In this blog post, we will explain the Jones and Davis Correspondent Inference Theory. We will discuss what is the inference and information gathering process, probability estimation and the base rate.

Jones and Davis Corresponding Inference Theory

The correspondent inference theory is the generalization that is made about someone’s future behaviour, based on other previous internal and intentional attributions.

Making dispositional attributions about someone’s behaviour without making mistakes, finding permanent traits, is not easy; because in many situations the behaviours do not reflect the real personality of the subject, but it is the external factors that determine the behaviour.

To infer something related to the personality of the subjects, it is convenient to take into account a series of factors such as:

Spontaneity: there must be a precondition that is a free and intentional choice.

The singularity: we must consider the behaviours that are singular or unusual since they are those that provide us with information about the causes of the behaviours.

Social desirability: behaviours that have high social desirability (that is, that are socially approved and therefore more common) provide little information compared to less common behaviours that are low in social desirability.

Thus, when someone is observed performing antisocial behaviour, the possibility of a corresponding inference increases. Social desirability could be considered as a particular case of singularity.

Degree of involvement: refers to the degree of involvement of the person who evaluates the situation. Two factors are related to involvement:

Hedonic relevance: when the situation positively or negatively affects the person making the attribution. For example, when I have to evaluate a child on an exam: her behaviour will affect me, and my evaluation will be biased by my affective involvement.

Personality: is the tendency we have to make evaluations of others based on their reactions to us, interpreting them as dispositional features. For example, if our relationship with someone is very good, our assessment of what that person does will be very positive, regardless of their behaviour towards others.

The inference process

Social knowledge often involves going beyond the available information and requires making impressions, making judgments, or making inferences. “Inference”: “Process in which one goes beyond the available information, seeking to reach conclusions about data that is not completely contained in the data itself.

The inference and information gathering process

For Fiske and Taylor, “Inference is the central theme of social cognition. It is a process and a product“:

As a process: it involves deciding what information is gathered around a topic or issue, collecting such information and combining it in some way.

As a product: it comes to be the result of a reasoning process.

Some authors defend that when making any inference there would be a “model or normative theory” that would proceed in 3 phases:

Collection of information.
A sampling of the information.
Use and integration of information.

Information gathering

Under the regulatory model, the social recipient should scrutinize and weigh all relevant information before reaching a conclusion. However, the evidence shows that the decision-making process about which is the relevant information is strongly influenced by previous expectations (by the knowledge that has been activated).

This may be appropriate in some circumstances (job interview).

However, knowledge can be activated by various causes (accessibility, applicability and salience), which do not necessarily have to be appropriate for the situation in which we find ourselves.

Sampling the information

Once the person has decided what information is relevant, the data must be sampled (not everything that is known needs to be revealed).

Fiske and Taylor: When people are supplied with a suitable sample, we tend to use it fairly correctly. However, when we have to select the sample, we make mistakes:

Letting ourselves be carried away by extreme examples (Rothbart’s research: in observation, group B establishes, due to the presence of some other serious crime, a stronger association between group and crime).

Not paying enough attention to its size.

Using biased samples (Although we are warned of the biased nature of the sample, we continue to use it: Hamill, Wilson, and Nisbett investigation, interviewing a prison guard).

Use and integration of information

We use rules and mechanisms to combine the information that we have to carry out 2 fundamental types of operations:

Establishment of relationships between events: Consists of making a diagnosis of covariation. According to Fiske and Taylor, this diagnosis should take into account the relevant data. Furthermore, there is an influence of the previous beliefs of the social perceivers: illusory correlation.

Probability estimation

Research on how we perform these probability calculations has shown the occurrence of several phenomena:

Calculation bias due to the use of the representativeness heuristic: The representativeness heuristic is used when a person establishes the probability of an event by virtue of the degree to which:

It resembles in its essential properties the population to which it belongs.
It reflects the prominent characteristics of the process by which it was generated.

The representativeness heuristic provides a quick solution.

In many cases, probability and representativeness are correlated, however, the properties of an event are affected by many factors that have no impact on representativeness (deducing a high probability from high representativeness may be wrong).

Ignorance of previous probabilities (base rate).

Failures in the calculation of joint probabilities: Sometimes, we need to calculate the joint probabilities (probability of 2 events occurring together). To calculate it, the probabilities of occurrence of each event are multiplied separately, which explains why the joint probability is always less than the probability of the most probable event.

However, under certain circumstances, people predict a greater probability of occurrence of joint events than of each event separately. The Conjunction Fallacy.

Difficulties in handling diagnostic and non-diagnostic information: Diagnostic information: That which is related to the task to be performed. It does not seem logical that information without diagnostic value influences inference. However, a dilute or dilute effect has been found: If some diagnostic information is added to non-diagnostic information elements, the inferences become less extreme.

Corresponding inferences: Jones and Davis

One of the main objectives of inferences is to make predictions (internal or personal attributions are used to predict behaviour). Jones and Davis focus on the study of “corresponding inferences”: they directly infer the personal disposition or characteristic of the subject from the observed behaviour. For a corresponding inference to occur, a precondition must be met: intention.

Assigning specific characteristics to the person, DEPENDS ON A SERIES OF VARIABLES:

The uncommon effects of the action: Any behaviour produces different effects (Ex: Leaving training). The effects are common when they remain with both types of activities (abandon or not abandon training).

The uncommon effects are those that do change: the number of differentiating characteristics between 2 behaviours that can be chosen by the actor. The lower the number of effects not common to the two types of activities, the greater the probability of a corresponding inference.

The expectations about the actor: When the actor is observed performing anti-normative or socially undesirable behaviours, the probability of a corresponding inference increases.

Expectations can be:

Individual: According to the person’s prior knowledge.
Categorical: They come from the knowledge you have about the category or social group to which it belongs. The lack of confirmation of expectations leads to seeking explanations of a personal or internal nature.
Hedonic Relevance: If uncommon effects are abundant and/or expectations are confirmed, the probability of corresponding inferences will be low or nil.

However, there are exceptions for affective reasons: There is “hedonic relevance” when the actor’s behaviour has consequences that positively or negatively affect the people who carry out the attribution (observers).

Example: insult or aggressive action. The hedonic relevance, by strongly influencing affective aspects, reduces the number of uncommon effects that the observer perceives (increases the probability of corresponding inference).

Conclusions

In this blog post, we explained the Jones and Davis Correspondent Inference Theory. We discussed what is the inference and information gathering process, probability estimation and the base rate.

The correspondent inference theory is the generalization that is made about someone’s future behaviour, based on other previous internal and intentional attributions.

If you have any questions or comments on the content, please let us know!

References

Howard, J. A. (1985). Further Appraisal of Correspondent Inference Theory. Personality and Social Psychology Bulletin, 11(4), 467–477.

Kelley, E., & Sweat, S. (1983). Correspondent Inference: Theoretical Framework for Viewing Clothed Appearances. Clothing and Textiles Research Journal, 2(1), 49–55. https://doi.org/10.1177/0887302X8300200109

Weary, G., & Reich, D. A. (2000). Attribution theories. In A. E. Kazdin (Ed.), Encyclopedia of psychology (Vol. 1, pp. 320–325). American Psychological Association.

Was this helpful?

Thanks for your feedback!

Clinical psychology

Wellness

Psychotherapy

Personality

Correspondent Inference Theory( The complete guide)