Abstract
The subject of the research of the article is statistical models and methods for assessing the quality of test items in English for applicants to the master's program. The study used the Rasch model for data analysis using Winsteps. When assessing the quality of the developed test items, statistical characteristics of the test items were obtained. We took into account such indicators as the validity and reliability of the test item, the difficulty of the item, the correlation coefficient, the statistics of agreement, the correspondence between the level of preparedness of the subjects and the difficulty of the test items. The agreement of the experimental data with the Rasch model was assessed using characteristic curves. Compatibility of test items was determined based on the Chi-square test, and the test results were compared with the expected values based on the Rasch model. Rasch IRT results provide very detailed information about the quality, validity, and reliability of test items. Based on the data obtained, it was revealed that the English language test corresponds to a high level of language proficiency and can be used as a component of the entrance exam to the master's program.The article is intended for a wide range of specialists in the field of education, as well as for those interested in the problems of higher education in our country. Keywords: the Rasch model, the validity and reliability of the test item, the difficulty of the item, the correlation coefficient.