Does the philosophy of knowledge need statistics?

Michael Hardin encourages attendees to examine the history, development and philosophy underlying the analytic process

If a SAS Global Forum presentation doesn’t include at least one line of code, is it still a presentation?

Absolutely. Just ask those who donned their thinking caps to hear J. Michael Hardin, Dean and Professor of Statistics from Culverhouse College of Commerce and Business Administration at the University of Alabama. Hardin offered fair warning that his 90-minute talk and slide deck included no code, only one equation and more pictures than words.

“I want to challenge you to think more deeply about some of the underlying assumptions that are taken for granted when we analyze data,” he said.

Hardin referenced the perspectives of early and late-day philosophers and scientists, including Aristotle, Plato, Galileo, Newton, Pearson, Einstein and Mulaik, to help audience members better understand some of the confusion surrounding big data and business analytics. “I feel that statistics needs philosophical thinking rather desperately, and the philosophy of knowledge needs statistics,” he said.

Highlights and more zinger quotes from Hardin's presentation follow:

  • Years ago, a statistician might have claimed that statistics deals with the processing of data. Today’s statistician is more likely to say that statistics is concerned with decision making in the face of uncertainty.
  • The emphasis today is on prediction, not as much on estimation. In this new world, hypothesis may not be needed.
  • If you have a model that explains everything, have you gotten to the truth? Hardin highlighted the perspectives of Plato vs. Aristotle and the existence of chairs vs. “chairness” to illustrate the challenges faced in considering various forms of data modeling and decision-making.
  • Hardin quoted American analytic philosopher Charles Stevenson: “One would not expect a book on scientific method to do the work of science itself…The purpose of an analytic or methodological study is always indirect. It hopes to send others to their task with clearer heads and less wasteful habits of investigation. This necessitates a continual scrutiny of what these others are doing, or else analysis of meanings proceeds in a vacuum.”
  • No philosophical theory of science for statistics is without its problems or criticisms.

For an enlightening view of statistical thinking, be sure to view the archive video of Hardin’s presentation.


About Author

Becky Graebe

Director, Communications

In addition to traditional employee communication efforts at SAS, Becky Graebe oversees an award-winning global intranet and a variety of enterprise social media channels. Her goal is to create a working environment where SAS employees around the world feel connected and inspired to share fresh ideas, solutions and expertise with colleagues and customers. Having studied at Southern Methodist University and earned her degree from Stetson University, she now serves on the Employee Communications Section board for the National Public Relations Society of America, is an active member of Triangle Women in Communications, and volunteers with Citizen Schools and the Wake County Support Circle Program.

Comments are closed.

Back to Top