Factor Structure and Longitudinal Measurement Invariance of the Tangney's Brief Self-Control Scale in Chinese Adolescents

Abstract

Self-control is an important trait for humans to perceive inner and outer perceptions while maintaining harmony with others in society. People with lower self-control are more likely to engage in undesired or irresponsible behavior. The Brief Self-Control Scale (BSCS) is an effective scale with a brief set of items which can effectively measure the level of an individual's control abilities. So far, it has been widely used in many longitudinal studies. However, the factor structure of the scale remains controversial, and far fewer studies have examined the longitudinal measurement invariance of the BSCS. This study aimed to revise the BSCS and test its factor structure for use in Chinese adolescents. Three samples of adolescents (N= 1,330/1,000/600, 11–19 years of age) were used. The item-total correlation and inter-item correlation coefficients were used to evaluate the quality of items. The exploratory factor analysis (EFA), confirmatory factor analysis (CFA) and the principle component analysis (PCA) of the residuals were performed to test the factor structure of the BSCS. Three nested models were used to test the longitudinal measurement invariance (LMI) of the BSCS. Pearson correlation coefficient and Cronbach's alpha coefficient were conducted to test the criterion validity and internal consistency reliability, respectively. According to the CFA of different dimensional models of the BSCS, the results did not support the two-dimensional model, and poor factor loading was found for Item 12. Based on this, combined with lower item-total correlation and item-item correlations, Item 12 was eliminated. Based on results of the EFA with both Kaiser eigenvalues and minimum average partial correlations, only one factor of the revised 12-item BSCS was extracted to make the fit indices of the confirmatory factor analysis acceptable. Meanwhile, the results of principle component analysis of the residuals supported the unidimensional assumption. The fit indices of three nested models supported the longitudinal measurement invariance, indicating that this scale has the same meaning over time. The internal consistency coefficient of the BSCS-12 was 0.81 and the test-retest reliability was 0.70. Good concurrent validity was also demonstrated. Overall, these findings suggest that the revised 12-item Tangney's Brief Self-Control Scale has a one-dimensional structure and has good reliability and validity in Chinese adolescents.