How Emotional Intelligence Is Measured
Emotional intelligence is measured in three main ways: ability tests that score answers against a key (such as the MSCEIT), self-report questionnaires about your emotional tendencies (such as the TEIQue), and mixed or multi-rater instruments (such as the EQ-i 2.0 and ESCI). These methods do not agree closely: in one study, self-rated EI correlated only r = .19 with ability test scores. Pick a measure by what you want to know, and never treat one method’s evidence as proof for another.
Editorially checked
Explore your emotions and relationships with the free EQTiger assessment.
By EQTiger Editorial. Published .
Reviewed by: pending qualified review
Every EQ test is answering one of three questions. Can you solve emotional problems when someone checks your answers? How do you see yourself emotionally? How do you come across, to yourself and to the people around you? The tests built for each question look similar from the outside, a page of items and a score at the end, but the numbers mean very different things.
The three methods at a glance
| Approach | What it measures | Typical instruments | Key paper |
|---|---|---|---|
| Ability EI | Maximal performance on emotion problems (reading faces, predicting how feelings change, choosing effective strategies), scored against a key | MSCEIT (141 items); STEU and STEM | Mayer, Salovey, Caruso and Sitarenios, 2003 |
| Trait EI | Self-perceptions and typical emotional tendencies, treated as part of personality | TEIQue (153 items, 30-item short form); Schutte SREIT (33 items) | Petrides and Furnham, 2001 |
| Mixed EI | Emotional skills combined with social skills, motivation, mood and workplace competencies | EQ-i 2.0 (self-report); ESCI (360-degree ratings) | Bar-On, 2006; Joseph and colleagues, 2015 |
The instrument names and item counts follow the comparison in O’Connor and colleagues’ critical review, which also advises choosing a measure by purpose and recommends that non-experts start with a trait measure (O’Connor and colleagues, 2019). The key papers are cited in the sections below.
How an ability test decides what counts as right
Emotion questions rarely have a textbook answer, so ability tests need a scoring rule. The MSCEIT offers two. Consensus scoring credits an answer by how many people in a large sample chose it; expert scoring credits it by how many emotion researchers chose it. In the developers’ study of 2,112 adults and 21 experts, the two groups largely agreed on which answers were best (Mayer and colleagues, 2003).
Critics are not fully reassured. O’Connor and colleagues note that relying on expert judgment to define correct answers is problematic for several reasons, and that common ability measures tend to have relatively weak reliability and validity (O’Connor and colleagues, 2019). The upside of ability tests is that you cannot simply claim to be good at them. The downside is that a keyed answer on paper is not the same as doing it at 6 p.m. with a deadline and a raised voice.
Why self-ratings and ability scores barely agree
Marc Brackett and colleagues gave the same people both kinds of test. Self-rated EI correlated with MSCEIT scores at r = .19, a real but weak relationship. People’s own estimates of how they had done on the MSCEIT were essentially unrelated to their actual scores (r = .10 and r = .03). Those in the bottom half of scores overestimated their performance, and those in the top half underestimated it (Brackett and colleagues, 2006).
That does not make self-report useless. It measures something real: your confidence, habits and emotional self-image, which matter for how you act. It just answers a different question from an ability test. Trait EI researchers make that explicit by placing their construct inside personality rather than intelligence (Petrides and Furnham, 2001).
What other people’s ratings add, and their limits
Multi-rater instruments ask colleagues as well as you. The ESCI is designed mainly to collect this kind of 360-degree feedback on twelve workplace competencies (Korn Ferry, 2017). Observer ratings capture how you come across, which self-report cannot.
They also disagree with each other more than people expect. In a meta-analysis of EI and leadership, agreement between different raters about the same leader’s EI averaged just .16 (Harms and Credé, 2010). Your manager, peers and direct reports see different slices of your behavior, so a 360 report is best read rater group by rater group, not as one averaged truth.
The overlap problem with mixed measures
Broad mixed instruments cover a lot of ground, and much of that ground is already mapped by other tests. Bar-On’s model includes scales such as optimism, self-regard and happiness (Bar-On, 2006). A meta-analysis of mixed EI found that ability EI, self-efficacy, self-rated performance, personality and general mental ability together accounted for most of what mixed EI measures (multiple R = .79), and that mixed EI’s link with job performance disappeared once those were controlled (Joseph and colleagues, 2015). A high mixed-EI score can therefore reflect a confident, conscientious, emotionally stable person as much as a distinct emotional skill.
