A Study of Six Cloze Formats for Junior High Schools in Taiwan and Its Pedagogical Implications

• Main Authors: Chang Yu-lun, 張瑜倫
• Other Authors: Huang, Tzyh-lai
• Format: Others
• Language: zh-TW
• Published: 1998
• Online Access: http://ndltd.ncl.edu.tw/handle/05939188059742642879

碩士 === 國立師範大學 === 英語學系 === 86 === The purpose of the present study is twofold. First, it aims to compare the capacity of six different cloze formats as a measuring instrument for both the readability of witten materials and English proficiency of students,particulary at the junior high school level in Taiwan, thus providing a practical guide for classroom teachers to select the most appropriate format according to different testing purposes. Secondly, it is designed to examine the effects of different deletion types on students' performance regarding language proficiency, text difficulty and mode of response. By such an analysis, students' weaknesses in answering certain deletion types under certain circumstances are uncovered. The reults will have positive implications for English teaching in junior high school. Twelve cloze tests employed in the study were developed from six different formats over two passages of differing difficulty levels. A total of 643 third graders at Jie- shou Junior High School in Taipei that took part in the study were randomly assigned into twelve groups. Each cloze group was administered one version of the cloze tests. In addition, words deleted in each cloze test were identifiedand categorized in advance to allow the examination of the possible effects of different deletion types on students' performance. The main findings of thestudy are as follows: 1. The six formats of cloze tests are capable of rankingpassages of differing levels of difficulty in the same order as the two standard readability formulas. 2. Of the six formats, the rational cloze, C-test and FC-test can further discriminate passages of differing levels of difficutly significantly. 3. While the six formats of cloze tests are comparable measures of English proficiency in terms of validity, reliability and discriminatory power among students, they creat tests that are distinct inlevels of difficulty. 4. For the fixed-ratio and rational cloze tests, the acceptable-word scoring procedure yields a more accurate assessment of students' EFL ability. 5. Different deletion types affect students' performance to a different extent. Moreover, subjects' proficiency level, textdifficulty and mode of response are crucial in determining the actual effect.Based on the findings, implications for text selection, cloze in English testing, cohesion in junior high school English teaching and the importance of vocabulary teaching are further discussed.