The major flaw in "No Child Left Behind"

So we went to a meeting about Becker, an elementary school in our neighborhood last night. We’re considering sending Stella there, due to the extreme overcrowding at our local elementary (> 120%, highest in AISD). Plus a lot of our friends are in the neighborhood. There was talk of a daul-language program, which was interesting, but one thing that struck me was their academically unacceptable rating and their small class size. Linder (the overcrowded school in our neighborhood) has been academically acceptable since the creation of “No Child Left Behind”. That seemed a bit odd to me. But first a word about averages.

Steve Crossland of the excellent Crossland Blog has often spoke about how useless averages are in real estate (and really in much of life). When someone asks you what the average home price is in Austin, you don’t actually tell them the average. You tell them the median. Thanks to the million dollar homes the average home price is probably around half a million. The median is probably in the high 100k’s. When we say average we generally mean median. And when we see averages we generally extrapolate them to the median in our head. When I was in school, if I saw an average grade of 75% in one of my classes, I would assume pretty much everyone got a C. Yet, depending on the size of the class it could have been 2 unprepared lunkheads who got no answers right, and the rest of the class aced the test. Averages tell you very little about individual performance.

With that in mind I decided to look at test scores in town. I decided to compare against 3rd grade, but the numbers are pretty consistent. Here’s Becker (all numbers are number of kids who failed the test):


Grade Average Class Size Reading Math Writing Science
3 12.8 4.1 4.22 3.2 3.07

I was shocked. So academically unacceptable in this case means that 3-4 kids failed the test in each class? That’s not so bad. And when you consider averages are at play here it’s possible that there are a few bad teachers dragging the rest down. So I decided to see what academically exemplary means. This is a school that does not have a lot of english profeciency problems, economic disadvantage, or section 8 housing:


Grade Average Class Size Reading Math Writing Science
3 21 0.21 0.42 0.21 1.05

So the difference is 2-3 kids per class? Not as much difference as you would expect from the terms “exemplary” and “unacceptable”. I finally decided to look at Linder to see how it stacked up:


Grade Average Class Size Reading Math Writing Science
3 22.4 4.93 6.5 2.46 7.39

So, Linder, which is an academically “acceptable” school actually has more kids failing per class than Becker which is “unacceptable”. The law of averages at work. So this means that school districts are rewarded for having over-crowded schools. It explains why suburbs that can’t seem to build enough schools to keep up with demand don’t seem to have the same problems their inner-city counter parts do. And it shows me that realistically most of the schools in Austin are quite good and after meeting some teachers last night, I have to say they’re really dedicated to voluntarily go to meetings until 7pm on a work night. I’m looking forward to sending Stella.

Here’s my spreadsheet if you want to check my work.