Hierarchical modeling of response accuracy and response times for reading skills

In the current study, it is aimed to estimate and interpret the person and item parameters related to the abilities and speed of students in reading skills and to evaluate the level of fit of the behaviors exhibited by them. The study group is comprised of 5232 students participating in the PISA 2015 application from 42 different countries. The data of the study were generated by using the data of students who answered items in the same order about reading skills in a form of the 36th booklet in the computer-based PISA 2015 application. The missing data in the data set, in which all the variables to be used for 9396 students who answered this form were brought together, were cleaned with the list-based data deletion technique, and the research data were provided by using the remaining 5232 student data. The first step in the analysis of the data was the selection of the variables that predict the levels of accuracy and time using genetic algorithms based on the regression technique. Variables that were found to predict the levels of accuracy and time with regression models analyzed with genetic algorithms were added to the hierarchical model as explanatory variables of the ability and speed parameters. In the second step, the data on response patterns, response times and explanatory variables were arranged to analyze the hierarchical model. In the third step, the levels of compliance of the behaviours reflecting the speed and abilities of the individuals were evaluated through person fit statistics. According to the results, a low negative correlation was found between speed and ability parameters. The number of students exhibiting aberrant behaviour regarding response patterns was 496 (9.48%), the number of students exhibiting aberrant behaviour regarding time patterns was 99 (1.87%), and the number of students exhibiting aberrant behaviour common to response and time patterns was 14 (.27%).