A Note on Cloze as a Measure of Comprehension

Abstract
To test the hypothesis that cloze measures local redundancy, rather than comprehension of major ideas, 180 seventh-grade boys and girls were tested by means of either cloze tests or comprehension questions on three kinds of reading material: original passage and two versions reduced by. 30% and 50%, respectively, by deletion of function words. Reduction of redundancy had a minor effect on comprehension as measured by questions, but cloze test scores on reduced passages were 22% and 25%, respectively, of scores on original material and were thought to be supportive of the hypothesis.