Conclusion
In conclusion, this study was the first to use ESEM to evaluate the dimensionality of the EIS. The findings extended the lack of consensus regarding the psychometrics of the EIS, for example, omega estimates failed to support the subscales stability. Furthermore, despite the advantages of ESEM over traditional CFA and EFA procedures, support for a unidimensional and four-factor model was not provided. Support for Lane and colleague’s (1999) six factor model was provided; however, the model is not appropriate for use with athletic samples with poor fit for both weak and strong invariance models. Thus, inability to detect differences across sports expertise may be a result of methodological rather than theoretical suppositions. Previous research has suggested alternative measurement models for the EIS; however, we have not provided an alternative estimation of the model as this only adds to the lack of consensus in the literature (Gignac, 2009; Laborde & Allen, 2016; Mayer et al., 2008). Alternatively, we call interested researchers to clarify and refine the EIS conceptualization, providing a clear rationale for the measure. The present findings suggest that the EIS is not a suitable measure of TEI in sport, and caution is warranted in future use with the scale.