Latent variables represent qualities that are not directly measured but only inferred from the observed covariation among a set of variables (Tabachnick & Fidell, 2001). Latent dimensions are hypothetical constructs that are used to explain behavior. Multivariate analyses are used to identify latent dimensions (e.g., Factor Analysis, Discriminant Analysis, Canonical Correlation), which are defined by the sets of weights that are assigned to the observed variables. Structural equation modeling is a statistical technique that allows one to examine the causal influences that exist among a set of latent dimensions. Latent dimensions are the essential elements of interest to social scientists because they represent the prime causal constructs operating in the mental world of individuals (Borsboom, Mellenbergh, & van Heerden, 2003).