Framing effect in evaluation of others' predictions

This paper explored how frames influence people's evaluation of others' probabilistic predictions in light of the outcomes of binary events. Most probabilistic predictions (e.g., ``there is a 75% chance that Denver will win the Super Bowl'') can be partitioned into two components: A qualitative component that describes the predicted outcome (``Denver will win the Super Bowl''), and a quantitative component that represents the chance of the outcome occurring (``75% chance''). Various logically equivalent variations of a single prediction can be created through different combinations of these components and their logical or numerical complements (e.g., ``25% chance that Denver will lose the Super Bowl'', ``75% chance that Seattle will lose the Super Bowl''). Based on the outcome of the predicted event, these logically equivalent predictions can be categorized into two classes: Congruently framed predictions, in which the qualitative component matches the outcome, and incongruently framed predictions, in which it does not. Although the two classes of predictions are logically equivalent, we hypothesize that people would judge congruently framed predictions to be more accurate. The paper tested this hypothesis in seven experiments and found supporting evidence across a number of domains and experimental manipulations, and even when the congruently framed prediction was logically inferior. It also found that this effect held even for subjects who saw both congruently framed and incongruently framed versions of a prediction and judged the two to be logically equivalent.