October 04, 2008

Computer Adaptive Assessment

Something I’ve noticed in my new job is that really, for many years, math teaches the same basic concepts again and again, just at higher levels of difficulty and complexity. Second graders learn algebra with simple number sentences. High-schoolers learn to use algebra with any number, and with multiple steps. If curriculum and standards are aligned as such, then computer adaptive assessment like the GRE would present a new option for measuring learning growth in individual students. If the test spans multiple grade levels, it will allow the struggling student to demonstrate what they DO know and challenge the gifted students to reach for the highest level possible. The scores would then better reflect growth from year to year because the standards are the same, but students should be pushed and taught to the higher levels of complexity and difficulty.

It could also cut back on “teaching to the test” and cheating because the students wouldn’t necessarily be on the same questions since the computer will adapt the questions according to the student’s answers and timing on initial questions. On a more touchy issue, I think this could also provide a better and more objective system for teacher accountability because it would more accurately reflect and individual student’s learning growth from year to year. If student does not show a certain measure of growth, questions should be asked why.

With anything, there are difficulties. The difficulties I can think of immediately are the cost of computerized testing for all students. However, one must consider technology’s growing and more mandatory role in education. There are also Murphy’s laws of technical issues during a test. Testing strategies and psychology would have to be changed: computer screen versus a pencil and paper. For this, solutions have already been in the works with online only end of course testing. New strategies would have to be taught and students would have to be pre-prepared for the different environment. Another issue I see is that the magnitude of testing a subject and it’s many different strands. Math is generally divided into 5 domains and each domain has its own set of subdomains, each with a set of skills and concepts. Not everything could be tested other the test would just be too large. However, the test could isolate the concepts that are generally accepted as necessary for the adult world--skills such as problem-solving (algebra), data analysis, and spatial skills. Another difficulty would be structuring the questions to best measure reasoning and procedural thought in an individual student. However, like the GRE, there could be a section of the test sent off for personal scoring.

Your thoughts are welcomed. I'm hoping to do some more research into this and maybe it could give me something to study/research/work on if my dreams of grad schools come to fruition in the next couple of years.

The idea in and of itself has been an encouragement to me as I've been adjusting to the change in career. I've realized that education is a multi-faceted career die. God's roll of the die for my life didn't come up teaching. [Yes I know, well, am pretty sure :-) God does not roll dice in structuring the details of a person's life] Now that I have a direction and an idea for hopefully changing education for the better, I'm not feeling quite so much that what I'm doing right now can make a difference like being a teacher in a classroom. It also helps that I'm doing tutoring part-time occasionally, also allowing me to work with more than one grade level and more than once concept. I mentally thrive on the challenge of stretching my brain to adjust to multiple grade levels.

Posted by Anna at October 4, 2008 07:08 PM | TrackBack