LANGUAGE IMPAIRMENT IN MULTILINGUAL POPULATIONS : A TEST BATTERY FOR SINGAPORE BILINGUAL KINDERGARTEN CHILDREN

Abstract

Specific Language Impairment is diagnosed when a child has difficulty acquiring language, but is developing normally in other areas. Research has shown that SLI can have negative effects on academic, cognitive, social, and emotional development, making it important to identify the problem as early as possible. Identifying SLI in unilingual children is not a simple procedure, but it is relatively straightforward because tests and norms are available, and research on normal and deviant language development has already increased awareness of language problems. Identifying SLI in bilinguals is much more complex. When children know two languages, their patterns of language development are often very different from that of unilingual children. This difference challenges the simple transfer of the assessment principles used for unilingual children to bilingual children for various reasons. First, issues of test validity arise, because most of the available screening tools have been developed for (white) English-speaking children living in the West, and bilingual children in Asia often come from rather different linguistic and social backgrounds. Second, reliability of the norms is questionable. Third, most of the tests are in English but bilinguals should be tested in both their languages, whether or not their first language is English.

This lack of appropriate tests in Singapore and other Asian countries makes almost impossible to differentiate children who are language impaired from those who have problems learning English as a second language, or those who are intellectually impaired.

This thesis addresses the problems of language assessment in Singapore with two studies. The first study describes how a bilingual language screening battery for kindergarten children was developed. Tests were chosen with due regard to the work of Duncan (1989), Bishop (1997) and Whitehurst & Fischel (1994), and included single word comprehension (BPVS) and production (Picture Naming), as well as sentence level comprehension (Syntax Comprehension Test) and production (RAPT). The original test were in English, and parallel tests were developed in Mandarin, the language of majority of the kindergarten population. The English and Mandarin version were then renormed separately on two groups of kindergarten children (aged 4:6 - 6:5): children who came from Mandarin-speaking homes, and those from English-speaking backgrounds. The results showed that almost all the tests had sound psychometric properties, and could provide some reliable local developmental data for the two languages. The second study comprises seven case studies to explore the predicative validity of the test battery, which is especially important for any screening tool. Five kindergarten children whose had low scores on the test battery were followed up into Primary school, and their academic performance was tracked. These children continued to show weak language skills after two years and school performance appears to be negatively affected Two additional case studies of children who obtained relatively high scores on the test battery, but had similar nonverbal abilities to the other case studies are also reported. The two children showed average to above-average performance in Primary school, suggesting that the test battery has predictive validity in both directions. Although a large scale follow-up study is needed to support these single case studies, there was clear evidence to show that there are children whose language problems persist, and that these children will require some forms of intervention if they are to cope with Primary school; i.e., many children do not "grow out of' their language problems. Taken together, this these two studies show that it is possible to develop a valid and reliable bilingual test battery with relative ease if certain general principles are followed. These include renorming the tests on the local population, norming the tests separately for children with different dominant languages, and assessing in both languages. Mandarin-English speaking Singapore children can now be reliably screened for language problems in kindergarten, and early intervention offered to prevent later academic, social and emotional problems.