Emotional Intelligence and Managerial Decision-Making: The Ability Model, the Popular Version, and What the Evidence Actually Supports

The term "emotional intelligence" entered academic psychology through Peter Salovey and John Mayer's 1990 article in Imagination, Cognition, and Personality, which defined it narrowly: a set of abilities contributing to the accurate appraisal and expression of emotion in oneself and others, the effective regulation of emotion, and the use of feelings to motivate thought and action. Five years later, Daniel Goleman's 1995 book Emotional Intelligence: Why It Can Matter More Than IQ took the phrase to a mass audience and, in the process, changed what it referred to. Goleman's version folds in persistence, impulse control, empathy, social skill and several other qualities that read, to a personality researcher, like established trait and competency dimensions rather than a cognitive ability.

This is not a minor difference in emphasis. Salovey and Mayer's model, later elaborated into a four-branch structure (perceiving emotion, using emotion to facilitate thought, understanding emotion, and managing emotion) and operationalized in the Mayer-Salovey-Caruso Emotional Intelligence Test, treats emotional intelligence as an ability: something measured by giving a person an emotion-related problem with a demonstrably better and worse answer, the same logic used to measure any cognitive ability. Goleman's model, and the commercial self-report inventories built on similar lines, treat emotional intelligence as a mixture of ability, personality trait and learned competency, measured by asking people to rate themselves on items such as how well they handle setbacks or read a room. A test that asks someone to solve a problem is measuring something structurally different from a test that asks someone to describe themselves, even when both tests use the same name for what they are measuring. Confusing the two, treating a self-report questionnaire result as evidence of an ability-test finding or vice versa, is probably the single most common analytical error in management writing on this topic, and it matters because the evidence base for the two models does not transfer cleanly between them.

A measurement instrument has construct validity when it demonstrably measures the thing it claims to measure, rather than something else that happens to correlate with it. This is where the mixed, self-report model of emotional intelligence runs into its most serious and least publicized problem. Moshe Zeidner, Gerald Matthews and Richard Roberts, and earlier Michael Davies, Lazar Stankov and Richard Roberts in a widely cited 1998 article in the Journal of Personality and Social Psychology, examined several available emotional intelligence measures against established personality and cognitive-ability instruments and found that a number of them overlapped substantially with existing, already-named personality traits, particularly dimensions resembling emotional stability and extraversion from the five-factor model.

The practical implication is uncomfortable for anyone selling emotional intelligence as a distinct, trainable managerial capability separate from personality. If a self-report EI score is largely redundant with a personality questionnaire a company could administer more cheaply and with a longer validation history, then a program built around measuring and developing "emotional intelligence" through self-report may be re-labeling existing personality variation rather than identifying and building a new skill. This does not mean self-report EI measures are worthless. It means a manager evaluating a vendor's emotional intelligence assessment should ask a specific question before buying it: has this instrument been checked against the Big Five and against cognitive ability, and how much of its variance survives once those are accounted for? Few vendors volunteer that comparison, and the answer, where it has been checked, is frequently that less is left over than the marketing implies.

Setting the inflated claims aside, what does the accumulated research actually show about emotional intelligence and job or managerial performance? Dana Joseph and Daniel Newman's 2010 meta-analysis proposed and tested a cascading model in which emotion perception supports emotion understanding, which in turn supports emotion regulation, with regulation most directly tied to performance outcomes. Their central, carefully qualified finding was not a single flat correlation but a moderated one: ability-based emotional intelligence related positively to job performance in roles with high emotional-labor demands, the kind of work where managing one's own displayed emotion and reading others' emotional states is a core task requirement, such as customer-facing management or healthcare supervision. In roles with low emotional-labor demands, the relationship was inconsistent and in some analyses close to zero or reversed. That is a specific, falsifiable, moderated claim, a considerable distance from "emotionally intelligent people perform better at work," and it is the more defensible version of the claim precisely because it specifies where the relationship holds and where it does not.

Ernest O'Boyle and colleagues' 2011 meta-analysis in the Journal of Organizational Behavior took a broader approach, coding studies into three streams, ability-based, self-report, and mixed-model EI, and examining all three against job performance using a larger pooled sample than earlier reviews. All three streams showed positive, but not large, associations with job performance, and the study's dominance analysis found that emotional intelligence, across its streams, contributed incremental explanatory value alongside the five-factor personality model and cognitive ability rather than substituting for either. Read together, these two meta-analyses support a specific, bounded claim: emotional intelligence, measured carefully and by a method appropriate to the model being tested, relates to job performance at a level worth taking seriously in roles where managing emotion is part of the job, as one input among several, not as a master variable that outweighs ability or conscientiousness. A manager who hears "EI explains most of what separates great managers from average ones" is hearing something the primary research does not say.

