Development and factorial validity of a moral disengagement in Sport Short Scale

According to Bandura, individuals are able to violate their personal standards, without self-sanction, by using the psychological operations of moral disengagement. For Bandura et al., moral disengagement is characterized by eight mechanisms belonging to one of the following four groups: (a) reconstructing conduct; (b) reconsideration of negative effects; (c) disqualification of the victim; and (d) obscuring of personal causal agency. Other researchers have measured moral disengagement in various contexts of everyday life using Bandura et al.'s scale and suggested that moral disengagement mechanisms would fall into two or three groups according to context. One context in which moral issues have a major role is sport. Three complementary studies were carried out on a total of 1305 young French adult athletes to develop and validate a Short French Questionnaire of Moral Disengagement in Sport (SFQMDS) and to test its invariance according to gender. STUDY 1: With reference to the existing literature, an initial French version of the SFQDMS was developed. French university students (n=220) then voluntarily completed the questionnaire. The validity of this preliminary version and the clarity of the items were examined and ascertained, and factorial analyses identified 10 items that loaded onto two factors (i.e., projecting fault onto others or sharing of responsibility; minimization of transgression and their consequences). Each factor displayed good internal consistency. STUDY 2: Confirmatory factor analysis (CFA) was conducted using AMOS 7.0 software. The sample included 1021 French university students (M(age)=21.52; SD=2.34). The first analysis of the data from 298 French students suggested that four items should be eliminated. The six-item model was then tested with a CFA of the data from 723 other participants (M(age)=21.51; SD=2.34) and exhibited acceptable fit indices: (χ² [8, 723]=1.54; p>0.09; GFI=0.97; TLI=0.97; CFI=0.97; RMSEA=0.03; RMSEA LO/HI=0.01/0.05). These results confirmed the bifactorial structure of the instrument, as well as its partial invariance across genders at the most complex level (i.e., strict) of its factorial structure. These statistical analyses demonstrated the excellent internal consistency and very good construct validity of the SFQDMS. STUDY 3: The third study examined the temporal stability of the SFQDMS and its theoretical validity with a sample of 221 French students (M(age)=21.00; SD=2.05). Our results were found to be st