 |
In this paper, we carry out induced value experiments that recreates the salient features of Discrete Choice Experiments (DCE) in an experimental economics laboratory setting. We first show that non payoff maximizing choices in a simple induced value DCE are numerous. We then study the potential of several simple devices to improve choices. Only a truth telling oath is effective in our setting whereas traditional devices such as monetary incentives or helping respondents (a calculator) have little or no effect on choices. The analysis of response time together with the implementation of the two oath procedures that target effort rather than truth-telling tell us that truth-telling in DCE is a serious concern. We suggest that this concern should be addressed in a systematic way in field surveys, by asking respondents to take an oath prior to being interviewed as we have done in this paper or by implementing weaker forms of commitment like a preliminary pledge or even a simple signed agreement to tell the truth.
