Examining Others in Depth: The Role of Distinctiveness, Consensus, and Consistency in Attribution

 According to an influential paper by Harold H. Kelley (1967), people tend to attribute causes by using logic that resembles how social scientists test their hypotheses. They analyze the variance of presumed causes and effects in a manner corresponding to the basic logic used in statistical analysis. (to be continued...)

                A friend of ours was picking up his son, Billy, from nursery school. As he walked in, he saw Billy run over to a little girl and hit her so hard that she fell down. The teacher, nothing the father’s shocked expression, explained, “Oh, Mr. Smith, it is not just Billy. Everybody always seems to be hitting Mary.” Our friend, although still troubled about Mary, was at last relieved to have some of the onus removed from Billy. Consider some other possible explanations the teacher might have offered:

                “I don’t know what has been getting into the children today. They are usually quite nice, but today they have all been fighting with one another.”

                “I can understood your concern, Mr. Smith. All the children tend to play well together, but Billy has been fighting with the others for some time now.”

                The point is, of course, that Mr. Smith, as a concerned father, would like to believe that his son is not hostile or cruel, but if he is, then Mr. Smith certainly wants to know about it. Having seen Billy do something upsetting, he is exploring to know what extent he should attribute the cause to Billy—a dispositional attribution—or the situation.

                According to an influential paper by Harold H. Kelley (1967), people tend to attribute causes by using logic that resembles how social scientists test their hypotheses. They analyze the variance of presumed causes and effects in a manner corresponding to the basic logic used in statistical analysis. Thus Kelley’s model is often referred to as the “analysis of variance” or “ANOVA” model of attribution theory. We can consider most events as comprising three variables: (a) People act on (b) entities, tasks or targets, that may be objects or other people, at particular (c) situations or times. In the nursery-school example, (a) Billy hit (b) Mary (c) today in the play area at school. Billy and the other children are the actors (Mary could for some purposes be considered an actor). Mary is the target. The third variable may be representated as “today” (or any other day) or the playground where Billy hit Mary. More generally, various children (actors) could engage in the behavior, at various times and locations, and any number of children could be targets.

                To assume that the behavior comes out of characteristics internal to Billy, that is, to make a dispositional attribution, we must consider three factors:

1.       Distinctiveness tells us whether and to what extent someone responds differently to a given target from how the person ordinarily responds. If Billy only hits Mary and not others, then perhaps Mary instigates him. If Billy only hits Mary, the distinctiveness of the act is high. Hitting people is “unlike Billy,” and we conclude that a dispositional attribution is not accurate. Thus to attribute the act to Billy, the disctinctiveness of the act for him must be low.

2.       Consensus tells us the extent to which others respond similarly to the target. If everybody else at nursery school is hitting Mary, then Mr. Smith might again reasonably assume that an attribution to Billy’s disposition is unjustified. If Billy were the only one hitting Mary, if a dispositional attribution were justified, then consensus would be low. Thus to attribute the act to Billy’s disposition, consensus with regard to the act must be low.

3.       Consistency tells us the extent to which the actor responds similarly toward different targets at various times. To make an internal attribution toward the actor, consistency should be high. Otherwise, his behavior might be considered a temporary reaction.

The explanation offered by the teacher (“Everybody always seems to be hitting Mary all of the time.”) seems to blame Mary for the act. In such a case, Mr. Smith, if he is a caring person, might be relieved that Billy is not to blame, but might wonder what it is that Mary does. Perhaps she teases or goads the others.

We have been elaborating on relatively simple relationships among the three variables in the Kelley ANOVA model, but it is also allows for the analysis of more complex relationships, including the effects of interaction. We could use it to analyze, “Billy seems to fight with everyone during naptime; Sally seems to fight on the playground” or “Billy hits Mary just before naptime; Mary fights with Charlie every day on the playground.”

                In one of the more important experimental tests of Kelley’s framework, Leslie McArthur (1972) provided participants with brief descriptions of another person’s response (for example, “John laughs at the comedian”) along with information regarding consensus (whether others laugh at the comedian), consistency (whether John has laughed at the comedian in the past), and distinctiveness (whether John laughs at other comedians as well). Consistent with Kelley’s framework, McArthur found that behavioral information high in consensus, consistency, and distinctiveness tended to generate external attributions (“Something about the comedian that made John laugh”). On the other hand, information that was low in consensus, high in consistency, and low in distinctiveness generated more internal attributions (“Something about John made him laugh at the comedian”). Kelley’s theory has been supported by other researchers as well (Ruble and Feldman, 1976; Frieze and Weiner, 1971; Zuckermann, 1978; Hansen and Donoghue, 1977; Ross, Greene, and House, 1977; Gilmore and Minton, 1974; Karaz and Perlman 1975), although some work (e.g., Nisbett and Borgida, 1975) has indicated that of the three sources of information, consensus probably is less important in determining attributions than the other two.

(Source: Raven, Bertram H., Rubin, Jeffrey Z. 1983. SOCIAL PSYCHOLOGY 2nd Edition. United States of America: John Wiley & Sons, Inc.)

× 『rui@96yR』【butterflyuu】 ×