Playing with numbers

If you are the least bit interested in standardized testing, read this New York Times piece from today's paper. In following the ins and outs of the manipulation of data done by New York City over the past few years, you get a very good picture of the impurity of testing.
Much of the misunderstanding around standardizing testing boils down to this very notion: is it "pure"? I think many who have never taught have in their head a notion that a test is a test, and that any test is somehow held apart, not subject to bias or manipulation or perspective. Those who perhaps have studied this a bit have gotten to the point of realizing that a teacher might have a perspective or bias that is inherent in a test that teacher gives (good teachers, incidentally, are well aware of this).
What too often is missed, however, is that all about standardized testing is also subject to human vaguries. Someone has to write the questions. Someone has to decided how things will be weighed (how much essay? how much on algebra?). Someone has to decide--and this one is particularly subject to political and very public whims--what a passing grade will be.
The notion that somehow the MCAS, the FCAT. the NYSTP, or any test, by virtue of being a bubble test, or statewide, or hired out, is somehow a more virtuous system is to miss the central point: it's still a human test.
And the stakes are much, much higher when you start piling property values, principals' jobs, school funding, and so much else onto what is already a fallible system. And there's much more reason and room to play with the numbers, as was done in New York City, when you've got so much else riding on it.