Wednesday, March 08, 2006
I'll put the number down at the bottom of this post.
Now of course that's a simple correlation, and as Frank Stephenson (the fellow who calculated the number -- his sources are linked on his post) points out there is of course the possibility of omitted variable bias. So he's inviting you to write him and ask for the data and see if you can improve on his estimate.
Of course, this is actually old news. But the number is interesting. And you're welcome to try to change the result.
(Which was negative:
Specifically, the correlations between real teacher pay and student performance on the math and reading NAEP for fourth and eighth grades ranges from -0.226 to -0.30.)