Abstract:
It is well known that preferences are not invariant across elicitation processes or contexts, suggesting that there may be systematic differences between choice and judgement tasks (Moore 2004). Thus, when predicting market shares or choices via choice simulators, such differences should be taken into consideration. Surprisingly, current choice simulators based on judgement tasks ignore them, even though the selection of choice simulator has a huge impact on the validity of predictions based on judgement tasks (Hartmann and Sattler 2004).
