Pediatric Arabic Closed-Set Word-Recognition Test: Development and Evaluation of Psychometric Characteristics.

Abstract

Speech audiometry materials are widely available in many different languages. However, there are no known standardized materials for the assessment of speech recognition in Arabic-speaking children. The aim of the study was to develop and validate phonetically balanced and psychometrically equivalent monosyllabic word recognition lists for children through a picture identification task. A prospective repeated-measure design was used. Monosyllabic words were chosen from children's storybooks and were evaluated for familiarity. The selected words were then divided into four phonetically balanced word lists. The final lists were evaluated for homogeneity and equivalency. Ten adults and 32 children with normal hearing sensitivity were recruited. Lists were presented to adult subjects in 5 dB increment from 0 to 60 dB hearing level. Individual data were then fitted using a sigmoid function from which the 50% threshold, slopes at the 50% points, and slopes at the 20 to 80% points were derived to determine list psychometric properties. Lists were next presented to children in two separate sessions to assess their equivalency, validity, and reliability. Data were subjected to a mixed design analysis of variance. No statistically significant difference was found among the word lists. This study provided an evidence that the monosyllabic word lists had comparable psychometric characteristics and reliability. This supports that the constructed speech corpus is a valid tool that can be used in assessing speech recognition in Arabic-speaking children.