Let's talk a little bit about the monolingual materials available to
us and why they are imperfect for a bilingual patient. Remember,
this is NOT a criticism of those materials. They were designed to
test for disorders in monolingual patients. When we use them to evaluate bilingual patients, we are trying to use
them outside of their intended purpose, since we don't always have
appropriate bilingual tools available. After discussing why these
materials aren't a perfect match for our patients, I will explain how
I handle write-ups.
Each language has a different phonology, morphology, and grammar
rules
This is why direct translations of English tests are imperfect
measures of language ability. Sometimes the difference is obvious:
if the Home Language doesn't mark plurals, then there is no way to
translate a test item probing for plural markers. Sometimes the
difference is more subtle: if the Home Language has a richer verb
morphology than English, they may acquire verb morphologies at
different rates and in different orders than typically developing
monolingual English speakers.
Some tests (especially older ones) use an academic dialect of
English
Especially for patients who learn English as a second language, it
can take several years longer to learn academic English than
conversational English. Also, remember that bilingual patients learn
the dialect of English they are exposed to—and that isn't always
academic English.
Idioms and cultural idiosyncrasies
Anyone who has visited another country (or region of the US!) or has
friends from another culture understands why this is problematic.
Sayings like "raining cats and dogs" and even items probing
for certain social interactions may not make sense in another
culture.
The test is "normed" on monolinguals
As you know, a standardized score compares a patient's accuracy on
test items to a normed sample of their peers. Well, for bilingual
patients, their peers are ONLY other bilinguals exposed to the same
languages. Monolingual language development and bilingual language
development are DIFFERENT. The English of a monolingual English
speaker develops differently than the English of a bilingual
Cantonese-English speaker, whether or not a language disorder is
present. What does this mean for testing? It means that normed
samples for monolingual tests cannot give an accurate "standardized
score" for bilingual patients. They're not supposed to.
If there were enough bilinguals to statistically influence the normed
sample of a monolingual test, that would likely produce bimodal
distribution and could muck up the standardized scoring process for
both monolingual and bilingual patients. Most testing materials
include a statistical analysis of the normed sample. A solid normed
sample will not include a significant split between monolingual
English speakers and bilingual speakers, but will include almost
exclusively one population or the other.
A Write-up Tip
If I am required to include in my
write-up scores from a monolingual test to a
bilingual patient, I omit standardized scores and age equivalents and add the following caveat: Interpret scores with (extreme) caution, as the patient is bilingual
and is not well represented in the normed sample; for this reason, standardized scores are not reported.
When a "perfect match" test is not available for our
patients, it is ethical to use testing materials that are available,
provided that we understand their limitations and use our best
clinical judgment to make up for those limitations. Luckily,
researchers in the bilingual scene are absolutely working on this!
For more information on how to help, click this link.
