How Emotional Intelligence Helps in a Difficult Performance Conversation

Emotional intelligence helps by keeping the conversation about changeable work

Emotional intelligence helps in a performance conversation by changing what each person does with emotion. Here, it means noticing emotional information and using it to guide a response: pausing before irritation becomes accusation, listening when someone disputes a detail, and returning to the work. The aim is a fair, testable next step, not a calm display. Keep three things distinct. Work facts include behavior, expectations, and impact that can be checked. Emotional cues, such as tension or defensiveness, show how the exchange may be landing; they do not prove why behavior occurred or whether the appraisal is correct. Process conditions include whether expectations were clear, standards consistent, and the meeting’s purpose honestly stated. Confusing these categories can turn frustration into a character judgment or make emotional skill carry an unclear process. The University of Minnesota’s “Difficult Performance Conversations” guide advises supervisors to check what happened, consider information that might contradict their view, notice personal triggers, listen to the employee’s account, and agree on expectations and next steps. These practices keep attention on changeable work and make room for relevant context. Judge the conversation by what becomes clearer and what happens next, not by whether anyone shows emotion.

Sources: Difficult Performance Conversations

What should the conversation stay anchored to?

Start with an observable action or result, the expectation that applied, and the effect on the work. Then ask what the employee saw differently or what information is missing. A performance conversation becomes easier to examine when it concerns something that could be checked, rather than a verdict about the person. “You are careless” asks someone to accept a broad judgment. “The client file went out without the revised figures; our agreed check was to compare it with the final sheet before sending” identifies an event and a standard. The second statement still may need correction, but both people can investigate it. Consider this explicitly illustrative case: a project coordinator submits a monthly report after the agreed deadline. The observation is that the report arrived Tuesday morning; the expectation was delivery by Monday afternoon; the impact is that a review meeting proceeded without the figures. Those details do not yet establish why it happened or whether the delay was the employee’s fault. The deadline may have changed, an input may have arrived late, or the person may have misunderstood which version was final. These are questions to check, not explanations to assume. The point is to keep the parts distinct: what happened, what standard applied, what consequence followed, and what the speaker thinks it means. The University of Minnesota’s “Difficult Performance Conversations” guide recommends preparing by checking what happened, considering possible explanations, looking for information that contradicts one’s view, and defining an objective focused on work. It also asks supervisors to notice when their own strong reaction may come from personal preference or bias. That is a practical use of emotional awareness: irritation can be a cue to slow down and inspect the evidence, rather than evidence that the employee is irresponsible. If a manager strongly prefers advance notice, for example, that preference should not silently become a performance standard unless the role or team agreement actually requires it. The same discipline helps the employee take part without having to accept a character label. They can say which fact they dispute, identify a missing dependency, or explain how they understood the deadline. Listening does not require the manager to agree with every account. It makes the conversation a chance to test the claim. If the report was late and the expectation was clear, that remains relevant; if the expectation or input was different from what the manager believed, the concern may need to be narrowed or revised. The CIPD’s “Performance management: A guide for people managers” places clear expectations and fair, timely attention to underperformance within effective performance management. In practice, specificity means naming the relevant expectation before asking for change. If there was no agreed deadline, the next step may be to set one. If a standard was clear and missed, the discussion can turn to what would make the next submission meet it and what support is needed. Avoid labels such as “resistant” when the actual issue is a disputed instruction or a question about feasibility. The employee can answer a concrete concern; an identity judgment invites an argument about who they are. A checkable account leaves room for accountability and for facts that change the appraisal.

Sources: Difficult Performance Conversations; Performance management: A guide for people managers

Why can feedback make the work harder to discuss?

Feedback can redirect attention from the task toward self-protection. A specific example and a clear next action give both people a way back to the work, though no phrasing can guarantee that result. The risk is easy to miss because a performance conversation often begins with a work concern and quickly becomes a judgment about the person. “You are careless” asks the employee to defend a character label. “The client handoff on Tuesday went out without the revised figures” identifies an event they can inspect. The second sentence is not automatically fair or complete; its value is that it gives the discussion something concrete to test.

The 1996 meta-analysis in “The Effects of Feedback Interventions on Performance” gathered 607 effect sizes from 23,663 observations. Feedback interventions improved performance on average (the reported standardized effect was d = .41), yet more than one-third of the interventions were followed by lower performance. Kluger and DeNisi proposed feedback intervention theory to explain why feedback can help in one circumstance and hinder in another: it can shift attention among task learning, task motivation, and self-focused concerns. In their analysis, effectiveness tended to fall as attention moved away from the task and closer to the self. That is a useful warning for a difficult review. Once the exchange becomes a debate over whether someone is competent, loyal, or a “good employee,” both people may spend less attention on what happened and what would change the result next time.

