Thursday, November 29, 2012

More on graduation rates

The DCist also has a piece up on the recently released DOE graduation rates (our original here).  The reason I'm posting is not that there's anything new, but because of the update that appears in their post:
UPDATE, 10:45 a.m.: As a commenter rightfully points out, comparing D.C. to the states is somewhat unfair—D.C. is an all-urban school system, after all. New York City, for one, recorded graduation rate this year of 60 percent, roughly the same as D.C.
I had a commenter in our original story expressing much the same thing. I hear similar things when statewide academic scores are compared, and DC is invariably near the bottom. And (statistics being the thing I do in the job that actually pays me) the argument is, in a way, correct.  It is, however, in other ways wrong.  I'll try to explain.

The "this isn't an apples to apples comparison" argument is that DC as a "state" has a single, urban district while all of the other states contain multiple districts, some urban, some suburban and some rural.  So, a fairer comparison would be to compare DC to only other urban districts.  When this is done DC looks more in line with its peers.

Its a similar argument for measuring a car brand's fuel efficiency by comparing their average fleet's average miles-per-gallon.  Take GMC and Ford for example.  When comparing the two GMC is going to look worse that Ford, but that's not necessarily because GMC does a worse job at designing efficient vehicles.  What's driving most of the the difference is that the GMC brand only sells trucks, vans and SUVs whereas Ford brand sells everything from sub-compacts to trucks.   If you want truly want to compare efficiency then compare vehicles with similar weight, engine displacement, etc. to each other.

Here's the rub. Subcompacts and trucks differ in MPG statistics for very fundamental reasons: it takes more energy to move more mass the same distance. And a vehicle that's designed to move two pallets of sheetrock is going to have more mass than a vehicle that's designed to only move two human bodies (think SmartCar).  That's why expecting them to have the same average MPG is unfair.  There is, however, nothing fundamentally different about people, or children in this case, at the basic level. A kid from an urban school is biologically no different from a kid from a suburban school who is biologically no different from a kid from a rural school.  "Urban schools underperform" is not a priori. A school district underperforms for a reason.  Comparing yourself with other underperforming districts and saying "hey, we're not *that* bad" is, frankly, insulting.

No comments: