Adaptation and validity of the Type D personality scale (DS-14) in a Mexican sample

Abstract

espanolEl objetivo del presente trabajo consistio en explorar las propiedades psicometricas de una adaptacion al castellano del cuestionario de personalidad tipo D (DS-14) en una muestra de mexicana (n = 702). Se evaluo la escala mediante un enfoque no parametrico basado en el escalamiento Mokken y se aplico el analisis factorial confirmatorio por el metodo de estimacion minimos cuadrados ponderados con media y varianza ajustada (WLSMV), consistencia interna con coeficiente omega (ω), invarianza de medicion por genero, y validez convergente con medidas de salud mental (MBI = burnout, GHQ = depresion, trastornos de sueno, disfuncion social y sintomas somaticos), parcializando los efectos de estresores laborales (JCQ = demandas y control laboral). Los resultados evidenciaron correlaciones parciales de bajas a moderadas de los items DS-14 con los puntajes de salud mental, asi como indices de confiabilidad aceptables (ω =.62-.86). Se hallo invarianza de medicion entre hombres y mujeres, y la estructura bifactorial original mostro indices de ajuste satisfactorios (CFI = .97, RMSEA = .04, SRMR = .05). Adicionalmente se vislumbro la viabilidad de una estructura unidimensional, y en particular para una version de 9 items (CFI = .99, RMSEA = .04, SRMR = .04), la que necesita estudios adicionales de replicabilidad. Se concluye que la escala DS-14 produce puntajes validos y confiables, y se discuten estos hallazgos y sus implicaciones para la investigacion futura. EnglishThe goal of this study was to explore the psychometric properties of a Spanish version of the Type D Personality Scale (DS-14) in a sample of Mexican participants (n = 702). The scale was evaluated through a non-parametric approach based on Mokken scaling. Moreover, confirmatory factor analyses were performed using the weighted least squares mean and variance (WLSMV) method, internal consistency estimation with omega coefficient (ω), measurement invariance by gender, and convergent validity with mental health measures (MBI = burnout, GHQ = depression, sleep disorders, social dysfunction, and somatic symptoms), and controlling job stressors (JCQ = job demands and job control). The results showed low to moderate partial correlations in DS-14 items concerning the mental health scores, and acceptable reliability rates (ω = .62-.86). Measurement invariance between men and women was found, and the original two-factor structure showed satisfactory adjustment indices (CFI = .97, RMSEA = .04, SRMR = .05). Furthermore, the viability of a one-dimensional structure, which needs future replicability studies, was glimpsed particularly in the 9-item version (CFI = .99, RMSEA = .04, SRMR = .04). It is concluded that the DS-14 scale generates valid and reliable scores, and these findings and implications are discussed for further research.