The file Data.csv has 5 columns, providing information about the subject, . Each row corresponds to one task completed by one subject. Hence there are 2730 rows (78 subjects * 35 CE or PE tasks).
We describe here each column:
- SubjectID: An arbitrary number identifying the subject. The eight subjects excluded from our analysis as explained in the first paragraph of Section 3 are subjects 4, 42, 45, 55, 60, 63, 65, and 70. 
- Task_Type: There were 3 parts in the experiment. Task_Type = 1 refers to the CE tasks described in section 2.1. Task_Type = 2 refers to the PE tasks on Ellsberg urns described in section 2.2. Task_Type = 3 refers to PE tasks on the Dutch stock index (see footnote 8 and supplementary material).
- Task_number: For Task_Type 1, Task_number gives the lottery number as in the first column of Table 2 (l_i with i from 1 to 11). For Task_Types 2 and 3, Task_number gives the act number as in the first column of Table 3 (f_i with i from 1 to 12).
- Order: Parts (=all tasks with same Task_Type) were randomized and tasks were randomized within parts. Order gives when the task was displayed. It is a number from 1 to 35. E.g.: Order = 27 means it was the twenty-seventh task the subject completed in the experiment.
- PE_CE: PE_CE gives the CE (for Task_Type 1) or the PE (for Task_Type 2 and 3). CEs are in euros, PEs in %.
