Kitamura T, Sugawara M, Aoki M, Shima S
Section of Mental Health for the Elderly, National Institute of Mental Health, Japan.
Psychol Med. 1989 May;19(2):507-11. doi: 10.1017/s0033291700012538.
The validity of the Japanese version of the 30-item General Health Questionnaire (GHQ) was examined against the semi-structured interview-based Research Diagnostic Criteria (RDC) as external criteria. The GHQ total score discriminated 'cases' and 'non-cases' satisfactorily but its recommended cut-off point was higher (7/8) than that of the original English version (4/5). Discriminant function analysis revealed that only 13 items contributed to the discriminatory power and that their discriminant function score was better than a simple summation of the 30 GHQ item scores in terms of validity.
以基于半结构化访谈的研究诊断标准（RDC）作为外部标准，对30项一般健康问卷（GHQ）日文版的有效性进行了检验。GHQ总分能够令人满意地区分“病例”和“非病例”，但其推荐的临界值（7/8）高于原始英文版（4/5）。判别函数分析显示，只有13个项目对判别力有贡献，并且就有效性而言，它们的判别函数得分优于30个GHQ项目得分的简单相加。
