Experimentally Validated Survey Evidence on Individual Risk Attitudes in Rural Thailand

This study validates a survey-based measure of general risk attitude by an incentive compatible experiment among more than 900 participants in rural Thailand. The survey measure of self-assessed risk attitude provides a useful approximation of the experimentally derived risk attitude. This holds when we add various socio-demographic control variables to the survey-experiment-relation which are available from the representative household survey and which are related to risk attitude in plausible ways. The survey measure also predicts individual behavior towards risk in other cases; the survey measure even outperforms the experimental measure in this respect.