Un viaje alrededor de alfa y omega para estimar la fiabilidad de consistencia interna

Abstract

Resumen: En este trabajo se presenta una guía conceptual y práctica para estimar la fiabilidad de consistencia interna de medidas obtenidas mediante suma o promedio de ítems con base en las aportaciones más recientes de la psicometría. El coeficiente de fiabilidad de consistencia interna se presenta como un subproducto del modelo de medida subyacente en las respuestas a los ítems y se propone su estimación mediante un procedimiento de análi-sis de los ítems en tres fases, a saber, análisis descriptivo, comprobación de los modelos de medida pertinentes y cálculo del coeficiente de consistencia interna y su intervalo de confianza. Se proporcionan las siguientes fórmu-las: (a) los coeficientes alfa de Cronbach y omega para medidas unidimensionales con ítems cuantitativos (b) los coeficientes omega ordinal, alfa ordinal y de fiabilidad no lineal para ítems dicotómicos y ordinales, y (c) los coeficientes omega y omega jerárquico para medidas esencialmente unidimensionales con efectos de método. El procedimiento se generaliza al aná-lisis de medidas obtenidas por suma ponderada, de escalas multidimensionales, de diseños complejos con datos multinivel y/o faltantes y también al desarrollo de escalas. Con fines ilustrativos se expone el análisis de cuatro ejemplos numéricos y se proporcionan los datos y la sintaxis en R. Palabras clave: Fiabilidad; consistencia interna; coeficiente alfa; coeficiente omega; medidas congenéricas; medidas tau-equivalentes; análisis factorial confirmatorio.Title: A journey around alpha and omega to estimate internal consistency reliability. Abstract: Based on recent psychometric developments, this paper presents a conceptual and practical guide for estimating internal consistency reliability of measures obtained as item sum or mean. The internal consistency reliability coefficient is presented as a by-product of the measurement model underlying the item responses. A three-step procedure is proposed for its estimation, including descriptive data analysis, test of relevant measurement models, and computation of internal consistency coefficient and its confidence interval. Provided formulas include: (a) Cronbach's alpha and omega coefficients for unidimensional measures with quantitative item response scales, (b) coefficients ordinal omega, ordinal alpha and nonlinear reliability for unidimensional measures with dichotomic and ordinal items, (c) coefficients omega and omega hierarchical for essentially unidimensional scales presenting method effects. The procedure is generalized to weighted sum measures, multidimensional scales, complex designs with multilevel and/or missing data and to scale development. Four illustrative numerical examples are fully explained and the data and the R syntax are provided. Key words: Reliability, internal consistency, coefficient alpha, coefficient omega, congeneric measures, tau-equivalent measures, confirmatory factor analysis.Hubo una época en que el coeficiente alfa de Cronbach (α, Cronbach, 1951) era ampliamente aceptado como indicador de la fiabilidad de u...

Psychometric properties of the MHC‐SF: An integration of the existing measurement approaches

“…The coefficient is interpreted similarly to other popular indices of reliability, i.e., a higher value indicates greater reliability. We decided not to use Cronbach's coefficient due to its assumptions, for example, requiring seldom met in psychological science essentially -equivalence (which is the assumption that all discrimination parameters, as e.g., factor loadings, are equal) or its limited use with multidimensional measures (including bifactorial models; Viladrich, Angulo-Brunet, & Doval, 2017). The coefficient, which is a factor analytic model-based reliability estimate, overcome these difficulties (Green & Yang, 2015;Rodriguez et al, 2016).…”

Sustainable Information in Shoe Purchase Decisions: Relevance of Data Based on Source

“…Table 2 summarizes the means, standard deviations, skewness, and kurtosis measures of all the characteristics or items in Section 2. According to the results, the items have sufficient dispersion, with standard deviations between 0.954 (lowest) and 2.563 (highest), which, according to Viladrich et al [58], are acceptable values. It can be seen that variables with an average higher than four are items m, n, o y, and p, i.e., the manufacturer follows international agreements on workers' rights, gives information on the environmental impact, pollution aspects can be compared with others, the information on the label can be verified online, and the ranking or rating of the product within its category (e.g., energy efficiency labelling for white line products).…”

Entwicklung eines Messinstruments zu den Grunderfahrungen des Informatikunterrichts

“…Basierend auf der in Studie 1 gefundenen Faktorenstruktur, mit den drei latenten Konstrukten GI1, GI2 und GI3 und den insgesamt zwölf Prädiktoren, wurde in R mit dem Paket " Lavaan 0.6-1" eine CFA durchgeführt. Als Schätzer wurde Robust Maximum Likelihood (RML) gewählt, da dieses Verfahren für ordinale Daten und bei mäßigen Verletzungen der Normalverteilung geeignet ist (Li, 2016) (Viladrich, Angulo-Brunet & Doval, 2017).…”