There's the idiot who wrote it for the New York Post and the idiot who wrote it up for slashdot.
Three of biggest examples of the idiocy, in reverse order of stupidity:
1) The story is not about kids scores going up by 40 points. The story is about the percentage of kids "passing" the test going up by 40 points. If a lot of kids are near-but-below the passing mark, scores could go up by just a couple of points and yield a 40% increase in the passing rate. (Obviously, an extreme way to make the point, but there clearly was nothing like a 40 point increase by the kids).
2) Sample size. That almost-40% number is just at one school. Any one who knows anything about statistics and/or testing knows that small schools are the most likely to see wide swings, showing the greatest gains in passing rate and the greatest losses. One of two schools cited in the story has about 65 kid/grade. That's a small school, even by the "small school" standard. The other school also has less than 100 kid/grade. If you don't understand this stuff, you should not be writing about testing or statistics of any sort.
3) Why doesn't the story report the overall impact? Why doesn't the story report the overall increase in the participating schools compared to the rest of the city? I don't mean controlling for factors, just a straight comparison? (I'll tell you why: a quarter of the school showed LESS improvement in math and a third showed less improvement in reading. At least the story mentioned that.) We all know that scores were up city-wide.
I'm not even getting into testing issues, pedagogical issues, developmental issues or the sustainability of these gains. I am just talking about the reporting.
Slashdot Mirror
OH MY GOD is that poorly written.
There's the idiot who wrote it for the New York Post and the idiot who wrote it up for slashdot.
Three of biggest examples of the idiocy, in reverse order of stupidity:
1) The story is not about kids scores going up by 40 points. The story is about the percentage of kids "passing" the test going up by 40 points. If a lot of kids are near-but-below the passing mark, scores could go up by just a couple of points and yield a 40% increase in the passing rate. (Obviously, an extreme way to make the point, but there clearly was nothing like a 40 point increase by the kids).
2) Sample size. That almost-40% number is just at one school. Any one who knows anything about statistics and/or testing knows that small schools are the most likely to see wide swings, showing the greatest gains in passing rate and the greatest losses. One of two schools cited in the story has about 65 kid/grade. That's a small school, even by the "small school" standard. The other school also has less than 100 kid/grade. If you don't understand this stuff, you should not be writing about testing or statistics of any sort.
3) Why doesn't the story report the overall impact? Why doesn't the story report the overall increase in the participating schools compared to the rest of the city? I don't mean controlling for factors, just a straight comparison? (I'll tell you why: a quarter of the school showed LESS improvement in math and a third showed less improvement in reading. At least the story mentioned that.) We all know that scores were up city-wide.
I'm not even getting into testing issues, pedagogical issues, developmental issues or the sustainability of these gains. I am just talking about the reporting.