著者

都澤 真智子 二村 英幸 今城 志保 内藤 淳

出版者

経営行動科学学会

雑誌

経営行動科学 (ISSN:09145206)

巻号頁・発行日

vol.18, no.1, pp.21-30, 2005-01-30

被引用文献数

1 2

---

This study reports the results of two meta-analyses of validation studies of a personality test using performance appraisals as criteria. The purpose of this study is twofold. Firstly we seek out the general levels of personality validity coefficients with Japanese samples and their generalizability across situations such as different jobs or organizations. Secondly we address issues of aggregating studies with different criteria, study purposes, and study periods. Five out of seventeen scales showed corrected validity coefficients higher than .10 in absolute value, with the highest coefficient of .21 on "Vitality". The results were similar with those of the previous meta-analyses on U.S. and European samples. Our second meta-analysis included only the studies that met the all three conditions of being conducted within a certain periods of time, using the same criterion, and being confined to research purposes. As a result, the validity coefficients became higher and more generalizable compared to the initial analysis. It is argued that the differences among studies that were often not dealt with in preceding meta-analyses have a significant effect.

著者

都澤 真智子 二村 英幸 今城 志保 内藤 淳

出版者

経営行動科学学会

雑誌

経営行動科学 (ISSN:09145206)

巻号頁・発行日

vol.18, no.1, pp.21-30, 2005-01-30 (Released:2011-01-27)

参考文献数

37

被引用文献数

2 2

---

This study reports the results of two meta-analyses of validation studies of a personality test using performance appraisals as criteria. The purpose of this study is twofold. Firstly we seek out the general levels of personality validity coefficients with Japanese samples and their generalizability across situations such as different jobs or organizations.Secondly we address issues of aggregating studies with different criteria, study purposes, and study periods.Five out of seventeen scales showedcorrected validity coefficients higher than. 10 in absolute value, with the highest coefficient of. 21 on “Vitality”. The results were similar with those of the previous meta-analyses on U. S. and European samples. Our second metaanalysis included only the studies that met the all three conditions of being conducted within a certain periods of time, using the same criterion, and being confined to researchpurposes. As a result, the validity coefficients became higher and more generalizable compared to the initial analysis. It is argued that the differences among studies that were often not dealtwith in preceding meta-analyses have a significant effect.