Toward an understanding of the fundamental attribution error: Essay diagnosticity in the attitude attribution paradigm

The observer error in attitude attribution was examined, focusing upon the perceiver's conception of the relationship between a writer's attitude and the quality of performance on an essay task. Subjects appear to have invested essays, written under assignment, with diagnostic value on the presumption of a correlation between the quality or strength of the essay and the writer's attitude. When subjects were given essays varying in direction, strength, and constraint, their attributions indicated a reversal of correspondent inference for weak essays produced under high constraint, replicating an earlier, conceptually important result (E. E. Jones, S. Worchel, G. R. Goethals, & J. Grumet, Journal of Experimental Social Psychology, 1971, 7, 59–80). It was suggested that the attribution error need not reflect general misunderstandings about, or the low salience of, situational factors, but rather is based on the perceiver's inclination to adopt a diagnostic judgmental set in the attitude attribution paradigm.     

Attibution of attitudes to speakers and listeners under assigned-behavior conditions: Does behavior engulf the field?

A conceptual replication of the assigned-behavior conditions of previous research in attitude attribution was conducted. Participants made attitude attributions about themselves and also about others who had either given, listened to, or not been exposed to either a proattitudinal or counterattitudinal essay. The position adopted in the essay significantly affected the attitudes attributed to those assigned to deliver the essay. Essay position had an equally strong effect upon attitudes attributed to those who merely listened to the essay, however. Thus, the fact that the actor had delivered a particular assigned essay apparently did not in itself convey information to the observers about the actor's attitudes. Experimental results suggested that essay direction affects attitudes attitudes attributed to others in an assigned-behavior setting because observers use their own attitudes to infer the attitudes of others. Since observers' own attitudes change after exposure to different essays, they attribute different attitudes to others.     

Author Race, Essay Quality, and Reverse Discrimination

This study examines the interaction between author race and essay quality on evaluations of essays by experienced teachers and student teachers. Subjects ( n = 160) were told that they were reading essays written by high school seniors applying for university admission and that their evaluations would result in the authors' admission and class assignment to remedial, normal, or advanced composition classes. Four preselected and prerated essays: one poor, two moderate, and one excellent were presented to the subjects in a booklet, each with a bogus admissions form, an essay rating scale, and a class assignment form. The attribution of author race was accomplished surreptitiously. Results showed that both the essay quality manipulation and author race attribution were successful. Subjects easily discriminated the difference in quality among the essays. They also showed a significant tendency to use reverse discrimination in rating black authors higher than authors whose race was not indicated. An interesting finding was that reverse discrimination was greatest for the moderate essays. Class assignments showed a similar trend, but differences in assignment of black and non-black authors were not generally significant. Discussion focuses on the distinction between ambiguity and ambivalence as psychological constructs in explaining prejudice and discrimination. The significance of these findings for minority education is also examined.     

Attribution and reconstructive memory

Two experiments were conducted to test the effects of making an attribution on later memory for the event that gave rise to the attribution. Subjects in Experiment 1 observed a scenario in which an actor's behavior was associated with high or low variance (distinctiveness) across situations, and high or low congruence (consensus) to the actions of others. Subjects either made attributions for the actor's behavior immediately following the scenario or not. One week later, subjects were asked to recall consensus and distinctiveness for the actor's original behavior. Subjects who made attributions were significantly better at estimating the high-high and low-low combinations of consensus and distinctiveness than were their no-attribution counterparts. It was suggested that making an attribution may allow for a reconstructed memory for the original event, but not enhance direct access to the original event information. A second experiment tested this concept further by having subjects view an edited version of the scenario in which either the distinctiveness or the consensus information was deleted and having subjects make attributions or not. One week later, subjects were asked to indicate their certainty that consensus and distinctiveness information was a part of the original scenario and to estimate the levels of consensus and distinctiveness. Subjects who made attributions were more confident and accurate in estimating the level of consensus or distinctiveness that was given in the original scenario than were no-attribution subjects. However, attribution subjects were also more confident that consensus information or distinctiveness information was contained in the scenario (when it was not) than were the noattributio subjects. Results of the two experiments suggest that eliciting attributions can distort subsequent memory for the event on which the attributions were based.     

An empirical test of the covariation analysis in causal attribution

The present investigation was designed to reveal the cognitive inference processes associated with both detection and utilization of covariation information in causal attribution. Male undergraduates were (a) informed that a test was easy or difficult and shown a videotape in which (b) the test-taker's performance was high or low, and (c) covariation between the test-taker's effort expenditure and trial-by-trial outcome was present or absent. High performance was attributed to the test-taker's effort and ability, whereas low performance was attributed to the difficulty of the test. However, recognition of the covariance relationship decreased the attribution of high performance to ability and of low performance to test difficulty and increased the attribution of low performance to effort. Effort attribution in the high performance condition was independent of covariation information. The results are discussed in terms of the relationship between covariation information and typical beliefs about the causes of achievement outcomes.     