One person, three results
Jordan, a nurse manager, takes three assessments in a month. A trait questionnaire puts Jordan high: Jordan genuinely believes they handle stress well. An ability test puts Jordan in the middle: strong on managing emotions items, weaker on reading facial expressions. A 360 survey splits: staff rate Jordan’s empathy higher than Jordan’s own manager does.
None of these results is wrong. They answer three different questions, and the mismatches are the most useful part. The gap between self-view and ability score suggests checking reads more often. The split between raters suggests Jordan shows different sides to different audiences.
Where EQTiger fits
EQTiger uses two item types in each of six dimensions: scenario questions with four response options, keyed by the item authors, and habit questions on a five-point frequency scale. That makes it a self-report and situational judgment tool. Its keys are not consensus or expert keys, it has no reference sample, and it has not been validated against any of the instruments above. The results are descriptive 0 to 100 EQTiger scores for reflection. The methodology explains the scoring, and the instrument guides describe the published tests in more depth, including the MSCEIT and TEIQue.
Try it: audit any EQ test in five minutes
- Find the method. Are you answering what you would do, what you usually do, or how others see you? That tells you whether it is ability, trait or multi-rater.
- Find the key. Who decided the best answers: the authors, a large sample, or a panel of experts? If the page does not say, note that.
- Find the comparison group. Does the score compare you with a described sample, or is it a raw scale? A number with no reference group cannot rank you against other people.
- Find the evidence. Is there published research on this exact instrument, or only on emotional intelligence in general?
- Write one use the result supports (for example, picking a skill to practice) and one it does not (for example, a hiring decision).
The short version
Measurement method changes the meaning of the score. Ability tests show what you can do when tested, self-report shows how you see yourself, and observer ratings show how you come across. For a fuller discussion of online tests specifically, see how accurate online EQ tests are, and read understanding EQ scores before comparing numbers.
Use the free EQTiger assessment to choose a starting point for reflection. Read what the assessment includes and the limitations before interpreting your result.
Research notes
O’Connor and colleagues, 2019: This critical review separates ability, trait and mixed EI, notes that more than 30 widely used EI measures exist, and recommends matching the measure to the purpose after checking its factor structure, reliability and validity evidence.
Peter J. O’Connor, Andrew Hill, Maria Kaya, Brett Martin (2019). The Measurement of Emotional Intelligence: A Critical Review of the Literature and Recommendations for Researchers and Practitioners. Frontiers in Psychology, 10, 1116. doi.org/10.3389/fpsyg.2019.01116
Mayer and colleagues, 2003: In the test developers’ study of 2,112 adults and 21 emotion experts, the general sample and the experts largely agreed on which MSCEIT answers were correct, the test showed reasonable reliability, and factor analysis supported the theorized structure.
John D. Mayer, Peter Salovey, David R. Caruso, Gill Sitarenios (2003). Measuring emotional intelligence with the MSCEIT V2.0. Emotion, 3(1), 97-105. doi.org/10.1037/1528-3542.3.1.97
Petrides and Furnham, 2001: Two studies (N = 227 and N = 166) located a trait EI factor inside established personality models and concluded that trait EI is best treated as a distinct composite of personality traits and self-perceived abilities.
K. V. Petrides, Adrian Furnham (2001). Trait emotional intelligence: psychometric investigation with reference to established trait taxonomies. European Journal of Personality, 15(6), 425-448. doi.org/10.1002/per.416
Bar-On, 2006: Bar-On described his emotional-social intelligence model and its self-report measure, the EQ-i, organized into five composite scales (intrapersonal, interpersonal, stress management, adaptability, general mood) and 15 subscales.
Reuven Bar-On (2006). The Bar-On model of emotional-social intelligence (ESI). Psicothema, 18(Suppl.), 13-25. www.psicothema.com/pi?pii=3271
Korn Ferry, 2017: Korn Ferry’s technical manual describes the ESCI, a 360-degree rating of emotional and social competencies by self and others, and reports its research base.
