Author:
Deflandre Anne,Kassabian Arine
Abstract
The aim of this manuscript is to discuss the reliability and the factorial, convergent and discriminant validities in measurement model of the 21-item Exercise Dependence Scale Revised (EDS-R) used for diagnosing exercise addiction. A study was conducted among 695 (364 male and 331 female) Lebanese bilingual young adults (aged 18–25) who filled the English version of this questionnaire, along with a general questionnaire. This study showed that, despite an acceptable fit index confirming the 7-factor structure and invariance across gender, exercise duration and age groups; the EDS-R lacks in reliability and convergent and discriminant validities in measurement model, due particularly to the “Reduction in other activities” subscale. The construction of the EDS-R is then to be discussed.