988368

Objectives: This study aimed to evaluate the psychometric properties of the SRT that was performed in 4-6 year-old Persian-speaking children. The SRT is a nonword repetition task that assesses phonological working memory and was designed by Shriberg in 2008.

Methods: The present research was a non-experimental study with a methodological design. The content validity of the task was evaluated by 15 speech language pathologists (SLP). The Lawshe coefficient was acceptable, and therefore without changing the number and phonological structure of the nonwords, the SRT was performed by 140 normal children in two groups (4-5 and 5-6 years old) that were selected randomly from 10 kindergartens of Tehran. To assess the test-retest reliability, the SRT was performed at one-week intervals. Finally, to determine the differential validity of the SRT, the task was performed in 30 children with speech sound disorders (SSD) that had been selected from speech and language clinics of Tehran. SPSS software version 21 was used to  determine the Pearson correlation coefficient, Cronbach's alpha coefficient and T-test.

Results: The CVR coefficient of the SRT was between 0.57 and 1. Cronbach's alpha coefficient of the total score was 0.83, and the Pearson correlation coefficient between successive runs was 0.87(P<0.001). There was a significant difference between the performances of two age groups (P=0.00). Also, the SRT score in SSD children was significantly lower than in normal children (P=0.00).

Discussion: The SRT appears to be a psychometrically  valid and reliable nonword repetition task for evaluating phonological working memory. The evidence of the differential validity of the SRT was approved in two ways: Differences between two age groups of normal children and differences between normal and SSD groups. The poor performance of children with SSD indicates that these patients have difficulty to store and retrieve phonological information in their working memory.