인물화에 의한 아동의 지능측정

Abstract

The purpose of this study was to pretest the possibility for standardization of the Goodenough-Harris Drawing Test as a new intelligence test in Korea. A historical background of the study about children's drawings was surveyed in general, and also the outline of the Goodenough-Harris Drawing Test was introduced. The test was administered to 1605 children from 1st to 6th grade of 4 elementary schools and 3 kindergartens in Seoul. For the preliminary study, at each age level from 5 to 12, 1265 children were selected so that the samples can be centered at the midyear, and if possible, they may represent the occupational distribution of Seoul population in 1960. The problems and their methods of the research were as follow 1. An analysis of the distribution of the raw scores on the Man and Woman Scale ; The means and standard deviations at each age group were calculated. The smoothed means and standard deviations were also calculated. 2. The reliability and the validity of the test ; The reliability was evaluated with inter-rater reliability and alternate-form reliability. The validity was evaluated with correlation between the drawing test and another intelligence test, Intellectual Maturity Test. 3. Item analysis ; The percentages passing each item of high group(upper 27%) and low group(lower 27%) was calculated. From this data, the value of item discrimination was found for a test of internal consistency. 4. Norms ; The smoothed valued were converted to standard scores(M = 100, SD = 15). The standard scores were converted to percentile scores. The results of the study are summarized as follow : 1. The increment of raw scores as age increases is sufficiently discriminable. 2. Girls do better on the drawing test than do doys, and this difference exists to a greater degree on the Woman Scale. 3. There are no large discrepancies between the scores of Korean children and American children. 4. Inter-correlations between two independent scorers range from .92 to .97, and correlations between Man and Woman Scale range from .67 to .80. 5. The correlations between the drawing test and the Intellectual Maturity Test range from .35 to .42. 6. The number of items which have low discriminating power on Man Scale are 22 items and 20 items on Woman Scale. 7. Two kinds of norms are provided, one is standard score I.Q.'s, the other is percentile scores. These could be adequately used for the children aged 5 to 12 in Seoul. This study is only the first step for standardization of the Goodenough-Harris Drawing Test. There are many problems remained, such as item validity, retest reliability, construction of quality scale, with a new large samples.