That as a competent keyless citizen he had proceeded energetically from the unknown to the known through the incertitude of the void
Member-to-group comparisons are prevalent in everyday life. A person might consider whether one politician would make a better president than other candidates, whether one home is more suitable than other prospective homes on the market, whether one food item is healthier than others, or whether one vacation spot is more desirable than others. In turn, the outcomes of such comparisons have important consequences for a person’s choices, decisions, moods, thoughts, and, ultimately, welfare. Indeed, rational models of choice are predicated on the idea that human beings can maximize their utility by identifying the best and worst options in a choice set.
However, recent work by Klar and his colleagues suggests that people are far from unbiased in their comparisons. Individual members of positively valenced groups (e.g., healthy foods, good politicians) are rated better and individual members of negatively valenced groups (e.g., unhealthy foods, bad politicians) are rated worse than the group average, in defiance of simple mathematical rules stating that the average of the individual members must equal the group average. In one of the first studies demonstrating these nonselective inferiority and superiority biases, Giladi and Klar had shoppers evaluate randomly selected pleasant-smelling or unpleasant-smelling soaps and found that any given pleasant soap was rated better than the rest of the group, and any given unpleasant soap was rated worse than the rest of the group. These effects have since been observed with other object categories, including desirable and undesirable acquaintances, restaurants, social groups, pieces of furniture, hotels, and songs, and thus appear to be highly robust and reliable.
In explaining these nonselective biases, Giladi and Klar proposed that when one member of a positive or negative group is compared with others (e.g., how does good restaurant A compare with other good restaurants B and C?), that member is evaluated against a standard that is one part local (restaurants B and C) and one part general (all other restaurants, including bad ones). Thus, although the member that is being evaluated should be compared only with the normatively appropriate local standard, it is actually compared with a hybrid standard that includes both the local and the general standard. Consequently, almost any member of a positive group will be rated better than others (because the general standard is more negative than the local standard), and almost any member of a negative group will be rated worse than others (because the general standard is more positive than the local standard).
In this article, I propose an additional reason (beyond the confusion of local and general standards) why almost any group member is rated more extremely than others in its group.
photo { Lise Sarfati }
