Interpreting Nonverbatim Cloze Responses to Evaluate Program Success and Diagnose Student Needs for Reading Instruction

Student nonverbatim cloze responses, appraised for their semantic and syntactic acceptability, were used to evaluate an elementary school's implementation of the whole language approach to reading instruction. Ninety-six hearing-impaired students, ranging in age from 7 to 15 years, completed cloze versions of stories written at the students' respective reading levels. In addition to crediting students for verbatim responses, judges rated nonverbatim responses for their semantic and/or syntactic acceptability. Acceptability of student performance on the full passage was established by the Borderline Group technique. Comparison of the cloze results to both probe and free response measures of comprehension indicated that the cloze procedure produced valid discrimination. Reliability was not measured directly, but evidence of stable measurement was also apparent. The diagnostic value of individual student scores is discussed and corresponding instructional strategies explained.