Attribution bias

From Wikipedia, the free encyclopedia

In psychology, an attribution bias or attributional bias is a cognitive bias that refers to the systematic errors made when people evaluate or try to find reasons for their own and others' behaviors. People constantly make attributions regarding the cause of their own and others' behaviors; however, attributions do not always accurately reflect reality. Rather than operating as objective perceivers, people are prone to perceptual errors that lead to biased interpretations of their social world.

Attribution biases were first discussed in the 1950s and '60s by psychologists such as Fritz Heider, who studied attribution theory. Other psychologists, such as Harold Kelley and Ed Jones expanded Heider's early work by identifying conditions under which people are more or less likely to make different types of attributions.

Attribution biases are present in everyday life, and therefore are an important and relevant topic to study. For example, when a driver cuts us off, we are more likely to attribute blame to the reckless driver (e.g., "What a jerk!"), rather than situational circumstances (e.g., "Maybe they were in a rush and didn't notice me"). Additionally, there are many different types of attribution biases, such as the ultimate attribution error, fundamental attribution error, actor-observer bias, and hostile attribution bias. Each of these biases describes a specific tendency that people exhibit when reasoning about the cause of different behaviors.

Since the early work, researchers have continued to examine how and why people exhibit biased interpretations of social information. Many different types of attribution biases have been identified, and more recent psychological research on these biases has examined how attribution biases can subsequently affect emotions and behavior.

History

Attribution theory

Research on attribution biases is founded in attribution theory, which was proposed to explain why and how we create meaning about others' and our own behavior. This theory focuses on identifying how an observer uses information in his/her social environment in order to create a causal explanation for events. Attribution theory also provides explanations for why different people can interpret the same event in different ways and what factors contribute to attribution biases.

Fritz Heider

Psychologist Fritz Heider first discussed attributions in his 1958 book, The Psychology of Interpersonal Relations. Heider made several important contributions that laid the foundation for further research on attribution theory and attribution biases. He noted that people tend to make distinctions between behaviors that are caused by personal disposition versus environmental or situational conditions. He also predicted that people are more likely to explain others' behavior in terms of dispositional factors (i.e., caused by a given person's personality), while ignoring the surrounding situational demands.

Jones & Davis

Building on Heider's early work, other psychologists in the 1960s and 70s extended work on attributions by offering additional related theories. In 1965, social psychologists Edward E. Jones and Keith Davis proposed an explanation for patterns of attribution termed correspondent inference theory. A 'correspondent inference' assumes that a person's behavior reflects a stable disposition or personality characteristic. They explained that certain conditions make us more likely to make a correspondent inference about someone's behavior:

Intention

People are more likely to make a correspondent inference when they interpret someone's behavior as intentional, rather than unintentional.

Social desirability

People are more likely to make a correspondent inference when an actor's behavior is socially undesirable than when it is conventional.

Effects of behavior

People are more likely to make a correspondent, or dispositional, inference when someone else's actions yield outcomes that are rare or not yielded by other actions.

Harold Kelley

Soon after Jones and Davis first proposed their correspondent inference theory, Harold Kelley, a social psychologist famous for his work on interdependence theory as well as attribution theory, proposed a covariation model to explain the way people make attributions. This model helped to explain how people choose to attribute a behavior to an internal disposition versus an environmental factor. Kelley used the term 'covariation' to convey that when making attributions, people have access to information from many observations, across different situations, and at many time points; therefore, people can observe the way a behavior varies under these different conditions. He proposed three factors that influence the way we explain behavior:

• Consensus: The extent to which other people behave in the same way. There is high consensus when most people behave consistent with a given action/actor. Low consensus is when not many people behave in this way.
• Consistency: The extent to which a person usually behaves in a given way. There is high consistency when a person almost always behaves in a specific way. Low consistency is when a person almost never behaves like this.
• Distinctiveness: The extent to which an actor's behavior in one situation is different from his/her behavior in other situations. There is high distinctiveness when an actor does not behave this way in most situations. Low distinctiveness is when an actor usually behaves in a particular way in most situations.

