Online market research surveys.

Entrepreneurs often rely on choice experiments to discover the needs and desires of their customers. These surveys are conducted more and more online because large cost reductions can be achieved this way. In online surveys, the answers given can be processed immediately and can therefore be used to optimize the next choice set to be shown to a specific respondent. As such the heterogeneity in the market can be described more precisely. In this project, such a sequential procedure will be optimized for the different statistical models that are used nowadays to describe market heterogeneity. Another advantage of online surveys is that one can collect some covariate information (such as age and gender) before generating choice sets. In this project, we will use this covariate information to improve the prior distributions that are needed to generate the choice sets. We also hope to counter the self selection bias using these covariates. As less control can be exercised on the sample selection, some market segments will be over- or underrepresented. By using covariate information at the design and at the analysis stage, we hope to diminish this selection bias.

Keywords:Design of experiments, Market research, Discrete choice model, Mixed logit model, Latent class model, Online surveys

Disciplines:Applied mathematics in specific fields, Statistics and numerical methods, Applied economics, Economic history, Macroeconomics and monetary economics, Microeconomics, Tourism, Marketing