島根農科大学研究報告

ダウンロード数 : ?
島根農科大学研究報告 7
1959-03-31 発行

大学生の成績の順位相関について

Rank-Correlation among the Four Class-Records in a College
新宮 忠雄
ファイル
内容記述(抄録等)
The effect of the entrance examination to those who apply a college is often discussed in our country, but we have few scientific data for this purpose. Our president, Dr. Takezaki, Yosinori suggested me to gather such data from the graduates of the Simane Agricultural College and to analyze them. In March of 1958 our college sent off the fourth-time graduates, so even only in our college still we can get twelve kinds of data for this purpose, for there are three classes each year.
To calculate the correlation coefficient between the grades at the entrance examination and those at the graduation one may think that it would be better to use the average scores of each student than to use the ranks in his class. Yes, that may be true, but its calculation is very tedious and yet the results are almost the same. See one example done for the data from (F2) at p.144. Therefore I decided to use their ranks in the class instead of the average scores. In selecting the sample I did as follows ;
Those who joined the class later or earlier were omitted from the ranking. It means that if a student was admitted into his class at the beginning of the second or third year from other college he was eliminated from the ranking. And also if a student could not graduate after four years by his absence from school or by the failure in the examination he was omitted from the member of his class. The reason for such elimination is simple. Students in one class should had the same or at least almost the same examinations at the entrance and in the college education.
At the end of the introduction I wish to express my great appreciation to Miss Takeda, Yukiko for her gathering data from the students' records.