Abstract
The main objective of the study was to verify the stability of the seven-factor structure of the Exercise Dependence Scale-Revised (EDS-R). The EDS-R by Hausenblas and Symons-Down (2002) is one of the commonly used measures to assess exercise dependence. This study was conducted to examine the factor structure of the EDS-R and determine if the seven-factor model is supported in the Filipino context. Additionally, this study aimed to evaluate if the seven criteria or dimensions of the EDS-R represent seven distinct factors as operationalized by the Diagnostics Statistical Manual of Mental Disorder IV-TR (DSM-IV-TR). The English version of this questionnaire was administered to 373 Filipino fitness enthusiasts (224 males and 149 females) aged 18 years old and above. Since the seven factors were assumed to be correlated, exploratory factor analysis (EFA) using parallel analysis and maximum likelihood method with oblique rotation (promax) were used. Using Bartlett’s test of sphericity and the Kaiser-Meyer-Olkin (KMO) test, the correlation matrix was not random, and the test was found to be factorable. The KMO test also indicated adequate sampling yielding .92. The cut-off point of the minimum residuals factor analysis was set at .40. Fit indices, parallel analysis, eigenvalue greater than 1, and the scree test were compared to determine the number of factors and assess the adequacy of the model. Parallel analysis and the maximum likelihood estimation were used to determine the number of factors. Results suggest that the seven-factor model was not supported. Parallel analysis supported a six-factor model which accounted for 64.5% of the variance. Factor loadings range from .41 to .97, except for Item 12 which has a factor loading of .29). Although the Tucker-Lewis Index (TLI) for the six-factor model is lower than .95, it is still better compared to the seven-factor model or the two-factor model. Since most of the correlations between the factors in the six-factor model range from moderate (r = .49) to high (r = .74), the three-factor model’s utility was supported. Only Factor 2 and Factor 5 are weakly correlated (r = .26). Despite this, the results suggest that the six-factor model better reflects the experiences of exercise dependence among Filipino fitness enthusiasts.
