The purpose of the present study was to investigate two variables on the concept learning in kindergarten chidren. One was the verbalization variable (verbalization, nonverbalization), the other was the variable of the order of the presentation of positive and negative instances: i. e. a) only positive instances, b) first positive ones and second negative ones, and c) positive and negative instances mixed together.
In Exp. I, the Ss were 96 kindergarten children with a mean age of 5 years 10 months. They were assigned to one of six experimental conditions for the training: verbalization-positive instances (V-P), verbalization-positive plus negative (V-P+N), verbalization-positive and negative mixed (V-PNmixed), nonverbalization-positive instances (NV-P), nonverbalization-positive plus negative (NV-P+N), nonverbalization-positive and negative mixed (NV-PNmixed).
The Ss were trained to learn a concept (a correct value on the relevant dimension) from the stimuli consisting of three dimensions (shape, color and size) with two values. The stimuli were presented successively and the responses by the Ss were reinforced. The learning criterion was an errorless performance on 8 successive trials, with a maximum of 96 trials.
Only the Ss who reached the criterion in the training were given a test, consisting of new stimuli, in order to compare the effects of six conditions for the training.
The main results were (1) As for that training, NV-PNmixed was the most difficult to learn (2) About the Ss who reached the criterion in the training, as for V, P+N, PNmixed were easier to learn than P. On the other hand, as for NV, PNmixed was the most difficult. V conditions were easier to learn then NV's in P+N and PNmixed (3) As for the test, V-P+N seemed to be effective but this result might be due to the factor of the Ss in V- +N who could reach easily the criterion in Pthe training.
In Exp II, the Ss were 95 kindergarten children with a mean age of 5 years 11 months. Except for the criterion in the training (errorless performance of 4 successive trials, with a maximum of 48 trials) it was lower than in Eip. I, the method being the same as in Exp. I.
Comparisons of the results in Exp. I and Exp. II showed that (1) When the criterion was high, the verbalization and the positive and negative instances affected the performance in the training.(2) When the criterion was Iow, only the verbalization affected it.(3) As for the test, the experimental variable had effect on the performance, only on the low crite rion.
It was found that the verbalization variable was significant.
It was considered that such a result was due to the effect of generalization by verbalization.
View full abstract