15

In my university, you must take TOEFL in order to graduate. It is not cheap. My English composition teacher told me that the test doesn't represent real English skills. But I got even more shocked when I discovered that those tests are created by a private corporation.

So I've been wondering is there any study confirming or denying the effectiveness of these tests - for example correlating test score with academic success?

alexchenco
  • 251
  • 1
  • 3
  • 1
    As long as universities *require* these tests, it should be obvious that they are necessary for a student's career and it is not really a question. You *might* want to ask: "Do these tests really test your English skills". – Lagerbaer Apr 28 '11 at 01:24
  • 1
    I don't really know but I'd say they are a parameter, an added layer to filter out students. Just yet another almost useless challenge on top of many others we face in universities. But it does have its value and it does filter out many people with far from enough english skills. – cregox Apr 28 '11 at 02:07
  • 1
    @Cawas Unfortunately, it doesn't. When I applied at UBC for gradschool, my supervisor told me they don't require a GRE anymore because students from some countries always had awesome scores but could barely speak English. The department therefore saw no point in requiring them anymore. And I didn't have to do a TOEFL either, I just demonstrated that I spent a year studying in an English speaking country. – Lagerbaer Apr 28 '11 at 02:20
  • 1
    @Lagerbaer I took a TOEFL and I can't see how someone who can barely speak, like my mother, could pass that test. And I see no direct relation between GRE and TOEFL. As for colleges accepting students without TOEFL, I think that's great. They are open minded in to the fact they just want some strong evidence you speak english, that's all they need. – cregox Apr 28 '11 at 02:23
  • 1
    @Cawas Well, it was alleged that the government of a particular country had influence over the testing centers in that particular country so the test results got heavily skewed. Doesn't the GRE also test English language? (Essay questions and such?). – Lagerbaer Apr 28 '11 at 02:34
  • @Lagerbaer as far as I know, and I'm no native in english, GRE is a school score that can be converted from schools in non-english speaking countries. It's not an english proficiency score in any sense. And, for government skewing TOEFL results, well, that's good reason enough to invalidate the particular skewed tests, not the whole TOEFL exam method. – cregox Apr 28 '11 at 02:38
  • @alex: Edited a bit so it's answerable. – Sklivvz Apr 28 '11 at 10:40
  • Problem with TOEFL (and also FCE/CAE/CPE), is that there are courses designed specifically to make people score high on these tests, rather than learn language. – vartec Apr 03 '12 at 14:42

1 Answers1

3

I think there are two questions here.

  1. Do the test results correlate to English skills?

  2. Do the test results correlate to academic success?

Here is a quote from an article on the subject at The Linguist Library, talking about one such test:

ETS has published studies showing a close correlation between the TOEIC test results and results from alternative ways of evaluating English speaking skills, such as interviews. There are other independent studies which resulted in a lower level of correlation between TOEIC results and results from other measurements of language skills. Obviously any evaluation of language skills cannot be absolutely accurate. In comparing the results of two different language tests, at best we can measure the correlation between two imprecise evaluations.

That shows some differing opinions on the first question (see the page for references).

ETS claims that TOEIC is designed to be a good predictor of writing and speaking skills, even though those are not tested in TOEIC. Perhaps this is so. However, even with TOEFL, which does measure writing and speaking skills, people with similar scores will often have quite different abilities to speak and write English. There are criticisms that some students with high TOEFL scores cannot communicate well in English and do poorly at English speaking universities. Similarly, studies of IELTS results and academic success show a surprisingly low level of correlation between test scores and academic results. See TL Forum 1999: Dooey.

This addresses the second question.

Should you trust the views of The Linguist Library? I am not familiar with their work, and I can't comment on their rigour.

p.s. I can understand a university declining admittance based on English skills if they could show that poor English test scores correlated very strongly with poor academic performance. However, declining a student the opportunity to graduate, when they have implicitly demonstrated basic communication skills by passing the other requirements of the course seems a bit harder to justify.

Oddthinking
  • 140,378
  • 46
  • 548
  • 638