ISR's Pasek and Conrad discuss the art and science of predicting election outcomes
"Numbers Cruncher" - Scientific American. 11/10/2016.
In another piece examining why the polls got it wrong in the presidential election, Josh Pasek and Fred Conrad look at assumptions, parameters, and other factors in polling models. Conrad says: “There are so many choices in building these models that it is an art in a lot of ways. It all becomes mathematical because it’s implemented in a model. But somehow intuitions are quantified.” While recognizing a "couple sources of systematic error that seemed to push everybody off a bit" in predicting the presidential race, Pasek says that in general the polls were a good barometer. Both researchers, and the others cited in this piece, agree some post-mortem self-examination will help make the election polling field stronger.