The influence of intelligence on the concept formation will depend on the differential effect of the intrinsic reinforcement in high and low intelligence groups, as well as the effect of the extrinsic reinforcement. And the effect of intrinsic reinforcement should be stronger in the high intelligence group than in the low intelligence group, which results in the superiority of high intelligence group in the conceptual learning. Two experiments were conducted to test the following hypotheses based on the above consideration, using four years old children as subjects.
1. The effect of training will be observed when there is no extrinsic reinforcement, if the intrinsic reinforcement exists through the training trials.
2. If the effect of the intrinsic reinforcement is stronger in the high intelligence group than in the low intelligence group, training with no extrinsic reinforcement will have greater transfer effect in the high intelligence group than in the intelligence group.
3. If the intrinsic reinforcement is like the ‘function pleasure’, the effect of the intrinsic reinforcement will be determined by the degree of fitness of the problem difficulty to the intelligence level of the subject. That is, the effect of the intrinsic reinforcement in the low intelligence group will be weaker as the problem becomes difficult.
In the first experiment, six groups were constructed to test Hypotheses 1 and 2 by the 3×2 factorial design in terms of (1) three conditions of training (reinforcement, nonreinforcement, and no training) and (2) two levels of MA (high and low). Ten number conservation problems consisting of various kinds of toys were given during the training, 10 trials per day for two days. Test of transfer was done by the same kind of problems, immediately after the training, 40 days after, and 180 days after.
In the next experiment, eight experimental groups and two control groups were run to test Hypthesis 3. Experimental groups were the combination of three variables; two levels of difficulty of training problems (the quantity and length conservation problems, and the number conservation problems alone), the conditions with and without reinforcement, and two levels of MA. High and low intelligence control groups were not given any training problems before the test. The number conservation test problems were given immediately following the training trials, and after 40 days.
The training with extrinsic reinforcement produced significantly positive transfer effect in both intelligence groups, when the test was done immediately after the training. In the test after 40 days, high intelligence group with nonreinforcement training, which was assumed to possess the intrinsic reinforcement, showed significantly higher scores than the control group, and in the test after 180 days high intelligence groups with reinforcement and with nonreinforcement performed significantly better than the control group. These results were in accord with Hypotheses 1 and 2.
In the second experiment, when the test was given immediately after the training, significant difference was found between low intelligence group with nonreinforcement training and control group for the easy problem, but not for the difficult one. In the test after 40 days, however, there was no significant difference between low intelligence group and control group regardless of the level of the problem difficulty. It seems that the effect of intrinsic reinforcement in low intelligence group is so weak even for the easy problem that it disappears after 40 days. This result was not in accord with Hypothesis 3.
抄録全体を表示