Cloze Technique and Multiple Choice Technique: Reliability and Validity

Publisher Website
Google Scholar
Abstract
Two multiple choice forms of two cloze tests were developed based on the most frequent responses given to the cloze forms by college students. These forms were administered to large samples of undergraduate students and were scored and item analyzed with original keys, empirical keys, and empirical keys with option weighting. Correlations were calculated between different scales and Scholastic Aptitude Test Scales, Co-operative English Test scales and High School Deciles for the samples. These validity indices for the scales and tests were compared. The results indicated a trend for the empirical key scales to be more reliable and valid than the other scales across different forms and samples. The multiple choice forms, although ostensibly less reliable, appeared to be more valid than the cloze forms. Option weighting did not increase the validity of the scales, and as such, was considered not useful with cloze tests.