Assessing Somatic Symptom Scale-8 (SSS-8) in Chinese youth and adolescents: using Classical Test Theory and Rasch Measurement Theory

Evaluating the increasing somatic symptoms in adolescents and youth is essential in epidemiological surveillance. This study aims to investigate the psychometric properties of the Somatic Symptom Scale-8 (SSS-8), a brief instrument assessing somatic symptoms in Chinese youth and adolescents using both Classical Test Theory and Rasch Measurement Theory using two independent samples (N = 398 and 16,231). Internal consistency was assessed by Cronbach's alpha and McDonald's omega coefficients. Construct validity, and convergent and discriminant validity were examined by Spearman's correlations of the items or scale with external instruments and internal symptom domains. Structural validity was evaluated using confirmatory factor analysis (CFA). Rasch analysis was performed including rating scale functioning, unidimensionality, item difficulties and person abilities, item and person reliabilities and separation, and differential item functioning (DIF). The Cronbach's alpha and McDonald's omega were both 0.82 (Sample 1) and 0.92 (Sample 2). The SSS-8 score correlated with two subscales of depression and anxiety (r = 0.56 and 0.62, p < 0.001), and positive mental health (r = -0.31, p < 0.001). All items strongly correlated with their own domains (r > 0.70). The CFA confirmed the high-order factor model and found partial invariance across gender. In Rasch analysis, step calibrations and category measures indicated adequate category functioning. The unidimensionality was proved as no second dimension occurred. The SSS-8 was found able to separate the participants into two levels of somatic symptoms. No significant DIF was detected across gender. The findings supported the use of the SSS-8 to assess somatic symptoms in this group.