Abstract
Ten different confirmatory factor analysis models, including ones with correlated traits correlated methods, correlated traits correlated uniqueness, and correlated traits uncorrelated methods, were proposed to examine the factorial structure of the French version of the Rosenberg Self-Esteem Scale (Rosenberg, 1965). In line with previous studies (Tomas & Oliver, 1999; Wang, Siegal, Falck, & Carlson, 2001), our results support the existence of a single global self-esteem factor underlying responses to the scale, in spite of the fact that method effects appear to be associated more with negatively worded items than with positively worded ones.