The study does not test a modern performance-review script, an emotional-intelligence course, or one particular meeting. It is a historical synthesis, and its authors note that task characteristics also matter. It therefore cannot show that behavior-specific wording will improve every conversation. The Chartered Institute of Personnel and Development’s 2022 “Performance feedback: An evidence review” similarly warns that feedback can be detrimental when it is not delivered in the right way, while emphasizing that effects depend on how feedback is given. Together, these sources support a narrower conclusion: feedback is not helpful simply because a manager has delivered it. Its content and delivery can affect where attention goes, and outcomes vary.

Emotional regulation matters at this point as a practical pause, not a demand that either person appear unbothered. A manager who notices irritation rising can resist adding old grievances, sarcasm, or a sweeping label to the current concern. An employee who feels accused can ask which example and expectation the feedback refers to, then say what part is inaccurate or missing. Neither person has to agree immediately for the conversation to remain useful. A short pause may help when voices rise or the same claim is being repeated; the discussion can return to the evidence and the decision that still needs to be made.

For example, if a report was submitted after an agreed deadline, the conversation can stay with the date, the expectation, and the impact on the next handoff. The employee may add that a dependency arrived late; that context could change who needs to act, while leaving the delay and its effect open for discussion. This is an illustration, not a documented case. It shows why a broad accusation makes the work harder to examine: it invites an argument about identity before either person has established what happened. Specific language is a route back to the task, not a technique for making criticism painless or forcing acceptance.

When attention moves toward whether the employee is good, bad, or under attack, return to one work behavior and ask what evidence would show improvement. State the observed event, the relevant expectation, and the impact; hear the employee’s account; then identify an observable next action. If the facts change the interpretation, revise the concern. If they do not, keep the standard clear and discuss what support or follow-up is needed. Emotional intelligence can help the participants notice when the exchange has drifted into self-defense and choose a response that brings it back to the work. It cannot make weak evidence strong, guarantee agreement, or promise a performance gain.

Sources: The Effects of Feedback Interventions on Performance: A Historical Review, a Meta-Analysis, and a Preliminary Feedback Intervention Theory; Performance feedback: An evidence review

How does fairness change what the employee can hear?

Respectful, even-handed treatment can make critical feedback easier to accept, but fairness of delivery and accuracy of the appraisal are separate questions. Columbia Business School’s abstract for “When is criticism not constructive? The roles of fairness perceptions and dispositional attributions in employee acceptance of critical supervisory feedback” summarizes two studies. In Study 1, greater interpersonal fairness in delivering negative feedback was linked to more favorable attributions about the supervisor, greater feedback acceptance, and better reactions toward the supervisor and organization. Study 2 reported that fair interpersonal treatment reduced negative attributions about the supervisor, which in turn increased acceptance and improved attitudes toward the supervisor. Published in Human Relations in 2001, the abstract gives no sample sizes or detail sufficient to quantify the effects. It supports a claim about reception, not a promise that courtesy produces agreement or better work.

Attribution is the link: when criticism arrives, the employee is also judging why it is being delivered. Is the supervisor clarifying a work problem, punishing, embarrassing, or defending a decision already made? A concrete reason, an opportunity to reply, and respectful treatment give the employee less basis to conclude that the exchange is personal. That may leave more room to examine the example, but it cannot establish that the example is accurate. Fair treatment affects how feedback reaches the employee; it is not evidence that the feedback is true.

The Columbia abstract distinguishes fairness in the procedures used to form feedback from interpersonal treatment during delivery. Procedural fairness concerns how the appraisal was reached: what evidence counted, whether relevant information could be raised, and whether a standard was applied consistently. Interpersonal fairness concerns how the person was treated while hearing the criticism. A courteous meeting cannot repair an appraisal built on a mistaken example or uneven standard; a warm tone cannot show that the process was fair. Neither dimension replaces checking the work evidence.

Emotional intelligence helps as a practical discipline: stay curious about the employee’s account while keeping the work concern visible. The University of Minnesota’s “Difficult Performance Conversations” guide advises listening for intent and context, asking how the situation arose, and making the exchange a dialogue. It also recommends specific descriptions of behavior or work outcomes and their impact. A supervisor might say, “I understood the handoff was due Tuesday, and the receiving team did not have the files then. What am I missing about the timing?” If the deadline changed, the example needs correction. If the facts hold, the supervisor can still state the expectation and discuss what must change.

