مرکز منطقه ای اطلاع رساني علوم و فناوري - Angoff (1-0), Nedelsky and Examination of Classification Accuracies of a Test by Determination Methods of Limit Values

Abstract :

This research includes the examination of classification accuracies of test successes of classteachership students by Angoff (1-0), Nedelsky standard determination methods, limit determination methods, ROC and interval forecast. In this research, different standard determination methods have been examined and the research is correlational in scanning model. In the research, a test consisting of 30 questions (selected coincidentally from Turkish and math questions in KPSS which was held in the years 2006/2012) was used as a data collection tool. At the end of this research, the highest cut-off scores for Turkish and mathematics tests were obtained by the Angoff (1-0) method, and it has been concluded that there exists a significant difference among the student percents that are deemed to be successful. As a result of the ROC analysis in order to determine the cut-off scores for Turkish and mathematics tests, it has been observed that the mathematics test is capable of classifying the successful/unsuccessful students accurately. It has shown that the cut-off scores determined according to interval estimation for mathematics and Turkish test conform to the cut-off scores determined in ROC analysis. As a result, Angoff (1-0) and Nedelsky for General Ability sub-tests and different cut-off scores for ROC and Interval Estimation Methods have been identified.