Sunday, October 21, 2007

Reliability in once-off tests 2

Topic: Validity and reliability - issues and discussion Date: 30 November 2006 10:07 PM
Subject: Reliability in once-off tests Author: Mondy, Steven

I like to think of a test as something that not only gauge someone's ability, but as a tool that helps a student see where they need to improve. I was doing some speed reading work in a class today that had simple multiple choice recognition items that helped students see a general overall comprehension score in relation to the time it took them to read a passage. By the time we got through 4 passages, students were able to see a pattern being established, which helped them make their own judgements on their reading ability.
In other classes, I tend to buck our school policy a little and let students mark their own tests, or their partners. One disadvantage would be that it could discourage students who are doing badly. On the other hand, it promotes a common mateship within a class, and a little bit of a competative spirit. Handled well, it can be quite a motivating tool, and reduce the negative image toward taking tests. I understand the need to withold tests on large scale apptitude type tests, but for classroom evaluation and learning, there are times when students need to see the test.
Does anyone else agree with this?
Steven

No comments: