心理学研究
Online ISSN : 1884-1082
Print ISSN : 0021-5236
ISSN-L : 0021-5236
間歇強化学習に關する実験的研究
石原 靜子
著者情報
ジャーナル フリー

1954 年 25 巻 2 号 p. 93-100

詳細
抄録

The two experiments were conducted to account for some phenomena of partial reinforcement. The first experiment was concerning the operation of a pattern of reinforcement. 21 albino rats were divided into three groups and trained on the straight alley with three trials per day for 10 days. In group I (experimental), the first trial was reinforced and the rest two were, not reinforced. In group II (experimental), only the third trial was reinforced. In group III (control), all three trials were reinforced. After these trainings, experimental extinction was performed. As the result, group II, non-reinforcement followed by reinforcement, showed the formal characteristics of partial reinforcement, i.e., the greater resistance to extinction. But group I, however, missed contrally such characteristics. The importance of this factor, pattern of reinforcement, was discussed, and then, the position of After-effect theory was theorized.
The second experiment aimed to measure the relative strength of habit acquired in partial reinforcement procedure, using the other method than experimental extinction. The apparatus was two straight alley type equipment which was similar to Yarkes' type discrimination box except being divided into two by a screen in choice room. For example, one of them was consisted of the stem and the left arm of the box and contained slightly left turned response. Stimulus to be discriminated was white or black card which hang at the entrance of each arm. (See Fig. 3) 12 albino rats were trained to each stimulus accompanying 10 reinforcement per day for 5 days. They were divided into two groups. In the experimental group, responses were continuously reinforced to one stimulus in one alley, but partially reinforced to the other stimulus in the other alley, which was presented in a preferred side of each rats. In the control group, all responses were reinforced. After these trainings, the screen was removed and two stimulus cards were presented simultaneously. The result indicated that in the experimental group, the stimulus which was under the continuous reinforcement was selected more frequently than the one under partial reinforcement. Up to the present, many studies agreed on the greater resistance to extinction in partial reinforcement, and this was theoretically the most controversial point. The result of the second experiment suggested the necessity of more careful investigation in this region.
This meant that two habits aquired in continuous and partial reinforcement were different in their strengths.

著者関連情報
© 公益社団法人 日本心理学会
前の記事 次の記事
feedback
Top