The Use of Statistical Methods in Management Research: Suggestions from a Case Study

Abstract

This article uses a case study of the methods used in a research paper in the Academy of Management Journal to present a critique, on two levels, of the way statistical methods are used in management research. First, I argue that the way statistical analysis is typically presented is unnecessarily obscure, and suffers from an over-emphasis on null hypothesis significance testing. Some practical suggestions are made for tackling these problems. These include an easy way to use p values to derive more user-friendly confidence levels, and a graphically presented bootstrap resampling method (implemented on a spreadsheet) to show the conclusions which might have followed from different samples from the same source – this leads to a way of assessing confidence levels for any hypothesis about a relationship between variables. Second, the article analyses the value and limitations of adopting a statistical approach in a given management research project. These arguments lead to a series of recommendations for the sensible use of statistical methods for researching management issues.

Keywords: Management research, Statistical methods, Null hypothesis significance tests, Bootstrap resampling, Confidence intervals