Cloze procedure as a proficiency test of Chinese as a second language
Date
1991
Authors
Xie, Wang
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The purpose of this study is to explore the usefulness of cloze procedure as an instrument to measure the proficiency of students learning Chinese as a second language.
Two aspects of cloze format were tested: (a) deletion method, both every fifth word deletion and random deletion and (b) textual material, both familiar and unfamiliar material.
A class of third-year university students, who were learning Chinese as a second language, provided the data for this study. There were a total of four cloze tests, and the English vocabulary sub-test from the Nelson-Denny Reading Test administered in the study.
The Kuder-Richardson 20 formula was applied to calculate the test reliability. The validity of the tests was determined by correlating each of the cloze tests with the students' final exam mark and final grade.
It was found that, regardless of the deletion method, cloze tests using familiar material possessed high reliability coefficients (.94 and .85). Also, cloze tests using familiar materials yielded satisfactory validities (.89, .83, 87, and .78).
There was no significant correlation between the English vocabulary test and the students' Chinese proficiency. A questionnaire survey also revealed that the students involved in this study had a clear understanding of, and a generally positive attitude towards the cloze tests. The results of the study indicated that cloze tests, using familiar materials, were valid and reliable predictors of students' proficiency in Chinese as a second language.