Emotional Intelligence: What a Questionnaire Cannot Tell You
PsyCove Editorial · 11 min read · Updated 2026-10-11
- Ability EI and trait EI are different constructs that correlate only weakly and predict different things. Know which one a test measures before you read the result.
- The four working domains are noticing your own state, regulating it, reading other people, and using that reading to maintain a relationship. All four are skills, not identities.
- Self-report inflates. Instruments combat this with reverse-worded items, which is why some questions read backwards — that is a deliberate design feature, not filler.
- The effects of emotional skill on job performance are real but modest, and they concentrate in roles with heavy emotional demands rather than across the board.
- Training studies show gains, but the durability and the transfer to real-life behavior are still open questions. Treat any program's numbers as its own claims.
Where the idea came from, and why it split
The term entered psychology in 1990, when Mayer and Salovey proposed it deliberately: emotions carry information, and the capacity to perceive them, use them to think, understand what they mean, and manage them is an intelligence in the same sense that reasoning is. That was a narrow, testable claim. Over the following decade it split in two, and the split explains nearly all the confusion in the field.
The ability branch kept the original definition and the original method: performance tests with scored answers, the largest being the MSCEIT. The trait branch — what Petrides and Furnham later formalized as trait EI — kept the word 'emotional' and redefined the thing as a cluster of self-perceptions about emotional functioning, closer to how conscientiousness is a self-perception about getting things done. The two correlate only weakly. A person can genuinely be good at reading emotional information and still think of themselves as emotionally clumsy, and the reverse is just as possible.
Most of the public argument about emotional intelligence is two different constructs shouting the same word.
The four working domains
Noticing yourself
The first domain is the capacity to identify what you are feeling with enough precision to act on it. This sounds trivial and is not: the gap between 'I'm fine' and 'I'm disappointed and slightly relieved' is where most of what gets miscommunicated originates. People who skip this step tend to manage other people's emotions while remaining unable to locate their own, which is why their calm often feels performed.
Regulating yourself
The second domain is what you do after noticing. Regulation is not suppression — suppression is one strategy among several and one of the worse ones over the long run. Regulation includes delaying the response, reappraising the situation, redirecting attention, and in some cases simply allowing a feeling to run its course without acting on it. The distinguishing feature is that it is aimed at something: a decision you want to be able to make.
Reading other people
The third domain is interpersonal perception: noticing that the person across the table is not in the state their words imply. This is a perceptual skill with real accuracy limits — studies consistently find people are better at reading high-arousal states like anger and much worse at low-arousal ones like resentment or quiet disappointment. Assuming you are a good reader of subtle emotion is itself an error this domain is meant to correct.
Using it to maintain a relationship
The fourth domain is where the other three get used: deciding what to say, when to raise something, how to give bad news without softening it into meaninglessness, and how to repair a small rupture before it becomes a structural one. This is the domain most visible to others and least measurable by self-report, since the relevant information is entirely in the outcomes.
What a self-report instrument can and cannot do
This is worth being blunt about, because the boundary is routinely crossed. A self-report scale can tell you what you believe about your own emotional functioning. It cannot tell you whether you are actually good at it, and the gap is not small — self-other discrepancies on emotional perception are substantial, and people with a positive self-view systematically overestimate accuracy on exactly the low-arousal emotions that matter most in close relationships.
The second blind spot is social desirability. Answering 'I usually stay calm' costs nothing socially and gains something, so the tendency is strong. The standard countermeasure is reverse-worded items: statements where agreement indicates the less desirable answer, such as agreeing that other people's moods are hard to read. Reverse items create confusion if you are not warned, and they do depress scores a little — both are reasons they are worth having.
- What self-report is good for: locating which of the four domains feels hardest for you, and tracking change over months as you practice.
- What it cannot do: certify a level, rank you against other people, or predict how you will behave in a high-stakes conversation.
- The number is a position on a scale, not a grade. There is no pass mark and nothing to pass.
- If you want to know how good you are rather than how you see yourself, you need other people's ratings or a performance task — and those are harder to arrange, which is why they are rare.
Can these skills be trained?
