[We’re pleased to welcome Jesper N. Wulff of Aarhus University in Denmark. Dr. Wulff recently published an article in the OnlineFirst section of Organizational Research Methods entitled “Interpreting Results From the Multinomial Logit Model: Demonstrated by Foreign Market Entry.”]
My article was motivated by the way researchers interpret results from discrete choice models with multiple outcomes in strategy and management. Previous research has already noted that scholars are rarely aware of the set of different challenges faced when using such models compared to the binary dependent variable case. One very critical difference is that the sign of the estimated model coefficients does not determine the direction of the relationship between a covariate and the probability of choosing a specific alternative. My literature review of the use of the multinomial logit model (MLM) in a subfield of strategy, foreign market entry, revealed that interpretation is often limited to coefficients and single summary measures.
In the article, I proposed that researchers base their interpretations of the results from the MLM less on coefficients and single summary measures, and more on complete graphical representations. I provided practical guidelines about what researchers concretely may do and used a practical example to illustrate the gains and losses of using a graphical approach. If we are interested in inferring the true nature of the relationship between a predictor and the dependent variable in an MLM, we must acknowledge that coefficients and sometimes even sometimes single summary measures are potentially misleading. Instead, I proposed that by shifting to full graphical representations we leave little in the dark and provide our readers with full and relatively intuitive information about the intrinsically non-linear and complex nature of the relationship.
I hope that the guidelines I present will result in richer and more complete interpretations of results from discrete models with multiple outcomes unleashing some of the valuable information that is too rarely used in organizational and social science research.
You can read “Interpreting Results From the Multinomial Logit Model: Demonstrated by Foreign Market Entry” from Organizational Research Methods for free by clicking here. Want to know about all the latest research like this from Organizational Research Methods? Click here to sign up for e-alerts!
Jesper N. Wulff is an assistant professor of management science at the Department of Economics and Business, Aarhus University. His research interests include discrete choice modeling, applied econometrics, and foreign market entry. He has delivered presentations at and acted as a reviewer for the Annual Academy of Management meetings.