How to Assess Emotional Intelligence in Hiring: Method Comparison
Why most EQ assessments fail
Emotional intelligence has become a hiring buzzword. Every second assessment claims to measure it. Most of them are useless. The reason is simple: they measure self-perception, not actual ability.
Someone who scores high on "I am good at reading people" is usually just confident. The correlation between self-reported EQ and actual ability to navigate emotions is weak. This is the same problem that plagued personality tests in the 1990s—people know what sounds good and answer accordingly.
The assessments worth using are different. They measure whether someone can recognize emotional dynamics in a scenario and choose a reasonable response, not whether they claim to be emotionally intelligent.
The methods compared
1. Self-Report Questionnaires (Don't Use Alone)
How it works: Candidate rates statements like "I am skilled at understanding other people's emotions" on a 5-point scale. The test aggregates scores into dimensions like self-awareness, empathy, relationship management.
Examples: EQTM by Goleman (note: Goleman is a journalist, not a psychologist), Hogan EQ, Emotional Quotient Inventory (EQ-i), EQMap.
Validity: Modest to weak. Self-perception correlates with job performance around 0.20–0.30 (on a scale where 1.0 is perfect prediction). Compare that to a structured interview rubric, which typically reaches 0.40–0.50.
Gaming: Easy. Candidates quickly learn that "I carefully consider other perspectives" is the right answer.
Cost: Low ($20–50 per candidate).
When to use: As a lightweight screen or for team development, not as a primary filter for high-stakes decisions. Never use alone for promotions or significant role changes.
2. Ability-Based Tests (Higher Validity But Rare)
How it works: Candidate views photos of faces and identifies the emotion; reads scenarios and selects the most emotionally intelligent response; watches video clips and interprets emotional subtext.
Examples: MSCEIT (Mayer-Salovey-Caruso Emotional Intelligence Test) — the most rigorous. Others include Situational Test of Emotional Understanding (STEU), Geneva Emotion Recognition Test (GERT) for face recognition specifically.
Validity: MSCEIT reaches about 0.25–0.35 correlation with job performance. Weaker than technical assessments but stronger than self-report. Face recognition alone (GERT, facial coding) adds little to hiring; most roles don't require reading microexpressions.
Gaming: Harder. You can't fake your way through emotion recognition if you're not looking carefully.
Cost: High ($50–150 per candidate). Requires licensed administration for MSCEIT.
When to use: For customer-facing roles (sales, service, support) where emotion reading matters. For management roles, pair with behavioral interviews about actual conflict management.
3. Scenario-Based Response Tests (Best ROI)
How it works: Candidate reads a workplace dilemma. Character A is frustrated, Character B is impatient, stakes are ambiguous. Candidate ranks 5 response options from most to least effective. What they choose reveals how they think about emotional dynamics.
Examples: ClarityHire EQ assessments, situational judgment tests adapted for emotional content, Hogan judgment assessments.
Validity: Moderate (0.25–0.40 correlation with performance). Comparable to ability tests but more practical because scenarios are job-specific. A sales-manager scenario is more predictive of sales-management performance than general emotion recognition.
Gaming: Possible if the candidate knows the framework ("name the emotion = good"). Reduced if scenarios are role-specific and non-obvious.
Cost: Moderate ($30–80 per candidate). Can be administered unproctored.
When to use: As a primary filter for roles with interpersonal load. Pair with behavioral interviews. Complement with work samples in technical roles.
4. Structured Behavioral Interview (Highest Validity)
How it works: Interviewer asks: "Tell me about a time you had to give difficult feedback to a peer. What was the situation? What did you do? What happened?" Then probe: what emotions were involved? How did you manage the conversation? What would you do differently?
Validity: 0.40–0.55 correlation with job performance. This is the gold standard across most hiring contexts.
Gaming: Harder than scenarios. Candidates can prepare a polished story, but follow-up questions (especially about failure) usually surface reality.
Cost: Moderate. Time investment is 30 minutes per candidate.
When to use: Always, for any role with interpersonal stakes. Use rubrics anchored to specificity, self-awareness, and reflection (not generic "leadership").
The research caveat
Daniel Goleman popularized "emotional intelligence" in the mid-1990s. The concept is real—people do vary in their ability to navigate emotions. But Goleman overstated the effect. His claim that EQ is a stronger predictor than IQ is not supported by rigorous research. Most meta-analyses find:
- Cognitive ability (IQ) predicts job performance across roles at 0.25–0.35 in correlation.
- EQ predicts at 0.15–0.25 in correlation.
- Cognitive ability predicts better in most roles; EQ matters most in high-contact roles (sales, management, customer service).
The MSCEIT researchers (Mayer and Salovey, actual emotion researchers) are careful about claims. Their test has validity but is modest. It's useful as one signal, not as the decider.
Also note: "EQ" is not unified. Face recognition ≠ conflict management ability ≠ self-awareness. An assessment measuring face emotion recognition will not predict whether someone can manage team conflict. Be specific about what you're measuring.
A practical assessment strategy for EQ
For customer-facing roles (sales, customer success, support):
- Scenario-based EQ test (20 minutes) — filters for basic emotional reasoning.
- Behavioral interview (30 minutes) — "Tell me about a difficult customer interaction. What did you notice about their emotion? What did you do?"
- Work sample — if the role requires it (e.g., handle a mock customer escalation).
For management roles:
- Structured behavioral interview focused on conflict and feedback (30 minutes).
- Optional: scenario-based test for alignment check.
- Reference calls to past reports: "How did this person handle feedback? Conflict? Their own frustration?"
For individual-contributor technical roles: Not a primary filter. Technical skill and cognitive ability predict better. Use EQ scenario tests if the role has frequent cross-functional or mentoring load.
What to avoid
- Selling EQ as the primary predictor. It's one signal. Technical skill, work samples, and structured interviews are stronger.
- Using only self-report EQ. It's self-selection bias. Use scenario tests or behavioral interviews.
- Over-indexing on face-reading ability. Unless the role requires reading microexpressions (therapy, high-stakes negotiation), this is not predictive.
- Assuming EQ is fixed. It improves with feedback and practice. Use assessments to understand where someone is, not to permanently slot them.
EQ matters. But measure it carefully. The difference between "this person seems emotionally intelligent" and "this person actually handles conflict well" is the difference between a hunch and data.
Use scenario-based tests or behavioral interviews. Pair them together. Skip the self-report questionnaires unless you're doing team development, not hiring decisions.
ClarityHire's emotional intelligence assessments combine scenario-based responses with integrated behavioral interview workflows, making it easy to pair tests with follow-up conversation data.