Investigating Stable and Dynamic Aspects of the Vietnamese Self-Compassion Scale using Generalisability Theory

Abstract

Objectives: Evaluating comprehensive reliability of the Vietnamese Self-Compassion Scale (VSCS) and its ability to distinguish between trait (stable) vs state (dynamic) aspects of self-compassion using Generalisability Theory (G-Theory) is necessary. This investigation contributes to both reliability and validity of research that uses the VSCS to measure self-compassion in Vietnamese adults. Methods: In a sample of 155 Vietnamese adults who completed the VSCS at three occasions that were each 2 weeks apart, a G-study was conducted to measure reliability and trait vs state aspects of each VSCS subscale and the short-form VSCS, and a D-study was conducted to examine the effects of removing subscales on overall scale reliability as well as evaluate trait vs state aspects of each item. Results: With G-coefficients of 0.93–0.98, both the complete and short-form VSCS (VSCS-SF) demonstrated excellent reliability in measuring trait self-compassion. Three of the six subscales—self-judgement, mindfulness, and kindness—also demonstrated excellent reliability, with G-coefficients of 0.82–0.85. Eighteen of the 26 items measured trait more than state. The remaining eight items reflected a mixture of trait and state, but this did not affect overall reliability. Conclusions: This study indicated that the VSCS, VSCS-SF, and three VSCS subscales reliably measured trait self-compassion, with scores generalisable across the Vietnamese population and occasions. Thus, overall self-compassion levels remained stable over time, which is useful for evaluating the effectiveness of an intervention because significant changes of self-compassion are likely to be long-lasting.