Reflection on Alderson, J. C. (2000). Technology in testing: the present and the future. System, 28(4), 593-603.

In this article, the author first reviews the advantages and disadvantages of implementing computer-based testing. Although a number of disadvantages exists in computer-based tests such as incapability in testing productive skills and computer illiteracy, the author is predicting a massive development in CBTs in a few years and some might even tackle these problems.

Advantages of CBTs include reducing group administrators of assessments, no time constraint for tests, immediate results after testing, high testing security etc. One of the great advantages of CBTs is the computer adaptive testing. The system can choose questions of appropriate levels according to the answers students picked. The more questions students get correctly, the more difficult questions would be provided. Immediate feedbacks provided by computer-based tests is also another pedagogic advantage of CBTs mentioned by the writer. While students might need to wait for months for the feedback after a traditional paper-pencil test, it is more meaningful for them to know how well they have performed after each question, or after they have finished the whole test, as they still remember what they have done. One more advantage that CBTs brings is its tailor-made user friendliness.

Even with different innovative features of CBTs, the author found that most CBTs now are actually similar to the paper-and-pencil-based testing. Although some innovations are noteworthy such as adjusting pace of tests and choosing more than 1 choice in Multiple Choice questions used in CBT TOFEL, not many variations could be seen nowadays.

I agree with the author that apart from making the test more user-friendly by CBTs, more features should be developed in order to raise its compatibility with language pedagogy. Even the author suggested program like E-rater, which is designed to give scores to productive output, writing, I would not trust the score given as it follows samples extracted from the previously scored open-ended essays. As long as the test takers know about the rating principle, the system can be fooled by giving some specific structures or vocabulary item which are not totally fit in the context. I seriously have doubt in measuring productive outcomes by CBT. Hopefully, the future development could give some insights about this area.