Phil Schrodt is a just retired political scientist and quantitative methodologist. His retirement letter is well worth a read: it’s part sweet, part scathing analysis of contemporary political science and academia more generally. His letter led me to his recent working paper titled Seven Deadly Sins of Quantitative Political Analysis. Following Chris Achen’s influential critique of garbage-can models in political science, Schrodt argues that political scientists abuse the linear model while chasing statistical significance and ignoring Bayesian alternatives. Pretty standard stuff, if you follow Andrew Gelman’s blog and the like, but especially entertaining and clear. In the 2013 post-script, Schrodt offer some reasons for pessimism in the light of a decade of fighting and losing this methodological war:
The institutional response to Achen (2003) was, of course, the notorious journal-length methodological suicide note, Conflict Management and Peace Science Vol. 22, No. 4, which, like J.R.R. Tolkien’s Gollum leaving the sunlit world for a lonely life in subterranean darkness, reads like “Precious, oh precious garbage can models. Evil Achens wants to take away garbage can models…no, no, we won’t lets them…precious garbage can models…”
I just love that image, and I highly recommend the rest of the paper as well.