Thursday, January 29, 2015

And today in totally meaningless math

Ah, I see that WalletHub is back at it again with another round of "we took this number and this number and did some math and pretended that means something!" No, seriously:
 WalletHub divided each city’s aggregated standardized test scores in reading and math for grades 4 and 8 by its total amount of education spending per capita.
Riiiight...that's real research there, folks. They're using a test score calculation with no relation to anything: they're comparing test scores across state lines, which means different tests, and they aren't norming the scores against...anything? Per capita education spending has a real relationship to the number of kids in your school v. your total population (this is why Boston would do "well": they don't have as many of their kids in public schools versus their total population). And the kids that are actually in your schools impacts both your spending and your test scores.
They do say that they adjust for special ed, low income, etc, but don't specify how, so I guess we just should trust them?

Unfortunately, it appears that GoLocal has decided that this is a swell news story and has decided to run with it (that's a Do Not Link url, for reasons that will become clear), and a colleague of mine has decided to repeat the same arguments regarding WPS spending that have been voted down for several years running because they have no merit.
Well. Election papers are available March 3.

Here's the thing: I've actually held off on posting anything about GoLocal's education coverage, because I'd just as soon not call attention to it, as it is so often just so very wrong. It was a sore temptation when this article about pilot schools came out which was wrong in every particular: that isn't the history of pilot schools in Massachusetts, that isn't the Worcester history on deisolation, deisolation has very little to do with our transportation spending, and pilot schools are innovation schools before we had innovation schools. But, these are arcane details, possibly, so I let it pass. That is, though, when I decided that I'd stop returning phone calls from GoLocal (a tough choice for me, because I return phone calls from everyone), as they're not only wrong, but they refuse to be corrected. Being unwilling to learn is where I draw the line.

So, should you need to clear your head with some meaningful math, here's some coverage of some actual research on school spending and student learning:
More spending on public schools can significantly improve the lives of disadvantaged children, asserts a working paper from the National Bureau of Economic Research. The amount a state spends on each low-income student can make or break his or her chances of finishing school and earning good wages later on. In short, its authors wrote, "investments in school spending are worthwhile."
That's coming from the National Bureau on Economic Research. The paper is here.
And in general? Please consider the source of your information.

No comments: