This past week we covered the topic of computers and language testing. There were a number of benefits and drawbacks that we discussed. In this blog post I highlight some of those.
An advantage for using computer based testing relates to practicality. It is saves on time and resources when a computer delivers the directions for the exam and the questions. Scoring is also a much more efficient process on the computer. Another advantage mentioned in the DuBravac (2013) chapter was that test items can be enhanced in online formats.Animation, video, audio, and pull-down menus can all be used to enhance more traditional types of testing items.It was also been discussed that computer rating systems are more reliable than human raters because the issue of rater bias is not a factor.
Despite these advantages, there are a number of unresolved aspects of computer based testing that make implementation questionable. For starters, automated scoring programs such as ETS Criterion can only assess certain aspects of production. In writing, for example, it is difficult to assess more discourse-level organization features than grammatical features using automated scoring. This is one area where I believe using a computer is still not as effective as a human rater. Doug Hesse has recently made a similar argument in an article written for the Washington Post.
Another issue with computer based testing has to do with fairness. A test should be fair and equally accessible to all learners. However, as was brought up in class, even in such a technology savvy age there are still some learners that have little to no experience using computers. So designing tests that rely upon computers could be unfair to these learners. Some learners also have anxiety using technology, so using a computer based test could raise the level of anxiety, which in a high stakes testing situation, is likely to already be at a high level to begin with.
No comments:
Post a Comment