Assessing psychometric properties and measurement invariance of the Sleep Quality Questionnaire among healthcare students
Mengyi Huang,Haiyan Ma,Karen Spruyt,Joseph M Dzierzewski,Chen Jiang,Jiaxuan He,Nongnong Yang,Yiwei Ying,Bolanle Adeyemi Ola,Runtang Meng
DOI: https://doi.org/10.1186/s40359-023-01276-2
2024-01-19
Abstract:Objective: The sleep of healthcare students is worth discovering. Mental health and self-rated health are thought to be associated with sleep quality. As such, valid instruments to assess sleep quality in healthcare students are crucial and irreplaceable. This study aimed to investigate the measurement properties of the Sleep Quality Questionnaire (SQQ) for Chinese healthcare students. Methods: Two longitudinal assessments were undertaken among healthcare students, with a total of 595, between December 2020 and January 2021. Measures include the Chinese version of the SQQ, Patient Health Questionnaire-4 (PHQ-4), Self-Rated Health Questionnaire (SRHQ), and sociodemographic questionnaire. Structural validity through confirmatory factor analysis (CFA) was conducted to examine factor structure of the SQQ. T-tests and ANOVAs were used to examine sociodemographic differences in sleep quality scores. Multi Group CFA and longitudinal CFA were respectively used to assess cross-sectional invariance and longitudinal invariance across two-time interval, i.e., cross-cultural validity. Construct validity, internal consistency, and test-retest reliability were correspondingly examined via Spearman correlation, Cronbach's alpha and McDonald's omega, and intraclass correlation coefficient. Multiple linear regression analysis was performed to examine incremental validity of the SQQ based on the PHQ-4 and SRHQ as indicators of the criterion variables. Results: CFA results suggested that the two-factor model of the SQQ-9 (item 2 excluded) had the best fit. The SQQ-9 scores differed significantly by age, grade, academic stage, hobby, stress coping strategy, anxiety, depression, and self-rated health subgroups. Measurement invariance was supported in terms of aforesaid subgroups and across two time intervals. In correlation and regression analyses, anxiety, depression, and self-rated health were moderately strong predictors of sleep quality. The SQQ-9 had good internal consistency and test-retest reliability. Conclusion: Good measurement properties suggest that the SQQ is a promising and practical measurement instrument for assessing sleep quality of Chinese healthcare students.