Abstract
Considerable evidence suggests that cloze techniques can create tests which measure aspects of students' second language competence. However, it remains unclear how variations in the cloze procedure affect measurement. This study compared results obtained from cloze passages constructed from the same text using four different procedures: fixed-ratio, rational, (rational) multiple choice, and C-test. The four procedures produced tests similar in reliabilities but distinct in levels of difficulty and patterns of correlations with other tests. These results are discussed in view of theoretically-based expecta tions for convergent and discriminate relationships of the four cloze tests with other tests.
Get full access to this article
View all access options for this article.