The actor's behavioral history,his current role,and the divergence between actor and observer responsibility attributions

The actor's behavioral (High or Low) and role (High, Medium, or Low) respectability were varied in an actor-observer attribution study. A scenario described the actor's background combination of behavioral and role respectability and an event in which the actor participated. Subjects assigned actor and situational responsibilities to the event from their own viewpoint and from their perceptions of the actor's viewpoint. Subjects' awareness of the attributional divergence between actors and observers (Jones, E. E. & Nisbett, R. E. In Attribution: Perceiving the causes of behavior. New York: General Learning Press, 1971) and the information (public versus private) utilized by each were reflected in the results.     

The group attribution error

There is a parallel between our tendency to infer the attitudes of an individual on the basis of his or her behavior, regardless of the external constraints (Jones & Harris, 1967; Ross, 1977), and our tendency to infer the attitudes of a group on the basis of the group's decision, regardless of the group decision rule. The present research focuses on this latter process. What we term the group attribution error consists of the tendency to assume that group decisions reflect members' attitudes. This assumption can be erroneous because group decision rules, in addition to members' attitudes, can influence group decisions. In Experiment 1, members of a community in which a water conservation law was or was not instituted were assumed to have correspondent attitudes, regardless of how the community decision was made. In Experiments 2 and 3, subjects inferred a greater correspondence between out-group decisions and out-group attitudes than between an in-group decision and in-group attitudes. The fourth experiment found that subjects committed the group attribution error because they attended as much to the outcome of a recall election as to the actual proportion of voters for and against the recall. Finally, Experiment 5 showed that subjects' inferences of jury members' attitudes were influenced not only by the final jury vote but also by the actual decision, which was determined by the vote plus the decision rule by which the jury was bound. The results are related to previous research on the fundamental attribution error, stereotyping, and polarized appraisals of out-groups.     

Selective self-disclosure with and without correspondent performance

If people are motivated to elicit attributions from others, are they capable of using attribution theory as a means to this end? To explore this question, subjects were asked to select information to disclose to hypothetical target persons in order to have those target persons make specific attributions about the subjects' traits and abilities. The disclosures conformed well to a model sensitive to the principles of augmentation and discounting. The model considers the causal relevance and value of personal information and whether the self-presenter is given the opportunity to engage in behavior that accomplishes some specific self-presentational objective. The results, in line with the model, showed that subjects having the opportunity to engage in such “correspondent” behavior, are subsequently eager to disclose impediments that highlight the likely presence of highly valued behavioral causes. Subjects without such an opportunity seek to disclose facilitating causal factors that would make an effective performance appear to be more likely should the opportunity later arise.     

Outcome expectancy,locus of attribution for expectancy,and self-directed attention as determinants of evaluations and performance

Are internal versus external attributions of responsibility for prior outcomes important determinants of subsequent performances? or is their effect limited to influencing the affective and evaluative experiences that are associated with the task outcomes? Recent theoretical statements appear to differ on this issue. The present study examined the question, while at the same time testing the influence of self-directed attention on the process under investigation. Subjects attempted a series of mazes in collaboration with an ostensible cosubject (actually a confederate). The pair experienced either three consecutive sucesses or three consecutive failures. Subjects were led to perceive the responsibility for these outcomes as residing primarily with themselves or primarily with their partner. Self-focus was manipulated (by a mirror) prior to attempting a fourth maze and completing a set of rating scales. Success-condition subjects performed better on the fourth maze in the mirror's presence than in its absence, whereas failure-condition subjects tended to perform more poorly in the mirror's presence than in its absence. The manipulation of internal versus external attributions did not influence behavior, but did influence subjects' affective and evaluative reactions to themselves and their partner. Discussion centers on the relationship between these findings and other recent findings in the areas of attribution and achievement-related behavior.     

The association between implicit theories of personality and the attributional process

The influence of implicit theories of personality (entity vs. incremental theorists; see Dweck, Chiu and Hong, 1995) on the stages of the Sequential Operations Model of attribution (Gilbert, Pelham, and Krull, 1988) was investigated. Two hundred eighty Norwegian participants were given a Norwegian translation of the implicit personality theories measure. Participants then read two essays, one advocating the pro-life position and the other advocating the pro-choice position on the abortion issue. The essay positions were ostensibly assigned rather than freely chosen by the author. After each essay, participants were asked to rate the essay position and the true attitude of the author. Entity and incremental theorists showed no differences in their ratings of the essay position; however, entity theorists made significantly stronger correspondent inferences about the author's attitude than did incremental theorists. These results support the contention that entity theorists engage in less attributional correction than incremental theorists.     

