| dc.description.abstract | Purpose: The long-term goal of this line of inquiry is to develop an ecologically valid and efficient measure of speech sound accuracy in connected speech that is suitable for use in clinical practice for children aged 3;6-4;11. In this study, we explored a novel sentence imitation task, the Story-Sentence Imitation Task for Speech (SSITS), as a measure of speech sound accuracy. We investigated whether the measure (1) leads children to repeat a sufficient proportion of target consonants, (2) demonstrates inter-rater, intra-rater, and test-retest reliability and scoring stability, and (3) demonstrates convergent validity with published measures of speech sound accuracy and intelligibility, and (4) whether a scoring training leads speech-language pathologists (SLPs) to follow the intended protocols for deriving the final score on the SSITS.
Methods: Eleven typically developing children with normal receptive language skills, between the ages of 36 and 59 months, participated. Speech sound production abilities freely varied so that participants’ speech production skills encompassed a wide range of accuracy. Children completed a battery of assessments including a single-word intelligibility measure, a word-level speech accuracy measure, and the SSITS. Nine children returned for a second study visit, in which they completed only the SSITS. The SSITS was presented as a story script accompanied by a wordless picture book, where the children repeated each utterance after the examiner. Target sounds were selected for scoring based on phoneme frequency in child conversational speech. Five SLPs scored all child videos at least once after completing a self-guided PowerPoint training on the SSITS.
Results: The SSITS demonstrated high feasibility, good inter-rater reliability, excellent intra-rater reliability and test-retest reliability, high scoring stability, and strong convergent validity with a measure of single-word speech sound accuracy and a measure of intelligibility.
Conclusion: This study adds to the evidence base surrounding the assessment of children’s speech sound production by demonstrating that a sentence imitation task presented as a narrative illustrated in a wordless picture book can have strong psychometric properties. | en_US |
