RUI: Collaborative Research: Assessments and Stances Regarding the Uncertainty of (Un)Desired Outcomes

Appalachian State University, Boone NC

Investigators

Abstract

Inaccuracies in expectations and predictions about future outcomes can ruin decision making and optimal planning. One pernicious type of inaccuracy is an optimism bias, in which preferred outcomes are viewed as more likely than warranted and/or unpreferred outcomes are viewed as less likely than warranted. Optimism biases lead to problematic individual behaviors like unwise risk taking and over-purchasing, human tragedies like deaths from engineering failures and societally important phenomena like stock market bubbles. Motivational goals are often assumed to underlie optimism biases but there is an inadequate understanding of when different goals impact optimism, including goals that have the potential to reduce optimism and promote accuracy. In fact, major theories assume that preferences for outcomes and expectations about the outcomes are independent, which is inconsistent with some empirical findings. The proposed research addresses this discrepancy. The research is structured by a novel Goal-Biased Expectations Framework, which posits contextual features that shape the balance between two orientations that a person can take toward forming expectations. The balance of orientations ultimately determines whether and how goals shape expectations. The framework is also useful for developing steps to debias or reduce overoptimism and promote accuracy among professional forecasters and the public. Debiasing overoptimism can yield improved predictions and ultimately better planning, preparation, and outcomes across many contexts (e.g., financial, health, security). Three lines of experiments will be conducted. One line will examine the scope of, and processes responsible for, a crucial difference in how measures of expectations are affected by outcome desirability or preference. Past research findings have returned quite different indicators of the extent to which outcome desirability has a biasing influence, but these differences might align with a distinction between two ways in which expectations are commonly solicited in studies and in everyday contexts. A second line will address how types of available information influence the desirability bias. One surprising possibility is that as information becomes more clear, people's orientation and goals shift in a manner that produces more bias rather than less. A third line will test a novel debiasing intervention for overoptimism. Participants in the studies will include sports fans, owners of homes in floodplains, and patients?three groups known to exhibit overoptimism. All of the proposed work will inform the validity of the Goal-Biased Expectations framework. The studies on this grant will involve collaboration with graduate and undergraduate students, and therefore will play important roles in their research training/education. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.