The third piece of information is distinctiveness. This is how similarly the person behaves towards stimuli that are similar. If the person does not behave in the same way towards a similar stimulus (e.g. they are not afraid of other dogs) distinctiveness is high. If they behave in the same way (e.g. they are afraid of all other dogs) distinctiveness is low.
Kelley said that if consensus was low, consistency was high, and distinctiveness was low, (LHL) we would attribute the behaviour to the person. For example, if person A had always been afraid of all dogs, and no-one else they knew was afraid of dogs, we would attribute their fear to them being timid or easily scared.
However if consensus, consistency and distinctiveness were all high (HHH), we would attribute the behaviour to the entity. For example, if person B was afraid of one particular dog, but no others, and many other people were afraid of this dog, we would attribute that to the dog having an aggressive temperament.
The last combination is if consensus is low, consistency is low, and distinctiveness is high(LLH). This behaviour would be attributed to a particular set of circumstances. For example if person C was afraid of only one dog, and had never been before, and no-one else was afraid of this dog. An explanation of this would be that the dog was usually calm but had recently been aggressive to person C.
This Co-Variation theory has been tested in experiments, and showed that when the three pieces of information are manipulated, people make the attributions predicted by Kelley.
However, despite giving accurate results, there are problems with the theory.
Garland et al showed that people use other information such as personality and context, instead of the information Kelley used. One study found that people would only use the consensus, consistency, and distinctiveness information when no information about the situation or context was available.
The theory is also very cognitively expensive as it takes a lot of effort to process the consensus, consistency and distinctiveness, even when they are available to use. Due to our tendency to act as “cognitive misers” we need a quicker way of processing the information.
Attributional Biases
An attribution bias is a perception in distortion or judgement about our own or other peoples behaviour.
The first type of bias is the Fundamental Attribution Error. This is defined as the tendency to underestimate the importance of situational causes and overestimate the degree that actions reflect dispositional causes , or in other words, people tend to choose dispositional reasons for behaviour and ignore situational reasons, even if both reasons are equally likely.
It has been suggested that the reason for the FAE is that our “cognitive miser” approach means we ignore situational reasons unless we are directly told about them. This means that once we have thought of a dispositional reason for somebody's behaviour we do not want to make things more complex by adding other reasons.
One experiment to test this was by Ross in 1977. He randomly assigned participants as questioners or contestants on a general knowledge quiz. The questioners created questions from their own specialised knowledge, so the participants struggled to answer them. The quiz was observed by others, and everybody was asked to rate the knowledge of the questioners and contestants. The observers and contestants rated the questioners as having high general knowledge, but the questioners did not say they had higher knowledge.
Ross argued that the observers and contestants had ignored the situational; factors (the questioners had used specialised knowledge and compiled the questions) and instead chosen dispositional factors (the questioners had high general knowledge)
Although the FAE has been proven in a number of studies, it has been argued that it is not fundamental because there are some situations it does not occur in (e.g. the discounting principle) and that it is more appropriate to call it a bias instead of an error.
The next type of bias is the actor/observer bias. This says that when we observe the behaviour of others, we are more likely to attribute their behaviour to a dispositional factor. (e.g. If we see someone trip over in the street, we would probably say that they are clumsy).
However, when we are the actor (we perform the behaviour) we are more likely to attribute it to a situational factor (e.g. if we tripped over, we would say there was a loose paving stone)
Jones and Nisbett suggested the reason for the bias in attribution is because the actor and observer focus on different information about the behaviour.
The actor has more direct information about the behaviour than the observer (e.g. we know the paving stone was loose). We also know more about our previous behaviour (e.g. we know that we do not trip over very often, whereas the observer might think this is our usual behaviour.) Also, the focus of attention is different; actors focus outwards towards the environment and are more likely to attribute behaviour to the environment, while observers focus on the actor, and are more likely to attribute behaviour to the actor.
There are some situations where the actor/observer bias does not work in this way, and these can be explained by the Self-Serving biases.
For example, the actor/observer bias says that when we are asked to explain why we are successful at something, we will attribute it to a situational factor (e.g. I passed the exam because it was an easy paper). However, people usually explain it using dispositional factors (e.g. I passed the exam because I am good at that subject) This shows that we make dispositional attributions when we are successful, and take responsibility for our success, which is called the Self-Enhancing bias.
The reverse of this is the Self-Protecting bias, which is when we attribute failure to situational factors instead of dispositional, and deny responsibility for our failures (e.g. I failed the exam because it was unfairly hard)
They are thought to exist as part of self-esteem; the self-protecting bias helps us to protect our self-esteem when we fail, and the self-enhancing bias helps increase our self-esteem when we succeed.
These biases also work at a group level; the success of the “in-group” is explained in dispositional terms (e.g. our team won the football match because they are a good team) and failure is explained in situational terms. Similarly, the success of the “out-group” is explained in situational terms, and their failure in dispositional terms (e.g. their team lost because they are a bad team)
An extension of these biases is the Self-Handicapping bias. In some cases, we accept responsibility for failure if we know we can change our performance to not fail again. For example people might say “ I failed that exam because I did not prepare for it”
When it is not possible for us to change our performance, we will try to give reasons for failing before the event (e.g. I am going to fail my exam because I did not revise for it) Handicapping ourselves at the start means that we have a way to explain our failure in situational terms afterwards.