The psychometric property of Rosenberg Self-esteem Scale in Thai youth

Abstract

Objectives  To investigate the psychometric property of the Rosenberg Self-esteem Scale (RSES).

Materials and method The data from psychosocial status of members in “To be number one” project in 2009 were analyzed. There were  2661 cases  consisted of adolescents attending the Friend’s corner, non participants and juvenile delinquents living in Bangkok and 11 provinces. Items were analyzed by using exploratory structural equation modeling (ESEM) confirmatory factor analysis  (CFA) in 20% of cases selected by random sampling and then confirmed by confirmatory factor analysis in the other group by using the Mplus program. The goodness of fit indices were Chi-square index, RMSEA, CFI and TLI. Descriptive statistics were used. Pearson’s correlations were using to analyze the concurrent validity among the mean scores of RSES, desired behavior questionnaire, life fulfillment and Hospital Anxiety Depression (HAD) scales.

Results  ESEM analysis was done in 551 cases. There were 8 items retained (RSES_8) and contained 2 domains i.e. proud and unproud. CFA (in 2110 cases) showed the goodness of fit indices at good level and demonstrated configure invariance among gender. The internal consistency for the total scale and each domain (proud and unproud) by using composite reliability tests were 0.87, 0.70 and 0.84 respectively. Pearson’s r correlations between RSES_8 - desired behavior, RSES_8 - life fulfillment scores and RSES_8 – HAD were 0.47, 0.27, -0.47 respectively.

Conclusion  The 8 item- RSES showed the good validity among teenagers, young adults and juvenile delinquents. The whole scale or each domain can be used as an outcome measurement for health related self-esteem activities.