Assessing suicide-related behaviors and psychological strains in the Chinese context with self-report instruments

Abstract

In three inter-connected studies, this thesis research used contemporary psychometric methods to evaluate the psychometric characteristics of the Suicidal Behaviors Questionnaire-Revised (SBQ-R) and the Psychological Strain Scales (PSS) used to measure (1) suicide-related behaviors, and (2) psychological strains in the Chinese context. Data for Chinese undergraduate students (Study 1: N = 590; Study 2: N = 11,412; Study 3: N = 10,961) was used for undertaking the analyses involved in this thesis research. Additionally, data for American undergraduate students (Study 1: N = 320; Study 3: N = 1,162) was used as the reference group for cross-cultural validation. Findings of previous studies suggest that the widely-used SBQ-R presents with adequate estimates of reliability and validity across several Western cultures. Yet, no Chinese version of the SBQ-R (C-SBQ-R) has been rigorously validated. In Study 1, the SBQ-R/ C-SBQ-R was investigated for its psychometric properties and measurement invariance across the American and Chinese samples. Findings showed that the SBQ-R/ C-SBQ-R was unidimensional with adequate estimates of internal consistency reliability, supporting use of the SBQ-R/ C-SBQ-R total score. Item 1 (suicide attempt history) responses demonstrated highest discrimination, severity, and information of suicide-related behaviors, which supported its use for developing subgroups or screening purposes. Also, a cutoff point of 7 for the SBQ-R/ C-SBQ-R total scores would maximize its sensitivity and specificity in screening for suicide-related behaviors. Test of measurement invariance of the SBQ-R/ C-SBQ-R total scores and differential item functioning analysis of the SBQ-R/ C-SBQ-R item scores across the American and Chinese samples showed negligible cross-cultural differences in the understanding of the construct. To exude confidence in the cultural relevancy of the findings, the Strain Theory of Suicide (STS; Zhang, 2005) developed in the Chinese context provided the conceptual foundation for the validity-testing analyses in Study 2 and Study 3. In particular, scores on the PSS based on the STS were used to assess evidence of convergent validity for the SBQ-R scores. Result from the bifactor item response theory analysis in Study 2 showed that the PSS was unidimensional, and thus the PSS total scores should be reported. Unidimensional subset of five PSS items were selected to construct a short form of the PSS. Study 3 examined the associations between psychological strains, negative affective distress, and suicide-related behavior parameters of the SBQ-R. Results from the Bayesian network analysis were in accordance with the STS proposing that psychological strains as an original cause of affective disorders and suicide. The findings of this thesis research have research and practical implications for the psychometric evaluation and clinical utility of these self-report instruments assessing suicide-related behaviors and psychological strains. In addition, most of the contemporary data analytic strategies used in this thesis research could serve as exemplars, which are useful to measurement validation and evidence-based practices in the discipline of social work and other disciplines in the behavioral and social sciences.