How to Read Emotional Intelligence Test Scores

How your four sub-scores were calculated

The assessment is twenty statements, five for each of four skills. Each statement is answered on a five-point agreement scale, from strongly disagree to strongly agree. The five answers belonging to a skill are averaged, and that average is rescaled to run from 0 to 100 so the four are comparable with each other and with other assessments that use a different response scale.

Some of the statements are reverse-keyed, which matters for reading your own answers back. "I can't name what I'm feeling as it happens" and "I'm often surprised to hear I upset someone" both count backwards: agreeing with them lowers the self-awareness score rather than raising it. This is standard practice and it exists to break the habit of answering every item the same way, but it means you cannot reconstruct your score by counting how often you agreed.

Two consequences follow for interpretation. Unanswered items are skipped rather than counted as zero, so an incomplete run gives you a score based on fewer statements rather than an artificially low one. And a sub-score built from five items is not precise to the individual point — the difference between a 62 and a 68 is one answer moved one step, which is well inside the noise of how you happened to read a sentence that day. This is why the output is bands. A 62 and a 68 are the same result; a 35 and a 75 are not.

What each of the four skills is, and what a low score there costs

Self-awareness is noticing what you feel while you are feeling it, rather than reconstructing it afterwards. A low score here usually does not present as an emotional problem at all. It presents as being tired and short with people without knowing why, or as being told you were upset about something when you experienced yourself as fine. The cost is that nothing downstream can work properly: you cannot regulate or explain a feeling you have not identified yet.

Self-regulation is what you do with the feeling once you have it. This is the skill most people mean when they talk about emotional intelligence, and it is genuinely separable from noticing — knowing you are angry and choosing what to do next are different capacities. A low score with high self-awareness is a specific and common pattern: the feeling is fully legible to you and it still arrives in your tone before you have decided to let it.

Reading others is accuracy about other people's emotional states, including the ones they are not saying out loud. It dissociates from the first two more than any of them do from each other. A low score here is expensive in a particular way, because misreading is invisible from the inside — you find out weeks later that someone was upset the whole time, which means there was no moment at which you could have corrected.

Social response is what you do with what you read, and it is where accuracy either becomes useful or does not. Solving someone's problem before acknowledging it, or apologising with a defence built into the apology, are failures of response rather than of perception. A low score here alongside a high reading score is the frustrating combination: you can see exactly what is happening and you keep doing the unhelpful thing about it.

The profile shapes worth recognising

High noticing with low regulation is the most common uneven profile, and it is the one where the popular advice is least useful. Being told to become more self-aware is redundant when self-awareness is already the strong side. The work is in the gap between recognising a state and acting differently in it, which responds to specific regulation strategies — changing the situation, redirecting attention, reframing what the event means — rather than to further introspection.

High reading with low self-awareness is the profile of someone who has usually been described as perceptive since childhood. It is coherent and it is not a contradiction: attending closely to other people's states is a skill that can develop precisely because attending to your own was not safe or useful. The characteristic cost is running on other people's signals until something gives, because the internal ones never entered the calculation.

High regulation with low reading tends to be read by others as calm and slightly unreachable. The regulation is real, and it is partly why the reading score stays low: a state that never gets expressed also never gets checked against what someone else perceived. This profile benefits less from emotional work than from mechanical habits — asking rather than inferring, and making the inference explicit so it can be corrected.

A flat profile in the middle band on all four is not a failure to produce a result. It is the most common shape, and it means the four skills are roughly level for you rather than that all of them are weak. In that case the informative comparison is not across the four but across time — the same assessment taken again after a period when something has changed.

What the research says a score like this predicts

The largest meta-analysis on the question pooled 1,246 effect sizes from 158 studies covering 42,529 students, and found emotional intelligence correlated with academic performance at about .20 overall (MacCann et al., 2020). That is a real association and a modest one — enough to matter across a population, nowhere near enough to predict an individual result.

The part worth knowing before you read your own number is that the same analysis split the field into three ways of measuring emotional intelligence, and they did not perform equally. Ability tests, which score your answers against expert consensus, correlated at .24. Mixed measures came in at .19. Self-report — asking you to rate your own skills, which is what this assessment and almost every free online EQ test does — was the weakest of the three at .12. After statistically controlling for general intelligence and the Big Five traits, self-report emotional intelligence added under one percent of unique explanatory power, against roughly two percent for the other two approaches.

That result is not a reason to skip a self-report measure, but it is a reason to be precise about what you are holding. A self-report score is a well-structured record of how you currently see four specific skills. It is not a demonstration of those skills, and the research says it is the version with the loosest grip on outcomes. Where self-report scores do predict workplace performance, meta-analytic work suggests much of that prediction runs through their overlap with self-efficacy, conscientiousness, emotional stability and cognitive ability rather than through anything unique to emotional intelligence (Joseph et al., 2015).

The associations are stronger on the outcomes closest to the thing being measured. Self-report emotional intelligence correlates more substantially with mental health, subjective well-being and life satisfaction than with performance measures — which is what you would expect from a scale asking how well you think you handle your own emotional life. Read your result as information about your relationship with your own emotions, which is what it is good evidence about, rather than as a forecast of how anything will turn out.

Why the same test gives you a different number later

Self-report emotional intelligence is measured by asking you to rate your own skills, which means the score moves with your current view of yourself as well as with the skills themselves. Mood is the largest single source of that movement: people in a low mood rate their own emotional competence lower across the board, and the effect is large enough to shift a band. Taking the test during a difficult week produces a real reading of that week, not a corrupted reading of your general capacity.

This is a limitation and it is also the most useful property the assessment has. Because the four skills are specific rather than global, and because emotion regulation in particular responds to deliberate practice more readily than personality traits do, a repeat measure after a genuine period of work is interpretable in a way that a single measure is not. Reviews of emotional intelligence training find improvements that persist, concentrated in identifying and regulating emotion rather than spread evenly across all four.

The practical guidance is to leave enough time that something could actually have changed. Retaking within days measures how your week is going. Retaking after a few months, ideally when you can remember what you were working on in between, measures something closer to what you wanted to know. Comparing your own two results is also the one comparison that does not depend on a reference sample, which is why it is the one this site is built around.