Emotional Intelligence Self-Assessment

Emotional intelligence is the ability to recognise, understand, manage and use emotions, your own and other people’s. Unlike measured intelligence, the skills involved are widely held to be developable at any age, which is most of the reason anyone bothers measuring them.

Two models, and they disagree about what is being measured

The ability model (Mayer, Salovey and Caruso) treats emotional intelligence as a genuine cognitive ability with right and wrong answers, tested the way a reasoning test is. Four branches:

1. Perceiving emotions: reading emotional signals in faces, voices and posture
2. Using emotions: recruiting the right mood for the task at hand
3. Understanding emotions: knowing how feelings blend, escalate and shade into one another
4. Managing emotions: regulating your own responses, and influencing other people’s without manipulating them

The mixed model (Goleman) folds in motivation and social skill and measures the whole thing by self-report. That is the five-domain version, and it is what this page uses, because a self-report questionnaire cannot do the ability model justice. Anything scored by asking you about yourself is measuring your self-image alongside the trait.

How this page scores it

Five statements, one per domain, each rated 1 to 5. Total 5 to 25, reported as a percentage of the 25-point maximum.

Worked example

Self-awareness 4, self-regulation 3, motivation 5, empathy 2, social skills 4.

Total = 4 + 3 + 5 + 2 + 4 = 18 of 25, or 72%, which lands in the solid band.

The total is the least useful number on the page. Empathy at 2 against motivation at 5 is the finding: a driven person who is not reading the room. Averaging that into 72% hides exactly the thing worth working on, which is why the result panel reports every domain separately and names the weakest one.

A caution about the 58% figure

If you search for this topic you will quickly meet the claim that emotional intelligence “accounts for 58% of job performance”. It comes from TalentSmart, a company that sells emotional-intelligence tests, and it has never appeared in a peer-reviewed journal.

The peer-reviewed picture is more modest and more interesting. Meta-analyses (O’Boyle and colleagues, 2011) find emotional intelligence does predict job performance, and does add something beyond general mental ability and the Big Five personality traits, but the incremental contribution is small. It matters most in jobs with high emotional labour: nursing, sales, teaching, customer-facing work, management. In jobs that are mostly solitary technical work, it barely registers.

None of that makes the domains below not worth developing. It makes the sales pitch worth distrusting.

What actually improves

Self-regulation and empathy respond best to deliberate practice, and both have a specific technique with evidence behind it. For regulation it is cognitive reappraisal, changing your reading of the situation before the response builds rather than suppressing it afterwards. For empathy it is active listening: reflecting back what you heard before replying, which feels artificial for about two weeks and then stops feeling like anything.

Self-awareness is the awkward one. People are poor judges of their own self-awareness, so the domain this questionnaire asks about first is the one a questionnaire is worst at measuring. If your self-awareness score is your highest, that is worth a second look rather than a note of satisfaction.