The Decision Sciences Journal of Innovative Education

Why McNemar’s Procedure Needs to be Included

in the

Business Statistics Curriculum

Mark L. Berenson and Nicole B. Koppel

Department of Information and Decision Sciences, School of Business

Montclair State University, Upper Montclair, NJ 07043

e-mail: berensonm@mail.montclair.edu

e-mail: koppeln@mail.montclair.edu

ABSTRACT

 In business research situations it is often of interest to examine the differences in the responses in repeated measurements of the same subjects or from among matched or paired subjects.  A simple and useful procedure for comparing differences between proportions in two related samples was devised by McNemar (1947) nearly sixty years ago.  Although the technique has enjoyed widespread usage in both behavioral and medical research, it has unfortunately not received proper attention in marketing, advertising, consumer research, public relations, operations management, or organizational behavior endeavors – disciplines wherein it is particularly desirable to evaluate the significance of changes in attitudes, opinions, or beliefs regarding some campaign or strategy.  Perhaps the major reason for this obscurity is simply a general lack of awareness as to its value and practical power.  The McNemar test is generally not currently being taught to business statistics students.  It is the purpose of this paper to demonstrate the usefulness and applicability of the McNemar procedure to these aforementioned disciplines so that it may be more appropriately employed as an important confirmatory data analysis tool within the basic business statistics curriculum.

Subject Areas:            Data Analysis, Business Statistics, Marketing, Advertising, Consumer Research, Public Relations, Operations Management, Organizational Behavior, and Computer Spreadsheets and Packages