“…Because the items of rating scales used in past CFA studies are categorical (ordinal), the fit of the different factor models require evaluation using estimators that are appropriate for use with categorical data, namely, unweighted least squares, weighted least squares (WLS), or robust WLS (Brown, 2006). Except for Lahey et al (2008) and Spence (1997), past studies that evaluated internal validity of DSM anxiety disorders assessed by rating scales conducted CFAs with estimators that were not appropriate for use with categorical ordinal data (e.g., maximum likelihood estimation [ML]; robust ML) (i.e., De Ross et al, 2002;Nauta et al, 2004;Spence, 1998;Spence et al, 2003) or did not report the type of estimator used in the CFAs (i.e., Chorpita et al, 2005;Ebesutani et al, 2010). Conducting CFAs with estimators that are not appropriate for categorical data has the potential of yielding test statistics and standard errors that are not accurate (Brown, 2006).…”