A Framework for Measuring the Importance of Variables with Applications to Management Research and Decision Models*

Authors


  • *

    The authors acknowledge, with thanks, the helpful comments received from an associate editor on a previous draft of this paper. The authors also thank Market Probe, Inc. for providing the bank customer satisfaction data, a portion of which is used for the last example.

  • Ehsan S. Soofi is a profeso o statistics at th School of Business Aministration, University of Wisconsin-Milwauke. He received his BA in mathematics from UCLA, MA in statistics from the Univesity of California, Bekeley, and PhD in applied Statistics from the Univrsity of Califonia, Riveside, Dr. Soofi is a Fellow of the American Statistical Association. He is an associate editor of the Journal of the American Statistical Association sice 1991, and a vice president of the International Association for Statistical Computing for 1999–2001. His reseach interest is in the areas of information-theoretic, Bayesian, computational, and graphical statistics, and their application in management research and decision problems. His aticles have appeared in the Journal of the American Statistical Association, Journal of the Royal Statistical Society, Biomtrika, Journal of Econometrics, Operations Research, and Marketing Science.

  • Joseph J Retzer is the Director of Marketing Sciences, Telecom Research Group, for Maritz Marketing Research Inc. (MMRI), Oak Brook, IL. Prior to joining MMRI, he was director of Product Development at Market Probe, Market Research, Inc. He holds a BA and MA Degree in economics, and a PhD in management science from the University of Wisconsin-Milwaukee. He has taught statistics, Economics, and management science at the University of Wisconsin-Milwaukee at both the graduate and undergraduate levels. His Research interests include statistical and econometric modelig of marketig models in both classical and Bayesian frameworks.

  • Masoud Yasai-Ardekani is a professor of strategic management at the School of Business Administration, University of Wisconsin-Milwaukee. He holds a BS in electrical engineering from the Imperial College of Science and Tecnology, University of London, an MS in adminitrative sciences, and a PhD in management studies from the Graduate Business Center, the City University, London. His research focuses on designs for strategic planning processes and their effectiveness, performance implications of strategy-environment alignments, strategic and structural responses to environments, and management of Strategic change and turnaround. He has published his research in journals such as Academy of Management Journal, Academy of Management Review. Strategic Management Journal, IEEE Transactions on Engineerig Management, MIS Quarterly, and Management Science.

Abstract

In many disciplines, including various management science fields, researchers have shown interest in assigning relative importance weights to a set of explanatory variables in multivariable statistical analysis. This paper provides a synthesis of the relative importance measures scattered in the statistics, psychometrics, and management science literature. These measures are computed by averaging the partial contributions of each variable over all orderings of the explanatory variables. We define an Analysis of Importance (ANIMP) framework that reflects two desirable properties for the relative importance measures discussed in the literature: additive separability and order independence. We also provide a formal justification and generalization of the “averaging over all orderings” procedure based on the Maximum Entropy Principle. We then examine the question of relative importance in management research within the framework of the “contingency theory of organizational design” and provide an example of the use of relative importance measures in an actual management decision situation. Contrasts are drawn between the consequences of use of statistical significance, which is an inappropriate indicator of relative importance and the results of the appropriate ANIMP measures.

Ancillary