Mills A C, Becker A M, Sampel M E, Pohlman V C
Saint Louis University School of Public Health, MO.
J Prof Nurs. 1992 Nov-Dec;8(6):351-7. doi: 10.1016/8755-7223(92)90098-j.
Nine years of data from first-time nurse candidates taking the National Council Licensure Examination for Registered Nurses (NCLEX-RN) were examined to identify predictors of successful performance and determine probabilities of success. Using logistic regression, four statistical models at strategic time periods during an accelerated baccalaureate nursing program were tested to see when intervention might be initiated to influence student performance on the NCLEX-RN. By the end of the first semester the model could predict failure for 94 per cent of those who failed, but it was less consequential in correctly predicting success of those who passed. Significant variables placing nurse candidates at risk included their first-semester grade point average, sex, and whether they were foreign educated. By the end of the second semester, with each full letter grade increase in cumulative grade point average, nurse candidates had a 46 times better chance of passing the NCLEX-RN; American-educated nurse candidates had a 4.5 times better chance of passing than their foreign-educated counterparts. At the end of the final semester, with each full letter grade increase on the final cumulative grade point average, nurse candidates had a 97 times greater likelihood of performing successfully on the NCLEX-RN.