles). The Ss were 72 kindergartners with a mean age of 6 years and 4 months, and divided into 6 groups. The task was 2-choice simultaneous discrimination with color (red and yellow) and form (circle and square) dimensions. The Ss in marble groups were instructed to receive marbles for each correct responses with E's saying "Hit", but not to receive in case of incorrect responses with E's saying "Miss". No instruction was given for the Ss in no-marble groups. Half the Ss were trained to a criterion of 10 successive correct responses and the rest were given additional 20 overtraining trials. Immediately after completing the original learning, they were given a reversal discrimination to reach 10 successive correct responses or 80 trials. The reward condition for each group did not change throughout the experiment. Table 1-shows mean number of trials to criterion in the original and reversal learnings. Only the main effect of learning was significant, indicating that reversal learning was faster than original learning (see Table 2). Although there was no statistical significance, inspection of Table 1 suggests that (a) the original learning tended to be slower with increasing magnitude of reward, (b) ORE occurred in all reward conditions, and (c) degree of ORE measured by the difference in trials between criterion and overtraining groups decreased with increasing magnitude of reward. The last result is opposite to the prediction from the incentive model. The inferior performance of the marble groups may be due to the dissociation between positive stimuli and reinforcing stimuli (marbles) or the distracting effect suggested by Spence (1970). The present results and previous studies (Hirai, et al., 1970; Sugimura, 1965) suggest that contrary to animal studies the incentive model is not confirm in case of children. The discrepancy between animal and children's results may be explained in terms of the difference in perseverative errors during reversal learning. The incentive model presupposes the existence of frustrative responses produced by nonrewarded trials which occur successively during perseverative errors. Frustrative responses are strengthened by overtraining since many animal studies have shown to occur more such errors after overtraining trials. If number of perseverative errors are small and no difference is found between criterion and overtraining groups, there is no difference in the strength of frustrative responses and then in the performance of reversal learning. Table 3 verifies these reasonings.
奈良教育大学の論文
1971-10-30

著者

杉村健
奈良教育大学心理学教室

幼児における過剰訓練逆転効果と報酬量

スポンサーリンク

概要

著者

関連論文

スポンサーリンク