Examination of the Extreme Response Style of Students using IRTree: The Case of TIMSS 2015

Abstract

In the literature, response style is one of the factors causing an achievement-attitude paradox and threatens the validity of the results obtained from studies. In this regard, the aim of this study is two-fold. Firstly, it attempts to determine which item response tree (IRTree) models based on the generalized linear mixed model (GLMM) approach (random intercept, random intercept with the fixed effect of extreme response and random intercept-slope model) best fit the Trends in International Mathematics and Science Study (TIMSS) 2015 data. Secondly, it purports to explore how the extreme response style affects students’ attitudes toward mathematics of students. This study is both basic research and descriptive research in terms of seeking for answers for two different research questions. For the sample of this research, 15 countries were randomly selected among countries participated in TIMSS 2015. The students’ responses to items measuring attitude in the student questionnaire were analyzed with the packages “lme4” and “irtrees” in R software. When the model fit indices were evaluated, the random intercept-slope model was found to be the best fit to the data. According to this model, the extreme response style explains a significant amount of variances in the students’ attitude toward mathematics. Additionally, students with a negative attitude toward mathematics were found to have an extreme response style. It was concluded that an extreme response style had an effect on students’ attitude.