Farklı ortak değişkenlerle test eşitlemenin ortak maddeli test eşitlemeyle karşılaştırılması

---

Thesis Type: Doctorate

Institution Of The Thesis: Gazi University, Eğitim Bilimleri Enstitüsü, Eğitim Bilimleri Enstitüsü, Turkey

Approval Date: 2023

Thesis Language: Turkish

Student: ELİF SEZER BAŞARAN

Supervisor: Mehtap Çakan

Open Archive Collection: AVESIS Open Access Collection

Abstract:

This study aimed to compare the equating errors in the covariate design (NEC), in which various covariates are used in equating TIMSS 2019 eighth grade mathematics subtests with Kernel equating methods (linear and equipercentile), from the common test (NEAT) and equivalent groups (EG) designs. The equating results in NEC were examined by using covariates related to mathematics achievement in three different groups and comparing with the equating results in other designs (NEAT and EG). Different covariates representing student and school characteristics were used such as: (i) science achievement, (ii) home educational resources, (iii) students' confidence in mathematics, (iv) students' value of mathematics, (v) students' enjoyment of learning mathematics, (vi) instructional clarity in mathematics lessons, (vii) disorderly behavior during mathematics lessons, (viii) student bullying, (ix) sense of school belonging, (x) school's emphasis on academic achievement, (xi) school discipline, (xii) sex of students, and (xiii) socioeconomic level. The study's sample consisted of 10100 students who answered the 1st, 2nd, 11th, 12th, 13th and 14th booklets of the mathematics test and the covariates in e-TIMSS. The study concluded that the comparison results of NEC designs with NEAT and EG were similar in Kernel linear and Kernel equipercentile equatings. The design that was the closest to the NEAT (post-stratification and chained equating methods) and the most different from the EG was the NEC design, in which science achievement was highly correlated with mathematics in all groups. However, when the variables which had weak or very weak correlations with mathematics achievement (such as school's emphasis on academic achievement, socioeconomic level, sex, bullying, instructional clarity) were used as covariates in NEC, the equated scores were very similar to the EG. The closeness of NEC designs in which the students' affective characteristics (confidence in mathematics, like learning mathematics, value mathematics) and home educational resources were used as covariates to NEAT and EG changed according to the groups. When the common item number ratio and reliability of the tests were different, the order of the NEC designs used the affective characteristics and home educational resources as the covariates, varied according to the student groups; however, the order of the other NEC designs was similar in the examined groups.