Using logistic regression-based procedures to establish predictability of NCLEX-RNRTM

Logistic modeling was employed to predict the degree to which use of a supplementary nursing curriculum and other student and program characteristics account for the success or failure on the National Council Licensure Examination for Registered Nurses (NCLEX-RNRTM). A comparison of the results is reported with demonstration of the pragmatic approach used. The benefits and pitfalls of both regression-based procedures are covered as it pertains to analysis of this binary outcome data. Finally, an interpretation of the results of significance testing for the prediction study ensues.; Nursing students enrolled in associate (54%) and baccalaureate degree (46%) nursing programs nationwide. All participants took the RN Comprehensive Predictor 3.0 (RNCP) six months before graduation. Approximately 60% of the students received a supplemental curriculum for review of the NCLEX-RN RTM. The other 40% did not.; Maximum log likelihood estimates of the parameters and their significance to the model were obtained using traditional and hierarchical logistic regression-based procedures. The overall predictive validity for each model was obtained using regression classifications. Prediction of success for NCLEX-RNRTM pass (81.5%) was slightly better than prediction of failure (74.6%). However, in light of the small number of predicted and observed failures, the overall classification for pass was nonetheless accurate more than three-quarters of the time (80.7%). High predictability for failure is unlikely primarily due to the motivational influence that receiving a low score has for an individual striving for a passing score on a high-stakes test. This stimulus for improvement can lead to remediation, which can change the outcome predicted.; The results of this study illustrate the pragmatic usefulness of applying hierarchical procedures for analysis of student variables nested within educational program characteristics, particularly when there is interest in providing accurate estimates at both levels. Although the regression procedures produced minimally different results, characteristics were common to both: RNCP, Program Pass and School Diversity. Policy implications for nursing programs striving to improve or maintain a high rate of pass on the nursing licensure exam strongly indicate the usefulness of the RNCP for identifying students at risk for failing as well as the influence of program quality and diversity.