Verbal accounts and attributions of social motives

The effects of verbal accounts offered by a threatener on targets' subsequent attributions of the threatener's social motives was studied. Following a standardized interaction in a Prisoner's Dilemma game the subjects' opponent offered one of three accounts for using threats: cooperative intent, establishment of transrelational equity, or ignorance. In a fourth condition the confederate offered no account for his actions. Attributions were assessed by having subjects rate each of four responses representative of the social motives of cooperation, competition, apathy, and deceit in five different situations. It was found that the type of account had specific attributional effects. A cooperative account led to a correspondent inference of a cooperative disposition, a transrelational equity account was apparently perceived as illegitimate and led to an attribution of a deceitful motive, and an excuse of ignorance was linked with apathy.     

Is dissonance phenomenologically aversive?

This study examined whether dissonance is, phenomenologically, an aversive state. Experimental subjects were induced to write counterattitudinal essays under a high-choice condition. One group of subjects was led to believe that a pill, which they had just taken in the context of a separate experiment, would cause them to feel pleasantly excited. A second group was led to believe that the pill would make them feel tense. A third group was given no information about the pill's potential side effects, while a fourth group expected to have no side effects at all. In this last condition, the results yielded the usual dissonance effect: subjects stated attitudes more congruent with the essay than did subjects in a survey control condition. When subjects were given an opportunity to attribute their arousal to an arousing, but nonaversive pill (i.e., the pleasant excitement condition), this effect was unchanged. In contrast, when subjects could attribute their arousal to an aversive pill (i.e., in the tensè and the no information conditions), this effect was virtually eliminated. These results are consistent with the notion that dissonance is an aversive state and that subjects will seize, when possible, an external attribution for this state.     

Social influence via threats

Influence via threats was investigated in correspondent relationships (characterized by commonality of interest) and in noncorrespondent relationships (characterized by conflict of interest). In addition, the degree of informational power attributed by the influencee to the influencer and to himself, and the severity of threats used by the influencer were manipulated. The results point to the existence of two qualitatively different processes of influence. In correspondent relationships, threats are viewed as conveying the influencer's belief about the choice of action that would benefit both parties to the relationship. Further, the more severe the threat, the more confident the influencer would appear to be of his choice of action. As a result, the influencee's willingness to rely on the influencer's judgment, as conveyed by the threat, becomes an important aspect of the influence process. Such willingness is affected by the influencee's attribution of informational power to the influencer and to himself. In noncorrespondent relationships, on the other hand, suspicion and mutual distrust preclude reliance on threats as sources of information. Under such circumstances, social influence via threats becomes a simple process of coercion.     

The effects of perceived similarity upon the use of reward and punishment

Subjects were led to perceive a trainee as either similar or dissimilar to themselves. During subsequent influence trials, subjects attempted to assess the causes of the trainee's performance and to employ rewards, punishments, or manipulations of the trainee's environment to optimize her performance. It was predicted that subjects would perceive the performance of dissimilar trainees as caused more by the dispositional factor of motivation than the performance of similar trainees, and that dissimilar trainees would thus be rewarded more upon success as well as punished more upon failure. The hypotheses were supported, and a model of reward-punishment behavior and attribution is discussed.     

Attribution of attitude as a function of behavioral direction and freedom: Reactance in the eye of the observer

In a variation of the Festinger and Carlsmith dull-task paradigm an actor-stooge was given free-choice, forced-compliance choice, or no-choice as to whether he should tell the “waiting subject” that the task was dull or that the task was enjoyable. Crosscutting this three-level manipulation of behavioral freedom was a two-level manipulation of behavioral direction. The actor-stooge either chose or agreed to inform the “waiting subject” that the task was dull or that the task was enjoyable. In agreement with Jones and Davis' theory of correspondent inference observer-subjects made positively correspondent inferences in the free-choice condition, i.e., the actor-stooge was attributed with a more unfavorable attitude toward the task when he chose to advocate the dullness of the task than when he chose to advocate the enjoyableness of the task. However, within the no-choice condition, previously unobtained negative correspondent inference occurred. The actor-stooge was attributed with a more unfavorable attitude toward the task when he was required to advocate the enjoyableness of the task than when he was required to advocate the dullness of the task. These later results were interpreted as indicating that the observers attributed reactance to the actor-stooge in the condition in which the most social pressure was exerted. A second experiment demonstrated that the negative correspondence effect was not due to assumptions regarding other subjects' preference and also demonstrated the dependence of the effect on the absence of a rationale for the no-choice assignment of behavioral direction.     

