Investigation of Classification Validity in Terms of Various Variables in TIMSS 2019 International Benchmarks of Students

Authors : Esra Oyar

Pages : 57-76

Doi:10.33709/ictimaiyat.1515877

View : 23 | Download : 7

Publication Date : 2024-10-31

Article Type : Research

Abstract :In this study, it was aimed to determine the variables affecting students\' proficiency classification by using the data of 4th-grade students participating in the TIMSS 2019 application in the fields of mathematics and science. For this purpose, it was tried to provide evidence for classification validity with the variables of school belonging, bullying, home resources for learning, self-efficacy for computer use, disorderly behavior in Math lessons, like learning Math/Science, confident in Math/Science and instructional clarity in Math/Science lessons. The study was conducted with a correlational design. The sample of the study consisted of 3887 students in both lessons, which remained as a result of the missing data deletion and assignment processes from 4028 students who originally participated in the application. Logistic regression and discriminant analysis were used to analyze the data. As a result of the study, it was determined that 41.6% for Mathematics and 43% for Science in logistic regression analysis and 42.5% for Mathematics and 45% for Science in discriminant analysis were correctly classified through independent variables. The results obtained from the study were discussed in the light of the literature and recommendations for both researchers and practitioners were presented.
Keywords : lojistik regresyon, ayırma analizi, TIMSS 2019, Matematik ve Fen Bilimleri, yeterlik sınıflamaları