Title: Detected Item Bias in Ukrainian National Test Author: Viktoriia Dobriak Affiliation: SSA Group, Kharkiv, Ukraine Abstract: National testing in Ukraine, known as the External Independent Testing (EIT), was introduced for the first time in 2008. Since then, such a form of educational level testing has been used every year. Another test - the Test of General Educational Agency Competence (TGEAC) - was also successfully implemented during this period. In general, research was limited only to estimation of the tests validity. It cannot be denied, however, that it is important to carry out an in-depth study of these tests for the existence of an unfair advantage of one subpopulation over another one. So far, rather little attention has been paid to the problem of detected item bias when analyzing these tests. The subject of our study is the result of the primary school students' educational achievements analysis, which we obtained via new monitoring investigation. The purpose of the monitoring was to determine the skill level of students in reading and mathematical comprehension. We considered the results of two tests in mathematics and reading. The students were divided in two groups - a reference group and a focal group - according to such criteria as gender, school type (ordinary or specialized), level of training (free time, additional preparation), domicile and the financial condition of a student's family. The analysis in order to find the items with systematic errors in the test results was implemented with the R software environment including the difR and difNLR libraries. The results of the study are represented in the form of a web-application that was developed using the Shiny package.