Development and psychometric evaluation of the Persian Grammar-Phonology Screening Test: A preliminary study.

Abstract

This study aimed to develop a screening test for children suspected of developmental language disorders (DLD). The study was conducted in six steps. In the first two steps, we collected 34 sentences for the sentence repetition (SR) subtest, and 40 non-words for the non-word repetition (NWR) subtest, respectively. Experts determined the content validity in the third step and 19 sentences and 19 non-words were selected. In the fourth step, we draw simple pictures for the SR subtest. In the fifth step, two pilot studies were conducted on 30 and 35 Persian-speaking children aged 4.5-5.5 years, respectively. After calculating the discrimination index of the items, we removed items with the discrimination index less than 0.3. Nineteen sentences and 14 non-words remained in this step. In the sixth step, 92 typically developing (TD) children aged 4.5-5.5 and 10 children suspected of DLD were recruited. Then, we removed some items and evaluated the construct validity and reliability of the final form of the Persian Grammar-Phonology Screening Test (PGPST).The final form of the PGPST consisted of 16 sentences, and 11 non-words were developed. The correlation between the two subtests was 0.29. Children suspected of DLD in SR (p = .001) and NWR (p = .007) performed significantly worse than the TD children. The correlation between the two administrations of the test was 0.74 and 0.71, respectively, for SR and NWR. Cronbach's alpha was also 0.84 and 0.86 for the SR and NWR subtests, respectively. The PGPST is a quick and brief screening test with acceptable validity and reliability.