Research suggests that bilingual children who speak one language at home and another in school are frequently misdiagnosed when a language delay is suspected. What exactly causes this delay? Is it insufficient exposure to the language of schooling? Or should other child-internal factors, such as specific language impairment (SLI), be considered?
To improve diagnosis, a newly developed test was scrutinized by researchers in the city of Utrecht, Netherlands. Known as a ‘non-word repetition task’ (NWRT), this test contains a list of non-existing words that children have to read aloud within two minutes. Because these words are non-existing, children cannot possibly rely on their own memory or recognition of words. This gives researchers insight into the real technical reading skills of the child and is therefore considered a reliable tool in diagnosing language impairments such as SLI.
One drawback of the NWRT for bilingual children is that the non-words are based on the language of schooling, which can be too limiting when testing children who speak another language in the home. Hence, for the newly developed test, researchers tried to design ‘quasi-universal’ non-words instead of language-specific ones. These non-existing words contain features of many languages and are designed to test the reading and pronunciation skills of speakers of different languages.
It is, however, very challenging to design a test that is independent of specific language. This study from Utrecht examined whether the researchers had indeed succeeded in doing so and thereby developing a more accurate tool for diagnosing bilinguals with specific language impairment.
They used a large group of 5-6 year olds, divided into smaller groups of monolinguals and bilinguals, including those displaying typical language development and those with SLI. The children were asked to complete both types of NWRT: the language-specific and ‘quasi-universal’ test. The study showed the success of developing quasi-universal non-words as this test was more accurate in diagnosing language impairment in bilingual children. While bilinguals’ performance in the two types of test was shown to be different, the monolingual children performed almost the same on both tests.
This newly developed test has only been scrutinized in the Dutch context and further testing across other languages is required. Yet the key take-home message is that this quasi-universal NWRT has great potential in diagnosing SLI as the source of linguistic delay in bilingual children.
Boerma, T., Chiat, S., Leseman, P., Timmermeister, M., Wijnen, F., & Blom, E. (2015). A Quasi-Universal Nonword Repetition Task as a Diagnostic Tool for Bilingual Children Learning Dutch as a Second Language. Journal of Speech, Language, and Hearing Research, 58, 1747-1760.