Heuristic Biases in an Introductory Statistics Course

• Activity Number: 518
• Type: Contributed
• Date/Time: Wednesday, August 6, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Education
• Author(s): Andrew Neath*+
• Companies: SIU Edwardsville
• Keywords: statistics education ; scientific reasoning ; fallacy of the transposed conditional ; representativeness ; base rates

Abstract

Our minds rely on heurisitic thinking and intuition, often with much success. However, a major bias in our heuristic thinking stems from an inability to properly account for the role that randomness plays in the world. It should be expected that formal statistical training would lead to a scientific approach free of such bias. I will argue that the traditional approach to teaching introductory statistics is, in fact, promoting methods based on the same illusions that corrupt our heuristic thinking. A consequence of this approach to teaching statistics is an issue currently faced in science where an unacceptably large proportion of "statistically" established findings fail upon attempts at replication.