Assessment of grammar comprehension: Adaptation of TROG for Serbian language

Abstract

In this study we present our adaptation and preliminary standardization of Test for Reception of Grammar TROG (Bishop, 1989) for Serbian language. TROG is a receptive test of grammatical structures, constructed primarily for an assessment of grammatical development and detection of children with Specific Language Impairment (SLI). Due to the lack of standardized tests for language development in our clinical community, TROG is selected for adaptation as a test which includes two components relevant for discrimination of children with language difficulties: a) measure of receptive abilities and b) distinguishing knowledge of grammar from semantic aspect of comprehension. Preliminary standardization was done on a sample of 335 participants between 4 and 7 years of age, divided into 8 age subsamples. Since dynamic of language change at early ages is faster, age samples covered range of 3 months at the ages 4;0-4;2, 4;3-4;5, 4;6-4;8, 4;9-4;11, and range of 6 months at the ages 5;0-5;5, 5;6-5;11, 6;00-6;05, 6;06-6;11. Analyses have revealed that the first version of Serbian TROG is discriminative for the differences between age samples, but discrimination is smaller then it was expected. The test discriminates three age-samples (4;0-4;8, 4;9-5;5, and 5;6-6;11). It is easier for the children older then 5 years, which causes statistical significance of discrimination to tilt within a narrow margin around 0,05. Reliability of the whole instrument is estimated very high - between 0,86 and 0,91, depending on the method of estimation. However, reliability estimated for particular blocks (grammatical structure) revealed that internal consistency of blocks is not homogeneous. This finding prevents reliable estimation of competence for particular structures, and makes difficult to define which contrast is understood by a child and which is not. Furthermore, internal inconsistency of blocks may also be additional source of low discrimination of test for children older then 5. Further improvement of test, balancing of items and inclusion of additional grammatical structures in order to get a wider range of age discrimination is required. This will be done in future research.

U ovom radu iznosimo opis adaptacije i rezultate probne standardizacije britanskog Testa razumevanja gramatike (TROG) za srpski jezik. TROG je receptivni test gramatičkih kontrasta, konstruisan za procenu gramatičkog razvoja i otkrivanje specifičnog jezičkog poremećaja (SLI) kod dece (Bishop, 1989). Test pokriva dve komponente važne za diskriminaciju dece sa teškoćama u govornom razvoju: a. meri receptivnu funkciju govora i b. meri poznavanje gramatičkih struktura. Probna standardizacija rađena je na uzorku od 335 ispitanika uzrasta između 4 i 7 godina, podeljenih u 8 uzrasnih segmenata. Analize su pokazale da je prva probna verzija testa diskriminativna za razlike među uzrastima, ali da je diskriminativnost "grublja" od očekivane. Pouzdanost testa računata na nivou celog instrumenta je visoka i kreće se između 0,86 i 0,91. U daljem radu planira se usavršavanje testa i ujednačavanje ajtema, kao i uvođenje složenijih jezičkih struktura kako bi se postigla diskriminativnost za veći raspon uzrasta.