Above Average Effect

General Discussion

These studies have immediate implications for recent treatments of the above average effect, specifically those that have treated this particular phenomenon as conclusive evidence that the judgments that people normally render in the social realm are fraught with unrealistic cognitive illusions and distortions (Taylor & Brown, 1988). The studies described here confirmed one assertion made by proponents of this viewpoint: In our studies, college students (who we take to be a nondepressed population) provided self-assessments that appear to be self-serving and biased.

However, the analysis of the above average effect presented here throws its status as conclusive evidence of unrealistic bias into doubt. It may be that the assessments our subjects provided were hopelessly distorted (for, indeed, how can almost everyone be above average), but it might also be the case that our subjects were largely accurate in the assessments they provided. More specifically, after considering and presumably defining ambiguous traits, subjects may have subsequently provided percentile rankings that were accurate given their own idiosyncratic definitions. Buttressing the notion that people may be accurate within their own peculiar definitions is the fact that when considering unambiguous traits, people exhibited hardly any self-serving pattern whatsoever. That is, once the criteria of judgment are clearly established, people have the ability to assess their own standing in relation to their peers accurately.

Under this analysis, the above average effect by itself fails to serve as clear evidence of bias and distortion in social judgment. In effect, when people consider their own definition, and perhaps the one most relevant to the daily tasks they face, they may be largely accurate. However, even if this is the case, we can still conjecture that the above average effect, and the self-serving use of trait ambiguity, may have alarming social costs. These costs occur when one individual considers his or her peculiar definition as the only one worthy of consideration and fails to recognize when other plausible definitions are relevant.

One cost may occur when an individual believes that his or her definition of ability is the only one relevant to achievement outcomes. Specifically, when people attempt to predict how well they will perform in comparison to their peers on some common achievement metric (e.g., grades, games won, salary, rate of promotion), they may reliably overestimate how much they will achieve because they consider only their own self-serving definition of the relevant ability domain. For example, consider a company that has three vice presidents – one from sales, one from accounting, and one from engineering – who are being considered for promotion to president. Each considers him- or herself to be an outstanding manager among his or her peers because he or she only considers a self-serving definition of “managerial skill.” The vice president for sales, for example, can cite his well-developed ability to persuade. The vice president from accounting can cite her ability to keep track of all details of a project. The vice president from engineering can cite her ability to coordinate different individuals into an organized project. If each of these people only considers one peculiar definition of leadership ability when estimating the likelihood of receiving the promotion, there will clearly be some overconfidence and error.

From this example, it is also apparent that self-serving definitions of ability can also be a springboard for social disharmony. Consider the promotion once it is made. To two of the vice presidents, it will appear more mysterious than to the third. The choice may seem capricious, or even malicious, and may lead to some rancor. For another example, consider a group of professors awaiting word on a merit raise in their salary. Social conflict can occur when one receives an average or less than average pay increase when one’s self-assessment is higher (and recall, 94% of professors believe they do above average work; Cross, 1977).

In sum, it is impossible to tell whether subjects in four studies were largely accurate or biased in their assessments of ability when their idiosyncratic trait definitions were taken into consideration. From the preceding analysis, however, one point does become clear. To the extent that people fail to recognize when other definitions of ability are relevant for success and achievement, estimates of their future well-being will be exaggerated.