The first thing came
into my mind when I hear computer-based second language assessment was TOEFL,
so my reflection about Jamieson’s article is mostly related to this test. First, this article described the changes in
computer-based language assessment and mentioned some effectiveness of
computer-based assessment such as less time needed for testing, faster score
and authentic contents. One statement caught my attention was that ‘an added
benefit of a computer-adaptive test is that theoretically test-takers are given
items that are well-suited to their abilities, thus making the test both more
interesting and a more accurate measure’ (p.230). Since this statement is introduced
in a safe way with the word ‘theoretically,’ I cannot help myself to say that
this theory- ‘well-suited’ test items will make test more interesting and
accurate is just a theory. Even though
such computer-based assessment offers authentic tasks and test reflects the changed
perspective toward language, using computer itself seems not authentic and test
is still remained as not interesting. What
makes test more authentic? Contents? Tasks?
The first time that I
had taken TOEFL was about 7 years ago and the latest time was about 1 year ago.
Comparing those two tests, what surprised me was that I only recognized few
differences in test contents and items. Even though I used the same book that I
studied 7 years ago to prepare the latest test, there was no such a huge gap
between them and I hadn’t faced to any difficulties while taking the test. Improving test format, contents and tasks
costs a lot. However, I had never felt that this test is highly related to real
language learning situation and never interesting. I had never forced to complete
500-700 writing within about 20 minutes after reading short passage and
listening short lecture in real life. Then what is this for? Where this task is
considered as authentic?
Hi Aram,
ReplyDeleteI enjoyed reading your reflection about Jamieson's (2005) article on computer-aided testing. I worry that computer-aiding test is still not being used in the most effective manner. It can be useful for quick feedback (vocabulary, spelling, grammar). However, I don't think computers are up to the task of measuring things like fluency, conversation, creativity, and authentic speaking. It amazes me that people are still trying to get computers to replicate the way teachers assess student learning. Learning is so complex, it seems impossible to accurately assess it using programming variables.
Hi Aram,
ReplyDeleteGreat reflections! I've never thought about TOEFL ibt from this perspective. To be honest, I kind of like it. I guess that's because I have taken so many structure-based English tests in China (and I was never able to figure out how to situate "on, of, in, at" at the correct place). And I particularly like the listening section (which is a big challenge to me yet I enjoyed a lot in striving for conquering it). But I agree with you that some (if not all) features are not authentic. I guess all the tests, to a large extent, can only be considered as ways that can evaluate people's real performance indirectly. That's why I like the idea of "language portfolio" quite much.