Bootstrap Confidence Intervals for the Mean Correlation Corrected for Case IV Range Restriction: A More Adequate Procedure for Meta-Analysis

• 成果类型: Article
• 署名作者: Li, Johnson Ching-hong; Cui, Ying; Chan, Wai
• 署名单位: University of Hong Kong; University of Alberta; Chinese University of Hong Kong
• 刊物名称: JOURNAL OF APPLIED PSYCHOLOGY
• ISSN/ISSBN: 0021-9010
• DOI: 10.1037/a0029946
• 发表日期: 2013
• 页码: 183-193
• 关键词: META-ANALYSIS; confidence intervals; Case IV range restriction; correlation; bootstrap procedure

摘要

In this study, we proposed to use the nonparametric bootstrap procedure to construct the confidence interval for the mean correlation r corrected for Case IV range restriction in meta-analysis (i.e., (r(c4)) over bar; Hunter, Schmidt, & Le, 2006). A comprehensive Monte Carlo study was conducted to evaluate the accuracy of the parametric confidence interval and 3 nonparametric bootstrap confidence intervals for (r(c4)) over bar. Of the 4 intervals, our results showed that the bootstrap bias-corrected and accelerated percentile interval (BCaI) yielded the most accurate results across different data situations. In addition, the mean-corrected correlation (r(c4)) over bar was found to be more accurate than the uncorrected estimate. Implications of the mean-corrected correlation (r(c4)) over bar and BCaI in organizational studies are also discussed.