Relative state and decision making under risk

Abstract

The recently developed (and mathematically formalized) relative state model of risk-taking suggests that individuals are motivated by two non-independent pathways to risk-taking: a need-based pathway, wherein risk-taking is the preferable course of action because a non-risky decision fails to meet an individual’s needs; and an ability-based pathway, wherein risk-taking is the preferable course of action because individuals are in such a good state that taking risks offers a higher expected value than non-risky alternatives. I tested the predictive value of the relative state model using a modified version of the Ecological Decision Task (ECO Task; Mishra & Lalumière, 2010), which asked participants to “forage” apples from trees which differ in the outcome variance associated with the outcomes, such that the yellow tree provides 10 or 11 apples each time, whereas the blue tree provides between four and 17 apples. Participants chose from the two trees in hopes of meeting some need-threshold that ranged from needing four apples to needing 31 apples. I separated participants into three conditions: a poor-state condition where participants needed many apples to meet their need; a moderate-state condition where participants needed a moderate number of apples to meet their need; and a good-state condition where participants needed a small number of apples to meet their need. I predicted that those in the poor-state and good-state conditions would choose the riskier (blue) tree more often than those in the moderate-state condition. I tested the results using a Kruskal-Wallis test (N = 293) and found support for the need-based (Obs. Diff. = 74.48, Crit. Diff. = 29.20), but not ability-based (Obs. Diff. = 5.41, Crit. Diff. = 29.06) pathway. Exploratory analyses elucidated many causal, decision-level variables, while also highlighting limitations with the study of risk-taking. I discuss implications for the relative state model, future research using the experimental task from the current study, and risk-taking research in the discussion.