Race and self-esteem revisited: reply to Hafdahl and Gray-Little (2002)

A. R. Hafdahl and B. Gray-Little (2002) raised several issues of study inclusion and statistical reporting, none of which challenged the overall results or took issue with the theoretical conclusions of J. M. Twenge and J. Crocker (2002). Excluding the few studies whose inclusion Hafdahl and Gray-Little questioned would not change the results. Twenge and Crocker's inclusion of twice as many data sources is the more likely cause of differences between the 2 meta-analyses. The authors argue that including standardized regression coefficients is essential for reader understanding and reiterate the position that academic self-esteem scales should not be included in an analysis focusing on general self-esteem. Examining the specific domains of self-esteem would entail another meta-analysis, preferably one including other specific domains as well.