性格自己診断検査の作製

Abstract

第一部「綜括的自己診断検査の作製」においてはKiblerの自己診断表から出発し, 数次のGP分析を経て, 50項目からなる性格診断検査を作製した。これを矢田部Kibler向性検査と名づける。この検査の特徴は各項目が殆んど完全な内的整合性を有し, 従つて安心して綜合得点を算出しうるところにある。第二部「特性別自己診断検査の作製」においてはGuilford-MartinのInventoryから出発し, 数次のGP分析を経て, 各尺度が殆んど完全な内的整合性を有する二つのInventoryを構成した。一つは16特性を各25問の尺度項目によつて測定するものであるが, 項目間に重複のあるもので, これを矢田部Guilford性格検査第一型式と名づける。二つは13特性を各12問の尺度項目によつで測定するのであるが, 項目間に重複のないもので, これを同じくその第二型式と名づけることにした。第三部「矢田部Guilford性格検査の因子分析的研究」においては, 上述の第二型式を200名の京大生に施行した結果から, その因子構造を分析した。一次因子は8個抽出されたが, それらはThurstoneの因子と極めて近親的であつた。ただしかれのImpulsiveに相当するものは明瞭でなく, その社交性は二つに分離された。S^1(D^1).社会的内向性E^1(E^1).情緒安定性M^1(V^1).男子性A^1(A^1).活動性T^1(R^1).思索性F^1(S_1^1).社交性第一(Frustratedness)?(S_2^1).社交性第二(Aggressiveness)(X_1^1)(I^1)この因子の獨立性は疑わしい第二次因子は4個抽出された。A^2.情緒安定性B^2.現実性C^2活動性D^2.社交性これらはBaehrの二次因子A^2, B^2, C^2に対応するが, D^2はBaehrには対応者なく, むしろThurstoneの一次因子社交性に近い。第二次因子間の相関はすべて完全相関を示した。その意味は未だ明らかでない。

In the first part of this study, we have constructed a self diagnostic personality inventory which permits a unitary score. The widely accepted AWAJI VERSION TEST obtains its unitary score, called version quotient, as the ratio of the real numbers to the possible numbers of extraverted responses. For such a unitary score to be valuable it is requisite that all test items must be internally consistent, at least not be contradictory. In this respect the AWAJI TEST may be considered far from perfect. The main purpose of the first study is to remedy this imperfection. We started from KIBLER SELF DIAGNOSTIC LIST, and after several applications of the good-poor analysis technique, we constructed a new inventory which has almost perfect internal consistency. After application of the correction formula : t=|p-q|-Φ/√σp2+σq2-2Φσpσq, most of its 50 items show significant differences beyond 5% level between the upper quartile and the lower quartile of subjects. In data on 400 college students, only two items, and in data on 400 junior high school pupils, only 10 items did not show significant differences. The purpose of the second part of this study was to construct a new self disgnostic inventory from which one can get consistent scores on several different personality traits separately. Starting from three inventories of GUILFORD and GUILFORD-MARTIN which give 13 scores for different traits, we prepared at first a test of 240 items for 17 traits and applied it to 200 younger college students. Each of the 240 items was tested against the 17 measures by the good-poor analysis technique. Then we selected most significant items for each trait, and constructed two new inventonies. FORM I consists of 25 items each for 16 traits with reduplication of items. FORM II consists of 12 items each for 13 traits, each item given only once. These tests, especially FORM II, have these desirable characteristics : very high internal consistency and independence between the measures of each trait. These 13 traits for FORM II could be given the same names as in GUILFORD'S tests, namely, 1. social introversion, 2. thinking introversion, 3. depression, 4. cyclothymia, 5. rhathymia, 6. general activity, 7. ascendance, 8. masculinity, 9. inferiority, 10. nervousness, 11. objectivity, 12. agreeableness, 13. cooperativeness. But the trait definitions are not exactly the same. Our "thinking introversion" is purer than GUILFORD'S, i. e., is devoid of emotional tones. Depression, cyclothymia, inferiority, and nervousness are much more separated from each other. The largest change is effected in the trait of "objectivity". We have endeavoured in vain to conserve a trait in our list which may represent the character of an individual whose concern is always with himself and shows some paranoid symptoms. At the end of our efforts we found instead a trait which expresses an inclination to phantasy and neurasthenia. For like reasons the name aggressiveness seems to fit its category better than agreeableness, and frustration better than cooperativeness. It should be noted that our items for measuring masculinity were prepared by a procedure different from that of Guilford. We analysed all 240 test items with respect to the real biological sex difference of the subjects and selected those items which showed significant differences between the answers of both sexes. In the third part of our study we did a factor analysis of our FORM II Inventory. According to Thurstone's centroid method, and after rotations of the axes, we found eight primary factors which have striking affinities with factors analysed by Thurstone himself on the 13 GUILFORD-MARTIN scores given by Lovell. Only the trait called "impulsiveness" by Thurstone was not detected, and the factor called "sociable" is divided into two traits in our case. The names and the main loadings of our primary factors are as follows : S1 (D1) SOCIAL INTROVERSION or DOMINANT (Social Introversion : 0.56 ; Ascendance : 0.54) E1 (E1) EMOTIONAL STABILITY (Cyclothymia : 0.53 ; Nervousness : 0.56 ; Phantastic or Objectivity : 0.49 ; Depression : 0.47 ; Inferiority : 0.39) M1 (V1) MASCULINITY or VIGOROUS (Masculinity : 0.70) G1 (A1) REALITY GRADE or GENERAL ACTIVITY (Phantastic or Objectivity : 0.57 ; Reality Grade or General Activity : 0.43) T1 (R1) THINKING INTROVERSION or REFLECTIVENESS (Thinking Introversion : 0.81 ; Rhathymia : 0.26 ; Nervousness : 0.25) F1 (S1-I) FRUSTRATEDNESS or SOCIABILITY-I (Frustratedness or Cooperativeness : 0.79 ; Phantastic or Objectivity : 0.23 ; Nervousness : 0.21) A1 (S1-II) AGGRESSIVENESS or SOCIABILITY-II (Aggressiveness or Agreeableness : 0.45 ; Inferiority : 0.50 ; Nervousness : 0.38) X1 (I1) The independency of this factor is doubtful. The names and loadings of secondary factors are as follows : A2 EMOTIONAL STABILITY (E1 : 0.78 ; S1 : 0.67 ; A1 : 0.56) B2 REALITY GRADE (G1 : 0.78 ; S1 : 0.49) C2 ACTIVITY (G1 : 0.63 ; F1 : 0.60 ; M1 : 0.57) D2 SOCIAL ADAPTABILITY (F1 : 0.96 ; M1 : 0.81 ; A1 : 0.68) While the first three of the secondary factors correspond approximately to those of Baehr, the fourth factor deviates from Baehr and approaches rather to Thurstone's primary factor "sociable".