this study aimed to examine the psychometric properties of the Evans and Craig interview protocol in Persian-speaking preschoolers.
Materials and Methods: Language samples were collected from 340 Persian-speaking children (aged 5–6 years) attending preschools in Semnan, Iran. From these, a systematic random sample of 207 children (51 with language impairment, 156 without language impairment) was selected for analysis. The instruments included the Persian version of the Test of Language Development–3 and the 15-minute Evans and Craig interview protocol. The interview consisted of three main questions regarding family, school, and leisure activities, each followed by probe questions to elicit extended responses. Cross-cultural validity, criterion validity, and responsiveness of the protocol were examined based on Beaton et al.’s guidelines. Syntactic and semantic indices derived from the language samples, such as mean length of utterance (MLU), number of conjunctions, proportion of complex sentences, and type–token ratio, were calculated. Composite scores from the Test of Language Development–3 were also obtained. All data were analyzed in SPSS, and group differences were assessed using appropriate statistical tests at a significance level of 0.05.
Results: The mean scores of MLU, number of conjunctions, proportion of complex sentences, and type–token ratio in typically developing children (overall language ability score = 102.25) compared with children with language impairment (overall language ability score = 75) were as follows: 6.11 vs. 5.26 (p = 0.034), 13.37 vs. 7.53 (p = 0.001), 25.71 vs. 25.87 (p = 0.63), and 0.44 vs. 0.46 (p = 0.30), respectively. Except for type–token ratio, the other three indices showed significant correlations with overall language ability scores on the Test of Language Development–3 (p < 0.05). Regarding responsiveness, type–token ratio was the only index that revealed significant changes at the six-month follow-up (p = 0.004). Cut-off points were determined as follows: MLU < 4.5, number of conjunctions < 7, proportion of complex sentences < 14%, and type–token ratio outside the 0.45–0.65 range, all of which may indicate language impairment in 5–6-year-old children.
Conclusion: The findings indicate that the Evans and Craig interview protocol can be applied in 5–6-year-old Persian-speaking children. Given its cross-cultural adaptation and psychometric validation, speech and language pathologists may employ this valid and practical tool for screening and identifying language impairments in preschoolers.