- 著者
-
古賀 行義
- 出版者
- 公益社団法人 日本心理学会
- 雑誌
- 心理学研究 (ISSN:00215236)
- 巻号頁・発行日
- vol.9, no.5, pp.953-977, 1934
(I) The term attitude has been used in various different meanings, but here I take it provisionally as the sum-total that a man feels, thinks and want to do. And from the statistical point of view, I dealt with opinions as it seems to be the verbal expression of attitude. What we measure out from these opinions will have something to do with attitudes, though we know a man's opinion may in some special situations be distorted from his attitude. As attitude is so complex a thing as it can never be measured with a single index, we must remember that what we do is only an aspect of it and therefore an abstraction. This measurment of attitudes towards sport will consist a part of the study of social attitudes, in youth, which I project to complete in future. As for the method of measurement I owed much to the honourab e work of L. L. Thurstone.<BR>(II) Having prepared 75 opinions which may or may not be justitled in accordance with attitudes towards sport, I laid them out to about 40 college students and asked to classify into 11 piles by their own merit for sport, and thus I selected 20 opinions by equal-appearing intervals on the scale. On the other hand, I showed these 20 opinions to 353 midd'e-school students and demanded them to pick up those opinions which they agree according to their own attitude. In this sort of case it is quite rare that those who strongly approve sport may pick up the opposite opinions, and in case there are two opinions which are for sport, it is reasonable to assume that those who endose one will also pick up the other. Anyhow all opinions have their unique reiabi ity. So we put the opinions to the same students again after two months to decide reiabiity coefficients. By these means we come at last to the following 5 statistical facts:……(1) n<SUB>1</SUB>……total unmber of individuals who endosed opinion no.1, (2) n<SUB>2</SUB> …… total number of individuals who endosed opinion no.2, (3) n<SUB>12</SUB>……total number of individuals who endesed both opinions, (4) ρ<SUB>1</SUB>…reliability of opinion no.1, (5) ρ<SUB>2</SUB> ……reliability of opinion no.2. The simi-arity of all pairs of opinions were calculated by Thurstone's formula:φ= n<SUB>12</SUB>/(ρ<SUB>1</SUB> ρ<SUB>2</SUB> n<SUB>1</SUB> n<SUB>2</SUB>)<SUP>1/2</SUP> If we assume similarity coefficient corresponds to ordivate of normal distribution curve, we can find the deviation from mean for all φ in terms of standard deviation. Then we get the average of scale-separations between all pairs of opinions.<BR>(III) According to the order of opinions which we got from the college students, here we arrange the scale value of each opinion, giving zero to the least favourab e one.<BR>75) Sport makes man multilated 0.000<BR>68) Sport makes reckless man 0.108<BR>30) Sport compels man unnatural exercise 0.555<BR>65) Sport serves as means of money-making or advertisement 0.705<BR>58) Sport cost much money 0.819<BR>42) Sport demands much skill and time 1.173<BR>50) Sport is necessary for city-dwellers 1.288<BR>54) Sport makes man exciting 1.333<BR>43) Sport is good as it is manly 1.358<BR>12) Spott makes a man's movement alertl 1.404<BR>15) Sport makes a man's uw ll adamant 1.453<BR>5) Sport cultivates upright character 1.686<BR>From this list 8 opinions were automatically excluded, as, even if they were dealt with, it is difficult to put them along others. For example, ‘Sport restrains champions’ will not noly be enclosed by both sides, but has as small reliability as ‘Sport prevents healthy development of mind and body’<BR>Nevt, by the preceeding sca'e valne we marked the opinions endosed by 1614 middle and higher clioo students, and averaging them to survey the state of distribution we found they are fairly skewed towards the approval-side.