You're Bound to Pass Sooner or Later, So Here's Your Diploma

Fake Tests, Fake Schools

The Texas Education Agency must be staffed with refugees from the Enron smoke-and-mirrors factory -- except that the Enron guys were reputed to be "the smartest guys in the room," whereas these people in TEA are . . . I can't even describe it. You be the judge:

A Texas legislator grilled a TEA hack last week about how Texas students who had failed every single question on a standardized ("TAKS") test could have been scored as "passing." These passing scores evidently caused the number of schools ranked as first- or second-rate in the state's four-tiered system almost to triple in a single year. How did this work?

Simple. Instead of old-fashioned grading, they used the new "Texas Projection Measure." Under this measure, "nearly half the 1 million TAKS tests that had been failed [were counted] as passing for the purpose of rating schools and districts." You might think this had something to do with counting a performance as passing on the ground that it at least was an improvement from the even more dismal performance of the year before. Alas, no. The measure "looked only at last year's scores and, based on a formula devised from thousands of prior results, projected that children who pass reading or math [tests] were likely to pass other tests in future years." In other words, we'll pass you because this formula says that if you passed a test in subject A, you're very likely to pass a test in subject B at some point in the future.

It gets worse. The formula is not even based on identifiable trends in low-performing students in situations where improvement has been documented in the past. No: the formula is based on overall trends among all children who take the TAKS test, from the ones who ace it down to the ones who flunk it. In general, kids do tend to do better on tests from year to year, thank goodness. However, if you consider only the kids who flunk a particular test, the "Projection Measure" is about 50% accurate in predicting that they'll eventually pass that same test. So the thinking apparently is, since about half the kids who flunk can be expected to do better at some point in the future, why not go ahead and count all of them all as "passing" right away?

The author of the article wonders about the same question that so often engages us here: lunacy or malice? Could Texas education officials really be stupid enough to buy this kind of metric, or are they cynically manipulating any and all data in furtherance of the one true aim, which is to maximize funding?

But, as one commenter noted, what's the real difference between this and the system of vague expectations of future merit that resulted in both Barack Obama's Oval Office and his Nobel Prize? Wishin' makes it so.

No comments: