A simple heuristic to forecast elections
Filed in Research News
PREDICTING ELECTIONS FROM THE MOST IMPORTANT ISSUE FACING THE COUNTRY
Andreas Graefe and Scott Armstrong have put forth a model that predicts the outcome of U.S. Presidential Elections based on the most important issue as seen by voters. It’s a good illustration of how one cue, if it’s a good one, can be effective for forecasting.
It’s an early draft and they would appreciate comments.
The paper is available here: http://pollyvote.forecastingprinciples.com/images/papers/PollyMIP.pdf
Photo credit: http://en.wikipedia.org/wiki/File:Election_MG_3455.jpg
Great find. I’m linking to this post asap via my blog SimoleonSense. Thanks for keeping together such a great resource.
-Miguel Barbosa
December 3, 2009 @ 2:03 am
Nice find!
December 7, 2009 @ 5:04 am
Andrew Gelman has commented on this manuscript (which should make the authors happy, since they told me their they’re seeking feedback). Read on: http://www.stat.columbia.edu/~cook/movabletype/archives/2009/12/predicting_elec_1.html
December 19, 2009 @ 7:47 pm
[…] Predicting elections from the most important issue facing the country (PDF – I found it via Decision Science News, the work of GG’s collaborator Dan Goldstein), though the bare bones as […]
December 22, 2009 @ 10:24 am