The Science of Perception: Kelley’s Covariation Model Explained

The intricacies of human behavior have always intrigued psychologists and scholars alike. Among the many theories devised to explain why people behave the way they do, Harold Kelley’s Covariation Model stands out as a pivotal framework within attribution theory. This model, introduced in the late 1960s, revolutionized the understanding of how individuals attribute causes to behaviors, providing insights into the process of social perception and judgment.

Introduction to Attribution Theory

Attribution theory seeks to explain how individuals infer the causes of their own and others’ behaviors. It is a fundamental aspect of social psychology, as it delves into the cognitive processes behind judgments and decisions.

Harold Kelley explains:

“Attribution theory is a theory about how people make causal explanations, about how they answer questions beginning with ‘why?’ It deals with the information they use in making causal inferences, and with what they do with this information to answer causal questions” (Kelley, 1973).

Attribution can be categorized into two main types: internal attributions (dispositional) and external attributions (situational). Internal attributions ascribe behavior to personal traits, attitudes, or abilities, while external attributions link behavior to situational factors or external circumstances.

Scientific Analysis

Kelley’s model presupposes that individuals unconsciously use a pseudo scientific process in attributing cause to an entity (Försterling, 1989). The attributor possesses three different kinds of information that correspond to different causal possibilities. These are:

• consensus information: Do other actors behave in the same way to a given stimulus?
• distinctiveness information: Does the actor, and do other actors, behave in the same way to other stimuli?
• Consistency information: Does the actor, and do other actors, behave in the same way to the given stimulus across time and situational contexts? (Jones & Nisbett, 1972, p. 84).

The observer makes use of whatever available information at the moment to make a reasonable causal inference.

Kelley posits that the entity (the specific object, task, or situation that is the target of the attribution) will be attributed to the cause with which it covaries (Försterling, 1989). Accordingly, to make any logical covariance judgements, an individual must have information about the effect from two or more points in time. Accordingly, the more known points in time, the better the information for making accurate attributions.

The attributor, with knowledge of multiple points in time can draw conclusions from patterns based on the entity displays consistency, distinctiveness, and consensus.

See Attribution Theory for more on this theory

The Covariation Model Explained

Kelley’s Covariation Model posits that individuals make causal inferences by examining the covariation of behavior across different contexts and situations. Causal inference in attribution theory refers to the process of determining the cause of a particular behavior or event. When we observe something happening, we naturally seek to understand why it occurred

Kelley’s model is built on the premise that for something to be considered the cause of a behavior, it must be present when the behavior occurs and absent when it does not. This leads to the examination of the three critical types of information.

• Consensus: How other people behave in the same situation.
• Distinctiveness: How the person behaves in different situations.
• Consistency: How consistently the person behaves in the same situation over time.

For example, if someone consistently arrives late to work (high consistency) and other people in the same situation arrive on time (low consensus), we might attribute their lateness to internal factors, such as laziness or poor time management skills. However, if everyone in the organization is frequently late due to traffic or other external factors (high consensus), we might attribute the lateness to situational factors.

Consistency

Consistency refers to the regularity with which a particular behavior occurs in similar situations. It answers the question: Does the behavior happen every time the situation arises? High consistency indicates that the behavior is reliably linked to the situation, suggesting a potential causal relationship. For instance, if an employee is always late to work when it rains, there is high consistency in their behavior.

Prescott Lecky explains that consistency in behavior is an expression of self, meaning that different organisms may act to the same stimuli in different ways.

Lecky wrote:

“There is a coherence in the behavior of any single organism which argues against an explanation in terms of chance combinations of determiners and points to an organized dynamic system which tends toward self-determination. The recognition of the organism itself as a determiner is merely a tardy acknowledgment of the very obvious fact that each individual evaluates the world in his own terms” (Lecky, 1945, p. 79-82).

We cognitively desire to find consistency. We are motivated to uncover consistency in our attributions.

Jones and Nisbett explain:

“A person who is aggressive in one setting should, to be ‘consistent,’ be aggressive in other settings. To see dependence and independence in the same actor may lead to greater subtleties of categorization, or it may lead to misperception of the evidence so that it becomes more consistent. In short, all the cognitive mechanisms of inconsistency reduction can be put to work in the service of dispositional accounts of action” (Jones & Nisbett, 1972, p. 90).

We both seek factors of consistency but also create consistency through subjective interpretations of the facts. The element of consistency is a primary cause for bias and stereotyping.

See Self-Consistency Theory for more on this concept

Distinctiveness

Distinctiveness refers to how uniquely a particular individual reacts to a specific stimulus or situation compared to their reactions in different contexts. In other words, it assesses whether the person’s behavior is typical for them across various situations or if it stands out as unusual. Kelley posits that during causal attribution we determine whether an act reflects “some distinctive characteristic of the actor, and thus help us to identify the nature of his particular dispositions, or does it primarily reflect the evocative power of the environment?” (Kelley, 1972, p. 27).

For example, if someone usually laughs at jokes but does not laugh at one specific joke told by a friend, this low distinctiveness indicates that their non-laughter may be due to factors related specifically to that joke rather than an overall disinterest in humor. Conversely, if they consistently do not laugh at jokes (high distinctiveness), one might conclude that there’s something about the person or their general disposition affecting their response.

In summary, distinctiveness helps observers understand whether an individual’s behavior is influenced by external circumstances (like the uniqueness of a situation) or internal characteristics (such as personality traits). When assessing behavior through Kelley’s model, higher distinctiveness can lead observers to make situational attributions while lower distinctiveness might prompt dispositional attributions.

