期刊論文
EXPERIMENTS OF VARIABLES CONFOUNDING WITH PERCENTAGE OF REINFORCEMENT IN CONCEPT LEARNING

中華心理學刊 民 57,10卷,105-119
Chinese Journal of Psychology 1968, Vol.10, 105-119


MAY-JANE WANG

 

Abstract

Concepc learning was studied with partial or continuous reinforcement, and reinforcement for placement or reason. 80 Ss placed cards right or left and were reinforced for placements. Another group of 40 Ss placed cards right or left and gave trial by trial reasons and were reinforced for reasons.

In Exp. I, during acquisition, half of Ss received 100% and half 50% reinforcement until each reached 50 times of reinforcements. In Exp. II, during acquisition, half received 100%, and half 50% reinforcement until each reached 100 trials. The reasons for placement were odd number of figures per card to the left, and cards wich even number of figures to the right. Shifting period had 100 trials with 100% reinforcement, and the reasons for placement were shifted to color: dark colors to the right and light colors to the left.

In acquisition and shif,ing periods, PRE was found. Reinfoccement for reason was superior to reinforcement for placement in performance. In Exp. I, a tentative conclusion is, the more Nt acquired when Nr was kept constant, ,he slower the excinction during shfting period, or the more the perseverative errors. In Exp. II, the more the Nr, the quicker the extinction with fewer perseverative errors. In both acquisition and shifting periods, with reinforcement contingent on either verbal or motor resposnes, the PRE could be found in concept learning. It could also be found to be sustained· through blocks of continuous reinforcement. The shifting condition is analogous to an extinction condition.

登入
會員登入
更新驗證碼