Prior Information in Frequentist Research Designs and Social (Non-epistemic) Influences: The Case of Neyman’s Sampling Theory
Year: 2021
Type: preprint
Abstract: We analyze the issue of using prior information in frequentist statistical inference by bringing out the sampling theory of Jerzy Neyman (a key figure in frequentist statistics), which has so far been largely ignored in philosophical discussions on frequentism. Our scrutiny of the different kinds of sampling designs supported by Neyman’s theory reveals a variety of ways to explicitly, objectively engage with prior information. We argue that Neyman’s approach to sampling enables researchers to let values classically classified as non-epistemic influence the procedure of collecting evidence and formulating statistical conclusions in order to not compromise the epistemic reliability of a procedure and even to improve upon procedures. With this, the discussed solutions of Neyman pose a methodological argument against the distinguishing of epistemic and non-epistemic values, and against the value-free ideal of scientific inference.
Authors: Adam P. Kubiak, Paweł Kawalec
Language: English
Subfield: History and Philosophy of Science
Field: Arts and Humanities
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W3146727367
