A Bifactor Exploratory Structural Equation Modeling Framework for the Identification of Distinct Sources of Construct-Relevant Psychometric Multidimensionality

Morin, Alexandre J. S.; Arens, A. Katrin; Marsh, Herbert W.

doi:10.1080/10705511.2014.961800

Cited by 629 publications

(1,028 citation statements)

References 88 publications

Supporting

Mentioning

980

Contrasting

Unclassified

Order By: Relevance

“…As noted above, each of these models is relatively efficient at absorbing unmodelled construct-relevant multidimensionality (e.g., Asparouhov et al, 2016;Morin, Arens et al, 2016;Murray, & Johnson, 2013). Thus, unmodelled cross-loadings could typically result in inflated CFA factor correlations, or inflated bifactor-CFA estimates of the loadings on the G-factor.…”

Section: Analysesmentioning

confidence: 99%

“…This comparison is critical given that excluding true cross-loadings present in the population model from a bifactor model has been shown to result in inflated estimates of the variance attributed to the G-Factor (i.e., as illustrated by inflated estimates of the indicators' loadings on the G-factor; Morin, Arens et al, 2016;Murray, & Johnson, 2013). Similarly, ignoring the presence of hierarchically-superior (global) constructs is also likely to result in inflated estimates of cross-loadings in ESEM, or inflated factor correlations in CFA (Morin, Arens et al, 2016). In other words, even when the objective is simply to assess the presence of an overarching global factor (bifactor model), ignoring cross-loadings is likely to result in biased estimates of this global factor which will be forced to absorb unmodelled cross-loadings.…”

Section: A Variable-centered Perspectivementioning

confidence: 99%

“…In practice, this "absorption" will typically lead to biased estimates of the model parameters (such as the factor correlations) through which these moment. When ICM-CFA models force these associations (i.e., cross-loadings) to be zero, the only way for them to be expressed is through the inflation of the factor correlations, in turn leading to biased estimates of the relations among constructs according to a variety of simulation studies (Asparouhov & Muthén, 2009;Sass & Schmitt, 2010;Schmitt & Sass, 2011) and studies of simulated data Morin, Arens et al, 2016) in which the true population value of the factor correlations are known beforehand (for a review see Asparouhov, Muthén, & Morin, 2015). Given that the meaning of a construct lies in the way it relates to other constructs, these results suggest that the exclusion of true cross-loadings may induce a fundamental bias in the definition of the constructs.…”

Section: A Variable-centered Perspectivementioning

confidence: 99%

See 2 more Smart Citations

Complementary Variable- and Person-Centered Approaches to the Dimensionality of Psychometric Constructs: Application to Psychological Wellbeing at Work

et al. 2016

Self Cite

View full text Add to dashboard Cite

Section: Analysesmentioning

confidence: 99%

Section: A Variable-centered Perspectivementioning

confidence: 99%

Section: A Variable-centered Perspectivementioning

confidence: 99%

See 1 more Smart Citation

Complementary Variable- and Person-Centered Approaches to the Dimensionality of Psychometric Constructs: Application to Psychological Wellbeing at Work

et al. 2016

Self Cite

View full text Add to dashboard Cite

“…ESEM offers the possibility to integrate features of CFA, structural equation modeling (SEM), and exploratory factor analysis (EFA) in a single framework. This decision is based on the results from simulation studies Sass & Schmitt, 2010;Schmitt & Sass, 2011) and studies of simulated data (Marsh, Lüdtke, Nagengast, Morin, & Von Davier, 2013;Morin, Arens, & Marsh, 2015) showing that forcing cross loadings (even as small as .100, Marsh et al, 2013) present in the population model to be exactly zero according to typical CFA specification forces these cross loadings to be expressed through an inflation of the factor correlations. In contrast, these same studies show that the free estimation of cross-loadings, even when none are present in the population model, still provides unbiased estimates of the factor correlations (also see Asparouhov, Muthén, & Morin, 2015;.…”

Section: Preliminary Measurement Models and Tests Of Measurement Invamentioning

confidence: 99%

Achievement goal profiles among adolescent males and females.

