Liselere Geçiş Sisteminde İlçelerin Gelişmişlik Düzeyinin ve Cinsiyetin Değişen Çeldirici Fonksiyonuna Etkisi

Abstract

Evidence for test fairness is collected with the analysis of changing differential distractor functioning (DDF) and differential item functioning (DIF). Transition system to Turkish high school is an achievement test aimed at selecting students in transition from secondary school to secondary education. Multiple choice items are used for objective evaluation in large scale exams. Distractors are one of the important factors to be examined while determining the performance of multiple choice items in the measurement tool. With the DDF analysis, the possibility of choosing the wrong option versus the key option is determined for individuals in subgroups with equal abilities. While DIF focuses only on the key option; DDF focuses on all options and provides more indepth information. With DIF studies, it cannot be determined which distractor showed DDF. There is not enough evidence for DIF with DDF studies. Therefore, DDF analysis was performed independently from DIF analysis. According to this study, the students participated in a large scale tests conducted in Turkey aims to examine gender and development level of districts with confounding variables elections for DDF. For this purpose, 2018 LGS has been analyzed in terms of DDF. Multinomial logistic regression method was used for the DDF analysis. In the examination with 90 items and 270 distractors, DDF was observed in 10 distractors in the comparison made according to the gender variable and in 9 distractors in the comparison made according to the district variable. Distractors with a high probability of showing bias were interpreted with expert opinion.