Department of Statistics Seminar
North Carolina State University

presents

Deborah G. Mayo

Department of Philosophy, Virginia Tech

"Statistical Controversies and the Philosophy of Science"

ABSTRACT

Despite the widespread use of statistical methods based on error probabilities (e.g., Neyman-Pearson tests and confidence intervals) these methods have been the subject of enormous criticism, giving rise to the popular subjective Bayesian Way in the philosophy of science. Given the new emphasis philosophers of science have placed on actual scientific practice, this disregard, if not outright condemnation, of procedures that are widely and successfully used in science is out of place. An adequate philosophy of science will only emerge if it is not at odds with the use of statistics in scientific practice. I propose reinterpreting standard statistical tests as tools for obtaining experimental knowledge. In my account of testing, e is evidence for a hypothesis H to the extent that H passes a severe test with e. The familiar statistical hypotheses (e.g., null hypotheses) serve to ask questions about the presence of key errors: mistaking real effects for chance, or mistakes about parameter values, causes, and experimental assumptions. An experimental result is a good indication that an error is absent if there is a very high probability that the error would have been detected if it existed, and yet it was not detected. Test results provide a good (poor) indication of a hypothesis H to the extent that H passes a test with high (low) severity.

Friday, March 6, 1998

3:35 - 4:35 pm

206 Cox Hall

Refreshments will be served on the second floor of Dabney Hall (left of Room 222) at 3:00 pm.