given that the data for the IVs and DVs came from different sources. Thus, we next explored whether the variables from each source were distinguishable from one another. To do this we ran a 2-factor model for the independent variables and a 2-factor model for the dependent variables. As shown in Table 1, the fit statistics for the dependent variable model were acceptable, but the model for the IVs indicated improvements could be made. Following our logic for sample 1, we reran the IV model as a second-order model that modeled both empowerment and LMX as higher order factors each composed of 4 subscales. The fit for this model (see Table 1) was acceptable. In total, these analyses suggest that the scales provided by the same raters were distinct from one another and that using the empowerment and LMX scales unidimensionally was appropriate.