Listening is not automatic agreement, and respect does not require softening a clear standard until it disappears. Minnesota’s guide recommends acknowledging both people’s views, describing what success looks like, confirming needed support, and agreeing on next steps. These are practice recommendations, not proof that one script works everywhere. They preserve answerable questions: what happened, what impact did it have, and what should happen next? Consider the reply without allowing it to erase verified impact.

When an employee says criticism feels unfair, treat that as a reason to inspect the example and how it was reached, not as proof of avoidance. Ask what they dispute and what information would clarify it. Check whether the standard was communicated, applied consistently, and connected to the stated impact. The Minnesota guide asks supervisors to consider information that contradicts their perceptions and distinguish a work problem from personal preference or bias. Such checks may change the conclusion; they should not be a courtesy ritual before returning to a verdict already fixed.

Fairness can make criticism easier to hear by reducing the suspicion that it is personal and giving the employee a real chance to respond. It cannot certify the appraisal. Make room for the reply, correct factual errors openly, and retain the expectation when evidence supports it. If the process or standard remains unclear, resolve that before asking the employee to accept the conclusion.

Sources: When is criticism not constructive? The roles of fairness perceptions and dispositional attributions in employee acceptance of critical supervisory feedback; Difficult Performance Conversations

Before emotion becomes the explanation, check the process

Emotional intelligence cannot make an unclear expectation, impossible workload, inconsistent standard, or concealed formal decision fair. Before treating tension as a problem of composure, check whether both people understand what is being discussed and what can still change. Start by naming the purpose. Is this a developmental conversation about improving future work, or an administrative discussion connected to a decision already under consideration? The distinction matters because an employee cannot respond meaningfully if the meeting is presented as open coaching while its outcome is effectively predetermined. In “Performance management: Could do better?”, the CIPD’s evidence review reports that developmental appraisals tend to use a wider range of performance examples and stricter ratings than administrative appraisals tied to pay or promotion. It recommends separating them because they involve different cognitive processes. That is guidance from a review, not a universal rule; its useful point is that purpose changes what evidence is relevant and what the employee should be told. Emotional skill here means being direct about the meeting’s purpose, what decision is open, and which standards apply. If the manager does not know which process applies, clarify it before asking the employee to interpret the discussion. Calm delivery cannot compensate for hidden stakes. The standard itself needs checking. The CIPD’s “People manager guide: Effective performance management” identifies clear expectations and fair, timely attention to underperformance as parts of effective performance management. Ask what expectation was communicated, when it applied, and whether it matched the work conditions. A target may have changed, information may have arrived late, or the employee may have faced incompatible demands. The University of Minnesota’s “Difficult Performance Conversations” guide advises supervisors to understand what happened, consider possible explanations, look for information that contradicts their perception, and assess whether their reaction reflects work impact or personal preference. It also recommends defining a work-focused objective and checking whether the person has needed resources and support. Ask whether the concern can still be stated as an expectation, an observable gap, and a work impact after irritation has been set aside. Workload belongs in that test. If a deadline was missed while priorities changed repeatedly or essential tools were unavailable, examine those conditions alongside the missed work. The employee may need to change a behavior; the manager may need to clarify priorities, provide access, or renegotiate scope. Minnesota’s guide includes support and resources in defining success, The employee needs a clear action, and the manager must identify what needs to be in place. Consistency is another process question. Compare the standard used here with the expectation communicated and the way similar work is evaluated. A difference may have a sound reason, but explain it in terms of the work rather than personal impressions. “Performance management: Could do better?” highlights fairness and usefulness in employees’ reactions to appraisal and recommends checking in after a performance conversation. The employee’s view matters, but neither perceived unfairness nor a favorable reaction proves the judgment right or wrong. Emotional awareness should not become pressure to accommodate every objection. Listen for information that could change the account, correct errors when evidence warrants it, and state what remains unresolved. If the employee questions whether the expectation was clear or evenly applied, ask what record or example would help examine it. If the concern is confirmed, revise the conclusion or process. If not, explain why the documented standard applies and what action is expected. A boundary may help more than another attempt to smooth the exchange. If the purpose is unclear, pause to establish whether the conversation is developmental or administrative. If evidence is incomplete, agree on what must be checked and when to resume. If the standard or workload appears inconsistent, record the discrepancy and use the relevant workplace process to clarify it. Minnesota’s guide recommends a reasonable pause when discussion becomes unproductive, then a revisit with clear expectations, support, and next steps. Use the appropriate workplace process for the circumstances. The test is whether the conversation can now describe a fair, specific work expectation and a next action both parties understand. If not, clarify the purpose, verify the standard and available support, and address the process gap before treating the employee’s reaction as the central problem.

