Emotional intelligence at work: what a score can tell you

Overview

An emotional intelligence score may influence who gets hired, how a manager develops their team or what an employee takes away from a feedback session. Yet the meaning of that score depends on what the assessment measures and on the evidence supporting its intended use. Understanding those differences helps determine what conclusions are reasonable and where further information is needed.

Managers may treat an emotional intelligence score as an objective description of an employee, even when the assessment captures only one aspect of their behaviour or abilities. This becomes a problem when the result influences decisions without being checked against other evidence. Interpreting a score therefore begins with knowing which type of assessment produced it.

What the assessments measure

Emotional intelligence assessments do not all measure the same characteristics. The distinction matters because it affects how their results can be interpreted. A 2019 review of widely used measures, led by Peter O'Connor of Queensland University of Technology, distinguishes three things an assessment may capture. These are abilities related to emotions, typical emotional behaviour and a broader mix of personal characteristics. ^[1]

Ability tests assess how well people solve problems involving emotions. They measure performance on specific tasks, although this does not necessarily reflect how someone manages emotions in everyday workplace situations. Answers are scored as correct or incorrect, with scoring often relying on expert judgement about emotions. ^[1]

Self-report questionnaires ask people to describe their usual emotional responses and assess their own abilities. These results provide insight into self-perception, although they may differ from how someone behaves or is perceived by colleagues. Some assessments also collect ratings from colleagues, usually alongside the person's own. Colleagues' ratings show how others perceive someone's behaviour at work rather than how that person sees themselves. ^[1]

Researchers classify the constructs as well as the methods. Ability emotional intelligence (EI) refers to abilities related to emotions, measured through tasks the person performs. Trait EI refers to how people typically behave in emotional situations and how they rate their own abilities, as reported by the person. Mixed models combine traits, social skills and competencies, usually through self-report and sometimes with ratings from others. ^[1]

The distinction between ability, trait and mixed EI is closely related to how each is measured. Ability tests assess performance on emotional tasks, while trait and mixed measures generally rely on self-report. Some mixed measures also include ratings from others. ^[1] As a result, different assessments may produce different scores for the same person because they measure different characteristics or draw on different sources of information.

Matching the assessment to the decision

An assessment may measure emotional abilities consistently without providing useful evidence about job performance. Its suitability therefore depends on the decision it is intended to inform. Reliability concerns the precision and consistency of scores, while validity concerns the evidence supporting their interpretation for a specified use. The Standards for Educational and Psychological Testing state that evidence should be provided for each intended interpretation and use of a score. ^[4] A general assurance that an assessment is reliable and valid is therefore not sufficient.

The available evidence differs across assessment methods, particularly in their suitability for selection and development. The comparison below follows the 2019 review. ^[1]

These differences matter most when an assessment is moved from one purpose to another. In coaching, a self-report questionnaire helps a person reflect on how they see themselves. In selection, the same questionnaire informs a decision about that person. The review notes that self-report measures can be answered in a socially desirable way, mainly when respondents believe that a supervisor or potential employer will see the results. ^[1] Evidence that a questionnaire is useful in coaching therefore does not establish that it is suitable for hiring, because people respond under different conditions.

The testing standards make the same point in general terms. When a score is interpreted for a use that has not been validated, the user needs to justify that use and collect new evidence if necessary. ^[4] In hiring, the review advises that an EI measure should form only part of the assessment process. A score cannot be treated as comprehensive information about an applicant, even when the instrument is appropriate for the abilities or characteristics you want to assess. ^[1]

How strongly is emotional intelligence related to job performance?

Meta-analyses generally find positive associations between emotional intelligence scores and job performance, although the strength varies slightly by assessment method. A meta-analysis combines findings from several studies. A 2022 meta-analysis by Caglar Dogru of Ufuk University reports positive relationships for ability, self-report and mixed EI, with similar estimates for all three. ^[3]

A correlation describes the relationship between two measured characteristics. A positive correlation means people who score higher on one tend to score higher on the other. Correlations range from -1 to +1, and zero indicates no linear relationship. Dogru reports corrected correlations with job performance of .28 for ability EI, .33 for self-report EI and .31 for mixed EI. ^[3] On that scale, values around .30 indicate a modest relationship.

