DEVELOPMENT OF ROBUST AND COST-EFFECTIVE PREDICTIVE MODELS FOR IMPROVING STUDENTS’ PERFORMANCE IN PROGRAMMING COURSES
In this paper, a robust and cost-effective mobile-oriented system for predicting students’ performance in tertiary education programmes in Federal University, Oye-Ekiti, Nigeria was developed. The factors influencing the performance of students in programming related courses were investigated. Statistical approaches such as frequencies, mean, standard deviation, correlation and multiple regression were used for descriptive analyses and model development. Thorough analysis of the obtained dataset showed that major factors affecting the performance of students in programming courses are erratic power supply, bad university facilities, student health and students’ attendance. The developed predictive models will assist University stakeholders, managers and students in cost-effective and robust decision making that could facilitate improved student performance in programming courses in any prototype developing economy.