Abstract
The present study is an attempt to analyze, with special reference to test-retest intervals and to choice level, the stability of choices given by elementary school children in grades 2, 4 and 6.
Subjects were requested 1) to choose three classmates in an attempt to analyze the difference between test-retest intervals, and 2) to choose five classmates to analyze choice levels. The criterion of choice was with whom they most preferred to work in sweeping their classrooms and play ground.
In the former analysis, the author analyzed the fluctuations of the choices given after an interval of 5 months and 10 months, comparing choices given on the first sociometric test (April, 1962), with those on the second (September, 1962)and on the third (February, 1963).
In the latter analysis, the five choices given on each of three occasions by these subjects were examined to deternine the stability of each of five chices.
In each case, choice fluctuation score(one point to one change)was computed as an index of the stability of choice given.
In order to evaluate the contribution of interval, grade, sex and choice level to choice stability, the data were submitted to analysis of variance.
The main findings on the obtained data are as follows:
(1) The choice fluctuation scores of 10-months interval were significantly higher than those of 5-months.
(2) The tendency of the choice fluctuation as the interval increased, varied among grade levels.
(3) As the grade level of subjects increased, the choice fluctuation scores decreased significantly.
(4) The choice fluctuation scores of girls were found to be significantly lower than those of boys.
(5) There was a significant tendency for the fluctuation scores to increase as the choice level decreased.
(6) The above tendency for the fluctuation score to increase differed among grades, and there was some indication that sex seemed to play a part at least, in the difference.
