Five theoretically salient predictors of cynicism were used in the survey: anomia, education, professionalism, length of service, and work alienation. CFRM seems to be useful for constructing composite indices; the analysis showed that CFRM allowed a more powerful test of hypothesized relationships between dimensions of cynicism and the predictor variables than had been previously possible. However, the comparison of alpha and CFRM-omega reliability estimates suggests that the assumption of parallelity between measures is problematic. Hypothesis testing and construct validity are two issues that remain pertinent when using CFRM. 3 tables, 3 appendixes, and 28 references (Author abstract modified)