Frontline Learning Research vol:2013 issue:1 pages:42-71
Many studies have explored the contribution of various different factors, and from diverse theoretical perspectives, to the explanation of academic performance. Many of these factors have been identified as having important implications not only for the study of learning processes, but also as tools for improving curriculum designs, tutorial systems, and students’ academic outcomes. The prediction of academic performance has been carried out mainly with traditional methodological approaches. However, these traditional approaches have not reached consistent accurate predictions or classifications when compared with artificial intelligence computing methods. This paper explores a relatively new methodological approach for the field of learning and education, using both cognitive and non-cognitive measures of students, as well as background information, in order to design predictive models using artificial neural networks (ANN). These ANN can identify those predictors that could best explain different levels of expected academic performance, as well as making accurate classifications of the expected level of performance for each student. A total sample of approximately 800 entering university students of both genders, ages ranging between 18 and 25 was used. Three neural networks models were developed: one to identify the lowest 33% levels of expected performance, one to identify the highest 33%, and a third ANN to classify all three levels of performance (low – middle and high 33%), respectively, in terms of their estimated future general academic performance. Two of the models (identifying the top 33% and the low 33% groups) were able to reach 100% correct identification of all students in each of the two groups, using the corresponding ANN. The third model reached a precision of 87%. These ANN models showed interesting differences in the pattern of relative importance of the predictive weights amongst the variables in the predictive models. Results also provide a better understanding of the variables with the greatest impact on individual learning processes, and of those factors that explain these processes for different academic levels.