FACTORS RELEVANT TO THE ELTS EVALUATION
Background
Linguistic and phonetic factors
We have already considered some of the linguistic aspects under educational and the sociology/sociolinguistic, above. But there is an even more central linguistic aspect and that is the issue of sampling. Like any proficiency test, ELTS is meant to represent the language: those whose results indicate satisfactory proficiency are deemed to possess enough English for their study. This is a very large assumption. No proficiency test is likely to take up more than, say, three hours of test time and yet that test is expected to predict whether a student can cope with a year or three years’ study, including very complex spoken and written material.
It was also assumed in the non-specialist components of the ELTS test that a proficiency test should target language use in the community, on the reasonable grounds that a student who coped easily with English communication in life outside the university, interacting with other students, communicating in the local shops, solving accommodation problems, seeking medical attention, going to the cinema, managing an English-speaking daily life, would be more likely to succeed in his/her studies. These are all large demands from a test. What this means is that the samples of English use that are put into the test must be very carefully chosen on the basis of their normality and frequency within the language. Anything odd or idiosyncratic, or even jokey, is best omitted. Judgements of the samples in the ELTS test required careful analysis on the basis of linguistic knowledge of the English language, the kind of analysis that is necessary for selecting the materials contained in a pedagogic grammar.
In addition, judgements were necessary of the spoken component of the test. These in part related to the voices used for the listening sections and in part to the ways in which the students’ own speaking was measured, how far the question of intelligibility was raised, and of course intelligibility to whom. Just as the touchstone for the written materials in the test was the standard language, so the question which raised itself with regard to the spoken components of ELTS was what sort of norms should be applied to the judgements made of the candidates’ own speech. To this extent the language problem which led to the evaluation of ELTS was indeed a problem involving language as much as it involved the other factors to which we have drawn attention.
