Assessing the predictive validity of exam scores for competency: A cross-sectional study of nursing and midwifery admissions in Ghana

Introduction Globally, there is growing concern about the predictive validity of standardised examinations. Admissions into nursing and midwifery training colleges in Ghana rely on the West Africa Senior School Certificate Examination (WASSCE), selection interviews, and non-meritocratic considerations – the controversial influence of ‘protocol’. The WASSCE scores appear to be the first criterion for assessing applicants’ competencies despite growing evidence questioning their validity. This necessitates evidence-based evaluation of the competencies of applicants. Objective The study aimed to assess differential competencies in five domains (foundational knowledge, subject-specific mastery, real-world application, general aptitude, and communication) among nursing/midwifery applicants from diverse academic backgrounds. The study also explored the predictive capacity of WASSCE scores and school backgrounds on subject-specific performance. Method In this cross-sectional study, we assessed 469 nursing school applicants in Ghana using stratified random sampling. Trained examiners scored the applicants on the five competency domains using a standardised instrument. We conducted analyses that included descriptive statistics, ANOVA, and linear regression, taking into account the sampling weights. The instrument demonstrated excellent reliability (Cronbach’s alpha = .96). Results The analysis shows that traditional indicators such as WASSCE scores, school grades, and the type of school attended did not predict performance effectively (all p > 0.05, R² < 0.01). The sample showed two critical and statistically significant differences: a strong gender imbalance (90.2% female, p < 0.001) and dominance of General Arts (60.3%) over Science (9.4%, p < 0.001), with Science applicants outperforming their Arts counterparts in foundational knowledge. The only statistically significant predictor was age, with older candidates performing worse (β = -0.174, p < 0.001). Most importantly, all groups demonstrated inadequate competency, with mean scores of 0.71 to 1.08 out of 2.00 in every domain. Conclusion This study shows that the WASSCE score and the school tier of students applying for nursing and midwifery programmes are ineffective in predicting foundational knowledge and practical skills of students. The study recommends changes to the methods used to admit nursing students in Ghana. We suggest including standardised assessments based on competencies to reduce the negative impact of examination malpractice and non-meritocratic “protocol” admissions.