Probabilistic Polling
Election polls elicit from individuals forecasts of their prospective voting behavior. Standard verbal polling questions do not enable respondents to adequately express uncertainty about whether they will vote and, if so, for whom. This paper puts forward the idea of probabilistic polling, which elicits from respondents numerical subjective probabilities of voting. The probability scale enables respondents to fully express uncertainty about their future voting behavior and enables analysts to aggregate voting expectations into predictions of election outcomes. Empirical experience with probabilistic elicitation of many types of expectations suggests that probabilistic polling is practical.
Authors: | Manski, Charles F. |
---|---|
Institutions: | Institute for Policy Research (IPR), Northwestern University |
Saved in:
Saved in favorites
Similar items by person
-
Manski, Charles F.,
-
Manski, Charles F.,
-
Manski, Charles F., (2010)
- More ...