Abstract

The increasing availability of educational data has created opportunities for the application of predictive analytics in identifying students at risk of poor academic performance before learning difficulties become severe. This study investigated the application of predictive analytics for identifying at-risk students in Computer Science classes in secondary schools in Benin City, Edo State, Nigeria. The study adopted a descriptive survey design with correlational and predictive modelling components. The population comprised Senior Secondary II students offering Computer Science and their teachers in public and private secondary schools. A sample of 420 students and 35 Computer Science teachers was selected using multistage sampling techniques. Data were collected using the Predictive Indicators Questionnaire (PIQ) and the Students' Academic Performance Record Sheet (SAPRS). The instruments were validated by experts, while the PIQ yielded a Cronbach's alpha reliability coefficient of 0.91. Data were analysed using descriptive statistics, binary logistic regression, and decision tree classification techniques. The findings revealed that previous academic performance, attendance, classroom participation, assignment completion, continuous assessment scores, digital literacy, and socioeconomic background significantly predicted students' academic risk in Computer Science. The predictive model achieved an overall classification accuracy of 91.4%, with high sensitivity (89.8%), specificity (92.6%), precision (90.5%), and an area under the ROC curve of 0.94, indicating excellent predictive performance. The study concluded that predictive analytics provides an effective evidence-based approach for the early identification of at-risk students and can support timely intervention, personalised learning, and improved instructional planning. It recommended the integration of predictive analytics into secondary school academic monitoring systems, investment in educational data infrastructure, and continuous teacher capacity development in learning analytics and artificial intelligence applications.