小塩(2016)

小塩真司
心理尺度構成における再検査信頼性係数の評価―「心理学研究」に掲載された文献のメタ分析から―
Evaluation of test-retest reliability in the development of psychological scales: A meta-analysis of correlation coefficients described in the “Japanese Journal of Psychology”
【論文PDF】

Reporting reliability coefficients is an important procedure in articles that describe the development of new psychological scales. However, it appears that Japanese psychology researchers have not yet arrived at a consensus about what constitutes a desirable magnitude of a reliability coefficient. In this study, I conducted a meta-analysis summarizing 65 test-retest correlations that were collected from 58 studies from the Japanese Journal of Psychology, which is a notable peer-reviewed psychological journal in Japan. The results of the meta-analysis, which involved the use of a random effect model, showed that a desirable mean test-retest correlation wasρ = .76 (95% CI = .70—.81). There was no significant relationship between the test-retest correlations and coefficients alpha. The number of items of a scale was positively correlated with the test-retest correlation coefficients. Researchers tended to mention some problems regarding test-retest coefficients only when they were less than r = .50. The desirable usage of reliability coefficients was also discussed.

Key words: test-retest reliability, scale development, meta-analysis, correlation coefficients, reliability, validity

新たな心理尺度を構成する際に,信頼性係数を報告することは重要な手続きの1つとなっている。しかしながら,日本の心理学者の間でどの程度の信頼性係数が適切であるのかについては意見が一致しているとは限らない。本研究では,過去の心理学研究に掲載された信頼性係数,特に再検査信頼性係数を収集し,メタ分析を試みた。変量効果モデルによる母相関係数を推定したところ,ρ = .76 (95% CI = .70—.81)という結果を得た。再検査信頼性係数と内的整合性間の関連はほとんどみられなかった。尺度の項目数は,再検査信頼性係数の高さに関連していた。論文の著者は,再検査信頼性係数がr = .50を下回る際に,その係数に問題がありそうだと記述する傾向が見られた。得られた結果をもとに,信頼性係数の使用方法について議論された。

キーワード:再検査信頼性係数,尺度構成,メタ分析,相関係数,信頼性,妥当性