Detecting At-Risk Students in an Animal Science Module with Performance Data (Test 1 and 2) from Two Different Cohorts in a Rural-Based University

Abstract

Students' dropout which is often reflected as at-risk students remains a major problem in Higher Education worldwide not excluding South Africa. Numerous factors have been associated with student's dropout such as firstgeneration student pressure, choosing the wrong institution and course, social status at the institution, academic readiness, financial and family commitments and poor academic support. A couple of researchers have also established that early identification of at-risk students with mitigation strategies can decrease dropout. Test-1 (T1) has often been used for quick identification, but this study investigated the potential of adding Test 2 (T2) early to improve exam failure prediction. The study reflected on the role of T1 and T2 in predicting at-risk students (exam failure) in a small class (Applied Animal Nutrition module at a rural-based University). A total of 51 students belonging to cohort-1 (23) and cohort-2 (28) participated in the study for 14 weeks. A hybrid mode of lecture (half online and half face-to-face) was adopted after the relieved of COVID-19 pandemic. A double blinded study was used to avoid biasing the study. T1 and T2 predicted that 19.6% and 35.3% of students were at-risk of failing their exams, respectively. However, the exam failure rate showed that T1 and T2 were able to predict only 15.7% and 29.4%, respectively. T2 was 15.5% more accurate in predicting at-risk students than T1. It was also noticed that cohor-1 performed better than cohort-2 hence need for further investigation.