Got this in the email yesterday:
For skeptics who say school vouchers don't work, Greg Forster, a senior fellow at the Foundation for Educational Choice, decided to go back and review the data…
I guess the press-release writer didn't read my blog entry Monday about confirmation bias. I haven't read the think-tank paper, but the first question that came to mind was "How many of the studies Forster reviewed are his own work or that of Jay Greene?" The second: "Did he cite any of John Witte's work?" The third: "Did he report effect sizes?" There would be an interesting exercise to try a meta-analysis on either of the main claimed "effects" of vouchers (especially claims of a competition effect), but I think it's best done by someone from who most readers would not expect to be strongly influenced by confirmation bias.
Update: At least a quick skim of the relevant paper suggests that the answers are yes, only one, and no (respectively). Since this particular "study" is being distributed by a right-wing think tank, I should note that I've seen plenty of confirmation bias from all ideological and philosophical quarters in education. If you're feeling the slightest bit smug about this entry, you need a quick knock on the head: when was the last time you cited only the study (or few studies) that came to mind that supported your own view?
It’s not “confirmation bias,” in and of itself, to cite valid studies performed by oneself and one’s colleagues.
By itself, no. But that’s not by itself.