Collective Emotional Intelligence: Can Teams Measure It Responsibly?

The claim is about the room, not just the people in it

When one group raises a risk early and another becomes defensive during the same kind of handoff, people often say that the first team has higher emotional intelligence. That description may point to a real difference, but it leaves open what is being measured. Individual skill and a team’s way of working are related questions.

Emotional-intelligence measures do not all ask a person to do the same thing. Ability measures present emotion-related problems and score the response against a specified standard. Trait measures use self-report to ask about typical behavior or perceived emotional self-efficacy. Mixed measures combine emotional and social competencies with other personal tendencies. O’Connor and colleagues’ critical review recommends choosing the construct and purpose before choosing an instrument. [1]

Collective emotional intelligence makes a group-level claim. It concerns expectations and interaction patterns that members create together: whether someone can name a concern, whether the group tries to understand a reaction before assigning blame, and whether a damaged exchange can be reopened. A team-referent item asks about ‘this team’, rather than asking each person to describe their own ability. That change is the beginning of measurement, not a cosmetic wording choice.

A useful definition has to survive an ordinary work moment

The phrase becomes useful when it points to what a team is likely to do in a recognizable situation. During planning, a member says that the deadline is forcing a quality risk. Does the group explore the concern, defer it without discussion, or treat the speaker as obstructive? After a tense call, does someone clarify what happened and restore the task, or does the team move on while the disagreement hardens?

These scenes do not reveal a single hidden team essence. They give a team behaviors to examine. The foundational group-EI account by Druskat and Wolff describes emotional competence through group norms, including trust, group identity, and a sense that the group can act effectively together. The Group Emotional Competence Inventory similarly organizes the construct around norms at member, group, and external-relationship levels. [2, 3]

This is why a team can contain people with strong individual self-ratings and still handle emotion poorly as a group. An interaction can reward silence, punish challenge, or make repair somebody’s private burden. The setting changes what people can safely do, so a person’s score cannot stand in for the team’s norm.

The measurement move is to change the reference

Compare two questions: ‘I understand my feelings when working with my teammates’ and ‘In this team, we are able to describe our feelings.’ The first invites a judgment about the respondent. The second asks the respondent to report a property of the group as they experience it. Neither is automatically more truthful. They answer different questions.

Aritzeta and colleagues used this change-of-reference approach in developing the Team-Trait Meta Mood Scale, or T-TMMS. Its items address team attention to feelings, clarity about emotions, and repair of negative moods. The study examined leadership teams in organizations belonging to the Mondragon Cooperative Corporation in the Basque Country and reported 152 teams. [4]

The paper contains conflicting participant totals: 1,746 in the abstract and 1,764 in the methods. That is enough reason not to repeat either figure as settled fact. It also illustrates a broader habit worth keeping: before a team treats a published number as authority, read the method and ask what population, setting, response process, and unit of analysis the number actually describes.

A team-level result needs more than a mean. The analyst must explain the referent, the instrument’s dimensions, and whether members’ answers are aligned enough to summarize as a group pattern. Agreement does not prove skill. Low agreement means the report may be showing different experiences inside the team, which can itself be the important finding.

Why the average of individual EQ scores answers the wrong question

A spreadsheet can average everyone’s individual responses in seconds. The result still describes the people who answered, not necessarily the way they coordinate. Two teams can have the same average while differing sharply in who speaks, who is heard, and what happens after conflict.

Workgroup instruments show the distinction. Jordan and colleagues’ Workgroup Emotional Intelligence Profile was designed around work-team functioning and includes areas such as awareness, expression, recognition of others’ emotions, empathetic concern, and management of emotions. It is an example of a measure built for a group context, though it follows an individual-referent approach rather than making exactly the same team-level claim as the T-TMMS. [5]

A self-report records a person’s view of typical behavior, not a neutral camera record. The critical review notes that respondents may misjudge their own emotional tendencies and that self-report can be affected by socially desirable responding, especially when important people may see the results. [1] In a company, confidentiality and purpose are part of measurement quality because they shape what people are willing to report.

Use the questionnaire to locate a conversation, then compare the pattern with recent work: a decision where dissent was raised, a handoff that went wrong, or a repair that did or did not happen. The team is studying its process, not calculating a league table.

What the evidence supports, and where it stops

Research supports the idea that teams can report shared emotional norms and that team-referent measures can be studied with standard measurement methods. The T-TMMS paper reports a three-factor structure of Attention, Clarity, and Repair and examines links with other team measures. It also says that the study was cross-sectional and based on self-report data, and calls for objective and longitudinal research. [4]

That design can show relationships in the studied setting. It cannot establish a universal cutoff for a ‘high-EQ team’, prove that a score caused better performance, or tell a manager what will happen in a different industry and culture. The sample came from one cooperative organizational context, so transfer to a newly formed, remote, international, or highly unequal team requires judgment and additional evidence.

The Standards for Educational and Psychological Testing place validity on the interpretation and use of scores, not on a test name alone. Their guidance covers intended use, fairness, accessibility, and testing in employment contexts. [6] The direct question for a team is: what decision will this result inform, and is that decision supported by the instrument’s evidence? If the decision is whether people should be hired, promoted, or rated, a developmental team questionnaire is the wrong tool.

Responsible measurement begins before anyone answers

First name the work problem. ‘Improve emotional intelligence’ is too vague. ‘Make it easier to raise a delivery risk in weekly planning’ gives the team something to recognize. ‘Repair cross-functional disagreements before the next handoff’ gives it an observable target. The instrument should earn its place by asking about habits related to that problem.

Then state the boundaries before collecting responses. Participation should be voluntary. Individual answers should remain private by default, and the conversation should use team patterns rather than expose who gave which response. Results should not affect hiring, promotion, performance scoring, access to work, or membership of the team. These commitments protect the conditions under which a self-report can be candid.

After the result, ask for evidence in the team’s own work. Where did the reported pattern appear? Who had to carry the emotional cost of keeping the work moving? What would a different action look like at the next meeting? A small practice might be to invite one dissenting view before closing a decision, or to return to an unresolved concern after a heated exchange.

EQ Test’s team pathway can support private self-reflection followed by a structured discussion of observable behavior. Its individual assessment is a non-validated developmental self-report, not an ability test or employee score. A company rollout should keep that distinction visible from invitation to debrief.

The result should earn its place in the next conversation

A responsible team measure produces a narrower conclusion than ‘our team has high EQ’. It may show that members generally perceive a shared habit, that their views diverge, or that the items do not fit the work problem. Each outcome leads to a different conversation.

If the team sees a shared difficulty with repair, it can choose one recent disagreement and ask what happened between the sharp moment and the return to work. If members disagree about whether concerns are welcome, the disagreement deserves attention before anyone summarizes the team with one score. If the measure and the observed meeting tell different stories, examine the reference period, wording, power dynamics, and actual opportunity to speak.

The conversation to initiate is simple: ‘When disagreement becomes tense here, what do we actually do next, and what would we like to do differently in the next handoff?’ The answer gives the team a behavior to observe. A questionnaire may help organize that inquiry, but the working evidence remains in how people raise concerns, respond under pressure, and repair what the work has strained.

Questions readers ask