Examining the validity and reliability of the proficiency exams applied in teaching Turkish as a foreign language (Bozok TÖMER example)

Course completion exams applied in teaching Turkish to foreigners are exams for students to move from their current level to the next level. In these exams, it is measured whether the learners have acquired the achievements and skills of the level they are in. Each institution has success scores that they set for the course completion exams. The learners who get the determined score pass to the next level. The learners who cannot get the determined score repeat the same level. Students taking valid and reliable exams depend on the tests prepared in these exams. The aim of this research is to describe the appearance of the course completion exams applied in Yozgat Bozok University TÖMER in terms of validity and reliability. The case study method, which is one of the qualitative research methods, was used in the research. The study object of the research is the reading comprehension tests of the course completion exams for the 2022-2023 academic year, applied at TÖMER at Yozgat Bozok University. Content analysis, descriptive analysis, test and item analysis were applied to the research data. In the research, "How is the view of the course completion exams applied in teaching Turkish as a foreign language in terms of validity and reliability?" The answer to the problem has been sought. In order to describe the validity, the compatibility of the items in the reading comprehension tests with the achievements, their distribution according to the Revised Bloom's Taxonomy and the principles of item writing were examined. In order to describe the reliability, the difficulty, discrimination, difficulty and reliability of the tests were examined. The findings obtained as a result of the research are as follows: In the validity analyzes of the reading comprehension test items in the course completion exams applied at Yozgat Bozok University TÖMER; It has been determined that it is largely compatible with the achievements, it is largely suitable for the lower level according to the Revised Bloom Taxonomy, and there are a total of 42 outliers depending on the writing principles. However, the items in a test must be error-free, 100% compatible with the achievements, and include metacognitive steps. Only in this way, students will be given exams with high validity. In the reliability analyzes, on the other hand, the A1 level reading comprehension test was easily difficult, discriminating at a good level, and was reliable as a whole; A2 level reading comprehension test is moderately difficult, well discriminating, and reliable as a whole; The B1 level reading comprehension test is moderately difficult, discriminating at a good level, and as a whole, it is reliable at an acceptable level; The B2 level reading comprehension test is of easy difficulty, moderately distinctive, and its reliability as a whole is not acceptable; It was determined that the C1 level reading comprehension test was easily difficult, discriminating at a good level, and reliable as a whole. However, a test must be of medium difficulty, well discriminating, and reliable as a whole. Only in this way will more reliable exams be applied to learners.