Analyzing Learning Effects in the Newsvendor Model by Probabilistic Methods
20 Pages Posted: 11 Oct 2019
Date Written: October 11, 2019
Abstract
In this paper, we use probabilistic methods to analyze learning effects in a behavioral experiment on the newsvendor model. We argue why we should believe that suggested orders follow a multinomial logit distribution, and use the single parameter in that model to extract information on learning effects. We revisit the data, analyzed previously by Bolton et al. (2012), and show that our model predicts the pull-to-center effect in these experimental data very well.
Keywords: behavioral OR, experimental economics, bounded rationality, probabilistic methods, pull-to-center effect
JEL Classification: C00, C70
Suggested Citation: Suggested Citation