The experiment consists of five treatments, which allow for testing the effect of the experimental setup. The main treatment will be described here. The difference between the treatments is described below in the clustering description.
Subjects face questions on a simple lottery (50% chance of gaining 900 points and 50% chance of gaining 100 points), which is similar to the lottery used in Chapman et al. (2019), and a complex lottery (20% chance of gaining 975 points, a 30% change of gaining 850 points, a 30% chance of gaining 150 points, and a 20% chance of gaining 25 points). As a control subjects are also asked to answer the question for a 'sure' lottery (100% chance of gaining 500 points).
Subjects go through 3 main sections. First of all, the valuation uncertainty elicitation. The method is based on a standard multiple price list (MPL), in which subjects make a choice between a lottery and a sequence of certain amounts (Holt & Laury, 2002). We extend the method by introducing a third option: the option to receive the lottery or the certain amount randomly. The sure amounts for which subjects choose the option of random assignment define valuation uncertainty interval. The lower bound of the interval is defined by the maximum certain amount at which subjects choose the lottery only, and the upper bound is defined by the minimum certain amount at which subjects choose the certain amount only. These two bounds can serve as “the extremes that define an incomplete preference relation, connecting with the existing approach to modelling the endowment effect that does not rely on loss aversion” (Chapman et al., 2019, p. 37).
Since going through the whole list is tedious, we facilitate subjects’ decision making by asking them to indicate only their switching points from the lottery to the random assignment, and from the random assignment to the certain amount, their preferred choices will be automatically be filled in.
The second and third sections of the experiment will elicit the subjects' Willingness To Accept and Willingness To Pay for each lottery according to the standard approach (Chapman et al., 2019). In these measures subjects will again face a MPL, but this time with the traditional two options. The questions are framed as selling (WTA) and buying (WTP) the lottery for a certain amount. Again, subjects only need to select their switching point, and the table will fill automatically. Finally, in between the different sections subjects will be asked to answer some questions about their cognitive ability (IQ questions), and the desirability of control scale (Burger & Cooper, 1979; Gebhardt & Brosschot, 2002).