Many Standardized Tests do not Measure What they Should

Reason 2/5: Why Making the Atlanta Public School Teachers Felons Won’t Make a Difference:

Ever looked at a high school’s English curriculum? The objective strands look like a list of “shoulds”: Child will be able to synthesize primary and secondary documents. Child will be able to recognize post-war poetry elements. Child will be able to use punctuation properly. Child will be able to properly conjugate helping verbs. There are scores of these objective strands.  And on a standardized post-test, because there are so many strands, and they all need to be represented, each one of those elements gets usually just one question. So apparently using proper verbs is equal to recognizing Post-War poetry elements; worse, if the child misses the single question about verbs, then apparently he knows nothing about verbs, and I’ve taught him nothing about verbs.

A strand that says the child can develop a thesis and defend it with strong writing cannot be accurately tested on a multiple choice exam. So wrongly, in order to assess that strand, the child is asked to find the thesis statement in an offered essay. Test writers often do not spend enough time writing these questions well, and honestly, there really might be two or three sentences that could work as a thesis statement. As any good writer knows, essays don’t really have just one sentence that guides the whole flipping essay. (Tell that to the testing boards.) So, based on these standardized tests, a kid who can select a thesis out of a choice of four answer options, but not ever write an essay is deemed just as educated as the kid who writes well.

One year a major writing test that our county system gives to tenth graders had a major flaw. Over forty-five percent of the kids went down the rabbit hole created by this flaw and failed. Instead of retesting the whole population with a valid test, because there were already scheduled retests in place for kids who fail, the board did nothing about this flaw; they let the failures go on the children’s records and simply retested those students. The next year a new question was developed and applied. 85 percent of the kids passed on the first attempt. Suddenly, our county was bragging about how much better we were all doing as educators. LOOK how high this year’s scores trump last year’s! Aren’t we grand?   These scores didn’t represent what they are supposed to, at all.

So not only are objectives unbalanced in terms of importance, not only is success defined inaccurately, but we misuse the data collected.

How does the APS judge expect teachers and children to take any of these tests seriously when we know we are all being misrepresented?

Next up: Reason 3/5 Why Making the Atlanta Public School Teachers Felons Won’t Make a Difference:

The Tests Dumb down the System in General