ABSTRACT
In chapter 4, several problematic aspects of post hoc model-fitting in SEM were highlighted. Indeed, so common is this practice and so frequently is it conducted with little regard for the substantive meaningfulness of the respecified models that concerned researchers have provided various means by which such models can be tested more stringently (see, e.g., Anderson & Gerbing, 1988; Cudeck & Henly, 1991; MacCallum, 1995; and MacCallum et al., 1992,1993). It was noted in chapter 4 that one approach to addressing problems associated with post hoc modelfitting is to apply some mode of cross-validation analysis; this is the focus of the present chapter. The application demonstrated herein is a follow-up to the analyses of chapter 6 in which the validity of a causal structure describing determinants of burnout for a calibration sample of secondary school teachers was tested. However, before a walk-through of this cross-validation procedure, it is important to review some of the related issues.
