Abstract
Despite the limitations of overgeneralizing cutoff values for confirmatory factor analysis (CFA; e.g., Marsh, Hau, & Wen, 2004), they are still often employed as golden rules for assessing factorial validity in sport and exercise psychology. The purpose of this study was to investigate the appropriateness of using the CFA approach with these cutoff values for typical multidimensional measures. Furthermore, we ought to examine how a model could be respecified to achieve acceptable fit and explored whether exploratory structural equation modeling (ESEM) provides a more appropriate assessment of model fit. Six multidimensional measures commonly used in sport and exercise psychology research were examined using CFA and ESEM. Despite demonstrating good validity in previous research, all eight failed to meet the cutoff values proposed by Hu and Bentler. ESEM improved model fit in all measures. In conclusion, we suggest that model misfit in this study demonstrates the problem with interpreting cutoff values rigidly. Furthermore, we recommend ESEM as a preferred approach to examining model fit in multidimensional measures.
Highlights
Despite the limitations of overgeneralizing cutoff values for confirmatory factor analysis (CFA; e.g., Marsh, Hau, & Wen, 2004), they are still often employed as golden rules for assessing factorial validity in sport and exercise psychology
Less than 0.1% of data was missing in all samples, and there were no issues with outliers, following examination of Q-Q plots
Respecification of the measurement models significantly improves model fit, and this is an option for researchers encountering misspecifications
Summary
Despite the limitations of overgeneralizing cutoff values for confirmatory factor analysis (CFA; e.g., Marsh, Hau, & Wen, 2004), they are still often employed as golden rules for assessing factorial validity in sport and exercise psychology. Hu and Bentler (1999) proposed cutoff criteria for all commonly cited fit indices by examining rejection rates on hypothetical models These proposed criteria, including Comparative Fit Index (CFI) and Tucker-Lewis Index (TLI) close to .95, Standardized Root Mean-square Residual (SRMR) of close to .08, and Root Mean Square Error of Approximation (RMSEA) of close to .06, are referred to as a matter of routine in studies using any kind of structural equation methods. Such personality assessments could perhaps perform better in a CFA by reducing their size and/or complexity, but if this is at cost of predictive or other forms of validity, it is not a virtuous academic pursuit
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Measurement in Physical Education and Exercise Science
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
