Ren-Hau Li1,2, Chih-Mei Kao3,4, Yi-Ying Wu5. 1. Department of Psychology, Chung-Shan Medical University, Taichung, Taiwan. davidrhlee@yahoo.com.tw. 2. Clinical Psychological Room, Chung-Shan Medical University Hospital, Taichung, Taiwan. davidrhlee@yahoo.com.tw. 3. Department of Psychology, Chung-Shan Medical University, Taichung, Taiwan. 4. Clinical Psychological Room, Chung-Shan Medical University Hospital, Taichung, Taiwan. 5. College of Management, Yuan-Zi University, Taoyuan, Taiwan.
Abstract
PURPOSE: This research investigated gender differences in the six factors of Li's 18-item version of Ryff's psychological well-being (PWB) scale by using factorial invariance procedures. This version improved on Ryff's shorter PWB scale in regard to the reliability of each subscale. METHODS: The sample comprised 653 adult participants, specifically, 271 men and 382 women. Factorial invariance tests were conducted using multigroup confirmatory factor analysis. RESULTS: The results showed that the Li's version of Ryff's PWB scale has configural invariance, factor loading invariance, intercept invariance, partial invariance of factor variances and covariances, partial invariance of latent means, and partial invariance of error variances. According to latent means, women had significantly less autonomy (AU) and more environmental mastery (EM) than men did. Moreover, based on interfactor correlations, the correlation of EM and self-acceptance was higher for men than for women, and the correlation of positive relations with others and AU was lower for men than for women. CONCLUSIONS: Gender differences in PWB can be found at the subscale levels of Li's version, which exhibits metric and scalar invariance.
PURPOSE: This research investigated gender differences in the six factors of Li's 18-item version of Ryff's psychological well-being (PWB) scale by using factorial invariance procedures. This version improved on Ryff's shorter PWB scale in regard to the reliability of each subscale. METHODS: The sample comprised 653 adult participants, specifically, 271 men and 382 women. Factorial invariance tests were conducted using multigroup confirmatory factor analysis. RESULTS: The results showed that the Li's version of Ryff's PWB scale has configural invariance, factor loading invariance, intercept invariance, partial invariance of factor variances and covariances, partial invariance of latent means, and partial invariance of error variances. According to latent means, women had significantly less autonomy (AU) and more environmental mastery (EM) than men did. Moreover, based on interfactor correlations, the correlation of EM and self-acceptance was higher for men than for women, and the correlation of positive relations with others and AU was lower for men than for women. CONCLUSIONS: Gender differences in PWB can be found at the subscale levels of Li's version, which exhibits metric and scalar invariance.