Development of Student's Dropout Early Warning System Using Analytical Hierarchy Process

Abstract

As a higher education institution, Politeknik Statistika STIS also faces the same problems as universities in general, those are student failing to compare that year courses thus have to repeat those courses or student dropping out. To overcome this problem, this research proposes a Dropout Early Warning System (DEWS) that can provide early warnings for dropouts and repeat a class. With this system, it is hoped that it can help institutions to identify students who have the potential to drop out or repeat a class. The purpose of making this system is to help academic supervisors and decision makers from Polstat STIS in knowing the potential for student. The potential for students to drop out and repeat a class is measured by a potential score obtained from the results of an assessment of 5 criteria consisting of GPA scores, gender, economic factors, violation points, and record of repeating class. Prediction results are presented in three categories consisting of low potential, medium potential, and high potential which are calculated from the results of weighting calculations using the Analytical Hierarchy Process (AHP). The system is tested and verified using Black Box test and the evaluation of the calculation method using confusion matrix. Based on the test results, the functions that exist in the system can function properly and can supply the needs.