Workplace tool
Emotional Intelligence
MSCEIT, EQ-i 2.0, and TEIQue
Not one test but two incompatible families measuring different things under the same name. Knowing which one you took is the whole game.
Why that grade: Both families predict job performance incrementally over cognitive ability and the Big Five, but effects are modest and the two families correlate weakly with each other, which limits what "an EI score" can mean without qualification.
- Cost
- TEIQue free for research; MSCEIT and EQ-i 2.0 practitioner-administered
- Time
- 15-45 min
- Items
- 141
- Published
- 2002
The MSCEIT uses 141 items, the EQ-i 2.0 uses 133, and the TEIQue uses 153. They are different instruments measuring different constructs.
Origin
John Mayer, Peter Salovey and David Caruso (ability EI); Reuven Bar-On and K. V. Petrides (trait EI)
What you get free: The TEIQue is available free for research and educational use. The MHS instruments require a qualified administrator.
The verdict
Worth taking seriously, in the ability form, if you work in an emotionally demanding role and want a measure of something you can improve. Establish which family your instrument belongs to before you attach any meaning to the score.
Strengths
- Ability EI is measured against scored answers rather than self-report, which makes it the hardest instrument on this site to game.
- It measures something trainable. Unlike traits, emotion-regulation skill responds to deliberate practice.
- The incremental validity over cognitive ability and personality is real, if modest, and concentrated in exactly the roles you would predict.
Limitations
- The name covers two families that correlate weakly. A self-report EI score and an ability EI score are not interchangeable, and popular writing treats them as though they are.
- Scoring ability EI requires deciding what the emotionally correct answer is, usually by consensus or expert judgement, which is a genuine philosophical weakness.
- Trait EI instruments overlap substantially with the Big Five, raising the question of how much distinct variance they add.
- The popular literature dramatically overstates the effect sizes found in the meta-analytic record.
Two families, one name
Ability EI treats emotional intelligence as a form of intelligence and measures it the way you would measure any other: with problems that have better and worse answers. The MSCEIT is the standard instrument here.
Trait EI treats it as a constellation of self-perceptions and dispositions, and measures it by asking you. The EQ-i 2.0 and the TEIQue are in this family. The two approaches correlate only modestly, which means that a person can be confident about their emotional skills and not especially good at them, a result the literature reports and everyday experience confirms.
What the numbers actually support
The defensible claim is that EI adds a modest but real increment to the prediction of job performance beyond what cognitive ability and personality already explain, and that the increment is largest in jobs with high emotional demands. Nursing, sales, teaching, and management all qualify. Backend engineering largely does not.
The indefensible claim, common in popular business writing, is that EI accounts for the majority of workplace success. Nothing in the meta-analytic record supports a figure like that.
What it reports back
Results are expressed as continuous traits.
- Perceiving emotion
- Reading emotional information from faces, voices, and situations. An ability, measured against scored answers.
- Using emotion
- Harnessing emotion to assist thinking and problem solving.
- Understanding emotion
- Knowing how emotions combine, escalate, and change over time.
- Managing emotion
- Regulating your own emotions and influencing others’ constructively.
- Trait emotional self-efficacy
- What the trait-EI instruments measure instead: your own belief about how emotionally capable you are. Related to, but not the same as, ability.
The evidence, with sources
Nothing in this section is stated without an attribution. Where a figure varies by version or sample, the range is given rather than the most flattering number.
Reliability
Ability and trait EI instruments generally report acceptable internal consistency, but they correlate only modestly with each other, indicating they are not measuring the same construct.
meta analysisJoseph & Newman (2010), Journal of Applied Psychology, 95(1), 54-78
Validity
Meta-analysis finds emotional intelligence predicts job performance with incremental validity beyond cognitive ability and the Big Five, with the strongest effects for mixed and trait measures in emotionally demanding roles.
meta analysisO’Boyle, Humphrey, Pollack, Hawver & Story (2011), Journal of Organizational Behavior, 32(5), 788-818A cascading model shows emotion perception, understanding, and regulation relate to performance in sequence, and that emotional labour demands moderate how much any of it matters for a given job.
meta analysisJoseph & Newman (2010), Journal of Applied Psychology, 95(1), 54-78The widely repeated figures that emotional intelligence "accounts for 58 percent of performance" and that "90 percent of top performers have high EQ" are vendor marketing, not research. They appear in no peer-reviewed journal. The meta-analytic correlations with job performance are approximately .20 for ability measures and .29 for mixed measures.
meta analysisO’Boyle, Humphrey, Pollack, Hawver & Story (2011), Journal of Organizational Behavior, 32(5), 788-818
Who it actually fits
- Leadership development and roles with heavy emotional labour
- Anyone wanting a measure of something they can actually improve
- Coaching contexts where a practitioner will interpret the result
Not a diagnosis