Sources: Performance management: A guide for people managers; Difficult Performance Conversations; Performance management: Could do better?

Can an EQ test tell you whether the appraisal is right?

No. An EQ result may prompt self-reflection about habits, but it cannot verify a missed deadline, establish why it happened, or decide whether performance feedback is accurate. Those questions require evidence about the work: the expectation that applied, the record of what happened, the conditions around it, and the employee’s account. Emotional intelligence may help someone ask better questions about that evidence. A score cannot take the place of it.

The label “EQ test” covers measures built for different purposes. The Swiss Center for Affective Sciences’ overview, “Current Models,” distinguishes ability approaches from trait approaches. An ability measure asks someone to solve emotion-related problems, such as identifying how emotions change or which response may manage a situation. A trait or self-report measure asks people to rate their perceived typical tendencies. Some broader competency approaches combine elements such as self-awareness and relationship management, sometimes using other people’s ratings. These results are not interchangeable: a person’s report of how they usually respond is different evidence from performance on an emotion task, and neither directly documents the quality of a particular work product.

That difference matters when a conversation turns into an informal test of character. Imagine a manager says that a report arrived after its agreed deadline and then concludes, “You lack self-awareness.” The first statement can be checked against the deadline and the work record. The second leaps from one event to a broad claim about the person. A self-report profile may help the manager reflect on whether they tend to become impatient or the employee notice a habit of avoiding clarification. It cannot establish that either interpretation caused this delay. The relevant next questions are narrower: Was the deadline clear? Did something change? What action would prevent a recurrence?

The 32-item Emotional Skills Profile is intended as a private educational reflection tool, with questions about recent behavior and emotional situations. Its purpose is to help a person examine habits and consider a development priority. It has no validated population norms or established ability-test validity, so the report should not be read as a standing of the employee against colleagues or as proof that one person is better equipped for the job. Nor should its reflection prompts become an unofficial performance score. The profile concerns a person’s responses to its questions; an appraisal concerns specified work and the standards applied to it.

The distinction is easy to lose when a number appears precise. A score can compress answers into a clear-looking result, but the format alone does not tell a manager whether a deadline was realistic, whether the employee had the required information, or whether the same standard was applied to comparable work. In a fair performance discussion, these are separate records. The employee can disagree with the description of the work, add context the manager did not have, or accept the facts while questioning the expectation. An EQ result cannot settle any of those disagreements.

Research linking emotional intelligence with job outcomes is relevant to the broader idea that emotional skills may matter at work, but it answers a different question. The 2022 “A Meta-Analysis of the Relationships Between Emotional Intelligence and Employee Outcomes” separately examined ability, self-report, and mixed measures. Across 68 independent samples comprising 23,269 people, the authors reported a positive corrected association between overall EI and job performance (ρ = 0.28). The reported associations also differed somewhat among the three streams. This is a relationship across studies and groups, not evidence that an individual’s profile caused their performance or that a particular appraisal is correct. Some included studies measured performance through self-report, which the review identifies as a limitation. Even a real group-level association cannot tell a manager why one deadline was missed.

Evidence that emotional competencies can develop also does not validate a performance judgment. The 2024 systematic review “Training emotional competencies at the workplace” combined workplace training studies and reported an average improvement across its analyses, alongside wide prediction intervals and substantial variation between studies. Many included studies had serious or critical risk-of-bias concerns, including issues with confounding and outcome measurement. The result makes it reasonable to treat practice as a possible route to skill development; it does not show that a short exercise will change the next meeting, that every training approach works, or that any assessment can rank an employee’s work.

A useful comparison is between a reflection prompt and a work record. If a person notices that they tend to react quickly when criticized, they might prepare to pause, ask which example is being discussed, and check their understanding before responding. That observation can shape how they participate. The evidence about the appraisal still comes from the expectation, the work itself, and relevant context. If the manager has misunderstood what was delivered, those facts should change the appraisal. If the evidence supports a missed standard, reflection may help the employee plan a response, but it does not remove the need to name the standard and agree on what happens next.

Keep any EQ profile private and developmental. In the conversation, use it only if the person chooses to draw on it as a reflection aid; do not ask for a score to justify feedback or infer that a low or high result explains the work. Assess the concern from the applicable standard, specific evidence, and the employee’s account. That keeps two decisions in their proper places: what the work record supports, and what emotional habit someone may want to practice.

