Return to index page

February 03, 2004


In statistics and epidemiology, meta-analysis is the process of distilling several studies into one. For example, there may have been 50 different studies into the efficacy, of say, Viagra, and a meta-analysis study would look at all of them, and weigh them by various factors such as the number of patients in each study to conclude, overall, how good Viagra really is.

Seems there's a similar trend in journalism, but rather than combining the conclusions of several articles we just re-analyze the same article over and over again. Here's an article in The New Republic Online by Andrew Sullivan, where he discusses David Frum's discussion of Daniel Casse's analysis of George Bush's alleged conservatism. Just how many nested layers of analysis can we get? I guess I'm contributing to the problem by commenting on Sullivan's article myself.

It's turtles all the way down, sonny!

Posted by david at February 3, 2004 12:13 PM | TrackBack
Comment spammers: see our Unauthorized Advertising Policy and rates
Post a comment