Meta-Analysis of Gender Differences in Self-Compassion

Abstract

While research suggests strong associations of self-compassion with mental health and well-being, gender norms may hinder the development of self-compassion by women on one hand, and men on the other. This study represents one of the first systematic analyses of potential gender differences in self-compassion using meta-analytic techniques, including whether such gender differences are moderated by age or ethnic minority status. Fixed-effects models were used to estimate the average effect size (ES) of gender differences in self-compassion scores across 71 journal articles and dissertations providing a total of 88 estimates. Results revealed that males had slightly higher levels of self-compassion than females, with a small ES observed (d = .18). This difference was larger in samples with a higher percentage of ethnic minorities. Researchers and practitioners should take these group differences into account in future studies and interventions focused on self-compassion, while not overemphasizing gender differences in self-compassion as being large in size.

Keywords::

• Self-compassion
• Gender differences
• Age differences
• Ethnic differences
• Meta-analysis

Notes

^* Email: [email protected]

^** Email: [email protected]

^† Email: [email protected]

^‡ Email: [email protected]

^$ Email: [email protected]

^1. Note that the size of the beta weights for Age and Percent Ethnic Minority are partially a function of scaling of these variables, which does not affect their statistical significance. Specifically, Percent Ethnic Minority was coded on a scale from 0.00 to 1.00, typically a decimal value; whereas Age was on a scale of years (15 to 73). If rescaled, Percent Ethnic Minority values could be multiplied by 100 to yield a beta weight of β = .0036, though the statistical significance of all predictors in the model would remain the same. (Another option would be to divide Age values by 100, to achieve similar scaling between the two variables.) However, we chose to retain original scales for these predictors for more straightforward interpretation.