Kelley proposed that people are more likely to make dispositional attributions when consensus is low (most other people don't behave in the same way), consistency is high (a person behaves this way across most situations), and distinctiveness is low (a person's behavior is not unique to this situation). Alternatively, situational attributions are more likely reached when consensus is high, consistency is low, and distinctiveness is high. His research helped to reveal the specific mechanisms underlying the process of making attributions.

Biased attributions

As early researchers explored the way people make causal attributions, they also recognized that attributions do not necessarily reflect reality and can be colored by a person's own perspective. Certain conditions can prompt people to exhibit attribution bias, or draw inaccurate conclusions about the cause of a given behavior or outcome. In his work on attribution theory, Fritz Heider noted that in ambiguous situations, people make attributions based on their own wants and needs, which are therefore often skewed. He also explained that this tendency was rooted in a need to maintain a positive self-concept, later termed the self-serving bias. 

Kelley's covariation model also led to the acknowledgment of attribution biases. The model explained the conditions under which people will make informed dispositional versus situational attributions. But, it assumed that people had access to such information (i.e., the consensus, consistency, and distinctiveness of a person's behavior). What about when we don't have access to such information, for example, when interacting with someone we don't know well? Lack of information results in a tendency to take cognitive shortcuts, resulting in different types of attribution biases, such as the actor-observer bias that will be discussed below.

Cognitive explanation

Although psychologists agreed that people are prone to these cognitive biases, there existed disagreement concerning the cause of such biases. On one hand, supporters of a "cognitive model" argued that biases were a product of human information processing constraints. One major proponent of this view was Yale psychologist Michael Storms, who proposed this cognitive explanation following his 1973 study of social perception. In his experiment, participants viewed a conversation between two individuals; we'll call them Actor one and Actor two. Some participants viewed the conversation while facing Actor one, such that they were unable to see the front of Actor two, while other participants viewed the conversation while facing Actor two, obstructed from the front of Actor one. Following the conversation, participants were asked to make attributions about the conversationalists. He found that participants ascribed more causal influence to the person they were looking at. In other words, participants made different attributions about people depending on the information they had access to. Storms used these results to bolster his theory of cognitively driven attribution biases; because we have no access to the world except for through our own eyes, we are inevitably constrained and consequently prone to biases. Similarly, social psychologist Anthony Greenwald described humans as possessing a totalitarian ego, meaning that we view the world through our own personal selves. Therefore, different people inevitably interpret the world differently and in turn reach different conclusions.

Motivational explanation

These views of attributional biases as being a sole product of information processing constraints received criticism from researchers who argued that humans do not just passively interpret their world and make attributions; rather, they are active and goal-driven. Building on this criticism, research began to focus on the role of motives in driving attribution biases. Researchers such as Ziva Kunda drew attention to the motivated aspects of attributions and attribution biases. Kunda in particular argued that certain biases only appear when people are presented with motivational pressures; therefore, they can't be exclusively explained by an objective cognitive process. More specifically, we're more likely to construct biased social judgments when we're motivated to arrive at a particular conclusion, so long as we can justify this conclusion.

Current theory

Early researchers explained attribution biases as cognitively driven and a product of information processing errors. In the early 1980s, studies demonstrated that there may also be a motivational component to attribution biases, such that our own desires and emotions affect how we interpret social information. Current research continues to explore the validity of both of these explanations by examining the function of specific types of attribution biases and their behavioral correlates through a variety of methods (e.g., research with children or using brain imaging techniques).

Recent research on attribution biases has focused on identifying specific types of these biases and their effect on people's behavior. Additionally, some psychologists have taken an applied approach and demonstrated how these biases can be understood in real-world contexts (e.g., the workplace or school). Researchers have also used the theoretical framework of attributions and attribution biases in order to modify the way people interpret social information. Studies have implemented attributional retraining to help, for example, students have more positive perceptions of their own academic abilities.

