Abstract:

Multiple-choice tests are one of the widely used objective measurement tools in the assessment of learners in open and distance learning. The most important reason for this is that multiple-choice tests are practical in terms of application and evaluation. However, having the guessing effect is one of the most important criticisms directed at multiple-choice tests. Several methods have been used to eliminate the effect of such guessing. One of these methods is penalizing the test-taker by reducing the number of correct answers based on the number of incorrect answers, which urges test-takers to skip the questions whose answers they do not know. This study analyzed students’ behavior of leaving multiple-choice questions unanswered in the Fall semester midterm and end-of-semester exams for 6 courses (3 verbal and 3 numerical) conducted face-to-face at Anadolu University Open education system in the 2019-2020 academic year. Based on item analysis data, the question-skipping status of students who scored high (upper group) and low (lower group) was examined by the variables of question difficulty level, exam period, and course type. The Spearman Rank Correlation and independent sample t-test were used to analyze the data in the research conducted in the correlational survey model. When appropriate, the data were interpreted using proportional graphs. The results showed that the rates of leaving questions unanswered were higher in the end-of-term exam than in the midterm exam, in numerical courses than in verbal courses, and the proportion of skipping questions increased in line with the difficulty level.

Keywords: