Applying Some Computational Simplifications for Binomial and Poisson Random Variables to the Teaching of Statistical Concepts

• Activity Number: 294
• Type: Contributed
• Date/Time: Tuesday, July 31, 2007 : 10:30 AM to 12:20 PM
• Sponsor: Section on Teaching Statistics in the Health Sciences
• Author(s): Mark C. Fulcomer*+ and S. David Kriska and Maritza Jauregui and Marcia M. Sass and Jallah M. Kennedy
• Companies: Richard Stockton College of New Jersey and Restat Systems and The Richard Stockton College of New Jersey and University of Medicine and Dentistry of New Jersey and University of Medicine and Dentistry of New Jersey
• Keywords: teaching statistics ; binomial distributions ; Poisson distributions ; computational formulae

Abstract

Among the simplest forms of random variables and with seemingly vast potentials for "real" applications, binomial and Poisson distributions are discussed in many introductory textbooks on probability and statistical methods. However, despite initial appearances, the uses of both distributions in actual practice are often limited by computational difficulties as N increases. This presentation describes the incorporation of simplifying "recursive" formulae into two Excel spreadsheet programs that combine graphics with the ability to handle "large" problems. Utilizing textbook and other health-related examples in classroom settings, these programs have enhanced illustrations of several important and often challenging statistical concepts, including sampling distributions, hypothesis testing, the Central Limit Theorem, and normal approximations to both distributions.