Teaching Hypotheses-Testing Concepts with the Binomial Test and a Pepsi-Coke Challenge

• 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): Phyllis Curtiss*+ and John Gabrosek and Kirk Anderson
• Companies: Grand Valley State University and Grand Valley State University and Grand Valley State University
• Keywords: binomial test ; Type 1 and 2 errors ; hypotheses testing ; Pepsi-Coke ; Activity

Abstract

In introductory statistics or nonparametric statistics, the binomial test is a great starting point to discuss important hypotheses testing concepts such as p-values, rejection regions, Type I and Type II error and power. Students often have difficulty wrapping their minds around these concepts and how they are affected by other decisions made in the test. Using an interesting example to which students can relate has the potential to engage students and serve as a bridge activity to more technical discussions of these concepts.

R. A. Fisher's famous tea-tasting experiment can be modified to explore hypothesis testing concepts using the binomial test. Can the lady really tell the difference when tea is poured into milk as compared to when milk is poured into tea? Updating this idea to current times, Levine and Rolwing (1993) discussed using a Pepsi-Coke challenge to introduce hypothesis testing concepts. With these ideas as a basis, we developed a class activity that will help students understand the concepts of hypotheses testing and help them apply them to other cases. Extensions will be discussed on changing the experiment for use with other distributions such as the hyper