DEVELOPMENT AND RESEARCH IN APPLIED LINGUISTICS
Language assessment
Development
The project described here (Elder 1997) exemplifies the real problem approach at the heart of applied linguistics. We can represent the process thus:
1. there is a social problem which needs resolution;
2. an applied linguist is invited as consultant; and
3. a solution (not the solution) is proposed.
In this case the problem was in education, not, as is so often the case, in English as a foreign or second language, but in the teaching of so-called modern languages in an English-speaking country, in this case Australia. There these languages are known as LOTEs (languages other than English) and for historical and geographical reasons (Australia’s immigration policies after the Second World War and its location in the Asian-Pacific region) schools offer a very wide range of languages. In the State of Victoria, for example, students have the choice (not of course in every school) of some thirty-six different languages at the school-leaving examination, known as the Victorian Certificate of Education (VCE).
Apart from English, the choice of subjects students may offer at VCE is open. Performance on the ‘best four subjects’ of the VCE is used for university selection. This is very competitive and operates on a points system, known as the Tertiary Entrance Requirement (TER). In order to encourage the learning/teaching of LOTEs, candidates offering a LOTE are given an extra 10 per cent. This of course makes the selection of as LOTE attractive, at least for those who are good at languages. And there’s the problem. Those who are ‘good at languages’ include students who have started the LOTE in school from scratch and those with a background in the LOTE from home exposure. These so-called ‘background’ speakers include on the one hand students from Italian and Greek migrant families who may by now have been resident in Australia for several generations and who may still maintain some use of the language at home. They also include students who at the other extreme have recently arrived in Australia from, for example, Hong Kong, Taiwan, Vietnam or Lebanon, where they may have already received some (possibly all) of their education in the medium of the LOTE they are now offering at VCE.
Does this home use (home both in Australia and, for the more recent arrivals, previous home in the country of origin) give them an advantage over those students who are candidates for the same LOTE but who have no home use, no family connection with the language, other than perhaps a parent who also studied the same language when he or she was at school? The considered view of the authorities has been that this does constitute an advantage for the so-called background speakers. Such a view appears to make sense: if you have studied in Mandarin (for example) for a number of years, you are already literate in the written script, you are familiar with a large number of Chinese characters, then you would seem to have a serious advantage over your peers (who may well be in the same class as you at school) who have studied Chinese for perhaps six years, whose literacy is limited and whose spoken Chinese is still formulaic.
For this reason, those who declare themselves (on the basis of a questionnaire) to be background speakers are penalized at the TER stage. This is in practice more complicated since it is not that their TER is changed but that the university admissions officers are permitted to boost the scores of those who are not background speakers.
This is obviously a language problem. The contribution of applied linguistics was first to determine a methodology for categorizing background and non-background speakers. A questionnaire was designed, the results of which were used to separate learners into four categories; reliability of designation was assured by multiple ratings. The second contribution was, on the basis of these categories, to examine the test results for bias. It was decided that if the tests were fair then background speakers would have no special advantage. In the case of the Chinese students, it was found that they did. The question then was whether or not it was legitimate for them to have this advantage. What emerged was that the results could be interpreted in different ways depending on the point of departure, psychometric, educational or socio-political. Psychometrically speaking the test was not fair and it was biased. Educationally the test was fair, on the grounds that the background speakers did know more. Socio-politically, the score adjustment procedure was very unfair since ethnic minorities who were already disadvantaged because their English in many cases was not native-like were now also being penalized for native-like LOTEs, of which they were background speakers.
Applied linguistics in this project needed skill in devising a methodology for collecting and analyzing the data. It also needed knowledge of the bilingual LOTE setting and an ability, based on experience and knowledge, to bring together the different points of view, psychometric, educational and socio-political.