The honest answer is: partly, with caveats. Reviews of intervention studies generally find that emotional skills can be moved by deliberate practice — that is not in serious dispute. What is in dispute is durability and transfer. Whether gains on a training's own measures survive months later, and whether they show up in actual conversations with actual people, is where the evidence is thin and the programs' marketing is loudest.
- Naming emotions precisely is the most reliably trainable skill, and also the one with the clearest everyday payoff — it is the first step in every other domain.
- Regulation strategies that require no external support (reappraisal, delaying, attention redirection) train better than strategies that depend on suppressing expression.
- Perspective-taking improves in the short term; whether it persists without practice is uncertain.
- Motivating yourself is the least trainable domain in these designs, which fits with what self-determination research says about intrinsic motivation being relatively stable.
- Anything that works must be practiced in real relationships. Skills trained in isolation do not transfer on their own.
The point of measuring this is not the score. It is finding out which of the four you keep skipping.
Emotional skill at work, and where it should stay out
Emotional skill predicts job performance — but modestly, and unevenly across roles. The pattern that shows up in the research is a U-shaped relationship with job complexity: it matters more at the low end, where you are coping with routine pressure, and at the high end, where you are managing people and ambiguity, and least in the large middle of self-contained technical work. The often-repeated claim that emotional intelligence predicts performance better than cognitive ability is not supported.
Two uses are not defensible. Selecting candidates on a trait EI score runs into the same problem as any other trait — modest effects combined with a false negative that falls entirely on the applicant. And sharing an individual's scores with their manager without explicit consent converts a self-description into a personnel record, which changes how honestly the next person answers the same questions.
What this does not tell you
It does not measure empathy in the moral sense, and a high score is not evidence of compassion. It does not predict relationship quality — that has more to do with how you repair than how you perceive. And it is not a diagnosis; nothing here identifies a disorder, and no score should be read as one.
It also does not rank people. The four domains are skills, and skills are unevenly distributed in everyone. The most useful thing a result can tell you is which domain to point your practice at — and the second most useful is that a low domain is not a fixed trait.
Frequently asked questions
Is emotional intelligence a form of intelligence?
Only in the ability branch, where it means the capacity to perceive, use, understand and manage emotion — and it does correlate with conventional cognitive ability, though far from perfectly. In the self-report branch, 'intelligence' is in the name only. Most online EI tests, including this one, measure self-perceptions rather than ability.
Why do some questions in this test read backwards?
They are reverse-worded items, and they exist to counteract social desirability — the tendency to describe yourself as calmer and more perceptive than you are. They lower scores slightly and can feel confusing, which is the trade-off. Answer them as honestly as the rest.
Can you improve emotional intelligence?
Partly. Reviews of intervention studies find that skills like naming emotions precisely and delaying response do move with deliberate practice. What remains uncertain is how durable those gains are and whether they transfer to real conversations without continued practice.
Does a high emotional intelligence score mean someone is kind?
No. The construct measures skill at perceiving and managing emotion, not compassion or warmth. High emotional skill can be used in service of other people or in service of your own advantage; nothing in the scale distinguishes those.
Is it worth taking an emotional intelligence test at all?
Worth it if you use the result to choose what to practice. Not worth it if you expect a verdict, a comparison with other people, or a hiring-grade assessment — self-report cannot deliver any of those three.
What is the difference between trait EI and ability EI?
Trait EI is self-reported emotional functioning measured by questionnaires, close in kind to personality traits. Ability EI is emotional processing capacity measured by performance tests with right answers, such as the MSCEIT. They correlate only weakly, so a test of one tells you very little about the other.
References
- Mayer, J. D., Salovey, P., & Caruso, D. R. (2004). Emotional intelligence: Theory, findings, and implications. Psychological Inquiry, 15(3), 197–215.
- Mayer, J. D., Salovey, P., & Caruso, D. R. (2008). Emotional intelligence: New ability or eclectic traits? American Psychologist, 63(6), 503–517.
- Petrides, K. V., & Furnham, A. (2001). Trait emotional intelligence: Psychometric investigation with reference to established trait taxonomies. European Journal of Personality, 15(6), 425–448.
- Petrides, K. V., Pita, R., & Kokkinaki, M. (2007). On the criterion and incremental validity of trait emotional intelligence. Cognition and Emotion, 21(1), 26–55.
- Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78.