抄録
The purpose of this study was to test the hypotheses advanced in our previous studies, which had been developed in order to be applicable to the effect of the experience of success on the succeeding child behavior in general.
Method: The appratus and main procedure were described in detail in a previous report. One major variable manipulated in this experiment was the number of pre-experience trials-6, 12, and 18 trials were used, and each of the pre-experience trial groups was combined factorially with three percentages of success, 33, 50, and 67%, and regularity of the sequence of the successful response. In addition, one 0% and three 100% success groups, differentiating for the number of pre-experience trials, were also used as the control groups. So the experiment was designed to consist of 22 groups. Each of these 22 groups contained 6 boys and girls of second year class of elementary school as the subjects. A S was given as a reward one toffee on each success trial, and the last play of pre-experience situation was always a success play. Following the last pre-experience play, non-successful situation began and no S was rewarded again although he was permitted to play as long as he desired. One response measure was the number of plays in the nonsuccessful situation, and another measure was S′s expectation scores.
Results: A transformation of the number of responses in the nonsuccessful situation to common logarithms resulted in a satisfactory normalization, and an analysis of variance was performed on these log. scores. The main results of this analysis showed essentially that the number of pre-experience trials, the percentage of success, and the pattern of sequence had all significant effects upon the mean number of plays in nonsuccessful situation. That is; the smaller number of pre-experience trials, the smaller percentages of success (excepting 0% group), and the irregular sequence resulted in greater number of plays in non-successful situation. There were also significant interactions among these three variables.
The results related to the expectation indicated that: a) Expectation scores increased differentially during the pre-experience as a direct function of the percentage of success. b) Expectation scores decreased differentially during non-successful situation as an inverse function of the percentage of success. c) There was no clear-cut statistical evidence that the number of pre-experience trials or the patterns of sequence and their interaction had any effect upon expectation.
Discussion: It was appeared that the results of the present experiment could warrant our theory which was able to explain the effects of success on the children's bar-pressing behavior in the non-successful situation. In the theoretical formula, however, it was supposed that AE/NS1 and D should be revised in ΣAE/NS1 and De′ representing a expectation by discrimination between the pre-experience and non-successful situation, respectively. Therefore, we can state this as a general formula as follows, BNbar=f{(E′bar & GE+ΣAE/Nbar)-De′} This theoretical formula may explain and anticipate effect of success on child behavior.
