“The Unlikelihood Effect: When Knowing More Creates the Perception of Less”, 2022-10-13 (; backlinks):
People face increasingly detailed information related to a range of risky decisions. To aid individuals in thinking through such risks, various forms of policy and health messaging often enumerate their causes.
Whereas some prior literature suggests that adding information about causes of an outcome increases its perceived likelihood, we identify a novel mechanism through which the opposite regularly occurs.
Across 7 primary and 6 supplementary experiments, we find that the estimated likelihood of an outcome decreases [subadditivity] when people learn about the (by definition lower) probabilities of the pathways that lead to that outcome.
This unlikelihood bias exists despite explicit communication of the outcome’s total objective probability and occurs for both positive and negative outcomes. Indeed, awareness of a low-probability pathway decreases subjective perceptions of the outcome’s likelihood even when its addition objectively increases the outcome’s actual probability.
These findings advance the current understanding of how people integrate information under uncertainty and derive subjective perceptions of risk.
[Keywords: health decisions, information processing, probability judgments, risk, uncertainty]
…People are expected to make increasingly complex decisions for themselves in consequential domains with uncertain outcomes. The coronavirus pandemic highlighted the personal and social importance of estimating health risks in particular, but also the challenges and aversiveness of doing so. Details regarding the vectors that cause such risks are increasingly available and salient, potentially reflecting a growing desire for information in an effort to manage uncertainty and accurately estimate the total risk. Substantial research including the support theory literature had previously shown that adding pathway information caused people to “build up” their risk estimate (and thus perceive a higher total risk). This research was motivated by the question of whether adding the individual numerical probabilities of those pathways would cause people to switch into “breaking down” the objective total probability of a risk, decreasing their subjectively estimated likelihood. Our work examines how people integrate information through multiple mental processes to yield subjective perceptions of risk that govern complex real-world choices.