The main purpose of this study was to see the r elation of English achievement and capacity of using constraint of contexts on immediate memory. Artificial English sentences were used as materials, each of which contained 10 words and graded by the order of apprwdmation to English, and the subjects wrote them down immediately after presentation, when two ways of scoring were employed, namely letter score (L. 5.) and placement score (P. 5.).The subjects were 80 freshmen of high school arfor each of them the following data were obtained as reference materials. Test experiments 1) rote learning test 2) words recalling test 3) bonanzagram test Reference data 1) school records of English 2) IQ To see the types of the ways of memorizing, contents of memory were devided into structural and non-structural.part and P. S. was regarded as the index of the former. Then calculating the the correlation coefficient between non-structural part and L. S:, there was no correlation for the subjects as a whole, but stratifying them according to English achievement, upper 1/4 showed negative value of correlation and when materials were restricted to high orders of approximation, upper 1/2 showed negative correlation. These results show that there is a difference in the capacity of using the constraint of contexts between an upper group and a lower group. But from the fact that the upper 1/4 showed high negative correletion, for low order materials, it is considered that especially excellent students can give a strained meaning to the material which has not a logical structure in itself. This fact shows one of the characteristics of learning foreign languages.
The present study was so designed as to compare scores of the measures of the cancellation testsuch as amount of performance before resting (A) and after resting (B), error rate,(A & B), efficiency (A & B), rate of the resting effect, and ratios of omitted numbers of 3, 6 and 8 which were set to be crossed out, in schizophrenic subjects (8) with those in normal subjects (N). Besides scores of the measures of the. Uchida-Kraepelin test such as type of work curve, amount of performance (A & B), error rate (A & B), and rate of the resting effect were added, and then the structure of factors based on correlation coefficients of these scores in Ss was compared with that in N. The results were as follows: 1) Differences of mean scores of amount of performance (A & B), error rate (A & B), efficiency (A & B) in S and those in N were statistically significant, but there were no differences between S and N in mean scores of rate of the resting effect, and ratios of omitted numbers. 2) Four factors were extracted and rotated in S and N respectively. In S they were tentatively named as good-bad of performance of both cancellation test, and Kraepelin test, good quantity of cancellation test and bad quality of Kraepelin test, good quantity of Kraepelin test and bad quality of can tion test, and ratios of omitted numbers of 6 and 8, and in N as good-bad of performance of both tests, good quantity of cancellation test and bad quality of Kraepelin test, resting effect, and ratios omitted numbers of 6 and 8. 3) There might be little value as a psycho diagnostic measure in rate of the resting effect of cancellation test, and it seemed it was not necessary to measures distinguishing before the resting from after the resting.