Consensus

Consensus refers to the extent to which others behave similarly in response to the same stimulus or situation. It assesses whether multiple individuals react in the same way when faced with a specific event.

When evaluating consensus, observers consider whether a behavior is shared among many people or if it is unique to one individual. High consensus occurs when most people respond similarly to a particular stimulus; for example, if everyone laughs at a joke, there’s high consensus regarding the appropriateness and funniness of that joke. In this case, an observer might attribute the laughter not just to personal traits but rather to characteristics inherent in the joke itself.

Conversely, low consensus exists when an individual’s reaction differs significantly from those of others. If only one person laughs at that same joke while everyone else remains silent, this suggests that their laughter may stem more from personal preferences or traits rather than any quality of the joke itself.

In summary, consensus provides insight into whether behavior can be attributed mainly to situational factors (high consensus) or individual dispositions (low consensus). By considering how widely shared reactions are among different people, observers can make more informed judgments about why someone acted as they did according to Kelley’s covariation model.

By considering these factors, we make inferences about whether the cause of a behavior is internal (dispositional) or external (situational).

Applying the Covariation Model

To illustrate the application of Kelley’s Covariation Model, consider the example of a student, Jane, who frequently falls asleep in her history class. To determine why Jane falls asleep, we would analyze the three dimensions as follows:

• Consistency: Does Jane fall asleep in history class every time she attends?
• Distinctiveness: Does Jane fall asleep only in history class, or does she also fall asleep in other classes?
• Consensus: Do other students in the history class also fall asleep?

If the analysis reveals high consistency (Jane always falls asleep in history), high distinctiveness (Jane only falls asleep in history), and high consensus (other students also fall asleep), we might attribute Jane’s behavior to the external factor of the history class being particularly boring or scheduled at a time when students are generally tired. Conversely, if there is low consensus (only Jane falls asleep), we might consider internal factors such as Jane’s lack of interest in the subject or personal fatigue.

More Examples of the Covariation Model

Example 1: A Student’s Poor Performance on an Exam

• High Consistency: The student consistently performs poorly on exams.
• Low Distinctiveness: The student performs poorly on all subjects.
• Low Consensus: Most other students performed well on the exam.

Attribution: Internal (dispositional) – The student is likely not very intelligent or doesn’t study hard enough.

Example 2: A Coworker’s Angry Outburst

• Low Consistency: The coworker is usually calm and collected.
• High Distinctiveness: The coworker only gets angry in specific situations, like when dealing with a particular client.
• High Consensus: Many other people also get angry when dealing with that same client.

Attribution: External (situational) – The coworker’s anger is likely due to the specific situation or the difficult client.

Example 3: A Friend’s Kindness

• High Consistency: The friend is always kind and helpful.
• High Distinctiveness: The friend is kind to everyone, not just a specific group of people.
• Low Consensus: Few other people are as kind and helpful as this friend.

Attribution: Internal (dispositional) – The friend is likely a kind and caring person.

Implications and Criticisms

Kelley’s Covariation Model has significant implications for understanding social perception and attribution. It offers a structured approach to discerning the causes of behavior, which can be applied in various contexts, from clinical psychology to organizational behavior. By identifying whether behaviors are due to dispositional or situational factors, the model aids in developing interventions, enhancing interpersonal relationships, and improving decision-making processes.

However, the model is not without its criticisms. One major critique is that it assumes individuals have access to sufficient and accurate information to make judgments. In reality, people often rely on limited or biased information, leading to erroneous attributions. Additionally, the model presumes that individuals systematically process information, which might not always be the case, as cognitive biases and heuristics can distort judgment.

A Few Words by Psychology Fanatic

Harold Kelley’s Covariation Model remains a foundational theory in social psychology, offering a robust framework for understanding how people attribute causes to behaviors. By examining consistency, distinctiveness, and consensus, the model provides valuable insights into the cognitive processes behind attribution, helping to unravel the complexities of human behavior. Despite its limitations, Kelley’s model continues to influence contemporary research and practice, underscoring the enduring relevance of attribution theory in the study of social perception and judgment.

Last Update: April 15, 2026

Associated Concepts

• Psychological Coherence: This psychological state refers to the ability to maintain stability and consistency in one’s thoughts, emotions, and behaviors. When a person experiences psychological coherence, their beliefs, values, attitudes, and actions are generally aligned and in harmony with each other.
• Self-Discrepancy Theory: This theory proposed by psychologist E. Tory Higgins, suggests that individuals have three specific representations of the self: the actual self, the ideal self, and the ought self.
• Explanatory Style: This concept, also known as cognitive or attributional style, refers to an individual’s habitual way of explaining and interpreting events or situations in their life. It is a concept derived from attribution theory and cognitive psychology. 
• Correspondent Inference Theory: This theory explains how people make inferences about others’ dispositions based on their observed behaviors. It suggests that people are more likely to attribute a behavior to a person’s stable, internal characteristics (such as personality traits) when the behavior is perceived as intentional, freely chosen, and has distinctive effects.
• Park’s Meaning Making Model: This model describes the process of assigning meaning to events to lessen conflict between experience and our global understanding and beliefs about the world and self.
• Cognitive Appraisal Theory: This theory, developed by Lazarus and Folkman, suggests that individuals create meaning through their appraisal of a situation, which is a key aspect of Park’s model.