In a new paper in the area of semantics, we did an analysis of a group of bilingual 7 to 9;11 year old children with and without language impairment. We were interested in seeing if children with LI showed similar patterns on receptive and expressive semantics tasks compared to bilingual children with typical development. In a previous study, we examined younger (preschool age) bilingual children on expressive and receptive semantics items on the BESOS. In that study, described here, we found that in English, children showed higher standard scores receptivity compared to the scores they were able to achieve expressively when they had very little English exposure. We proposed that it was important to understand this potential for a receptive-expressive gap in children with LI.In the current paper, we looked at standardized receptive and expressive scores on the semantics subtest of the BESA-ME. The BESA-ME is an experimental test, it is the upward extension of the BESA. The semantics subtest influence many of the same item types as for the younger kids (functions, similarities & differences, categorization), and some other types as well (definitions, associations). On the experimental version of the test, the items are translation equivalents and there are equal expressive and receptive items.
The study included 37 children with language impairment. The gold standard we used to identify kids with LI included: parent report, teacher report, language samples, and whether they were identified by the school SLP as LI. Children had to fall in the impaired range on at least 3 of these to be identified with LI. The group with typical development was tested using the same measures. They matched the LI children on the basis of sex, age, exposure to English and Spanish.
Our results demonstrated higher scores in Spanish than in English overall. In Spanish, there was no receptive-expressive gap for the group overall. When comparing the typical and LI groups we found that the LI group scored significantly lower overall, and that they had a larger receptive-expressive gap than their typically developing controls. Averaging English and Spanish scores the typical children had a 5 point gap on average and those with LI had a 12 point gap on average.
If you look at the data, the typical kids have no gap on average in Spanish, but they do have one in English (10 points). The kids with LI have a gap in both languages. The Spanish gap is smaller (6 points) and the English gap is larger (19 points).
So, what does this mean? I think that for bilinguals, it may not be unusual to see a small (between 5-10 points, using standard scores) in the weaker language. Even if the receptive score shows that they are scoring within the normal range, the expressive score may not have caught up yet. But, for LI, we may see both things going on. Lower scores (our kids on average scored more than 1.5SD below the mean in their best language) AND a receptive-expressive gap in both languages, with a larger gap in the weaker language.