Emotional intelligence at work: what the research supports and what it does not
Emotional intelligence matters at work, but less than the hype claims. What the evidence shows, where it is weak, and how to build the useful skills.
- Components of emotional intelligence in Goleman's model
- 5
- Studies in Mattingly and Kraiger's meta-analysis of EI training
- 58
- Corrected correlations between EI and job performance (O'Boyle et al., 2011)
- 0.24 to 0.30
Emotional intelligence (EI, often called EQ) is one of the most popular ideas in management training. It is also one of the most oversold. Claims that it matters more than IQ, or that it explains most of what separates great leaders from average ones, go well beyond the evidence.
That does not make it useless. Some parts of emotional intelligence predict job performance, especially in roles with heavy emotional demands, and the skills can be trained. This guide separates the evidence from the marketing and turns the useful parts into habits you can practise.
Three different things share one name
Much of the confusion comes from the fact that “emotional intelligence” refers to at least three different models. Wikipedia’s summary of the field sets them out:
- Ability model (Mayer and Salovey). EI is a set of mental abilities: perceiving emotions, using emotions to help thinking, understanding emotions and managing them. It is measured with tests that have better and worse answers, such as the MSCEIT.
- Trait model (Petrides). EI is how you see your own emotional abilities, measured by self-report questionnaires and treated as part of personality.
- Mixed model (Goleman). EI is a broad set of competencies. In his Harvard Business Review article “What Makes a Leader?”, Daniel Goleman lists five: self-awareness, self-regulation, motivation, empathy and social skill.
When someone quotes a striking result about EI, ask which model they mean. A finding about a self-report questionnaire tells you little about an ability test, and the reverse.
What the research supports
EI does relate to job performance, at a modest level. A 2011 meta-analysis by O’Boyle and colleagues found corrected correlations of roughly 0.24 to 0.30 between EI and job performance, and the trait and mixed models added some prediction beyond cognitive ability and the Big Five personality traits. Joseph and Newman (2010) found that perceiving and regulating emotions matter most in jobs with high emotional demands, such as customer-facing and care roles.
EI skills can be trained. A meta-analysis by Mattingly and Kraiger pooled 58 studies of EI training programmes and found a moderate positive effect on EI scores. Published studies showed bigger effects than unpublished dissertations, which is a common sign that the true effect may be somewhat smaller.
What the research does not support
- “EQ matters more than IQ.” The idea traces back to the subtitle of Goleman’s 1995 book, “Why it can matter more than IQ”. The research does not establish this as a general rule. For academic results, EI does not generally predict higher grades once IQ is taken into account.
- EI as a separate, new ability. Critics such as Edwin Locke argue EI is intelligence applied to emotions. A 2010 meta-analysis found that the predictive value of ability EI dropped to zero after controlling for the Big Five and IQ, and a 2017 meta-analysis found that trait EI is very close to a general factor of personality.
- Mixed-model EI as a precise measure. A 2015 meta-analysis reported that the predictive power of mixed EI disappears once achievement motivation, self-efficacy and self-rated performance are controlled for. In other words, part of what these questionnaires capture is ambition and confidence.
- EI tests for hiring decisions. One study of 111 business leaders found no correlation between their MSCEIT scores and how their employees rated them. Using an EI score to screen candidates is hard to justify.
Goleman’s own HBR article describes research across nearly 200 large companies, but it does not publish the method or effect sizes, so treat its stronger claims with care.
The useful core for managers
Strip away the branding and you are left with a set of observable skills that help in any people-facing job:
- Notice your own state. Recognise when you are frustrated, anxious or tired before it shows up in an email or a meeting.
- Pause before reacting. Build a short gap between a trigger and your response.
- Read other people. Pay attention to tone, energy and what someone is not saying.
- Check rather than assume. Ask “You seem quieter than usual, is something up?” instead of guessing.
- Manage the room. Choose how you show up when the team is under pressure, because your mood spreads.
You can practise each of these without taking any test, and you can judge progress by how conversations go.
A worked example
An example (illustration only). Daniel runs a small marketing agency. A client emails at 6pm rejecting a campaign his designer, Mei, spent two weeks on. Daniel’s first impulse is to forward the email to Mei with “We need to fix this tonight”.
Applying the skills above, he notices he is angry and tired, so he waits until morning. He rereads the email and sees that the client objects to two images, not the whole campaign. He calls Mei rather than emailing, opens by saying the client has concerns about two images and that the rest landed well, and asks how she feels about reworking them. Mei admits she has been worried about her workload, so they agree a revised deadline with the client.
Nothing here needed an EI score. It needed Daniel to notice his state, check the facts and pay attention to Mei.
Common mistakes
- Treating a profile as a fixed label. “I’m low on empathy” becomes an excuse instead of a skill to work on.
- Confusing niceness with EI. Managing emotions well sometimes means delivering hard news clearly, not avoiding it.
- Buying training on the strength of the hype. Ask providers which model they use, how they measure change and whether the change shows up in behaviour at work.
- Using EI against people. Telling a colleague they are “too emotional” shuts down a conversation that may need to happen.
When to focus on it
Put the most effort into these skills when your role involves conflict, customers, sensitive news or managing people through pressure. That matches where the evidence is strongest. For technical problem-solving, cognitive ability and expertise still do most of the work.
Emotional intelligence is a reasonable name for a set of trainable people skills. It is not a secret ingredient, and the most useful thing you can do with it is practise the behaviours rather than chase a score.
Framework
Five habits that make EI practical
- NoticeName your own emotional state before it drives your behaviour.
- PauseBuild a gap between a trigger and your response.
- ReadWatch tone, energy and what is left unsaid.
- CheckAsk about what you see instead of assuming.
- ChooseDecide how you want to show up, because your mood spreads.
Checklist
Put it to work
Quick quiz
Check what stuck
Where does EI predict job performance most strongly?
What happened to ability EI's predictive value after controlling for IQ and the Big Five in a 2010 meta-analysis?
What did a meta-analysis of 58 EI training studies find?
Recap cards
Tap a card to check yourself.
Sources and credits
