AUTHOR=García-Batista Zoilo Emilio , Guerra-Peña Kiero , Garrido Luis Eduardo , Cantisano-Guzmán Luisa Marilia , Moretti Luciana , Cano-Vindel Antonio , Arias Víctor B. , Medrano Leonardo Adrián 

TITLE=Using Constrained Factor Mixture Analysis to Validate Mixed-Worded Psychological Scales: The Case of the Rosenberg Self-Esteem Scale in the Dominican Republic

JOURNAL=Frontiers in Psychology

VOLUME=Volume 12 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2021.636693

DOI=10.3389/fpsyg.2021.636693

ISSN=1664-1078

ABSTRACT=The Rosenberg Self-Esteem Scale (RSES) is one of the most used instruments for assessing self-esteem. Nevertheless, there is a dispute over the dimensionality of the scale. The objective of the current study was to examine the psychometric properties of the RSES scores for the Dominican population. In order to evaluate the possibility of misresponses due to the combination of regular and reversed-coded items we employed factor mixture analysis. The dimensionality and factorial analyses were performed on the total (N = 632) and the screened (N = 579) samples. Four models were evaluated: the one-factor model; a model with a substantive self-esteem factor and a wording method factor, using random intercept item factor analysis; a two-factor restricted factor model with the regular and reversed items loading on different factors; and a two-factor unrestricted factor model where the items were free to load on both factors. Besides, the internal consistency reliability and the criterion validity were assessed. The results demonstrated that RSES is, mainly, a one-factor scale. The one-dimensional model showed optimal internal consistency reliability and criterion validity. It is important to consider the sociocultural differences of the evaluated groups to prevent biased results on dimensionality, psychometric properties and scores of the RSES.