All empirical data and the resulting parameters are subject to error. In this paper we explicitly use the standard, but hardly profound, “model” <disp-formula>$$\hbox{Observed Value} = \hbox{True Score} + \hbox{Error}$$</disp-formula> as a lens for better viewing empirical studies in search of empirical generalizations...

We develop a consumer response model to evaluate and plan pricing and promotions in durable-good markets. We discuss its implementation in the U.S. automotive industry, which “spends” about $45 billion each year in price promotions. The approach is based on a random effects multinomial...