OBJECTIVE: To determine the reliability and validity of a computer-mediated, 50-word intelligibility test designed to be a global measure of severity of speech disability in children with repaired cleft lip and palate. DESIGN: A prospective between-group design was used with convenience sampling of patients from a university craniofacial center. PARTICIPANTS: A total of 38 children between the ages of 4 and 9 years participated. Of these, 22 had repaired cleft lip and palate; whereas, 16 had no clefts. Twenty adults served as listeners. MAIN OUTCOME MEASURE(S): Speech intelligibility scores were calculated for repeated administrations of a single-word test based upon the number of correct orthographically transcribed words by four groups of five listeners per child. Measures of parallel forms, interlistener, and intralistener reliabilities were estimated; measures of construct validity were also determined. RESULTS: All measures of reliability were adequate. Parallel forms reliability of the test based upon mean scores from five listeners per child was high (r = .97). Thirty-seven of 38 children had differences between forms of 11 percentage points or less. Construct validity of the test was shown by (1) significantly lower speech intelligibility scores for children with cleft lip and palate than controls and (2) a moderately high correlation (r = .79) between intelligibility scores and percent consonants correct for all children. CONCLUSIONS: A computerized, single-word intelligibility test was described that appears to be a reliable and valid measure of global speech deficits in children with cleft lip and palate. Additional development of the test may further facilitate standardized assessment of children with cleft lip and palate.
OBJECTIVE: To determine the reliability and validity of a computer-mediated, 50-word intelligibility test designed to be a global measure of severity of speech disability in children with repaired cleft lip and palate. DESIGN: A prospective between-group design was used with convenience sampling of patients from a university craniofacial center. PARTICIPANTS: A total of 38 children between the ages of 4 and 9 years participated. Of these, 22 had repaired cleft lip and palate; whereas, 16 had no clefts. Twenty adults served as listeners. MAIN OUTCOME MEASURE(S): Speech intelligibility scores were calculated for repeated administrations of a single-word test based upon the number of correct orthographically transcribed words by four groups of five listeners per child. Measures of parallel forms, interlistener, and intralistener reliabilities were estimated; measures of construct validity were also determined. RESULTS: All measures of reliability were adequate. Parallel forms reliability of the test based upon mean scores from five listeners per child was high (r = .97). Thirty-seven of 38 children had differences between forms of 11 percentage points or less. Construct validity of the test was shown by (1) significantly lower speech intelligibility scores for children with cleft lip and palate than controls and (2) a moderately high correlation (r = .79) between intelligibility scores and percent consonants correct for all children. CONCLUSIONS: A computerized, single-word intelligibility test was described that appears to be a reliable and valid measure of global speech deficits in children with cleft lip and palate. Additional development of the test may further facilitate standardized assessment of children with cleft lip and palate.