The purpose of this study was to identify academic variables that could be used to forecast success for minority, baccalaureate (BSN) students enrolled in higher education in the south. A second purpose was to develop a methodological framework for predicting success on the post-1988 versions of the National Council Licensure Examination for Registered Nurses (NCLEX-RN), that enhanced external validity, and generalizability.

The best predictors of success for minority BSN nursing students were the Mosby assesstest, school comprehensive exit exams, GPA microbiology, the college cumulative GPA, nursing course cumulative GPA, and ACT composite. The sample included 216 Black graduates of three BSN programs located in Louisiana, who took the July 1988 through July 1991 versions of NCLEX-RN. The findings indicated that high ACT scores are not necessarily predictive of NCLEX success, and the ACT and pre-admit GPAs should be used cautiously in eliminating minority students from nursing education.

A methodological design that incorporated discriminant analysis, factor analysis and a four stage variable selection process, employed prior to a stepwise procedure was used. The equation derived was externally valid, stable across schools, and correctly classified 96% of the students used in the variable analysis. Only two of forty-two subjects (4.8%), known to belong to the fail group, were misclassified. The cross-validation, correct classification rate for BSN graduates of a different school was 76%, and showed 26% improvement over what was expected by chance alone.