EGITIM VE BILIM-EDUCATION AND SCIENCE, vol.47, no.210, pp.259-280, 2022 (SSCI)
The aim of this study was to examine Turkish course exam questions in terms of originality, page layout, item types, item writing criteria and cognitive level. The research was conducted with the document analysis method, and to determine the data source, the maximum variation sampling method, one of the purposive sampling methods, was used. The data source of the research consisted of exam questions prepared by Turkish teachers employed in secondary schools in a medium-sized city in the Black Sea region. Exam questions from each of the school levels determined on the basis of High School Entrance Exam achievements, from each district within the provincial borders, and from each type of school at the secondary school level were included in the research. Within the scope of the research, 747 questions in 51 exam papers were analysed. Descriptive analysis was carried out for the research data, and the obtained findings were summarised using frequencies and percentages. When the research findings were examined, it was concluded that the exam papers had similarity rates ranging from 11% to 91%, and that several exam papers and questions were taken from different internet sources as they were without making any changes. Other obtained findings were that more than half of the Turkish course exam papers did not have the page layout features that should be found in an exam paper, that the most frequently used item type in the exam papers was the multiple-choice item type, and that teachers did not pay attention to the item writing criteria. Another striking finding of the study was that about four-fifths of the questions examined were prepared for the understand level and one-fifth for the remember and apply level, and that while very few questions were prepared for the analyze and create levels, no questions were prepared for the evaluate level. Moreover, it wasconcluded that the majority of the questions examined were prepared for the middle cognitive level, while higher-order questions were included only at a rate of eight per thousand. The obtained results are discussed within the scope of the literature and a number of recommendations have been made.