抄録
To examine whether or not the frequency polygons of the body weight in 296, 715 Japanese youth from 6 to 21 years old in 1971 are the bell-shaped curves, the tests of normality with reference to k-statistics are carried out on the frequency distributions of the body weight data. There are two types of departure in the distribution fromthe normal. One type is the distribution of the data which is asymmetrical or skewed. The other is characterized by either an excess or a deficit of data concentrated near the center of the range. In the normality tests, there are four types of the frequency distribution which depart from normal. They are called positively and negatively skewed, leptokurtic and platykurtic, respectively. The purposes of this study were: 1) to recognize what sorts of the frequency distribution are contained in the body weight data of the Japanese youth by the tests of normality; 2) to determine which growth curves (velocity and acceleration) are similar to the skewness and the kurtosis coefficient curves from 6 to 21 year-old subjects. The conclusions are: 1) the values of g_1 for the body weight data of every Japanese youth group are significantly positive and therefore indicate an excess in the number of items smaller than the mean; 2) the phenomena of the maximum positive skewness appear in the elementary school period; 3) the male and the female junior high school subjects have the minimum positive skewness in the frequency distribution; 4) the male and the female subjects have the maximum deceleration in the acceleration curves when they have the minimum positive skewness; 5) only eleven and fifteen year-old girls show the mesokurtic distributions in a measure of kurtosis; 6) the values of g_2 are significantly negative in the male subjects and significantly positive in the female subjects during the college period; 7) in the eight year-old boys who have the maximum positive skewness, 58.4% of the area under the frequency distribution curve lies less than the mean and 31.4% of the area under the curve lies between - 1.5 and - 0.5 standard deviation.
