Predicting Achievement with Artificial Neural Networks: The Case of Anadolu University Open Education System

Abstract

This study aims topredict the final exam scores and pass/fail rates of the students taking theBasic Information Technologies – 1 (BIL101U) course in 2014-2015 and 2015-2016academic years in the Open Education System of Anadolu University, throughArtificial Neural Networks (ANN). In this research, data about the demographics,educational background, BIL101U course mid-term, final and success scores of626,478 students was collected and purged. Data of 195,584 students, obtainedafter this process was analysed through Multilayer Perception (MLP) and RadialBasis Function (RBF) models. Sixteen different networks attained through thecombination of ANN parameters were used to predict the final exam scores andpass/fail rates of the students. As a result of the analyses, it was found outthat networks established through MLPs make more exact predictions. In theprediction of the final exam scores, it was determined that there is a lowlevel of correlation between the actual scores and predicted scores. In theanalyses for the prediction of pass/fail rates of the students, networksestablished through MLPs ensured more exact prediction results. Moreover, itwas determined that the variables as mid-term exam scores, university entrancescores and secondary school graduation year were of highest importance inexplaining the final exam scores and pass/fail rates of the students. It wasfound out that in the higher institutions serving for Open and DistanceLearning, pass/fail state of the students can be predicted through ANN underfavour of variables of students which have been found as most the importantpredictors.