Evaluation as a function of correspondence and hedonic relevance

An experiment was conducted to test and refine some of the implications of Jones and Davis' (1965) model of the attribution process in person perception. Subjects read anecdotes about actors who performed either in-role (low correspondence) or out-of-role (high correspondence) behaviors which were positive, negative, or neutral in the subject's estimation (hedonic relevance). Respect and admiration for the actor were influenced by both hedonic relevance and correspondence (did the action reflect an inner attribute?), but liking and friendship were affected by hedonic relevance only. The implications for the evaluation process in person perception were discussed.     

Trait attributions and defense against insult from a dissimilar other

It was hypothesized that subjects would prefer to blame a character assault on negative ability characteristics of a dissimilar attacker as opposed to negative motivational characteristics in order to escape responsibility for the attack. It was reasoned that because we generally think of ourselves as having less potential influence over the abilities as opposed to the motivations of another person, it might be possible to diminish one's responsibility for another's behavior by attributing that behavior to that person's ability characteristics. Subjects in this experiment responded either to an insulting or noninsulting stimulus person who was either similar or dissimilar by selecting from a list of both motivational and ability bipolar trait dimensions, those dimensions they would most prefer to use in rating the stimulus person. As predicted, subjects responding to an insulting and dissimilar stimulus person showed a significantly greater preference for ability trait dimensions than subjects in the other conditions combined and also disliked the stimulus person more. The significance of these results for defensive attribution processes and phenomena such as racism and sexism are discussed.     

Objective self-awareness,standards of evaluation,and moral behavior

In a performance setting, subjects were given an opportunity to cheat without fear of detection on puzzle problems. Subjects were led to believe that successful performance was due to ability in some conditions, but to luck in other conditions. In fact, most of the problems were insolvable, so that success was impossible without cheating. Self-awareness was induced in half the subjects by having them sit in front of a mirror and listen to a tape recording of their own voice as they worked on the puzzle problems; the remaining subjects were not exposed to a mirror and listened to a tape of someone else's voice as they worked on the problems. It was predicted that cheating frequency would be higher under ability attribution conditions than under luck attribution conditions, and that this effect of performance attribution would be greater among self-aware subjects than among non-self-aware subjects. Results confirmed these hypotheses. Discussion centered on the differential use of morality and competence standards for behavior when in a state of self-awareness.     

Inferential processes in the forced compliance situation: A Bayesian analysis

The attribution made by an observer (O) to an actor in the forced compliance situation was regarded as a probability revision process which can be described by a Bayesian inference model. Os' perceptions of the forced compliance situation were analyzed in terms of the input components into the Bayesian model: prior probabilities of the relevant attitudes and the diagnostic values of the behaviors which the actor may choose. In order to test propositions made by attribution theory about such perceptions (Kelley, 1967;Messick, 1971), Os viewed actors under conditions of Low Inducement (LI) and High Inducement (HI). Before observing the actor's decision, Os estimated the prior probabilities of the relevant attitudes and the conditional probabilities of compliance and refusal given each of the attitudes. After observing the actor's decision, Os estimated the posterior probabilities of the attitudes. As expected, in the LI condition, compared to the HI condition, compliance was seen as less probable and more diagnostic about the actor's attitudes, and the posterior probability of the corresponding attitude was higher. Contrary to expectations, within both conditions, compliance, compared to refusal, was seen as less diagnostic and more probable.     

Interactive effects of writing and reading a persuasive essay on attitude change and selective exposure

Subjects wrote an essay either supporting their own position on an issue or irrelevant to that issue. Subsequently, they received a communication that either supported or opposed their position, and that either claimed to present the only reasonable position or acknowledged the viability of alternative positions. Then, they reported their own beliefs in the position being advocated, and finally were given an opportunity to obtain subsequent information that either supported or opposed this position. Attitude change in the direction of a proattitudinal communication occurred only when both (a) this communication asserted that only the position advocated was defensible and (b) subjects had previously written a proattitudinal essay themselves. Moreover, either of these factors was sufficient to prevent attitude change in the direction of a counterattitudinal communication. Data supported predictions based upon a joint consideration of commitment and reactance effects. Subjects' subsequent preference for supporting over opposing information was greater when they had written a supporting essay than when they had not, greater when they had read an opposing communication than when they had read a supporting one, and greater when the communication they read was presented as stating the only reasonable position, no matter which position it advocated.