One of the primary challenges facing philosophy of science today is to explain, if not to justify, scientific methodologies for learning about the world. Dr. Mayo is developing a model of experimental testing that resolves key questions in philosophy of statistics and answers the current challenge facing philosophers of science to demonstrate the rationality of science. Dr. Mayo's testing model is based on standard Neyman-Pearson (NP) statistics. Despite its widespread use in science, NP theory has been the center of disputes raised by nearly all philosophers of statistics and confirmation, and many scientists as well. Dr. Mayo aims to resolve these criticisms. She is setting out a framework in which statistical tests link data to hypotheses via a series of models. She shows how piecing together local statistical tests permits errors to be detected and corrected, and argues that, suitably reinterpreted, NP methods provide an adequate account of what permits us to learn from experiments. Dr. Mayo uses the resulting account of experimental inquiry to grapple with key obstacles to theory testing: theory ladeness of data, relativity to background assumptions, and underdetermination of theory by data. The challenge to overcome such obstacles, raised by Thomas Kuhn and many others, has influenced the thinking in many fields. The audience for this project includes philosophers, statisticians, scientists, historians, sociologists, and students of interdisciplinary studies in science and society.
