Predictor Analysis of Middle and Low Achieving Students in Cell Biology Online Course in Health Nutrition Undergraduate Program

Abstract

\n Background\n\nLack of direct contact with the students in online courses can result in lower lecturer’s awareness of their engagement and progress. Examination scores generally have the highest proportion in determining students’ grades. Predicting examination scores from the earliest point of the course may be useful in designing timely and appropriate interventions.\nAim\n\nTo analyze the predictors of midterm and final examination scores in an online cell biology course of health nutrition in the undergraduate program at the Faculty of Medicine, Public Health and Nursing Universitas Gadjah Mada, Yogyakarta Indonesia.\nMethods\n\nThe learning materials including texts, PowerPoints, videos and miniquizzes were uploaded in a course management system before conducting online meetings. In total, 10 different topics were delivered. Pretest and posttest, both not used for grading, were given at the first and last online meeting. Components of the final score were classroom assessments, group assignments, midterm and final examinations. Students (n\u2009=\u2009154) were divided into three groups based on their final examination scores, i.e., low, middle and high score. Each component of the final score was reported as mean\u2009±\u2009standard deviation and the mean difference between groups was analyzed. Linear regression analysis was performed to reveal the main predictors of the midterm and final exam scores. Two-step cluster analysis determined by the earliest-obtained scores was performed to identify low achieving students.\nResults\n\nStudents with higher final examination scores had significantly higher scores for pretest, posttest, pre and post-midterm class assessments, midterm exam and group assignments (p\u2009<\u20090.05). Premidterm class assessment was the main predictor of the midterm exam score. Midterm exam score was the strongest predictor of final exam score and clustering using midterm-premidterm scores identified 75% of low achieving students. To assist students with low and middle achievement, several modifications were considered such as providing longer and multiple access to the learning and formative test materials, facilitating a communication platform with fellow students and sending personal encouraging messages.\nConclusion\n\nScores obtained in various activities during the course potentially predict student grades. The activities should be optimized to improve students’ achievement especially for the less privileged students.