PREDICTION OF STUDENTS SUCCESS FOR SMALL UNIVERSITIES USING MULTIVARIATE APPROACH

One of the primary goals of universities is to increase the success of students, while maintaining the quality of courses and consequently the quality of their graduates. To achieve this goal various educative tools have been proposed. These include improving motivation of students, using modern e-learning tools or transforming teaching and learning into a game. Another tool supporting this effort could be a system providing accurate information about students’ progresses within their studies as well as prediction of their future results. This prediction can be useful for students as a guide in their studies and for lecturers as a tool to identify weak as well as strong students.

The prediction of results for course is often handled via data oriented methods within universities having large number of students. In case of smalled universities and colleges having limited data the choice of the method is rather complicated. We are presenting results of specific method for prediction of students success for modern type of study program strongly oriented on information technology combined with economic field of study. This model considers small number of students and makes use of their various characteristics like simply determined year of study or more complex variable describing previous study results. Moreover, some of those may be of a quantitative character like student’s field of study.