That analysis and a 2017 meta-analysis led by Chao Miao of Wilkes University also examine attitudes to work, including job satisfaction and organisational commitment. Miao and colleagues additionally examine turnover intentions, meaning employees' intentions to leave their organisation. ^{[2] [3]} Both analyses use the term self-report EI for the category described earlier as trait EI.

Both meta-analyses adjusted their correlations to account for limitations in measurement. Dogru also considered whether the samples represented a narrower range of scores than the wider population. ^{[2] [3]} These adjustments affect the estimated strength of the relationships, so the figures should be interpreted as corrected estimates rather than correlations observed directly in individual studies.

The estimates differ across the two analyses, particularly for ability EI and job satisfaction, and for self-report and mixed EI with organisational commitment. Dogru notes that Miao and colleagues' estimates for organisational commitment are higher and does not explain the difference. ^{[2] [3]} The results support a positive link but leave its size uncertain. A figure cited in support of a workplace decision should therefore name the analysis and assessment category it comes from.

Interpreting an individual score

These correlations describe patterns across groups of employees. They do not establish how a particular individual with a given score will perform. The International Test Commission also advises against extending a result to characteristics the assessment does not measure. It asks users to consider measurement error and the limitations of scores. ^[5]

In a development conversation, the use of a score rests on professional judgement rather than on these findings. Begin with the employee's own examples of situations involving emotions, then read the report together and examine where it corresponds with those examples. A discussion about a disagreement, for instance, needs to distinguish understanding another person's emotions from the behaviour shown during that disagreement. The report and the examples can then inform one specific behaviour to observe in future conversations.

Agree what that behaviour would look like and when you will review it. If the report and the employee's account differ, explore the difference before drawing a conclusion about the person. This gives the discussion a concrete focus and ties any claim about progress to the evidence you collect.

What to ask an assessment provider

A provider should make the evidence behind its score interpretations available for others to examine. The International Test Commission advises using an assessment only for purposes that have relevant validity evidence. It also advises against judging an assessment only on testimonials or on advice from someone with a commercial interest. ^[5]

A provider's documentation should state the intended use of the assessment, the people it is designed for and how its scores should be interpreted. Where scores are compared with a reference group, that group should be clearly described and relevant to the people being assessed. The standard error of measurement expresses the uncertainty around a score arising from measurement error. The testing standards ask for it to be reported in the same units as the score. ^[4]

For a hiring decision, request evidence relating to the proposed selection use and the relevant applicants, rather than relying on a study of development participants. The provider's materials should show which score interpretations have validity and reliability evidence, so you can decide whether the documentation supports the decision. ^{[4] [5]}

The Inner View has a commercial interest in this subject. We provide emotional intelligence assessment to organisations and offer emotional intelligence coaching as a separate service. For organisations that develop their own measures, we also provide a test design and validation service. Put the same questions to us.

What the research does not show

Miao and colleagues' evidence was dominated by cross-sectional studies, which measure characteristics at one point in time. Such studies cannot establish that EI caused the reported work attitudes. Dogru identifies further limitations, including some studies that relied on employees' own reports of job performance. These features limit the conclusions that can be drawn about causation and workplace behaviour. ^{[2] [3]}

Some relationships could not be analysed because too few studies were available, including ability EI with organisational commitment and turnover intentions in the 2017 analysis. The entries marked "Not analysed" therefore mean that no estimate was reported, rather than that there was no relationship. ^[2]

The three research papers do not test whether training raises emotional intelligence, or whether such training improves those work outcomes. Associations between scores and outcomes cannot establish those effects. Claims about a development programme need evidence from research that evaluates the programme and the outcomes it promises.

Taken together, these findings give emotional intelligence scores a useful but limited place in decisions about people. Assessments differ in what they measure, the published associations with work outcomes are positive and modest, and several questions remain unanswered. For anyone who commissions or receives an assessment, the practical consequence is a short set of checks. Before using an emotional intelligence score, establish what the assessment measures, whether the evidence supports its intended use and how the result compares with relevant workplace behaviour. Those checks determine how much confidence the score deserves in a particular decision.

Emotional intelligence assessment

The Inner View offers an emotional intelligence assessment.