Diagnostic accuracy of language sample measures with Persian-speaking preschool children

Abstract

This study examined the diagnostic accuracy of selected language sample measures (LSMs) with Persian-speaking children. A pre-accuracy study followed by phase I and II studies are reported. Twenty-four Persian-speaking children, aged 42 to 54 months, with primary language impairment (PLI) were compared to 27 age-matched children without PLI on a set of measures derived from play-based, conversational language samples. Results showed that correlations between age and LSMs were not statistically significant in either group of children. However, a majority of LSMs differentiated children with and without PLI at the group level (phase I), while three of the measures exhibited good diagnostic accuracy at the level of the individual (phase II). We conclude that general LSMs are promising for distinguishing between children with and without PLI. Persian-specific measures are mainly helpful in identifying children without language impairment while their ability to identify children with PLI is poor.

• Diagnostic accuracy
• language sample analysis
• Persian
• pre-school children
• primary language impairment

Acknowledgements

We thank the children and their parents who took part in this study and Kowsar Rehabilitation Centre, managed by Shahryar Shariat and Tila Shokrani, for providing the setting for data collection. Thanks also to Ann Nockerts and Jon Miller for modifying the SALT computer program in order to accommodate the morphological structure of Persian, and Mahboubeh Nakhshab and Leila Ghasisin for their help with handling data. We also thank Paul Fletcher and Ghada Khattab for commenting on the first author’s PhD thesis on which this study based, and Nick Riches for commenting on an earlier draft of the article.

Declaration of interest

The authors have no conflict of interest to declare.