Taking into account the strength of an alternative hypothesis.

A common phenomenon in judgment under uncertainty is that alternative hypotheses are underweighted or ignored. This article addresses when and how the strength of the alternative is taken into account when there are 2 hypotheses. A learning manipulation was used to invoke 2 representations of 2 illnesses in a medical diagnosis task. One representation tended to lead to consideration of the alternative when, for example, requesting new information, reporting confidence, and making diagnoses. The other representation tended to result in ignoring or underweighting the alternative, but a simple change in how confidence was probed increased consideration of the alternative. Costs and benefits of each representation are discussed. Implicitly or explicitly, people are constantly assessing the likelihood of hypotheses, or unknown states of the world. Examples include estimating how likely it is that your favorite team will win its next game, that it will rain tomorrow, that a defendant is guilty, or that a new acquaintance is thoughtful. Every hypothesis has at least one alternative. For instance, when assessing your team's chances of winning, the alternative hypothesis is that the opposing team will win. If the competing hypotheses are mutually exclusive and exhaustive (i.e., exactly one of the hypotheses is true), confidence in the truth of one hypothesis should be influenced by the relative strength(s) of the other(s). Thus, when asked about the likelihood of your team winning, you should consider the quality of both teams. Ignoring the strength of the opposing team is potentially problematic because the stronger you think the opposing team is, the lower your confidence should be in your team.