Abstract
The purpose of this study was to clear the development of classificatory behavior. For this purpose, Ss were asked to find out as many classificatory criteria as possible, which is useful for allotting the instances with some attributes to c onjunctive categories.
As the main factors which might have influence on the difficulty of classificatory behavior, the number of the whole dimensions and relevant dimensions were considered.
The subject for this experiment were 215 male and female of elementary school and junior high school pupils. Each subject was tested individually. Two experiments were conducted.
Experiment I; Ss were 6, 8, 10 and 12 years old. Two sets of blocks were used. These had four attributes (color, type of figure, height of figure and slope), each with two different values, The task of the Ss was to classify the blocks into two, four, and eight groups, by finding out a classi - ficatory criteria.
Experiment II; Ss were 8, 10, 12 and 14 years old. The material for this experiment consisted of a set of 32 cards on each of which pictured five different figures of two-color-printing.
The task was also to classify the 32 cards into two, four and eight groups.
The fac is obtained are as follows;
(1) When the number of relevant dimension is one (twefold division), there was statistical difference between 6 and 8 years old in finding out the criterion.
(2) When the number of relevant dimension is two (fourfold devision), there' was statistical difference between 6 and 8, 8 and 10 years old in finding out the criterion.
(3) When the number of relevant dimension is four (eightfold division), there was difference between 8 and 10, 10 and 12 years old in finding out the criterion. The results suggest that the number of relevant dimensions would be more influential factors than number of whole dimensions in the classificatory behavior.
