Harold Kelley

Introduction

Harold Harding Kelley (February 16, 1921 – January 29, 2003) was an American social psychologist whose work shaped two of the field's central research programmes: the study of how people explain behaviour, and the formal analysis of how people in relationships depend on one another. His covariation model gave attribution research its first rigorous logical framework, specifying how observers combine information about consensus, distinctiveness, and consistency to decide whether an action says something about a person, a stimulus, or the circumstances. His collaboration with John W. Thibaut produced interdependence theory, a systematic account of social interaction built from the patterns of control and outcome that situations create. Later in his career Kelley helped establish close relationships as a legitimate topic for mainstream social psychology. A 2002 review of citation impact ranked him the forty-third most cited psychologist of the twentieth century.

Early Life and Education

Kelley was born in Boise, Idaho. When he was ten his family moved to the rural town of Delano, California, where his father established a vineyard. He attended Bakersfield Junior College before transferring to the University of California, Berkeley, where he earned a B.A. in psychology in 1942 and an M.A. in 1943. He married his high school sweetheart, Dorothy, and the couple had three children and, eventually, five grandchildren.

Like most social psychologists of his generation, Kelley's graduate education was interrupted by the Second World War. He served in the Aviation Psychology Program of the U.S. Army Air Forces, helping to construct and validate selection tests and to analyse the performance of air crews — work that gave him a lasting appreciation for measurement problems and for behaviour studied under real constraints. Advised by his aviation mentor Stuart Cook to continue his education, he enrolled at the Massachusetts Institute of Technology in the Research Center for Group Dynamics, then directed by Kurt Lewin. Kelley received his Ph.D. in group psychology from MIT in 1948. When Lewin died in 1949 the Center moved to the Institute for Social Research at the University of Michigan, and Kelley went with it for another year.

Academic and Professional Career

Kelley's first academic post was an assistant professorship at Yale University, which he took up in 1950. There he joined Carl Hovland's programme on communication and attitude change and, with Hovland and Irving Janis, produced Communication and Persuasion (1953), a landmark volume in the Yale attitude-change tradition. In 1955 he moved to the University of Minnesota as a professor, affiliated with the Laboratory for Research in Social Relations.

In 1961 Kelley joined the University of California, Los Angeles, where he spent the rest of his career until retiring in 1991, serving as departmental vice chair and as chair of the social psychology area. He remained professionally active as an emeritus professor. His UCLA years produced the attribution work for which he is best known and, later, his research on close relationships.

The defining collaboration of Kelley's professional life was with John W. Thibaut, beginning in 1953 and lasting until Thibaut's death in 1986. It started when the two were invited to write a chapter on group problem-solving for the Handbook of Social Psychology; the chapter, updated in 1968, led in turn to The Social Psychology of Groups (1959) and, much later, to Interpersonal Relations: A Theory of Interdependence (1978). Kelley's own modest description of the result was "a stable focus on phenomena at the group level," whose combinatorial nature, he said, brings order to numerous interpersonal and intergroup phenomena.

His contributions were recognised with the American Psychological Association's Distinguished Scientific Contribution Award (1971), the American Sociological Association's Cooley-Mead Award, and distinguished-scientist awards from the Society of Experimental Social Psychology and the Society for the Psychological Study of Social Issues. He was elected to the American Academy of Arts and Sciences and, in 1978, to the National Academy of Sciences. He died of cancer at his home in Malibu, California, on January 29, 2003.

Major Contributions to Psychology

First Impressions and the Warm–Cold Variable

Before turning to attribution, Kelley made an influential experimental contribution to person perception. Working from Solomon Asch's finding that certain traits reorganise an impression as a whole, Kelley told students that a guest lecturer was either "warm" or "cold" while holding every other aspect of the lecture constant. Students who expected a warm person rated him more favourably and participated more in discussion. The study showed that a single trait label could restructure an entire impression and steer subsequent behaviour toward the person, a finding that anticipated later work on expectancy effects.

The Covariation Model of Attribution

Kelley built on Fritz Heider's insight that people act as intuitive scientists seeking the causes of behaviour, and asked what information such a scientist would need. His answer, set out in a 1967 chapter and elaborated in 1971 and 1973, was that observers reason from three kinds of covariation:

  • Consensus: do other people behave the same way toward this stimulus? Low consensus points toward the person.
  • Distinctiveness: does this person behave the same way toward other stimuli? High distinctiveness points toward the stimulus.
  • Consistency: does this person behave the same way toward this stimulus on other occasions and in other settings? Low consistency points toward transient circumstances.

The classic internal attribution — "it's him" — follows from low consensus, low distinctiveness, and high consistency; the classic external attribution — "it's the thing" — from high consensus, high distinctiveness, and high consistency.

