Sekizinci Sınıf Öğrencilerinin TIMSS 2015 Matematik Başarılarının Bilişsel Tanı Modeli ile İncelenmesi

Abstract

This research was conducted with the aim of determining the learning deficits and areas that are weak and strong in the mathematics of the eighth grade students. In this context, it has also been examined whether there is an effect of item types on students' response. For this purpose, analysis studies were carried out with the DINA model from the Cognitive Diagnostic Models (CDM). In the analysis process, two booklets were applied TIMSS 2015 eighth grade mathematics evaluation and student data obtained from 69 items and Q matrix showing the relation with learning areas of the items were used. Cognitive attributes representing learning areas were determined by three field experts by examining the eighth grade mathematics curriculum and TIMSS 2015 assessment frameworks. After this study, Q matrices were created by associating each item with 14 properties determined by experts. Findings from the data indicated that the g (guess) and s (slip) parameters were low and the δ (item discriminant) parameter was high. When the parameters values of the items were examined, it had seen that the averages of g parameter were 0.24 and 0.19, the averages of s parameter were 0.21 and 0.20. Mean values of δ for the tests were determined to be 0.55 and 0.60. When the conclusions regarding the frequency of observations of attributes were examined, it was determined that about half of the students who responded to the tests did not have the specified attributes. When it comes to learning areas, it comes from the "fractions and decimal representations", which are at the top of the attribute that students had at least in the learning field, and the "algebraic expressions and operations" attribute in the area of learning algebra. It has been determined that the weakest attribute of the students in geometry learning is the "rotation geometry". When the analysis results of the item type were examined, it was seen that the s parameters obtained from the open ended items were higher. It was observed that ix this result was slightly increased especially in the partial scoring items In the case of multiple choice items, the parameter g is determined to be higher. When the findings of multiple choice and open ended items that measure the same attribute were examined, students with an attribute in a multiple choice item had a very low probability of false answers, whereas students with the same attribute in the open ended item had a high probability of false answers