Investigation of two language tests' (tedi̇l and todi̇l) differentiation of children with typical and atypical language development

The aim of this study is to define a data based cut-off criteria for two standardized, reliable and valid language development tests in Turkish (TEDİL: Test of Early Language Development-Third Edition and TODİL: Test of Language Development-Primary-Fourth Edition) and to investigate the tests' identification accuracy of typically developing children and children with language impairment. In the study, a total of 2683 children's data were used for both TEDİL (n=1431) and TODİL (n=1252), which is quite representative of the population. According to the analysis, the data-based cut-off point was found -1.00 sd or 85 quotient for both TEDİL and TODİL's subtests and spoken language score, and for the following analysis the mentioned data-based criteria points have been used. On the other hand, due to the limitations of the sensitivity and specificity, an alternative technique was recommended to find out positive and negative likelihood ratios. In other words, in diagnostic accuracy studies, not only high sensitivity and specificity ratios but also a minimum level of positive and negative likelihood ratios and minimum difference between sensitivity and specificity is required. In this study, these issues are also considered. According to the results, the tests' identification degree of normal developing children and children with language problems was found adequate. Namely, TEDİL and TODİL have the potential for discriminating the subgroups of primary (language delay and potential SLI) and secondary language problems (as intellectual disabilities and autism spectrum disorders). However, they have been found inadequate in the identification of typically developing bilingual children whose first language is not Turkish.