Korn Ferry (2017). Emotional and Social Competency Inventory: Research guide and technical manual. Korn Ferry. www.kornferry.com/content/dam/kornferry/docs/article-migration/ESCI_Technical_Manual_nav_04052017.pdf
Brackett and colleagues, 2006: Across three studies, self-rated emotional abilities and MSCEIT performance scores were not strongly correlated, and only the performance test predicted observed social competence, and only for men.
Marc A. Brackett, Susan E. Rivers, Sara Shiffman, Nicole Lerner, Peter Salovey (2006). Relating emotional abilities to social functioning: A comparison of self-report and performance measures of emotional intelligence. Journal of Personality and Social Psychology, 91(4), 780-795. doi.org/10.1037/0022-3514.91.4.780
Harms and Credé, 2010: Across 62 samples, EI and transformational leadership correlated at a validity estimate of .59 when the same person rated both but only .12 when ratings came from different sources, and different raters agreed little on a leader’s EI (.16) or transformational leadership (.14).
P. D. Harms, Marcus Credé (2010). Emotional intelligence and transformational and transactional leadership: A meta-analysis. Journal of Leadership & Organizational Studies, 17(1), 5-17. doi.org/10.1177/1548051809350894
Joseph and colleagues, 2015: Mixed (self-report) EI correlated ρ = .29 with supervisor-rated job performance, but the link became nil (β = -.02) once ability EI, self-efficacy, self-rated performance, personality and general mental ability were controlled, because those constructs overlap heavily with mixed EI content (multiple R = .79).
Dana L. Joseph, Jing Jin, Daniel A. Newman, Ernest H. O’Boyle (2015). Why does self-reported emotional intelligence predict job performance? A meta-analytic investigation of mixed EI. Journal of Applied Psychology, 100(2), 298-342. doi.org/10.1037/a0037681
3 questions
Questions
What is the most accurate test of emotional intelligence?
No single test is most accurate for every purpose. Ability tests such as the MSCEIT measure reasoning about emotion against a key. Trait questionnaires such as the TEIQue measure how you see yourself. A 2019 review recommended non-experts start with a trait measure and choose by purpose.
Can you measure emotional intelligence with a free online quiz?
A free quiz can prompt reflection, but most are self-report with no published evidence about their own scores. Look for a clear account of what is measured, how answers are scored, and what the result cannot tell you.
Why do my EQ scores differ between tests?
Different tests measure different things (ability, self-perception or broad competencies), use different scoring keys and score scales, and are affected by mood and context on the day. Results from different instruments are not interchangeable.
Related
- No. 01Understanding EQ ScoresAn EQTiger score is a descriptive 0 to 100 number built from your answers to scenario and habit questions in six dimensions. The bands (Emerging below 40, Developing 40 to 59, Established 60 to 79, Strong 80 and above) are design cut points, not norms, so a 70 does not mean you did better than 70 percent of people. Read the score as a summary of one sitting that points you to a skill worth practicing.
- No. 02Components of Emotional Intelligence, Model by ModelThe components of emotional intelligence depend on which model you use. The ability model has four branches: perceiving, using, understanding and managing emotions. Bar-On’s model has five composite scales, the Goleman-Boyatzis competency model has twelve competencies in four clusters, and trait EI has fifteen facets. The lists overlap on self-awareness, regulation and reading others, and differ on whether optimism, motivation and assertiveness count.
- No. 03EQ vs. IQ: What the Evidence Actually ShowsIQ measures general reasoning ability; EQ (emotional intelligence) measures reasoning about emotions or, in self-report versions, how people see their own emotional tendencies. Ability EI overlaps with IQ (about .35 with verbal intelligence) and can be modeled as a branch of general intelligence. EI adds a modest amount of prediction for job performance beyond IQ and personality, but the popular claim that EQ matters more than IQ is not supported as a general rule.
- No. 04What Is Emotional Intelligence?Emotional intelligence (EI) is the ability to perceive emotions accurately, use them to help thinking, understand how they work, and manage them in yourself and others. Psychologists Peter Salovey and John Mayer introduced the scientific term in 1990, but the word now also covers self-rated emotional traits and broad workplace competency lists. EI is linked to health, well-being and job performance at modest correlations, mostly in correlational studies.
Your next 9 minutes
Understand your emotions.
36 questions about everyday moments. A fresh starting point for reflection.
Free to take. Full report $6.95. One-time payment. No subscription.