Limitations of the theory

There is much inconsistency in the claims made by scientists and researchers that attempt to prove or disprove attribution theories and the concept of attributional biases. The theory was formed as a comprehensive explanation of the way people interpret the basis of behaviors in human interactions. However, there have been studies that indicate cultural differences in the attribution biases between people of Eastern and Western societies. Also, some scientists believe that attributional biases are only exhibited in certain contexts of interaction where possible outcomes or expectations make the forming of attributions necessary. These criticisms of the attribution model reveal that the theory may not be a general, universal principle.

Types

Researchers have identified many different specific types of attribution biases, all of which describe ways in which people exhibit biased interpretations of information. Note that this is not an exhaustive list (see List of attributional biases for more).

Fundamental attribution error

The fundamental attribution error refers to a bias in explaining others' behaviors. According to this error, when we make attributions about another person's actions, we are likely to overemphasize the role of dispositional factors, while minimizing the influence of situational factors. For example, if we see a coworker bump into someone on his way to a meeting, we are more likely to explain this behavior in terms of our coworker's carelessness or hastiness, rather than considering that he was running late to a meeting.

This term was first proposed in the early 1970s by psychologist Lee Ross following an experiment he conducted with Edward E. Jones and Victor Harris in 1967. In this study, participants were instructed to read two essays; one expressed pro-Castro views, and the other expressed anti-Castro views. Participants were then asked to report their attitudes towards the writers under two separate conditions. When participants were informed that the writers voluntarily chose their position towards Castro, participants predictably expressed more positive attitudes towards the anti-Castro writer. However, when participants were told that the writers' positions were determined by a coin toss, rather than their own free will, participants continued to express more positive attitudes towards the anti-Castro writer. These results demonstrated that participants did not take situational factors into account when evaluating a third party, providing evidence of what was later coined the fundamental attribution error.

Actor-observer bias

The actor-observer bias (also actor–observer asymmetry) can be thought of as an extension of the fundamental attribution error. According to the actor-observer bias, in addition to over-valuing dispositional explanations of others' behaviors, we tend to under-value dispositional explanations and over-value situational explanations of our own behavior. For example, a student who studies may explain her behavior by referencing situational factors (e.g., "I have an exam coming up"), whereas others will explain her studying by referencing dispositional factors (e.g., "She's ambitious and hard-working"). This bias was first proposed by Edward E. Jones and Richard E. Nisbett in 1971, who explained that "actors tend to attribute the causes of their behavior to stimuli inherent in the situation, while observers tend to attribute behavior to stable dispositions of the actor."

There has been some controversy over the theoretical foundation of the actor-observer bias. In a 2006 meta-analysis of all published studies of the bias since 1971, the author found that Jones' and Nisbett's original explanation did not hold. Whereas Jones and Nisbett proposed that actors and observers explain behaviors as attributions to either dispositions or situational factors, examining past studies revealed that this assumption may be flawed. Rather, the theoretical reformulation posits that the way we explain behavior depends on whether or not it is intentional, among other things.

Self-serving bias

A self-serving bias refers to people's tendency to attribute their successes to internal factors but attribute their failures to external. This bias helps to explain why we tend to take credit for our successes while often denying any responsibility for failures. For example, a tennis player who wins his match might say, "I won because I'm a good athlete," whereas the loser might say, "I lost because the referee was unfair." 

The self-serving bias has been thought of as a means of self-esteem maintenance. In other words, we feel better about ourselves by taking credit for successes and creating external blames for failure. This is further reinforced by research showing that as self-threat increases, people are more likely to exhibit a self-serving bias. For example, participants who received negative feedback on a laboratory task were more likely to attribute their task performance to external, rather than internal, factors. Therefore, the self-serving bias seems to function as an ego-protection mechanism, helping people to better cope with personal failures.

