THINKING THROUGH UNCERTAINTY: THE EFFECT OF RANDOMIZED AIDS ON THE DISJUNCTION EFFECT

Abstract

This study presents the first exploration of whether randomized aids could alleviate failures in contingent reasoning, specifically the disjunction effect observed by Shafir and Tversky (1992) in one-shot prisoner’s dilemma. I conduct a randomized controlled experiment whereby treatment subjects consider a random suggestion on what action they should choose - to defect or to cooperate - prior to making their decision in every game. The results support the hypothesis that treatment helps individuals to think through uncertainty; namely, while treatment had a negligible effect when the other player’s strategy was known, the effect of treatment became more significant only when the other player’s strategy was unknown, reducing cooperation by 9 percentage points relative to the control group. Consequently, the disjunction effect was attenuated in the treatment group. The results suggest that the randomized aid facilitates some form of contingent reasoning, possibly by simplifying the complexity of contingent thinking or prompting individuals to reason consequentially.