Here’s a new article that came out this week in the Asha Leader that Lisa Bedore and I wrote. In this short piece we presented some of our BESOS data to demonstrate how screening in two languages can be combined to increase diagnostic accuracy.
So, how do you determine language impairment in a group for which there is no gold standard?
One of our projects (well, two of them and old one (an NIH contract to develop a test for 4 to 6 year old Spanish-English bilinguals) and a new one (an R-21 from the NIH to extend our test (the Bilingual English Spanish Assessment) to children ages 7 to 9) focus on test development. The best way to test a new test for the purpose of identification is to look at how well it separates the two (or more) groups in question. Here, we focus on determining language impairment and typical language development in bilinguals. In order to test the new test’s classification accuracy you have to classify the children based on a gold standard.