A warning to college profs from a high school teacher

[caraher]

13. Appropriateness depends a lot on the purpose of the test

Last edited Tue Feb 12, 2013, 10:00 AM - Edit history (1)

For certain kinds of basic knowledge that you need in engineering you're absolutely right, a standardized test makes a lot of sense. I don't think political science, for instance, really fits the same mold! There are no rival schools of thought regarding the basic scientific and mathematical facts underpinning work in engineering, but the same is not true in other areas.

And even in a scientific or technical field, it's still worth bearing in mind that the ability to pass such tests only demonstrates a facility with a common set of basic facts and skills and not overall competence. "Common sense" and a certain kind of creativity are also crucial in separating a glorified technician from a truly talented engineer, and good engineering schools will cultivate those through lab courses, research or co-op experiences, the value of which cannot be captured on standardized tests.

In physics, it's very well-known that physics GRE test scores are a terrible predictor of graduate school success, particularly in experimental specialties. And the general test is no better. Incoming graduate students from programs that teach to the test routinely have perfect or near-perfect scores but do not outperform students from other programs who score maybe at the 50th percentile. There's enough of a mismatch between what it takes to score well on a long exam of short multiple-choice items and the real work of physics research that the test mostly measures test preparation itself rather than readiness for graduate study and aptitude.

It's not that standardized tests are bad in every setting; it's that they are taking over in inappropriate situations. It's like the GOP economic program where "more tax cuts" are the reflex solution to every problem. Sometimes a tax cut is a good idea; the problem comes in when it's considered the solution to every problem. Similarly, we've reached a point where the operational definition of effective eduction is invested entirely in standardized test scores, to the exclusion of any other measure, and that is problematic!