児童画の特徴とその評定の試み

抄録

Problem: Easel pictures painted by children freely were analysed into items according to the categories of painting characteristics and each of them was rated from two points of the quantity of painting expression (e.g., the length of strokes, area of a mass, number of dabs, etc.) and the effect of expression (e.g., distance from left edge of paper to painted things, recognizable forms, etc.). Then, the questions to be answered were as follows: What are the quantity and the effect of painting expression? How do they change with age? What factors are found in characteristics of children's painting? Have its characteristics been analysed and rated accurately? How much is its rating affected by errors? To what extent is the agreement among caters gained?
Procedure: Two age groups of young children, consisting of 22 four-year-olds (IQ 119 on the average), and 29 six-year-olds (IQ 111 on the average) were selected from a kindergarten and a nursery school of middle class districts in Kyoto City. Pictures painted 4 times each with interval of more than 7 days, were analysed into 71 items and were rated into 4 grades. The categories of items were defined as clearly and unambiguously as possible.
Results: 1. 4-year-olds were rated as more distinctively in items on color and on non-concreteness of form than 6-year-olds. The latter were more distinctively in items on direction and on concreteness of form than the former. Those developmental differences should be taken into consideration when personality studies and clinical researches with children's paintings are conducted.
2. The correlation clusters were found when correlation coefficients of inter-items were tabulated adequately, and 3 factors were found on factor analysis. The first factor was interpreted as the emotionality of expression, the second as the effectiveness of expression, the third as the color and form quality of expression.
3. Validity. Examining the internal consistency of items estimated from the intercorrelations of each item to the total by 6-year-olds, the majority of the items except several turned out to be valid.
Reliability. The internal correlations of each item on 6-year-olds were mostly recognizable as significant, but several were negligible. As reliability coefficient of item battery, r_tt by the method of Spearman-Brown was .74, which is reliable.
4. 48 pictures painted by 12 children of 4-year-old and 6-year-old; were rated on 31 items into 4 grades by 3 raters, whose opinions on children's painting were different, and after 19 months' interval they were rated again by one of the racers, Agreement among 3 raters' ratings and the agreement between the first and the second ratings of one rater were 61% and 76% respectively on the average.