The connection between emotional intelligence and managerial decision-making is often asserted vaguely, as though "being good with emotions" simply makes decisions better in some unspecified way. The more precise and more useful connection runs through a documented mechanism: the affect heuristic, described by Paul Slovic, Melissa Finucane, Ellen Peters and Donald MacGregor in a 2007 synthesis in the European Journal of Operational Research. Their account proposes that people attach a fast, automatic affective tag, good or bad, pleasant or unpleasant, to options and outcomes, and that under time pressure, high stakes, or genuine uncertainty, this affective tag substitutes for a slower, more effortful weighing of costs, benefits and probabilities. The tag is not irrational in the sense of being random; it draws on real prior experience and association. It is, however, a different cognitive process from deliberate analysis, and it operates fastest exactly when a manager has least time to check it.

This is where emotional intelligence, in its ability-model sense, has a plausible and testable connection to managerial decision quality that does not depend on inflated claims about overall performance. A manager who can accurately perceive their own rising frustration or anxiety during a high-stakes negotiation, the first branch of Mayer and Salovey's model, has a chance to notice that an affective tag is operating before it fully substitutes for analysis. A manager who can regulate that emotional state, the fourth branch, has a chance to slow down enough to check whether the felt sense of "this is a bad option" survives contact with the actual numbers. Neither step guarantees a better decision. Both describe a specific, mechanistic route by which an ability genuinely captured by the Mayer-Salovey-Caruso framework could plausibly improve decision quality under pressure, which is a more defensible and more useful claim than asserting that emotional intelligence improves decisions in general.

Sigal Barsade and Donald Gibson's 2007 review in the Academy of Management Perspectives, summarizing a substantial body of organizational research on affect, extends the picture beyond any single manager's internal state. Moods and emotions in the workplace are not confined to the person experiencing them; they are observably contagious across a team, and a manager's visible affect during a meeting measurably shapes the judgment, risk tolerance and cooperative behavior of the people in the room, not only their own decision. A manager who walks into a meeting visibly anxious after the plant scenario above is not only managing their own affect heuristic; they are setting an affective tone that can shift how subordinates weigh the same information.

This reframes what "managing emotion" means for a manager making decisions under pressure. It is not solely a private, internal act of self-regulation performed for the sake of one's own judgment. It has a visible, structural effect on the decision environment other people are reasoning within, because the people around the manager are themselves subject to the same affect heuristic, picking up ambient emotional cues and letting them color their own read of the situation. A manager's emotional self-management during a crisis decision is, among other things, a form of decision-environment design for everyone else present, whether or not the manager intends it that way.

This case is fictional and was constructed for teaching. The organization, people and figures are invented.

Return to the two branch managers from the opening of this lesson. Shilpa Bank\'s regional office issues an overnight directive eliminating a long-standing overtime allowance at every branch, effective in two weeks, with no advance warning to branch staff. Farhana, managing the Uttara branch, reads the directive, notices her own immediate anger at being given no lead time to prepare her team, names that feeling to herself explicitly, and takes twenty minutes before calling a meeting to draft three likely objections and a direct answer to each. Kamal, managing the Mirpur branch two kilometers away, reads the same directive, feels the same anger, and calls an all-staff meeting within the hour while still visibly agitated, announcing the change without further explanation and leaving when a teller asks a clarifying question.

Three distinctions carry more practical weight than any single piece of advice this lesson could offer. First, "emotional intelligence" is not one construct; the ability model tested by performance tasks such as the MSCEIT and the mixed, popularized model built from self-report inventories rest on different measurement logic and different evidence, and a claim proven for one does not automatically transfer to the other. Second, the genuine evidence, once separated from the popularizing claims that followed Goleman's 1995 book, supports a modest and moderated relationship between EI and job performance, strongest in emotionally demanding roles and weaker or inconsistent elsewhere, alongside an unresolved overlap between self-report EI and existing personality traits that any serious evaluation of a commercial EI assessment should ask about directly. Third, the connection to decision-making under pressure is not a vague appeal to "soft skills"; it runs through a specific, named mechanism, the affect heuristic, in which a fast emotional tag can substitute for deliberate analysis exactly when the stakes and the time pressure are highest, and emotion perception and regulation describe a concrete, plausible way a manager can notice and check that substitution before it fully drives a decision.

