Testing ESL Children: What we can do now

One of the challenges in assessment of bilingual children is deciding whether or not they have language impairment. On one hand SLPs might decide to wait for children to learn more English before they assess them. On the other hand it’s important to identify children who have language impairment early so that we can intervene.

As of yet there are no standardized tests for bilingual children. There are some standardized tests for children who speak other languages. But, often these tests are inappropriate because they do not apply to children who speak two languages. There are some folks working on development of such tests for Spanish-English speakers (including me), these are few and don’t apply to all language pairs or all ages. At least not yet. So, what can we do NOW for the kids who are referred for assessment of language ability? What do we do to make decisions about language ability in the absence of standardized tests or even in the absence of personnel who speak the child’s language?

There are a number of recommendations by researchers and educators in this area. First, it’s important to get some background:

• What language(s) does the child speak? Who (in the household) speaks the child’s language
• What is the child’s language history? What language were they exposed to at each age?
• What about their educational history? What level of education do they have prior to coming to your school? What language(s) was the curriculum delivered in?

These kinds of questions will help you decide what to expect and what skills they may have in each language. Often children will have knowledge that is distributed across the two languages. In terms of vocabulary they may know school-related things in English. They may not know these same things in their home language however if they are not exposed to the same kinds of things there. Similarly, they may know things in their home language because of their experiences there that they may not know how to express in their second language—even if it seems that they have pretty advanced skills.

I like to interview parents about how they describe their child’s language skills. Here, I think it’s important to ask for examples and to get detailed information. We’re finding that simply asking about concern does not mean the same thing to everyone. Language and communication skills are very important to mainstream American parents because these are skills that are critical for school success. Parents from immigrant groups may be more focused on other aspects of development and may not focus on language development as a “concern.” But, they likely know exactly how their child expresses him- or herself. It’s up to the SLP to then take this information and decide whether it is commiserate with expected developmental milestones. So, I think it’s important to ask parents to describe their child’s speech and language skills. Some questions to ask include:

• Can your child produce all the sounds of the language? Do they mispronounce any sounds? All the time, some of the time, on certain words. Here, make sure to ask for examples. It may be that children can produce all the sounds, but in some longer words or in more challenging combinations of sounds that they have difficulty. Here, we have to think about overall patterns of correct vs. error and it’s impact on communication relative to age.
• Does your child use a lot of different words for things? Here, it’s important to gauge a child’s vocabulary size and their connections among vocabulary words. Both vocabulary breadth and depth are important. Remember too that this information needs to be combined with testing in English to get an idea of whether the vocabulary is of sufficient variety across the two languages.
• Does your child make grammatical errors? Here, we want to get examples of the kinds of grammatical errors that children make in their home language as well as those that they make in English. We want to look for markers of language impairment in both their languages. This is especially challenging when we don’t have enough knowledge of the language, but there are researchers who are looking at different languages and documenting the kinds of errors that children with language impairment make in those languages. The work of Larry Leonard and many of his colleagues have made significant strides in describing these errors in different languages. We would expect bilingual children with language impairment to make many of the same errors that monolingual children make. Note however that they may additionally make ESL errors or errors in their home language that are influenced by contact with English. A key issue here is that typically developing children seem to have better strategies even if they are still in the process of learning English. For example, work that Jacobson and Schwartz have done with Spanish-English bilingual children with and without language impairment demonstrated that both groups made tense errors. The difference was that typical children tried to mark tense but sometimes got it wrong (e.g., past tense: runned for ran) while the children with impairments did not mark tense (e.g., past tense: run for ran). Children in the process of learning English as a second language will make errors, but typical children make errors that do not interfere as much with the message.

Some language sampling is in order here. As I’ve talked about before, I like narrative samples. They’re quick, fairly universal, and a good way to determine whether the child can be an effective communicator in their language and in English, even if they make errors. Children with language impairment will use more restricted vocabulary and grammatical structure in both languages. Their stories will be more sparse and immature. On the other hand typical children should not demonstrate these same difficulties. Yes, they will likely make errors, and in their second language they’ll often use more restricted vocabulary. Yet, the quality of their stories is different. They take the listener into account more, they will often find a way of making themselves understood.

What about testing?

After doing parent and teacher interviews and looking at children’s cum folders to educational history and progress in acquisition of English and the home language (depending on the language of instruction), some testing is appropriate. What tests should be given? Many of our standardized tests do not apply as most of you know. But, I think that there are some tests that can give us useful information. Some caveats though.

1. Ignore the standard scores—these will not help you very much.
2. Give all the items—sometimes children will know things beyond the ceiling.
3. Try some trial testing—really!! Trust your observations and clinical expertise.
4. Take both languages into account simultaneously.

What kinds of tests should be given?

• I like to give expressive vocabulary tests for one. Yes, I’ve spend a lot of my career critiquing them and we know they often do a poor job of identifying language impairment. But, they can provide useful information. Do these conceptually—that is accepting either the home language or English. Look at the kinds of errors the child makes. Are these errors that show they don’t know at all? Or are these errors that can be explained given their experience? I find that children with language impairment will tend to give more “don’t know” type responses but that children with typical development will describe, name it’s function, tell you where to get one, or give a word that’s “close” phonologically. These are very good strategies that young children with language impairment employ rarely.
• I try to gauge children’s comprehension. Even if children do not know English well, they may comprehend more than they can express. So, either a comprehension test of some sort, or a comprehension subtest might be helpful. Again, try to get at these same skills in their home language as well.

Dynamic assessment—you know I had to say this. Till now, we don’t have a lot of information about how DA works with children in the process of learning a second language, but we do have some. I talked about this in a previous post, and Christine Fiestas did her doctoral dissertation exploring this issue in the area of narratives. I’ve also done a couple of case studies that I’ve presented at conferences. This might have to be the topic of another post, but for now let me give you some things to think about:

• DA uses a pretest-teach-posttest approach. You can decide to teach in any domain, I’ve done studies in the area of vocabulary and narratives. And, in collaboration with Ron Gillam and Lynda Miller, I’ve worked on developing a procedure for dynamic assessment of narratives that is available through Pro-Ed.
• If you do this in English, the child needs to be fairly proficient in English—not necessarily dominant in English, but it’s going to be useless if they can’t follow you at all.
• Look at pretest-posttest differences. You should expect some change, but not a change that would be equivalent to an English-only child to make.
• Consider that you are not going to see change across the board. In DA of narratives children might increase their sentence length, or use words in more ways, or be more precise and descriptive, but probably they won’t change in everything.
• Observe their strategies. Typical children should be more flexible in use of different strategies as you teach them, they also can monitor their errors better than children with language impairment.
• Trust your instincts! SLPs are trained to observe language learning, I think this is what we need to do. Do not feel constrained by test scores or test protocols, you need to get information, get it any way you can.

Advertisement