|
Activity Number:
|
294
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Education
|
| Abstract - #308923 |
|
Title:
|
Comparing Two Proportions with the Binomial Distribution
|
|
Author(s):
|
James B. Orris*+
|
|
Companies:
|
Butler University
|
|
Address:
|
140 Holcomb Research Bldg., Indianapolis, IN, 46208,
|
|
Keywords:
|
hypothesis testing ; binomial distribution ; p-value ; two proportion test
|
|
Abstract:
|
The hypothesis testing module in most introductory statistics courses shows how to use the normal distribution to test a sample proportion vs. a hypothesized value. Some courses/textbooks also show how to get the exact p-value using the binomial distribution. This paper shows how students can extend this technique to compare two proportions by calculating a binomial distribution for each proportion and then calculating the joint probabilities from which the p-value can be determined. The method discussed in this paper is useful as a teaching technique because it reviews probability distributions and probability concepts. The calculations will be illustrated with an Excel spreadsheet that makes a good project for students. An Excel VBA program will also be presented that does the calculations and compares the results to the traditional normal curve method.
|
