What emotional intelligence tests can and cannot measure

Two different tests share one name

Search for an emotional intelligence test and you are offered two things that have almost nothing in common except the label.

One is an ability test. A study comparing the models describes the performance-based ability model as one in which EI is viewed as a form of intelligence that is based on emotional aptitudes, and is regarded as a mental ability that involves reasoning about our emotions. You are given emotional problems, and your answers are scored as better or worse the way a reasoning test scores answers.

The other is a questionnaire. You read statements about yourself and rate how accurately each describes you. Nothing you answer is right or wrong, and what the instrument records is the description you gave.

The test on this site is the second kind, and most of this guide is about what that means. The short version: a self-report questionnaire measures self-description. That is a real thing, worth knowing about yourself, and it is not the same thing as skill.

The ability side: the MSCEIT

The best-known ability test is the Mayer-Salovey-Caruso Emotional Intelligence Test. The Consortium for Research on Emotional Intelligence in Organizations describes it as an ability-based test designed to measure the four branches of the EI model of Mayer and Salovey. Those four branches, in the consortium's own wording:

• Perceiving Emotions: the ability to perceive emotions in oneself and others as well as in objects, art, stories, music, and other stimuli.
• Facilitating Thought: the ability to generate, use, and feel emotion as necessary.
• Understanding Emotions: the ability to understand emotional information.
• Managing Emotions: the ability to be open to feelings, and to modulate them in oneself and others so as to promote personal understanding and growth.

The same page says the MSCEIT provides 15 main scores: a total EI score, two area scores, four branch scores and eight task scores. The word task is the giveaway that something different is happening from a personality questionnaire. There are tasks, and tasks have performances.

What it is not is a free browser quiz, and that is worth carrying into a search results page. If a test finishes in three minutes, costs nothing, and never once asks you to solve anything, it is not measuring ability, whatever the heading over it says.

The self-report side, and what it can honestly see

The self-report tradition has its own flagship. The consortium's entry for the Trait Emotional Intelligence Questionnaire describes it as a self-report inventory that covers the sampling domain of trait EI comprehensively, comprising 153 items measuring 15 distinct facets, 4 factors, and global trait EI, with a 30-item short form designed to measure global trait emotional intelligence on its own.

A self-report instrument is not a weaker ability test. It measures a different thing: how a person describes their own emotional habits. That description is informative, and for some purposes it is the only thing available. It is also everything the instrument has, and it carries everything that shapes how someone talks about themselves, including the week they have had, how a particular word lands on them, and what they would prefer to be true.

The 42 statements on this site are that kind of item. Rating "Express my affection physically" as very accurate records that you think so. It does not record what the people on the receiving end noticed, and nothing on the page pretends otherwise.

Why the two kinds do not agree

If both kinds measured the same thing, their scores would line up. They do not. The study comparing the three models of emotional intelligence reports that although the three models essentially assess the same construct, any correlations between them appear to be weak.

That finding is the most useful thing to take from this page, and it cuts in both directions. A high self-report profile is not evidence of skill when a real feeling arrives. A modest one is not evidence that the skill is missing. They are separate measurements, and neither stands in for the other.

It also sets a sensible expectation for comparing free tests. Two pages can both call themselves EQ tests, hand you two different numbers, and neither be broken, because the label covers at least three different constructs measured in at least two different ways, and there is no conversion between them.

Where the 42 statements here come from

The items on this route are not ours. They are reproduced word for word from the International Personality Item Pool, which states that the items and scales are in the public domain, which means that one can copy, edit, translate, or use them for any purpose without asking permission and without paying a fee. That is why the page can show you the exact items it used rather than describing them.

The specific set is the IPIP key page titled "The Items in Each of the IPIP Scales Measuring Seven Components Potentially Related to 'Emotional Intelligence'" (Barchard, 2001). The wording of that title is more careful than most writing on the subject, and it is worth reading twice: components potentially related to emotional intelligence. The seven scales are positive expressivity, negative expressivity, attending to emotions, emotion-based decision-making, responsive joy, responsive distress and empathic concern. This site takes six items from each and gives every scale a plain-English name on the result, so that the reader is not asked to translate jargon before reading their own answers back.

Nineteen of the 42 items are reverse-keyed, and the keying is close to balanced inside each of the seven facets. That has a visible consequence which looks like a fault and is not. If you answer every statement the same way, out of hurry or habit, every meter lands near the middle rather than at one end. A questionnaire that cannot be pushed by an answering style is doing its job, and the cost is that the meters sit closer together than a test with a thumb on the scale would show.

Why there is no EQ number on the result

Three reasons, in order of how much they matter.

• There is no norm sample. A score becomes a number about a person only when it is compared with a reference group tested the same way, and the professional Standards for Educational and Psychological Testing set out what norms, reliability, validity and fairness require before that comparison means anything. A browser questionnaire has none of it, so a percentile here would be decoration with a number's authority.
• The seven facets move independently. Averaging them produces one figure that conceals the interesting part: a person can watch their own feelings closely and still almost never show them, or catch a room's good mood and still decide everything on the facts.
• No facet's high end is the good end. Showing more sadness is not more emotionally intelligent than showing less; deciding from feeling is not better than deciding from reasons. A total implies a direction, and on these facets there is not one to imply.

So the result is seven meters, each read only against your own other answers, with written notes for the facets that stood out. It is not an EQ score, not a percentile, not a pass mark, and not a diagnosis of anything.

What seven meters are actually good for

Treat a meter as a question rather than a finding. The useful move is to take one facet and watch it for a week, because a habit is easier to see in a diary than in a rating.

• Pick the facet that surprised you, high or low, rather than the one that looks worst.
• Each evening write one line about a moment that facet covers: a feeling you noticed late, a feeling you showed or swallowed, a decision you took on a hunch, someone whose trouble stayed with you afterwards.
• After seven days read the lines together and ask whether the meter matched the week, or whether the week was simply arranged so that the habit never came up.
• If a facet matters to you and the week disagreed with the meter, trust the week. It has more evidence in it than 42 statements answered in nine minutes.

One caution to finish on. This is one sitting, self-reported, on one kind of day, and a rough week moves several meters at once. It records how you described yourself, which is a genuinely useful thing to have written down and is not a measurement of what you can do when a real feeling arrives.