The Self-Compassion Scale (SCS) is the most widely used measure of self-compassion. The scale is constructed of six factors measuring positive and negative components of compassion. Support for this factor structure has been subject to debate and alternative factor structures have been proposed. We tested the proposed factor structures against existing models of the SCS including one derived from an exploratory factor analysis of our data. Respondents (n = 526) completed the original version of the SCS online at two time points, at baseline (time 1) and 2.5 months later (n = 332, time 2). Exploratory factor analysis (EFA) was carried out on time 1 data and confirmatory factor analyses (CFA) were conducted on time 2 data and retested using time 1 data. The EFA yielded a five-factor model. CFA was used to compare the following models: Neff's original six-factor correlated and higher-order models; a single-factor, two-factor, five-factor model (as suggested by the EFA) and a bi-factorial model. The bi-factorial model was the best fit to the data followed by the six-factor correlated model. Omega indices were calculated and yielded support for the bi-factorial model of SCS. In conclusion, this study supports the use of the six-factor scoring method of the SCS and the use of an overarching self-compassion score.
Keywords: Bi-factorial; Factor analysis; Factor structure; Omega; Self-compassion scale.