Kelleys covariation-principle (ANOVA-model)
It says that humans make causal explanations by evaluating three kinds of informations. These are:
- Consensus information - do all or only a few people respond to the stimulus in the same way as the target person?
- Distinctiveness information - does the target person respond in the same way to other stimuli as well?
- Consistency information - does the target person always respond in the same way to this stimulus?
Let's say, for example, that Mr. X is afraid of a certain dog and Mr. Y witnesses this incident.
- Case 1: The consensus is low, the distinctivness is high and the consistency is low
Mr. X is the only one who is afraid of that dog, he is only afraid of this dog but isn't always afraid of this dog. Conclusion: Mr. Y will probably assume that the situation (for instance the dog jumping up the unsuspecting Mr. X) is the reason that Mr. Y is afraid.
- Case 2: The consensus is high, the distinctivness is high and the consistency is low
Everyone is afraid of that dog, Mr. X is only afraid of this dog but isn't always afraid of this dog. Conclusion: Mr. Y will probably assume that the dog (for instance the dog has been turned wild by someone) is the reason that Mr. X is afraid.
- Case 3: The consensus is low, the distinctivness is low and the consistency is high
Only Mr. X is afraid of that dog, he is afraid of every dog and he is always afraid of this dog. Conclusion: Mr. Y will probably assume that the reason for Mr. X to be afraid is simply that he is frightend by dogs.
- Case 4: The consensus is high, the distinctivness is high and the consistency is high
Everyone is afraid of this dog, Mr. X is only afraid of this dog and he is always afraid of this dog. Conclusion: Mr. Y will probably assume that the dog (since everyone is afraid of it, it must be a pretty bad dog) is the reason for Mr. X to be afraid.
- And so forth...
Since nobody knows everything, false attributions are easily made. For instance, if Mr. Y in case 4 above doesn't know that everyone is afraid of this certain dog, he could as well assume, that Mr. X is afraid, because he has had a bad experience with that dog and was bitten by it. Mr. Y would therefore attribute Mr. Y's fear of this dog to Mr. Y instead of to the dog.
Also, humans tend towards to under-emphasize the role and power of situational influences and over-emphasize dispositional factors. For instance, if a soccer player has a scoring chance but slips and doesn't score, a lot of fans attribute this failure to the player and under-emphasize situational factors like a wet and slippery ground and so on (wikipedia:Fundamental attribution error).
 Configuration-principle: causal schemata, discounting principle and augmentation principle
As said, the ANOVA-model only works correctly, if every necessary information is at hand. In real life this is seldomly the case, meaning that attributions often must be made upon a single observation. And this can only be done with the assistance of causal schemata. These are opinions, presumptions and theories build on experience about how certain causes interoperate to cause the observed effect. Two examples for causal schemata are the discounting principle and the augmentation principle. The discounting principle says that a cause that can explain a certain effect looses significance if other plausible causes exist. An example would be a pupil that fails a test. While the pupils laziness is a plausible explanation, it looses its significance if one learns that the pupils mother is sick, since this is also a plausible explanation for the pupils failure. The augmentation principle says, that a certain cause can be revalued, if an effect occures despite of difficult circumstances. If for example the pupil does well on the test, even though he was sick, the plausible cause that the pupil studied hard would gain importance.
 Weiners motivational attribution theory
Weiners theory is based on the belief, that peoples motivations and also their emotions are largely based on their experiences and to what cause they attribute those experiences. If, for example, a pupil gets a bad mark in school, there are at least two attributions possible: either the pupil attributes it to a lack of cognitive resources or to the fact that he or she didn't study hard enough. In the first case, the cause is internally and stable, meaning that the pupil probably will resign and not try very hard to get better marks. Also, the pupil will be ashamed, "because he is dumb". In the latter case, the cause is internal and unstable, meaning that the pupil might try to study harder next time. In this case, the pupil will feel guilty because the cause for his failure was under his control. So it is his fault that he got the bad mark. But if he studied hard and not only he, but the majority of his class got bad marks, he will attribute his bad mark to the test, which was too difficult, and neither feel guilty nor ashamed. But he could be angry with the teacher, who is responsible for the difficult test.
Formally, Weiner differentiates between result-dependent and attribution-bound affects. Result-dependent affects are pretty general emotions like happiness, joy, sadness or frustration who are directly linked with the achievement or non-achievement of a goal. Attribution-bound affects are the result of causal attributions and also imply future expectations.
Event --> result-dependent affects --> causal attribution --> responsibility evaluation --> affect (annoyance, sypmathy) --> behavior
Weiner developed a three-dimensional pattern of causal dimensions (location x stability x controlability) and it was later shown that every single dimension is connected with specific affects:
Location: Whether the cause is Internal or External (self or others, one's own efforts or luck): Pride / Disappointment Stability: Whether the cause is temporary or lasting (efforts in this test or one's IQ: hope / hopelessness Controlability: Whether one has control over the event or not (One's level of preparation vs. badly set paper: Guilt / Shame
Especially the dimension of controllability (the evaluation of responsability) is important for the evaluation of future expectations, the felt emotion and in summary for the resulting behavior.
 Attribution theory in clinical psychology
 Case study
Mr. J. claimed that a "warm form" sexually excites him until he has an orgasm. This experiences were a burden for him and after he revealed them, he was diagnosed as being schizophrenic, was institutionalized and treated with Thorazin, a antipsychotic drug. Even though his psychiatrists were convinced that he was mentally ill, careful observations showed, that Mr. J. in fact unintentionally stimulated himself by moving his legs. After he was taught to correctly attribute his excitement to his leg-movement, his "delusions" disappeared and they didn't reappear for at least the following six months of observation.
The leg-movement was shown to him as a plausible alternative cause and, as the discounting principle would predict, he didn't believe in his originally assumed cause as strong as before.
 learned helplessness
Learned helplessness describes a condition, which is marked by learning deficits, negative emotions and passive behaviour. This condition is caused, when humans or animals learn that their (re)actions and the the wanted effects are independent from each other in one situation and develop the expectation, that they won't have control in similar situations. Learned helplessness can be used to explain human depressions and since learned helplessness is based on incorrect attributions, a cognitive therapy that tries to correct incorrect attributions might help to cure depressions.
While studies that tried to get a definite answer produced promising results, it is still unclear, whether incorrect attributions are the/a cause for depressions. Attribution therapy produces promising results with depressed patients, though.
 Fun facts
An example given to a friend of mine who studies psychology by his professor: if you are on your first date with someone, you should do something that's a bit exciting. Chances are good, that your "partner-to-be" will attribute his/her pounding heart internally and then think, that it is pounding because he/she is in love with you. But if your meeting is too exciting, your "partner-to-be" will probably attribute his/her excitement externally (that is: to the situation) and the trick won't work. Applied psychology. ;-)
 External Links
This is only a very short explanation of the basics of attribution theory. For further information take a look at
- Einige grundlegende Annahmen und Konzepte der Attributionstheorie by Wulf-Uwe Meyer (PDF, 30 pages, ~189KB and unfortunately in german)
- Attributionstheorie bad webdesign but worth reading. also in german
- wikipedia:Attribution theory