The effect of the 1-in-X numerical format on choices

The 1-in-X numerical format (e.g., 1 in 200) has been found to increase subjective probability evaluations and behavioral intentions in hypothetical scenarios compared with the N-in-NX format (e.g., 5 in 1000). However, it remains unclear whether this format can also bias choices between truly incentivized options. In four online studies (N = 1039), participants were presented with a small endowment (i.e., 1£) and an actual choice between two options: a sure loss of a part of such endowment and a lottery with the chance to lose the entire endowment, presented using either the 1-in-X or the N-in-NX format. In Studies 1–3, where the two options were equivalent in expected monetary value (EV) and the lottery was described with varying degrees of concreteness, participants preferred the lottery option to a lesser extent when the chance of losing the endowment was presented using the 1-in-X format compared with the N-in-NX format. The same effect was replicated in Study 4a when the lottery option had a higher EV than the sure loss, showing that the 1-in-X effect can also lead individuals to deviate from maximizing EV. However, the effect vanished in Study 4b when the difference in EV between the two options increased. Implications for risk communication and a possible interpretation of the results are discussed accordingly.