Causal Schemata and the Configuration Model

Kelley recognised that people rarely possess all three kinds of information. For single observations he proposed that perceivers fall back on causal schemata, learned beliefs about how causes combine: the multiple-necessary-causes schema (a marathon win requires fitness, motivation, and training) and the multiple-sufficient-causes schema (a failed drug test may have more than one adequate explanation). He also formulated principles for weighing competing causes, including the discounting principle (a plausible second cause lowers confidence in the first), the augmentation principle (a cause looks stronger when an inhibitory factor is also present), and the compensation principle (causes combine additively in producing an effect).

Interdependence Theory

With Thibaut, Kelley developed a formal analysis of interaction based on the situation's payoff structure rather than on personality traits or internalized norms. Its core concepts include outcomes, the rewards and costs of a given interaction; the comparison level, the standard against which a person judges whether an outcome is satisfactory; the comparison level for alternatives, the worst outcome a person will accept given other options; and the matrix of interdependence, which specifies how each person's behaviour affects both parties' outcomes. Later formulations, including the Atlas of Interpersonal Situations (2003), added the idea of transformation of motivation — the psychological process by which a person moves from the immediate, self-interested "given" matrix to an effective matrix shaped by norms, long-term goals, and concern for the partner.

Attribution and Close Relationships

In the 1970s Kelley turned to how couples explain each other's behaviour, arguing that satisfying relationships are characterised by attributions that credit the partner and excuse the partner's failings, whereas distressed couples make "distress-maintaining" attributions. His 1979 book Personal Relationships and the 1983 collaborative volume Close Relationships gave the emerging field both a theoretical framework and an agenda, covering attraction, love, commitment, power, and conflict. The work helped found the International Society for the Study of Personal Relationships.

Major Publications

  • Communication and Persuasion (with Carl I. Hovland and Irving L. Janis, 1953)
  • The Social Psychology of Groups (with John W. Thibaut, 1959)
  • "Attribution Theory in Social Psychology" (1967)
  • Attribution in Social Interaction (1971)
  • Interpersonal Relations: A Theory of Interdependence (with John W. Thibaut, 1978)
  • Personal Relationships (1979)
  • Close Relationships (with Ellen Berscheid, Andrew Christensen, John H. Harvey, Ted L. Huston, George Levinger, Evie McClintock, Letitia Anne Peplau, and Donald R. Peterson, 1983)
  • An Atlas of Interpersonal Situations (with John G. Holmes, Norbert L. Kerr, Harry T. Reis, Caryl E. Rusbult, and Paul A. M. Van Lange, 2003)

Impact and Legacy

The covariation model made attribution research operational: it turned a philosophical question about causation into a set of testable predictions, and for roughly two decades it defined the field's agenda. Its vocabulary — consensus, distinctiveness, consistency — remains standard in textbooks, and the distinction between internal and external causes that Kelley systematised underlies later work by Bernard Weiner on achievement attributions and by Lee Ross on the fundamental attribution error. Interdependence theory has proved similarly durable; its situation-based taxonomy is still used to classify social situations in research on cooperation, trust, and power.

The model has also attracted substantial criticism, much of it now standard. Leslie McArthur's 1972 test found that attributions broadly followed Kelley's rules but that the three information types were not weighted as his model implied: consistency dominated, distinctiveness came second, and consensus was largely ignored. More fundamentally, critics have argued that the covariation model is better read as a normative ideal than as a description of everyday reasoning. Studies by Lalljee and colleagues and by Ahn and colleagues found that people asked what information they would want rarely request covariation information, and prefer information about mechanisms and reasons; reviewers have concluded that there is little evidence that perceivers spontaneously gather consensus and distinctiveness data before explaining an action. Bertram Malle has pressed the deeper objection that the internal–external dichotomy suits unintentional events but does not capture how people explain intentional action, where explanation proceeds by way of reasons rather than causes. Reviews by Försterling and by Sutton and McClure, however, found that when covariation information is explicitly supplied, judgements do follow Kelley's principles reasonably well — which is why the model survives as a partial rather than a complete account.

Kelley anticipated several of these objections himself. The causal schemata, the configuration model, and the discounting and augmentation principles were all developed precisely because perceivers usually lack full covariation data. Interdependence theory has drawn a related line of criticism from Margaret Clark and Judson Mills, who argued that Thibaut and Kelley's exchange framing misdescribes close relationships: in communal relationships benefits are given in response to a partner's needs and without tracking of debts, not in expectation of comparable return. Kelley's later concept of transformation of motivation was in part a response to this class of objection, since it allows the effective preferences guiding behaviour to diverge from the situation's immediate payoff structure. Both debates remain open, and Kelley's own practice of revising his formalisms in the face of counterevidence is part of why his frameworks are still in use.

Learn More

References

  • Kelley, H. H. (1967). Attribution theory in social psychology. In Nebraska Symposium on Motivation (Vol. 15, pp. 192-238). University of Nebraska Press.

Last updated: 2026-10-20

Menu
Loading