Empirical Quality of Final Exam Questions in a Learning Management System-based Course

Abstract

This study aims to analyze the empirical quality of the question items in the final examination of the graphic media development course. The research method used was a descriptive method with a quantitative approach. The data collection technique was the analysis of the answer sheets of student work on the final examinations. The population was made up of students with a total of 90 students. The data were analyzed using the Microsoft Excel 2010 application to determine the level of difficulty, discriminating power, distractor effectiveness, validity, and reliability. The empirical quality research results from a total of 25 multiple-choice items for the item difficulty index show that there was 1 (4%) item in the too-difficult category, 7 (28%) items in the moderate category, and 17 (68%) items in the too easy category. The item discrimination index is 7 (28%) items in the bad category, 13 (52%) items in the moderate category, and 5 (20%) items in the good category. Then, for the effectiveness of the distractor, there were 15 (60%) effective items and 10 (40%) ineffective items. Meanwhile, for the validity of the items, there were 24 (96%) items in the valid category and only 1 (4%) item in the invalid category. The reliability of the test using Cronbach's alpha was in the very high category, namely 0.812. Thus, it can be concluded that the semester final examination for graphic media development courses is appropriate to be used as a tool to measure student competency.