Assessing phonological short-term memory in Azeri-Persian bilinguals: Psychometric properties of a non-word repetition test.

OBJECTIVE

This study aimed to develop and evaluate the psychometric properties of the Azeri Non-word repetition test (ANWRT) for Azeri-Persian bilingual children aged 7-12 years in Iran.

METHODS

In this cross-sectional study, 300 children (240 typically developing and 60 with developmental dyslexia (DD)) participated. The ANWRT consists of 20-words across four syllable lengths (1-4 syllables), constructed according to Azeri phonotactics. Face and content validity were confirmed by experts and child feedback. Construct validity was assessed by group comparison, criterion validity by correlation with forward digit span, and diagnostic accuracy using ROC analysis. Reliability was examined through test-retest and internal consistency.

RESULTS

Children with DD scored significantly lower than typically developing peers (p < .001). Criterion validity was supported by a significant correlation with forward digit span in the DD group (r = 0.287, p = .028). The test showed strong diagnostic accuracy (AUC 0.621-0.850) with optimal cutoffs (11.5 for Grades 1-2; 14.5 for Grades 3-6). Reliability was excellent (test-retest ICC = 0.93; Kuder-Richardson = 0.857; McDonald's ω = 0.837).

CONCLUSION

The ANWRT is a reliable and valid tool for assessing phonological short-term memory in Azeri-speaking children and effectively discriminates developmental dyslexia, offering a culturally appropriate instrument for bilingual populations.