Use of Nonrounded Bids Triggers System 2 Full Information Processing in Discrete

An aspect of bid vector design in choice experiments that has received limited attention is the effect of using non-rounded (or precise) rather than rounded bids. Insights from the marketing and psychology literatures suggests that there are many factors potentially influencing the effects of using non-rounded bids, including image and level effects, accuracy cues, processing fluency, and stimulating more cognitively-based System 2 thinking. Using a split-sample experimental approach in a choice experiment, we find evidence that non-rounded bids may trigger use of System 2 cognitive processes, thereby increasing the number of significant attribute coefficients as well as reducing model variance.