Attribution Theories: Harold Kelly & Bernard Weiner

1. Harold Kelly’s Covariation Model

    Kelley’s (1967) covariation model is the best-known attribution theory. He developed a logical model for judging whether a particular action should be attributed to some characteristic (dispositional) of the person or the environment (situational). He says that something can only be the cause of a behavior if it is present when the behavior occurs but absent when it does not occur.

The term covariation simply means that a person has information from multiple observations at different times and situations and can perceive the covariation of an observed effect and its causes.

He argues that people take into account three kinds of evidence to discover the causes of behavior:

[For this discussion let’s use the example of Tom. His behavior is laughter. Tom is laughing at a comedian.]

i. Distinctiveness: the extent to which the person reacts in the same manner to other different stimuli or events. It asks whether the behavior is distinct or unique. In the case of this example, if Tom only laughs at this comedian, the distinctiveness is high. If Tom laughs at everything, then distinctiveness is low.

ii. Consensus: the extent to which other people react to a given stimulus or event in the same manner as the person we are evaluating. If everybody in the audience is laughing, the consensus is high. If only Tom is laughing, the consensus is low.

iii. Consistency: the extent to which the person in question reacts to the stimulus or event in the same way on other occasions, over time. It asks whether the behavior occurs at a regular rate or frequency. In the case of our example, If Tom always laughs at this comedian, the consistency is high. If Tom rarely laughs at this comedian, then consistency is low.

Now, if everybody laughs at this comedian and they don’t laugh at the comedian after/before him and also this comedian always makes people laugh at all his shows, then we would make an external attribution, i.e., we would assume that Tom is laughing because the comedian is funny.

On the other hand, if Tom is the only person who laughs at this comedian, if Tom laughs at all comedians, and if Tom always laughs at the comedian, then we would make an internal attribution, i.e., we assume that Tom is laughing because he is the kind of person who laughs a lot.

So, what we’ve got here is people attributing causality based on correlation. That is to say, we see that two things go together, and we, therefore, assume that one causes the other.

2. Bernard Weiner’s Casual Attribution Theory

    Attribution theory assumes that people try to determine why people do what they do, i.e., attribute causes to behavior. A person seeking to understand why another person did something may attribute one or more causes to that behavior. A three-stage process underlies an attribution: (1) the person must perceive or observe the behavior, (2) then the person must believe that the behavior was intentionally performed, and (3) then the person must determine if they believe the other person was forced to perform the behavior (in which case the cause is attributed to the situation) or not (in which case the cause is attributed to the other person).

Weiner focused his attribution theory on achievement (Weiner, 1974). He identified ability, effort, task difficulty, and luck as the most important factors affecting attributions for achievement. Attributions are classified along three causal dimensions: locus of control, stability, and controllability.

i. The locus of control dimension has two poles: internal versus external locus of control.

ii. The stability dimension captures whether causes change over time or not i.e., stable or unstable. For instance, ability can be classified as a stable, internal cause, and effort classified as unstable and internal.

iii. Controllability contrasts causes one can control, such as skill/efficacy, from causes one cannot control, such as aptitude, mood, others’ actions, and luck i.e., controllable and uncontrollable.

These aspects are independent of the internal/external dimension. Some internal causes tend to be quite stable over time such as personality traits or temperament, which other internal causes tend to change like motives, health and fatigue. Similarly, some internal causes are controllable like temper while some are uncontrollable like illness. Same is true for external causes of behavior: some are stable like social norms or laws while some are unstable like meeting someone who’s in a bad mood.

Causal attributions determine affective reactions to success and failure. For example, one is not likely to experience pride in success, or feelings of competence, when receiving an ‘A’ from a teacher who gives only that grade, or when defeating a tennis player who always loses…On the other hand, an ‘A’ from a teacher who gives few high grades or a victory over a highly rated tennis player following a great deal of practice generates great positive affect. Students with higher ratings of self-esteem and with higher school achievement tend to attribute success to internal, stable, uncontrollable factors such as ability, while they contribute failure to either internal, unstable, controllable factors such as effort, or external, uncontrollable factors such as task difficulty.