Prediction in a Non-Partisan Election

Abstract
A recent municipal election in Denver gave scientific forecasters two headaches: how could polling yield a valid prediction when most of the potential candidates were unknown to the electorate; and what effect would party labels have in a supposedly non-partisan election? This article describes the manner in which these problems were overcome, and also suggests the application of similar techniques to related fields.

This publication has 0 references indexed in Scilit: