A cross-validation study for predictors of scores on state board examinations

Regression equations were computed and cross-validated for predicting nurses' scores on each of the five state board examinations (SBEs). Scores on five National League for Nursing (NLN) achievement tests of 101 nurses who graduated from a baccalaureate program between 1968 and 1972 were used as predictor variables. Stepwise regression analysis was used to compute the equations which were, in turn, used to predict the SBE scores for an independent sample of nurses who graduated in 1973. Cross-validation correlations between the predicted and actual obtained scores of the 1973 graduates ranged from .64 to .81. For each equation, the NLN test scores in Nursing of Children and Obstetric Nursing were consistently the best indicators of performance on the SBEs. Additionally, a factor analysis indicated that the SBEs do not measure independent entities, but that all five examinations have high loadings on the same factor. The results suggest that nursing programs should develop and validate prediction equations to assist nurses in preparing for SBEs.