Does Self-Objectification Entail an Opposition Between Appearance and Competence? The Likert Version of the Self-Objectification Questionnaire (LSOQ)

We propose a new method to test the reliability of Fredrickson et al.’s self-objectification questionnaire (SOQ). This scale being based on a ranking, traditional reliability estimates are inappropriate. Based on generalizability theory, we suggest to compute the reliability of each subset of questions related to physical appearance vs. physical competence separately in order to average them. We applied this method to a sample of female US undergraduates (n = 395) and evidenced that the reliability of the scale is very low (corrected Cronbach’s alpha = .31). We also noted that a large proportion of the sample (32%) failed to complete the scale correctly. In a second study (n = 93), we propose a Likert adaptation of the scale and show that the two dimensions of the SOQ are independent. In Study 3 (n = 195), we confirm results of Study 2 and demonstrate that the general structure of the Likert version has satisfactory model fit statistics. These observations lead us to discourage the use of the original version of the SOQ and rely on the Likert version of the Self-Objectification Questionnaire (LSOQ, see appendix).