Bilingual Assessment and Language Dominance: We need to MOVE ON

A long time ago (about 25 to 30 years ago) I learned that bilingual children should be tested in their dominant or home language. The prevailing view then was that if you tested in the weaker language you wouldn’t be letting the child demonstrate what they knew. I think that this part is true. The other part of this perspective is that there wouldn’t be anything in the weaker language that wouldn’t be represented in the stronger language. I don’t believe that this part is true. It’s the 21st century… we know better. 

Or at least we should. It’s frustrating to me that the idea of testing only in the dominant language is still around since we’ve known better for at least 20 years. In the last 10 years especially there is more and more and MORE information about how bilingual children’s language knowledge is distributed across two languages. So, why are the same old ideas around? Why are these ideas continually being retaught? Why haven’t school district policies kept up? (can you sense the frustration?)

What do we know?

  • Children’s vocabulary knowledge is distributed. If you test in only one language (even the “dominant” language, you may well underestimate what children know. We’ve know this since Umbel, Pearson, Fernández,  & Oller published a paper called Measuring bilingual children’s receptive vocabularies in 1992 (yes 23 years ago!). There’s been other work on this since then and the findings are similar for school age children and adults. The degree of overlap between two languages may vary depending on mode (receptive vs. expressive) and task (naming vs. generation). But the point is that people may know words in the non-dominant language that they don’t know or can’t recall in their dominant language.
  • Children need to learn the sounds of both their languages. Sounds that overlap are more likely learned sooner, and those that do not may be learned later (relative to monolingual norms). We know that bilinguals keep their systems separate and that there is some influence not only of L1 on L2, but also of L2 on L1. Some of this we’ve known for 10 years. And there have been many many more studies detailing mutual influences and variability of children’s sound systems across two languages. The point here is that you really need to test both languages testing only one may lead you to the wrong conclusions.
  • We’ve seen evidence of “mixed” dominance now for a while. One study looked at French-English bilinguals. If you look at their data carefully, you can see that there are a number of children who were high in vocabulary in 1 language and high in grammar in the other language. So, their dominant language was different depending on the domain. We also see this in Spanish-English bilinguals. This is way way more common that not. In fact, for a talk we did recently at TSHA, Lisa Bedore and I looked at Spanish dominant, English dominant and balanced bilingual children’s performance on the BESA. We compared phonology, morphosyntax, and semantics scores in each language. Among the balanced bilingual children only 6 of 103 actually demonstrated similar performance in both languages in all three domains– so 97 of 103 had mixed dominance. For the dominant English speakers 29 of 191 children performed better in Spanish on at least one subtest. And for the Spanish dominant speakers, 59 out of 229 children performed better in English on at least one subtest.

So, stop it already and move on. Determining a child’s “dominance” and testing only in that language was a good place to start. But, we’ve learned a lot more in the last 25 to 30 years and it’s time to let go of the old way and do what the evidence shows we should do.

, , , , , ,

  1. #1 by Carol on March 27, 2015 - 6:20 am

    If it’s any comfort, I got my degree from the University of South Carolina in 2012 and we had been taught that we should assess bilingual children in both languages using a bilingual slp if at all possible, and if not possible, then with the assistance of an interpreter who had been trained in how to administer the assessments being used. And in the school district I work for, we follow that policy.

  2. #2 by Michael Lopez on March 27, 2015 - 7:30 am

    Liz, great posting! I’ve faced the same challenges with trying to push people to think more critically about not only the importance of assessing the development of each language, but also being reflective about the “how” and “why” questions. As you know all to well, there may be slightly different approaches taken depending upon what specific question you are trying to answer (and the quality of the assessment measures you have to do so). Unfortunately, there are no simple answers…. Fortunately, your own measurement work has helped to add to the growing movement to develop a strong array of measurement options for people to choose from. Nice work!!!

  3. #3 by Elizabeth D. Peña on March 27, 2015 - 9:52 am

    Thanks Carol! I appreciate that.
    And Michael– good point. Certainly what is measured, how, and how it is put together should be driven by the question. Conceptual scoring is one example. Its use would not likely be informative when asking questions about second language growth.

  4. #4 by Cécile on March 30, 2015 - 1:32 am

    Hi, I totally agree with you.
    But unfortunately,as far as I’m concerned, it’s just a theory for now, because I am a French SLP working with bilingual children in an English speaking country, I can’t assess children in both languages because there is no test available made for these bilingual children.

    So my choices are either dynamic assessment (and I’m not trained for it, I feel very unconfortable with this way of assessing children), or qualitative assessment, based on parent questionnaire and my old monolingual tests. One of my indicators is wether or not the child’s difficulties present on both languages.

    Not enough, to my mind! Any suggestion would be great, thank you.

  5. #5 by Elizabeth D. Peña on March 30, 2015 - 6:24 am

    For French, even if there are not standardized tests, I think there are published papers that that indicate the kinds of errors children with language impairment make in French e.g.,
    So, I think language sampling in both languages and looking for indicators of impairment would be the best way to go if there are not tests. One can start in English– we’ve shown that this can work to an extent, and then if English testing shows no impairment then you’re done. If results show impairment, then one would need to verify by checking the other language. Just remember the kinds of errors that kids make in the other language may be different from the patterns we see in English.

  6. #6 by Cecile on March 30, 2015 - 5:15 pm

    Thank you Elizabeth. That’s pretty much what I’ve started to do for the last few assessments, but I’m happy to read the articles you mentioned and I hope it will help me to be more accurate and confident.
    Thank you very much for these links.

  1. Language dominance and assessment in bilinguals | Amaris Studio

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: