The Sociocultural Attitudes Towards Appearance Scale-3 (SATAQ-3) : étude de validation de la version française

Abstract

Body image literature suggests that the sociocultural environment contributes to the development and maintenance of body dissatisfaction and eating disorders by conveying thin standards of thinness. Three main sources of sociocultural influences were highlighted: family, peers and media. These three sources of pressure can lead to the internalization of messages about the importance of thinness and beauty. Among the various sources of sociocultural influences, media appear to be the most powerful communicator of these standards. Thompson et al. [Intern J Eat Disord 35 (2004) 293-304] developed a multidimensional questionnaire to assess the sociocultural influence of media on appearance: The Sociocultural Attitudes Towards Appearance Scale-3 (SATAQ-3). The SATAQ-3 is a 30-item scale composed of Likert-items in five points (1: completely disagree; 5: completely agree) including four subscales: information, pressures, internalization-general and internalization-athlete. This scale is used in Anglo-Saxon research, notably in clinical populations presenting eating disorder pathologies. People with eating disorders generally show higher scores on these four subscales. The present study is to validate the French version of the SATAQ-3. The sample consisted of 818 teenagers from private and public middle schools aged an average of 16.8 ± 1.2 years old, with an average body mass index (BMI) of 20.7 ± 2.9. Participants completed two questionnaires: the body shape questionnaire (BSQ) assessing body dissatisfaction, in particular worries about weight and body shape, and the SATAQ-3. We carried out the translation of the SATAQ-3 in parallel back-translation. First, two French-speaking people translated the English version into French. Then, two English-speakers translated the French version without the assistance of the original version. Thereafter, the two English-language versions (original version and version obtained by back-translation) were compared in order to ensure that there were no errors of meaning in the French version. A factorial analysis of the SATAQ-3 was conducted on the 818 subjects through extraction of the main components method with the extraction's rule of the curve for the Eigen values; the data was transformed through Varimax rotation. The factorial analysis showed five factors with Eigen values greater than 1 (11.3; 2.6; 2.1; 1.6; 1.07). The factors were then composed of the items presenting substantial saturation, greater than or equal to 0.5. Since only one item (20) presented saturation on factor 5, the four-factor solution was retained. These four factors accounted for 17.7, 16.3, 9.7 and 14.8% of the overall variance, respectively. This four-factor solution explains 58.5% of the total variance. The oblique rotation analysis underlined moderate to strong correlations between the four factors (0.3 to 0.66). The four factors of the factorial analysis correspond to the four subscales of the initial American study: pressures (Cronbach's alpha=0.89), information (Cronbach's alpha=0.87), internalization-athlete (Cronbach's alpha=0.82), internalization-general (Cronbach's alpha=0.92). The four subscales are positively correlated with the BSQ (0.26 to 0.55). The correlation for age and BMI indicates a very low association between these two variables and the SATAQ-3 subscales, with only one significant correlation (SATAQ-3 pressures and BMI=0.14; p<0.05). Results of a regression analysis showed that body dissatisfaction (BSQ) was predicted by pressures (β=0.37, t=10.2; p<0.0001), internalization-general (β=0.29, t=8.1; p<0.0001). This model accounted for 38.1% of the variance of body dissatisfaction. This study shows elements of satisfactory validity for the French version of the SATAQ-3. Indeed, as for the original scale validation, the current factorial analysis revealed four factors corresponding to the four subscales developed by the authors along with similar alpha values. Research relating to the risk factors in the development of body image disturbances and eating disorders are important given their implications for future strategies of intervention and prevention. Anglo-Saxon research identified media as a powerful and influential communicator concerning sociocultural standards of beauty. Because of the lack of French studies on this subject, the newly validated SATAQ-3 could provide a means for exploring new issues related to eating disorders in this population. Implications of the SATAQ-3 are multiple, particularly to detect "risk groups" and to prevent body image disturbance and eating behaviours. The current results provide a means for research on media influence, notably in French clinical samples with eating disorders.