Fig 1.
Each lottery ticket had a price tag; 9.00 Token in this example. Purchasing a lottery ticket had the following two results: either the lottery ended head, then the payment was 10.00–9.00 + 17.14 = 18.14, or the lottery ended tail, and the payment was 10.00–9.00 + 0.22 = 1.22. Participants could indicate their decision by clicking on one of the two buttons below the text. All lottery tickets were presented to participants sequentially in randomized order. Only one decision was paid out.
Table 1.
Overview of the 18 lotteries-of-interest.
Table 2.
Mixed-effects regressions.
Fig 2.
The main experiment results do not indicate a difference in purchasing rates (y-axis) for the three price-ending treatments.
A: purchasing rates across both price levels. B: purchasing rates for the low price level. C: purchasing rates for the high price level. CTRL = control prices, JB = just-below prices, RND = rounded prices.
Fig 3.
Purchasing rates across the exploratory variables.
A: purchasing rates divided per numerical ability–i.e., the number of questions the participant correctly answered during the incentivized part of the questionnaire. B: purchasing rates subdivided by highest obtained educational degree. C: purchasing rates subdivided by participant’s age. D: purchasing rates by gender (the categories “other” and “do not wish to say” are not depicted due to low sample sizes). CTRL = control prices, JB = just-below prices, RND = rounded prices.
Fig 4.
Results of the posthoc power analysis.
Each dot represents the percentage of correctly detected effects (i.e., statistical power) across 2500 simulated experiments. Each simulated experiment contained 266 fictive participants who had a 45.3% probability of purchasing the control ticket and a variable probability of purchasing the experimental ticket. The solid black line represents a best-fit curve through these points.