Correspondence: Randall C. Campbell, Department of Finance and Economics, Mississippi State University, Mississippi State, MS 39762, USA. Tel: +1 662 325 1516; Fax: +1 662 325 1977; E-mail: rcampbell@cobilan.msstate.edu

¹is an assistant professor of economics at Mississippi State University. He obtained his PhD in Economics at Louisiana State University in 1999 and his primary research and teaching interests are in applied econometrics. He previously worked as an economist in the market research department at Southwestern Bell, and has published articles in both marketing and economics journals.

Received 4 January 2008; Revised 4 January 2008; Published online 10 March 2008.

Abstract

We conduct a set of Monte Carlo sampling experiments to examine the use of discrete explanatory variables in market research for new products. We compare three alternative estimators on the basis of mean square error. As part of the experimental design we vary the number of price points and the distribution of the error term. In addition, we consider demand functions that are linear, linear with structural breaks and nonlinear. In each case, our results show that using two discrete price points leads to the lowest estimation risk.