Emotional Intelligence: What It Is

The term emotional intelligence names a specific capacity, one that several theoretical models describe and that gets measured differently depending on which model is used.

What emotional intelligence is

Peter Salovey and John Mayer, who coin the term in 1990, define emotional intelligence as a set of abilities for recognizing and expressing one’s own emotions and other people’s, regulating them, and using them to guide thought and action. The authors place it within the framework of personal intelligences described by Howard Gardner, but narrow it: it concerns the recognition and use of emotional states, not general knowledge of oneself and others.

The American Psychological Association’s dictionary picks up this origin: emotional intelligence is a type of intelligence concerned with the capacity to process emotional information and apply it to reasoning. That definition sits close to how cognitivist psychology describes the mind more broadly — a system that processes information — applied here to emotions. Knowing what counts as an emotion in the first place is the step before that, and the Recap on human emotions covers it.

Practical example: a student who notices they feel anxious before an oral exam, realizes the anxiety comes from a fear of getting something wrong, and uses that awareness to review calmly instead of spiraling, is exercising exactly the abilities Salovey and Mayer describe: noticing the emotion, making sense of it, regulating it.

The four-ability model, revised in 1997

In 1997 Mayer and Salovey revise their original definition into four abilities, according to the APA Dictionary of Psychology: accurately perceiving and appraising emotions; accessing emotions and using them when they help thinking; understanding emotional language; and regulating one’s own emotions and other people’s to support growth and well-being.

These four abilities remain the reference point for test-based measurement tools: questions with a right answer and a wrong one, much like any other cognitive task. The self-report questionnaires used for the other models work differently, as the next sections show.

Goleman’s model

Salovey and Mayer’s ideas are carried to a wide audience by Daniel Goleman, a science journalist and psychologist, through a hugely successful book that revises the original definition to include a range of personality variables, according to the APA Dictionary of Psychology. In Goleman’s mixed model, as described by the review from Bru-Luna and colleagues (2021), emotional intelligence comprises five elements: recognizing one’s own emotions, managing them, self-motivation, recognizing other people’s emotions, and managing relationships. These are the competencies Goleman links to managerial performance and leadership.

A related but independent model comes from Reuven Bar-On: five components — intrapersonal, interpersonal, adaptability, stress management and general mood — covering a broader set of non-cognitive abilities and competencies.

The trait model, and how it’s all measured

Petrides and Furnham propose a different model: a stable personality trait, spread across fifteen dimensions grouped into four factors (well-being, self-control, emotionality, sociability), according to the systematic review by Bru-Luna and colleagues (2021). It is measured through self-report questionnaires on typical behavior.

Model	Authors	Nature	How it’s measured
Ability	Mayer and Salovey	Cognitive capacity	Right-or-wrong tests
Trait	Petrides and Furnham	Personality trait	Self-report questionnaire
Mixed	Goleman	Abilities and competencies	Self-report questionnaire
Mixed	Bar-On	Abilities and competencies	Self-report questionnaire

Despite the differences between models, most measurement tools share a few traits: they produce a total score together with scores for each dimension. The dimensions also overlap quite often, since most instruments end up touching on how emotions get noticed, handled and put to use, according to the same review. Trait-based instruments stand out as good predictors of coping style when adults or children face everyday stress.

Emotional intelligence, IQ and success at work

According to the review by Cary Cherniss (2000), IQ on its own explains a limited share of the variance in job performance — estimated at between 4% and 25%, depending on the study cited. In one study Cherniss reports, following a group of science PhDs over forty years, social and emotional skills turned out to be four times more predictive than youthful IQ scores of the career success and prestige those same people reached later on. For a wider view of the classic theories of intelligence and the tests built to measure it, the distinction between the two research traditions helps make sense of these numbers.

Cherniss reports that both Goleman and the trio of Mayer, Salovey and Caruso treat the construct, taken on its own, as a modest predictor of job performance. They describe it instead as the foundation for concrete emotional competencies — what Goleman calls emotional competence, distinct from the underlying ability — and it is the competence that correlates with stronger performance. Cherniss also flags an academic critique from Davies, Stankov and Roberts (1998), who argue the construct adds nothing empirically new beyond what other scales already capture, and he concedes that, as of 2000, research on the predictive validity of the measurement tools was still thin.

Practical example: two managers with identical IQ test scores can perform very differently at work if one notices a team’s frustration building before it turns into open conflict, and the other doesn’t. By Cherniss’s own distinction, it’s that concrete emotional competence that makes the difference, not the IQ score on its own.