Saturday, January 19, 2008

Incorrect Pundit Predictions Underscore Hypothetical Nature of Empirical Statements

The dramatic difference between the results of the New Hampshire primary (1) and the projections made by pundits before the voting commenced (2, 3) underscores the hypothetical nature of all empirical statements. Empirical statements, including those made by pre-vote polling, cannot express necessary truths, only probabilistic truths. Ayer notes that “no proposition, other than a tautology, can possibly be anything more than a probable hypothesis.”(4) This is because, for any situation, there will always be some possible experience that will lead us to believe that our original conclusion was mistaken. Put more simply: we can never be sure that an empirical assertion is one-hundred percent accurate as the truth of such an assertion depends on facts that could be otherwise. Anywhere short of expressing necessary truths, empirical statements can be highly probable but may never be certain.

Such was the case with the NH polling data. Regardless of how many polls corroborated the prevailing belief that Hillary Clinton would lose by double-digits, the evidence bore no necessary relation to the eventual, quite opposite outcome of events (strictly speaking, this is a problem with the inference from the evidence, and not the veracity of the evidence).

While it is unlikely that the contrary to this principle was being employed before the NH primary (i.e., a belief that empirical statements expressed a necessary truth), acknowledgment of the probabilistic nature of empirical statements may have implications for, among other things, the decisions on which candidates to include in debates. The exclusion from a debate of such candidates whose polling support does not meet the arbitrary standards for inclusion -- such as Fox News’ decision to exclude Ron Paul (5) and NBC’s decision to exclude Dennis Kucinich (6, 7)—is less justified if the network’s sole criterion for inclusion is the candidate’s popularity in the polls.

(1) Healy, Patrick and Michael Cooper. “Clinton Stuns Obama; McCain Wins.” New York Times. Janurary 8, 2008.

(2) Frankovich, Kathy. “N.H. Polls: What Went Wrong?” CBS News. January 14, 2008.

(3) Howell, Deborah. “Pollsters in the Primary Storms.” Washington Post. January 20, 2008.

(4) Ayer, A.J. Language, Truth, and Logic. New York: Dover. 1952. P. 38.

(5) Catchpole, Dan. “Paul left out in the cold in NH.” Seattle Post-Intelligencer Blog. January 4, 2008.

(6) Hasselback, Benjamin. “NBC Tries to Pick the Next President.” Associated Content. January 15, 2008.

(7) Harris, Paul. “Let Them be Heard.” The Guardian. January 17, 2008.

No comments:

Creative Commons License
Analytic Politics is licensed under a Creative Commons Attribution-Noncommercial 3.0 United States License.