Predicted vs. Stated Preference: Application to Conjoint Analysis in Innovation Development

Commissioned by: The Institute of Economics, Zagreb
Project duration: January 29, 2019 – January 31, 2020
Project manager: Sonja Radas

Many areas of economics use subjective data gathered from respondents via traditional surveys as well as through new web-based information exchanges and markets. One of the areas that rely heavily on such data is innovation development. During the development process, input is elicited from potential users with the purpose to generate adoption forecasts and help choose optimal final specifications for the product. Traditionally all the data elicitations required individuals to report their personal choices and preferences, but this proposal introduced a new way to use subjective data in innovation research by asking individuals to predict choices and preferences of others. We embedded this new approach in conjoint analysis, a very popular methodology in innovation research. We postulated that the new approach would improve data reliability and forecasting accuracy. The new methodology relied on Bayesian Truth Serum.

RESEARCH AREA: Innovation, research and development