Akademik Veri Yönetim Sistemi
EGITIM ARASTIRMALARI-EURASIAN JOURNAL OF EDUCATIONAL RESEARCH, cilt.8, sa.31, ss.115-133, 2008 (SSCI, Scopus, TRDizin)
Problem Statement: In this study, different doze procedures were investigated in Turkish. Six different cloze test formats, including regular cloze, regular cloze with initial letter-cue, dashed cloze, dashed cloze with initial-letter cue, matching format, and maze format were used to determine how these different cloze formats measure students' reading levels. Purpose of Study: The main purpose of the study is to see how six different cloze formats measure students' reading levels on Turkish texts at medium difficulty level. The following four questions are investigated in this study: 1. What are the reliability levels of the different cloze test formats? 2. What are the students' reading levels on these formats? 3. Are there any differences among the students' scores on doze test formats? 4. What are the students' opinions regarding the useful strategies to find the answers and the perceived difficulties of the cloze test formats? Methods: Sixth grade students (N=288) selected from four different schools in the district of Yenimahalle participated in the research. Passages were selected from sixth grade textbooks. To analyze the data, percentage, frequency, mean of correct responses, and K-R 20 scores were calculated, and students' scores on different cloze formats were compared using one-way ANOVA test. Results: The results revealed that:, (a) all cloze test formats had high reliability levels, (b) students had the highest scores from the MC format, and the lowest scores from RC and DC formats, (c) omission was the most frequently seen error type, (d) significant differences were found among the cloze test formats, and e) inactivated background knowledge, mismatching 'words and blanks, and finding the correct words among distractors were seen as most experienced difficulties Conclusions and Recommendations: The results of the study showed that all cloze test formats can be used in Turkish language to determine readability levels of texts and reading levels of students. It is recommended that before selecting a cloze format, evaluators should know about the characteristics of cloze test formats such as difficulties of the test, ease of construction, and placement and interpretation of scores, so that they can select the best cloze test format and make an appropriate decision on texts and reading levels of students.