Bayesian truth serum and its applications to conjoint analysis: a reliable way to assess user preferences for new products, services and policies – BayInno


Program: FP7 Marie Curie International Outgoing Fellowships
Commissioned by: Research Executive Agency, European Commission 
Project duration: April 1, 2015 - April 1, 2018
Project manager: Sonja Radas, PhD


On April 1, 2015, the Institute of Economics, Zagreb began with the implementation of the project "Bayesian Truth Serum and Its Applications to Conjoint Analysis: A Reliable Way to Assess User Preferences for New Products, Services and Policies (BayInno)", obtained as part of the FP7 Marie Curie International Outgoing Fellowships program. 

The project will run for three years and will be conducted by Dr. Sonja Radas from the Institute of Economics, Zagreb, as project leader, in cooperation with Dr. Dražen Prelec from the Massachusetts Institute of Technology (MIT), Cambridge, USA. 


Project goals: 
Many areas of economics use subjective data, gathered from respondents with traditional surveys as well as through new web-based information exchanges and markets. For example, prediction markets rely on data from participants who provide probabilistic estimates of market-related and other events. Although "hard" data such as unemployment figures, actual consumer spending, exports etc., are an essential input for economic forecasting, this is often supplemented by qualitative judgments and surveys.   

One solution to the problem of data truthfulness is presented by Bayesian Truth Serum (BTS) – a new methodology developed by professor Dražen Prelec from MIT and published in Science in 2004. By improving the quality and reliability of subjective data, the BTS method opens completely new possibilities for subjective judgment to be incorporated into economic research, most notably innovation research. Collecting data from future users and potential stakeholders is crucial to ensure successful product adoption.   

For this purpose there is an existing methodology within innovation research: this is conjoint analysis. Conjoint analysis has seen a variety of modifications to ease respondent load, or to improve accuracy of estimation. Some of the cutting edge versions of conjoint analysis such as polyhedral conjoint were developed at MIT. However, these sophisticated algorithms could not address the question of basic quality and truthfulness of the subjective data that is used as input for conjoint analysis. The power of BTS method lies in the fact that for the first time there is a tool which can successfully address this basic issue, and this represents a large improvement over the state of the art. 

This proposal intends to take a radically new approach to address reliability of conjoint data and consequently accuracy of estimation by taking the BTS methodology developed within the field of cognitive sciences and applying it to innovation research within economic sciences. This approach imparts a significant interdisciplinary character to this proposal. The research objectives of this project are twofold: to incorporate the standard BTS methodology into conjoint analysis, and to extend the newly developed methodology to model respondent heterogeneity by making necessary adaptations to the BTS method. The new BTS-adjusted methods will be completely new in the field of innovation research and will likely have large potential for application.

 

This project has received funding from thr European Union’s Seventh Framework Programme for Support for training and career development of researchers (Marie Curie) under rant agreement No 622868.
Vrh