PREDICTING NCE CHEMISTRY STUDENTS’ PERFORMANCE IN REDOX TITRATION: A MULTINOMIAL LOGISTIC REGRESSION APPROACH

Abstract

This study examined the predictive influence of examination scores, class attendance, and practical laboratory performance on students’ achievement in redox titration. Using a quantitative ex-post facto design, data were collected from 174 NCE 200-level chemistry students in Oyo, Nigeria. Participants were classified into low, average, and high performance categories using Jamovi (v2.7). Multinomial logistic regression indicated a moderate model fit (McFadden R2 = 0.196), with omnibus tests confirming all predictors were statistically significant (p < .001). High performance was significantly predicted by examination scores, attendance, and practical work. While the model distinguished high from low performers, it was less effective for the average category. These findings underscore the importance of integrating attendance monitoring and practical work into chemistry curricula. This study contributes a novel methodological approach to addressing academic performance challenges in Nigerian higher education.