* Corresponding author. Department of Economics, University of Southern California, Los Angeles, CA 90089-0253, USA. Journal of Econometrics 89 (1999) 15—39 Modeling survey response bias — with an analysis of the demand for an advanced electronic device Cheng Hsiao*, Bao-Hong Sun Department of Economics, University of Southern California, Los Angeles, CA 90089-0253, USA National Taiwan University, Taipei, Taiwan, ROC Graduate School of Industrial Administration, Carniege-Mellon University, Pittsburgh, PA15213, USA Abstract Survey data are often subject to a number of biases. In this article several models are proposed to detect and adjust for survey response bias. The approach is then applied to the analysis of a market survey data on the demand for an advanced electronic device in a developing country. Substantial differences in take rates and price elasticities are found between the estimates derived from the conventional random utility maximization framework and the biased response model. 1999 Elsevier Science S.A. All rights reserved. JEL classification: C25; C42; C52 Keywords: Qualitative choice; Survey methods; Response bias; Take rates 1. Introduction Sample survey is widely used by marketing researchers. However, both sample errors and nonsample errors could occur. While one may view sample errors as the unavoidable cost of using a segment of the population to project the characteristics of the whole population, the nonsample errors could adverse- ly affect the validity of the inference. Since non-sample errors are typically 0304-4076/99/$ — see front matter 1999 Elsevier Science S.A. All rights reserved. PII: S 0 3 0 4 - 4 0 7 6 ( 9 8 ) 0 0 0 5 3 - 0