Litalien

Morin

McInerney

2017

Developmental Psychology

Self Cite

View full text Add to dashboard Cite

Achievement goal theory has long been a dominant model in the study of student motivation. However, a relatively small number of researchers have investigated gender differences in achievement goals or have considered the possible role that social and extrinsic goals may play in student academic motivation. Adopting a person-centered multiple goals perspective based on personal investment theory, this longitudinal study investigated whether males and females shared similar goal profiles, and whether the predictors (facilitating conditions) and outcomes (learning processes, task perseverance, and future aspirations) of these profiles were equivalent across genders. Profiles were extracted from 8 types of academic goals, based on a large sample of Hong Kong high school students (N = 7,848). Findings revealed 5 distinctive profiles for both males and females. Although the relative size of these profiles differed across samples of male and female students, the results show that 4 of these profiles were mostly equivalent across genders. Predictors of membership into these profiles were also equivalent across genders, whereas their relative outcomes were specific to gender.

show abstract

“…En constructos psicológicos complejos, como suelen ser los rasgos de personalidad, la unidimensionalidad pura es una entelequia, incluso a nivel del item (para una discusión en profundidad acerca de este tema, vid. eg., Morin et al, 2015). Un constructo puede tender a la unidimensionalidad psicométrica, en la medida en que las conductas que lo representan tiendan a producirse juntas.…”

unclassified

Análisis exploratorio de modelos de ecuaciones estructurales sobre la escala de resiliencia de Connor y Davidson (CD-RISC) en Chile y España.

Cisternas

2015

Salud soc.

View full text Add to dashboard Cite

RESUMENIntroducción: La resiliencia es el proceso de negociación, gestión y adaptación a las fuentes significativas de estrés o trauma. La CD-RISC es uno de los instrumentos para la evaluación de la resiliencia más utilizados, con diversos estudios sobre sus propiedades psicométricas en una gran variedad de países. Pese a ello, aun no existe consenso acerca de la estructura interna de la escala. Objetivo: modelar simultáneamente la comunalidad y las especificidades presentes en la escala, a fin de alcanzar una visión completa de la estructura dimensional de la CD-RISC en muestras distintas. Método: analizar las respuestas a la CD-RISC en muestras independientes originarias de Chile (n=376) y España (n=300), a fin de comprobar la estabilidad y solidez del modelo sobre muestras demográficamente similares pero pertenecientes a dos países distintos. Resultados: la CD-RISC en su totalidad es fuertemente unidimensional. El factor general fue estable entre ambas muestras, y explicó la mayor parte de la varianza fiable de los ítems. Después de parcializar el efecto de la dimensión general, las facetas no contribuyeron apenas a la medida, y fueron métricamente inestables entre ambas muestras. La excepción fue la faceta Espiritualidad, que resultó el único factor específico estable en ambas muestras. No obstante, su pobre definición en la CD-RISC (sólo dos ítems) impide llegar a conclusiones sobre la naturaleza substantiva del factor espiritualidad. Conclusiones: Las dimensiones de resiliencia no son demostrables desde un punto de vista psicométrico, por lo que tiene poco o ningún sentido utilizar las subescalas en forma de puntuaciones específicas, en cuanto que empíricamente no se diferencian entre sí. PALABRAS CLAVE: Resiliencia; CD-Risc; Estructura factorial. ABSTRACTIntroduction: Resilience has to do with dealing, negotiating and adapting to significant sources of stress or trauma. CD-RISC is one of the most common tools around the world to assess resilience, with a plethora of studies about their psychometric properties. Yet, there is no agreement regarding the internal structure of this scale. Objective: To apply simultaneously in two different samples the commonality and features of the CD-RISC in order to have a general vision of the dimensional structure of this scale by assessing its stability and solidity. Method: Answers to the CD-RISC from two different samples (Chile, n=376; Spain, n=300) were analyzed. Results: The CD-RISC is a strong one-dimensional scale. The general factor proved to be stable in both samples. That explains a reliable variance in all items. After being partial about the effect of the general dimension, the facets showed a different reality. They were metrically unstable in both samples. The Spiritual Facet was an exception, being the only stable specific factor in both samples. However, the poor definition of the spiritual factor (only two items) in the CD-RISC is not enough to reach conclusions about its substantive nature. Conclusions: From a psychometrical point of view, dimensions of...

show abstract

A Bifactor Exploratory Structural Equation Modeling Framework for the Identification of Distinct Sources of Construct-Relevant Psychometric Multidimensionality

Cited by 629 publications

References 88 publications

Complementary Variable- and Person-Centered Approaches to the Dimensionality of Psychometric Constructs: Application to Psychological Wellbeing at Work

Complementary Variable- and Person-Centered Approaches to the Dimensionality of Psychometric Constructs: Application to Psychological Wellbeing at Work

Achievement goal profiles among adolescent males and females.

Análisis exploratorio de modelos de ecuaciones estructurales sobre la escala de resiliencia de Connor y Davidson (CD-RISC) en Chile y España.

Contact Info

Product

Resources

About