We added age, sex (coded: 1 = men, 2 = women), ethnicity (coded: 1 = Hispanic, 0 = Non-Hispanic), race (dummy variables for each racial group coded: 1 = member, 0 = non-member), and education-level (coded: 1 = completed high school, 0 = did not complete high school) as covariates of each latent factor. Doing so reduced overall model fit for each model (Table 4). Yet even with these added covariates, the three factor model still exhibited relatively better model fit than the one and two factor solutions (Table 4).
