The primary purposes of the study were (1) to investigate whether three subtests of the Comparative Guidance and Placement Program (CGP): Mosaic Comparisons, Year 2000 and Letter Groups measure skills uniquely different from traditional verbal and mathematical skills; and (2) to test whether the New Jersey College Basic Skills Placement Test subtests are measuring the same skills as similarly named subtests of the CGP. The methodology employed was a confirmatory factor analysis using the COFAMM computer program. Data from 822 students who had taken both batteries were used to test a hypothesized four-factor model (Reading, Sentences, Mathematics, and Mosaic Comparisons). This model was found to fit the data; that is, subtests with the same names measured the same skills. Basic Skills subtests, Mosaic Comparisons, Year 2000, and Letter Groups, however, each measured something uniquely different from Reading, Sentences, and Mathematics. Although technically complex, this methodology is easily and inexpensively applied to this type of problem. It can be particularly useful in criterion-referenced test development for testing whether a priori subscales are actually measuring different skills. (Author/CP)