Modification of The Rosenberg Scale to Assess Self-Esteem in Children

Background: Rosenberg’s self-esteem scale (RSES) is widely used to assess global self-esteem (SE) in adults and adolescents but is not validated for children <12years. This study assessed the internal consistency, validity and factor structure of a modified RSES for schoolchildren (CRSES) aged 7-12years. Methods: A total of 711 children aged 9.0±1.5 years completed the CRSES; a subset (n=417) also completed a life satisfaction (LS) scale. Data were submitted for confirmatory factor analysis (CFA) and tests of factorial invariance by sex. Two-way ANOVA compared scores by age-group and sex; whilst Pearson’s correlations examined the relationship between LS and SE. Results: Following the use of modification indices, CFA revealed that the fit for the global SE model met the criteria for all goodness of fit statistics: χ(27, n=711)=77.22; χnormed= 2.860 CFI=.961; RMSEA=.051 with 90%CI =.038-.065; SRMR=.037; and displayed respectable reliability (α = .79). The model was also factorially invariant by sex. SE scores did not vary sex (p>.05); but were significantly reduced in children aged 9-10years and 11-12years compared to children aged 7-8years. The global SE score was significantly correlated (r=-.51; P<.001) with LS. Conclusion: The CRSES can reliably examine global SE in children aged 7-12years; extending the use of the RSES to allow tracking across the life course.