Abstract
The first experiment was conducted to investigate the cognitive process of size among profoundly retarded children by setting matching tasks. Materials were 3 pairs of white square boards (3cm, 7cm, and 12cm) and 41 Ss who were able to put a board upon another were used in this experiment. The task was 38 problems belonging to seven families being classified on the basis of similarity and number of matching. Ss were asked to do all the problems regardless of achievement. The results showed that thirty-four Ss of fourty-one were able to match size but only seven Ss made correct responses to all the problems so that it was found that there were small steps to go through in matching size. The easiest one was under the conditions that Ss had one board at hand to match with three boards on the desk and that the contrast between them was the greatest. The most difficult which was more difficult than three to three matching problems was the one that Ss had three boards to match with one or two boards on the desk. Between families qualitative aspects rather than quantitative aspects had more effects on deciding the grade. Within the families, however, the rate of success de-creased as the number of the board increased so that the number of the board became a fundation of task difficulty. It was also found that the size of board perse or the contrast between them had effects on the achievement in the lower grade. The purpose of the second experiment was to examine as to wether it was necessary to train profoundly retarded children by the grade gained in the first experiment. It was found that nineteen Ss of twenty-three became to match boards in the problem a step above the one through which each S could pass in the first experiment while only a S could pass through the most difficult problem after training. As a result it was revealed that small steps were necessary for the children to form the skill for matching size. But as the grade moved up, the effects of training decrease, further investigation would be needed in terms of training methods for matching size.
