“The Axiom of Research” and “The Axiom of Action”
By Tim Wilson on in Analysis with One Comment
I attended a one-day seminar today on “The Role of Statistical Concepts and Methods in Research” taught by Dr. Tom Bishop of The Ohio State University. Dr. Bishop heads up a pretty cool collaboration between Nationwide (all areas of the company, including Nationwide: Car Insurance) and OSU, and this seminar was one of the types of minor artifacts of that collaboration.
Dr. Bishop had me on page 5 of the seminar materials when he introduced “The Fundamental Axioms of Research,” which he stated are twofold:
- The Axiom of Variation — all research data used for inference and decision making are subject to uncertainty and variation
- The Axiom of Action — in research, theory is developed, experiments are conducted, and data are collected and analyzed to generate knowledge to form a rational basis for action
The rest of the seminar was a healthy mix of theory and application, with all of the “theory” being tied directly to how it should be applied correctly. Dr. Bishop is a statistician by training with years of industry experience, so it was pretty cool to hear him emphasize again and again and again that you can get a lot of value from the data without running all sorts of complex, Greek-letter-rich, statistical analyses. The key is to apply a high degree of rigor in understanding and articulating the problem and the approach.
Lots of good stuff!