When people are committed to one side in an athletic contest, say an American versus a Russian figure skater in the Olympics, we frequently observe a biased interpretation of the events. The opponent's small misstep is judged as fatal to her performance, while the same error by the athlete from one's own country is dismissed as insignificant. Could this distortion of physical evidence occur not only when a preference for one side is long standing and deeply held, but also when it is tentative, private, and fully reversible? The research undertaken in the present grant demonstrates that the same distortion that drives biased judgments in the case of strong beliefs seems to lead to substantial and unrecognized distortion of evidence in the case of newly developing preferences.
Researchers at Cornell University are examining this phenomenon in the ordinary situation of purchase decisions. They have labeled it "predecisional distortion," meaning the biased evaluation of new information to favor whichever product is currently (and tentatively) preferred while the decision is still in progress. In a consumer choice, as soon as one brand is slightly favored, new product information tends to be distorted to seem more favorable to that tentatively leading brand than it should. This, in turn, makes the currently leading option more likely to become the decision-maker's final choice. The phenomenon extends beyond consumer choice, having been observed also in managerial decisions (both by public auditors and by salespersons), and in the verdicts of mock jurors. Predecisional distortion of new information is also systematic, with distortion increasing as the commitment to the tentatively leading alternative increases. Interestingly, individuals who are in a good mood are more subject to this bias.
What causes and, therefore, makes individuals prone to this bias in situations where they are genuinely trying to think accurately? The researchers propose that predecisional distortion is driven by the simple expectation that new information should be consistent with existing beliefs, combined with the lack of awareness that distortion is occurring. That is, people expect coherence between new and old information, so much so that they actually distort new information to make it more "coherent" with existing beliefs.
The Cornell researchers have investigated a number of methods for ameliorating the distortion of information, including accountability to others for the correctness of the decision, increasing the importance of the choice, and providing financial incentives for accuracy. While these interventions sometimes help, none has reduced distortion by more than 50%. So far, it has proved impossible to eradicate, although additional methods to eliminate it are being explored. However, one major barrier to correcting the bias is that individuals seem unaware that they are engaging in predecisional distortion until after the fact. It is difficult for people to fix a mistake if they can't recognize that it is happening.