The unresolved issue worth carrying forward is this: none of the research reviewed here tells a manager how to build the regulation skill the ability model describes, only that it is conceptually distinct from personality and that it plausibly matters most in high-pressure, high-stakes moments. An organization that wants managers who pause before reacting to bad news needs a development approach built around that specific behavior, noticing and checking an emotional reaction before it becomes a decision, rather than a workshop built around a self-report score whose relationship to actual on-the-job behavior remains, on the evidence reviewed in this lesson, genuinely uncertain.

The checklist below is a reflection tool built for this lesson, not a validated psychometric instrument, and it does not produce an emotional intelligence score. The Mayer-Salovey-Caruso Emotional Intelligence Test is a licensed, professionally administered performance measure; nothing presented here substitutes for it. The value of the checklist is in noticing, in one's own recent decisions, whether the specific behavior this lesson describes, perceiving an emotional reaction and checking it before acting on it, is actually happening.

Identify a real, specific managerial decision, made by you or by someone you have observed closely, that was made under meaningful time pressure, uncertainty, or emotional strain. Examples include a decision to discipline, promote, or dismiss someone; a crisis response to a safety, financial, or reputational problem; or a high-stakes negotiation. You may disguise the organization and the individuals involved.

Prepare a report of 2,500 to 3,000 words that:

1. describes the decision, the time pressure and uncertainty actually present, and the information that was available but possibly not used at the moment of deciding;
2. analyzes the decision using the distinction between the ability model and the mixed model of emotional intelligence, stating explicitly which model's logic, if either, best explains what happened and why;
3. identifies whether and how the affect heuristic, as described by Slovic, Finucane, Peters and MacGregor, appears to have operated in the decision, naming the specific affective tag involved and what it substituted for, if anything;
4. evaluates what a self-report emotional intelligence score would and would not have predicted about this specific decision, given the construct-validity concerns and the modest, moderated meta-analytic evidence covered in this lesson;
5. recommends a specific practice the decision-maker, or someone in a comparable position, could use in future to notice and check an emotional reaction before it substitutes for analysis, stating exactly what the practice requires and when it should be used.

Use at least five academic sources, including at least three from this lesson's reference list. State explicitly where your evidence is direct observation, secondhand report, or inference, and do not present a correlation as proof of a cause.

Assessment rubric (100 marks)

1. Barsade, S. G., & Gibson, D. E. (2007). Why does affect matter in organizations? Academy of Management Perspectives, 21(1), 36-59.
2. Davies, M., Stankov, L., & Roberts, R. D. (1998). Emotional intelligence: In search of an elusive construct. Journal of Personality and Social Psychology, 75(4), 989-1015.
3. Goleman, D. (1995). Emotional intelligence: Why it can matter more than IQ. Bantam Books.
4. Joseph, D. L., & Newman, D. A. (2010). Emotional intelligence: An integrative meta-analysis and cascading model. Journal of Applied Psychology, 95(1), 54-78. https://doi.org/10.1037/a0017286
5. Mayer, J. D., & Salovey, P. (1997). What is emotional intelligence? In P. Salovey & D. Sluyter (Eds.), Emotional development and emotional intelligence: Educational implications (pp. 3-31). Basic Books.
6. Mayer, J. D., Salovey, P., & Caruso, D. R. (2002). Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT) user's manual. Multi-Health Systems.
7. O'Boyle, E. H., Jr., Humphrey, R. H., Pollack, J. M., Hawver, T. H., & Story, P. A. (2011). The relation between emotional intelligence and job performance: A meta-analysis. Journal of Organizational Behavior, 32(5), 788-818. https://doi.org/10.1002/job.714
8. Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition, and Personality, 9(3), 185-211.
9. Slovic, P., Finucane, M. L., Peters, E., & MacGregor, D. G. (2007). The affect heuristic. European Journal of Operational Research, 177(3), 1333-1352.

Verification notes: all nine sources were confirmed via web search result snippets giving matching author, year, journal, volume and page or DOI records (and, for the Joseph and Newman and O'Boyle et al. meta-analyses, a hosted PDF copy in the former case) rather than invented from memory. The Davies, Stankov and Roberts (1998) personality-overlap finding and the O'Boyle et al. (2011) three-stream pattern are described in general terms consistent with multiple independent summaries rather than quoted with an invented precise decimal correlation, since an exact figure could not be confirmed from the available snippets for either. The Joseph and Newman (2010) emotional-labor moderation finding was confirmed with enough specificity (high versus low emotional-labor jobs, inconsistent or near-zero relationship in the latter) to be stated directly. The Shilpa Bank case and the plant-contamination decision scenario are original teaching material for this series; the organizations and individuals are invented.

Next in The Human Capital MBA series: Article 29 examines communication, delegation, coaching and feedback as managerial practices.