Development and Validation of the Self-Compassion Scale for Youth

Abstract

We present a series of studies on the development and validation of the Self-Compassion Scale—Youth version (SCS-Y), which is intended for use with early adolescents in middle school. Study 1 (N = 279, M_age = 12.17) describes the selection of 17 items out of a pool of 36 potential items, with three items each representing the subscales of self-kindness, mindfulness, common humanity, self-judgment, isolation, and two items representing over-identification. Using state-of-the-art psychometric analyses ideal for examining multidimensional constructs like self-compassion—bifactor exploratory structural equation modeling (bifactor-ESEM)—findings supported the use of a general self-compassion score and six subscale scores. Study 2 cross-validated the factor structure of the SCS-Y with a second sample of youths (N = 402, M_age = 12.43). Study 3 found support for the test-retest reliability of the SCS-Y (N = 102, M_age = 12.52). Study 4 (N = 212, M_age = 12.18) established construct validity for the SCS-Y by demonstrating that SCS-Y scores were significantly associated with mindfulness, happiness, life-satisfaction, depression, resilience, and achievement goal orientation in expected directions. Overall, findings suggest that the SCS-Y is a reliable and valid measure of self-compassion for use with youths.

Notes

1 Because the overidentification factor was estimated from two indicators, thus creating a locally underidentified factor in bifactor-CFA, this factor was locally identified by using essentially tau-equivalent constraints (Little, Lindenberger, & Nesselroade, 1999) by putting equality constraints on the factor loadings of these two items to help locate the construct at the true intersection of the items.

2 The Spearman correlation between the two overidentification items was r = .491, p < .001. We also calculated the Spearman-Brown prophecy (see Marsh, Nagengast, & Morin, 2013) formula to verify whether reliability would improve if the number of items increased. According to these calculations, having three items on this factor would result in a Cronbach alpha coefficient that is higher than the recommended threshold (α = .742).

3 The Spearman correlation between the two overidentification items was r = .584, p < .001, and Cronbach alpha would be α = .803, according to the Spearman-Brown prophecy formula, with three items instead of two.

Additional information

Funding

The third author was supported by a Horizon Postdoctoral Fellowship from Concordia University and by funding from the Social Sciences and Humanities Research Council of Canada (435-2018-0368).