Hostile attribution bias

Hostile attribution bias (HAB) has been defined as an interpretive bias wherein individuals exhibit a tendency to interpret others' ambiguous behaviors as hostile, rather than benign. For example, if a child witnesses two other children whispering and assumes they are talking about him/her, that child makes an attribution of hostile intent, even though the other children's behavior was potentially benign. Research has indicated that there is an association between hostile attribution bias and aggression, such that people who are more likely to interpret someone else's behavior as hostile are also more likely to engage in aggressive behavior. See the following section on aggression for more details on this association.

Consequences for behavior

Aggression

Extensive research in both social and developmental psychology has examined the relationship between aggressive behavior and attribution biases, with a specific focus on the hostile attribution bias.

In particular, researchers have consistently found that children who exhibit a hostile attribution bias (tendency to perceive others' intent as hostile, as opposed to benign) are more likely to engage in aggressive behaviors. More specifically, hostile attribution bias has been associated with reactive aggression, as opposed to proactive aggression, as well as victimization. Whereas proactive aggression is unprovoked and goal-driven, reactive aggression is an angry, retaliatory response to some sort of perceived provocation. Therefore, children who are victims of aggression may develop views of peers as hostile, leading them to be more likely to engage in retaliatory, or reactive, aggression.

Research has also indicated that children can develop hostile attribution bias by engaging in aggression in the context of a video game. In a 1998 study, participants played either a very violent or non-violent video game and were then asked to read several hypothetical stories where a peer's intent was ambiguous. For example, participants may read about their peer hitting someone in the head with a ball, but it is unclear whether or not the peer did this intentionally or not. Participants then responded to questions about their peer's intent (e.g., "Do you think your peer hit someone with the ball on purpose?"). The children who played the violent video game, as compared to participants who played the nonviolent game, were more likely to say that their peer harmed someone on purpose. This finding provided evidence that just playing a violent video game could cause children to develop a short-term hostile attribution bias.

Intergroup relations

Research has found that we often exhibit attribution biases when interpreting the behavior of others, and specifically when explaining the behavior of in-group versus out-group members. More specifically, a review of the literature on intergroup attribution biases noted that people generally favor dispositional explanations of an in-group member's positive behavior and situational explanations for an in-group's negative behavior. Alternatively, people are more likely to do the opposite when explaining the behavior of an out-group member (i.e., attribute positive behavior to situational factors and negative behavior to disposition). Essentially, group members' attributions tend to favor the in-group. This finding has important implications for understanding other social psychological topics, such as the development and persistence of out-group stereotypes.

Attribution biases in intergroup relations are also observed as early as childhood. In particular, elementary school students are more likely to make dispositional attributions when their friends perform positive behaviors, but situational attributions when disliked peers perform positive behaviors. Similarly, children are more likely to attribute friends' negative behaviors to situational factors, whereas they attribute disliked peers' negative behaviors to dispositional factors. These findings provide evidence that attribution biases emerge as early as childhood.

Academic achievement

Although certain attribution biases are associated with maladaptive behaviors, such as aggression, some research has also indicated that these biases are flexible and can be altered to produce positive outcomes. Much of this work falls within the domain of improving academic achievement through attributional retraining. For example, one study found that students who were taught to modify their attributions actually performed better on homework assignments and lecture materials. The retraining process specifically targeted students who tended to attribute poor academic performance to external factors and taught them that poor performance was often attributable to internal and unstable factors, such as effort and ability. Therefore, the retraining helped students perceive greater control over their own academic success by altering their attributional process. 

More recent research has extended these findings and examined the value of attributional retraining for helping students adjust to an unfamiliar and competitive setting. In one study, first year college students went through attributional retraining following their first exam in a two-semester course. Similar to the previous study, they were taught to make more controllable attributions (e.g., "I can improve my test grade by studying more") and less uncontrollable attributions (e.g., "No matter what I do, I'll fail"). For students who performed low or average on their first exam, attributional retraining resulted in higher in-class test grades and GPA in the second semester. Students who performed well on the first exam were found to have more positive emotions in the second semester, following attributional retraining. Taken together, these studies provide evidence for the flexibility and modifiability of attributional biases.