Thursday, February 12, 2009

Most states now use constructed-response test items

Thirty-eight states now use constructed-response items as part of their statewide testing, including Kentucky. Of those, six vary their approach by grade, with one using only multiple choice for elementary and middle school testing and five using only multiple choice and writing prompts for high school. Important elements of performance can be measured with multiple choice, but not not students' ability to explain how they've tackled and solved problems. I assume that most states now value that capacity and design their tests to reflect that commitment. (To download a one-page chart of states, click here.)

1 comment:

  1. It seems from this information that most states have realized that there is more to testing students' performance than simply through multiple choice tests. It seems that there is some sort of consensus that there needs to be something in addition to a, b, c, or d. I think the quandary legislators face is how to objectively compare what KY students are learning with students in other states, and if written responses can adequately do that.

    Something that is routinely done in European countries, particularly France, is the administration of oral tests. Having studied abroad there with Centre College, I've experienced this method of evaluation and say that often it is more difficult, especially to BS something if you really don't know the answer.

    Given that in parts of Kentucky, especially eastern Kentucky, most knowledge is conveyed from generation to generation orally, I would predict that students in this area are likely to perform better if tested orally, than if they were required to write something, or if they could only chose a, b, c, d. As evidenced by my own imprecise writing, (I'm from the mountains) I could much more easily explain something orally than to convey it through writing. Granted, writing is utterly essential to proper functioning in modern society, it may not give an accurate portrayal of what a student really knows.

    I know the gut reaction to this is "how can we do such a thing?" "It wouldn't be objective!" "How can we keep test administrators from cheating?" But I think the overall net benefit would be to add another dimension to test if Kentucky students are comprehending the material that is being taught them. The scoring of such responses could be recorded to be audited, much the same as writing portfolios are now.

    I, by no means, am suggesting doing away with the writing component; I actually think it is much needed to have a fuller understanding of what our students are learning. But while we're asking these important questions, why not toy with the idea of using oral tests as well. Just something to ponder.


Updates and data on Kentucky education!