An Introspective Analysis of Context Cues Employed in a Cloze Test.

The diagnostic potential of a structured cloze test used in the reading measure of the New York State Regents Competency Testing Program was examined. Two passages at the third and fifth grade levels were administered to 12 low achieving fifth grade students, first in a standardized cloze format requiring the students to generate responses and next in a structured cloze format requiring the recognition of the correct response. Results indicated that both recognition tasks produced higher mean scores than the generation tasks. The analysis of types of context cues indicated that repetition resulted in the highest percentage of correct responses with the recognition task. The third grade passage contained fewer transition words to signal relationship, which may partially explain the comparable performances even though the passages represented different readability levels. (MKM)