Abstract: A sorting test was applied to 100 normals and 128 schizophrenics. The test consisted of (1) 25 pictorial cards; (2) 16 Chinese characters printed on separate cards; (3) 23 wooden blocks of different colours, forms and sizes; and (4) 10 cards, each bearing four meaningful contents, three of which are closely related, while the fourth one is different.Each subject was presented with the first three sets of test materials in succession and asked to classify them and also to give reasons for his classification. The rating of the experimental results was made in accordance with the quality of the classifications. He was then asked to point out the exceptional content from each card of the fourth set of material, and the rating was made according to the quality of answers given by him. The results were as follows:(1) For both normals and schizophrenics, the quality of classification was, in general, related to the educational level of the subjects tested, better results being obtained from those who had more advanced educational background.(2) Comparison between the two groups revealed the following differences: (a) the percentage of correct classifications was definitely higher in normals than in schizophrenics; (b) the test time was shorter for the normals than for the schizophrenics; (c) many sortings made by the schizophrenics were not only poor in quality, but also showed much deviation from the normal way of classification.

Key words: NULL