Sources: Current Models; A Meta-Analysis of the Relationships Between Emotional Intelligence and Employee Outcomes; Training emotional competencies at the workplace: a systematic review and meta-analysis; Emotional Skills Profile

Can these skills improve with practice?

Workplace emotional competencies appear trainable on average, but training evidence does not promise that one conversation will improve or that a particular intervention works equally well for everyone. The practical case for practice is narrower: rehearse specific choices, try them in real exchanges, and review what happened. That gives a person something concrete to adjust before the next difficult discussion without turning one outcome into a verdict on their ability.

In the 2024 systematic review and meta-analysis “Training emotional competencies at the workplace,” researchers combined 50 studies with 3,233 employed participants for pre-to-post comparisons, and analyzed 27 controlled studies separately. The controlled-trial analysis found an average standardized mean difference of 0.46 in measured emotional competencies. This suggests measured competencies can change through workplace interventions; it does not show that every participant improved or that a particular feedback script caused the change.

The spread matters. Results varied substantially; the authors report high heterogeneity and low methodological quality overall. Many studies had serious or critical risk-of-bias concerns, including uncontrolled designs and outcomes based on self- or peer-report. It does not test the short sequence here as a package or show that improved competency scores automatically mean better performance ratings.

A useful practice cycle begins before the meeting, with one example that can be described without a character label. Note the observable behavior or work result, the expectation it relates to, and the impact that can be checked. The University of Minnesota’s “Difficult Performance Conversations” guide advises examining what happened, considering possible explanations and contradictory information, and identifying the work objective. That preparation helps the speaker notice when irritation or preference is supplying more certainty than the evidence. The aim is to explain what is known and what remains interpretation.

During the exchange, notice your own activation before it chooses your next sentence. A rising urge to list every past failure or demand immediate agreement can pull attention from the specific issue. A brief pause makes room to ask what fact the other person disputes and whether the standard was clear. Minnesota’s guide recommends listening to understand, delaying a reactive response, and pausing to revisit the conversation if it becomes unproductive. Listening is not a promise to accept every explanation; it checks whether the account needs correction before deciding what follows.

Then agree on one observable next behavior and a review point. If a handoff was late, the next step might specify when an update should arrive and what support or dependency needs resolving. At follow-up, compare the expectation with what happened and ask whether the plan was workable. This practical application is not a tested formula from the meta-analysis.

Practice should match the evidence being used. Ability tests ask people to solve emotion-related problems; self-report measures describe perceived or typical tendencies; mixed competency measures may combine emotional skills with broader work behaviors. They do not provide interchangeable evidence about one appraisal. A reflective profile can help its user choose a habit to examine, but cannot establish whether a performance concern is accurate. Check instead: did the person describe the example clearly, hear a relevant correction, and leave with an assessable next step?

For the next conversation, choose one behavior to practice: prepare a precise example, notice the moment you want to defend or withdraw, ask a question that could change your understanding, or agree on a follow-up. Review what the exchange clarified. If the behavior remains unclear, the process unfair, or support unavailable, address that condition before asking either person for greater composure. Improvement is a pattern across conversations, not a conclusion from one meeting or EQ result.

Sources: Training emotional competencies at the workplace: a systematic review and meta-analysis; Difficult Performance Conversations

Leave with one behavior and a real follow-up

Close by turning the discussion into an agreement that can be checked. Name the specific behavior or work result, the expectation it needs to meet, and when you will review it. The University of Minnesota’s “Difficult Performance Conversations” guide recommends describing success, confirming needed resources, setting next steps, and revisiting progress through ongoing check-ins. If a handoff arrived without the agreed status note, the employee might commit to adding it before the next transfer. The manager might clarify which fields are required and make the template available. At follow-up, they can inspect the next handoff and ask whether the clarified standard or support resolved the obstacle. Both people have an action; emotional display is not the measure of progress. CIPD’s “Performance management: A guide for people managers” likewise places clear expectations within ongoing management. If either person cannot continue productively, agree on when to resume and what information to bring. Minnesota’s guide allows a pause and return within a reasonable time. A pause with a date preserves the issue; leaving without a next step can let it drift. Decide what observable evidence would show change and what support you owe, then set the review point. For private reflection on your emotional habits, the 32-item Emotional Skills Profile at /assessment offers a development prompt; it cannot determine whether the work concern is accurate.

Sources: Difficult Performance Conversations; Performance management: A guide for people managers; Emotional Skills Profile

Questions readers ask