Group Exercises and Role Plays: What Assessors Actually Write Down

Everything that feels unfair about a group exercise comes from not knowing what is being written down. An assessor sitting outside the discussion is not forming an impression of you. They are running a documented procedure: record behaviour as it happens, classify each piece of it against a named competency, then rate it later against a scale with written anchors. Once you know that, the behaviour that scores stops being mysterious, and so does the reason that talking the most tends to score badly.

What is actually on the assessor's sheet

Assessor training across the industry follows roughly the same four steps, usually taught as observe, record, classify, evaluate. During the exercise the assessor writes near-verbatim notes of what candidates say and do, often with rough timings, and deliberately does not judge yet. Afterwards they sort those notes into the employer's competency framework, and only then do they put a number on each competency, against a scale whose points are described in behavioural terms rather than as good, average and poor. At the end of the day, assessors meet for a wash-up or integration session and read their evidence to each other before ratings are agreed.

One assessor typically tracks two candidates rather than the whole room, which explains something candidates worry about a lot. You do not need to be the most memorable person present. The person marking you is looking at you specifically, for the whole exercise, and will notice a well-timed intervention in minute eighteen even if nobody else does.

Why dominating the discussion loses marks

Taking over feels like leadership and reads as something else entirely. There are three structural reasons it goes wrong, and none of them depends on the assessor disliking you.

• You are rated against a standard, not against the group. Assessment centre scoring is criterion-referenced: a group that works well can produce several strong scores, and talking over someone does not lower their rating. It only adds negative evidence under your own listening and teamwork lines.
• Airtime is not a competency. Nothing on the sheet is called contribution volume. If your fifteen minutes of talking maps to two classifiable behaviours, you scored two behaviours and burned the time in which you could have produced others.
• Most frameworks pair influencing with collaboration. Employers rarely assess persuasion on its own, because persuasion that costs you the team is not the behaviour they want. Interruptions and talk-overs get logged, and they are read against a second competency you never thought you were being marked on.

The contributions that reliably score

These are the moves that are easy to classify, which is exactly why they earn marks. Plan the wording before you go in, because the hard part is not thinking of them, it is producing them while the discussion is moving.

Bringing someone in, by name and for something specific

"Sam, you had the cost figures on the second option. What do they say?" This is credited almost immediately and hardly anyone does it. Make it a request for something concrete rather than a general invitation, because "does anyone else want to add anything" is a courtesy and does not demonstrate that you have been tracking who holds what.

Summarising at the moment of confusion

The useful summary is not a recap of everything said. It has three parts: what the group has agreed, what is still open, and what should happen next. Delivered when the discussion has gone in circles twice, it demonstrates listening, structuring and drive in one intervention, and it is usually the thing the group needed anyway.

Owning the clock, visibly

Three moments do most of the work. In the first two minutes, propose how long to spend on each stage and check the group agrees. At the halfway point, say where you are against that plan. Five minutes from the end, name what still has to be decided and propose how to close it. Groups that drift to a stop with nothing agreed do so because nobody said the last one.

Disagreeing well

This is the highest-value behaviour in the exercise and the one candidates avoid. The technique is to disagree with a criterion rather than with a person: "Option B looks strong on speed, but the brief caps us at the existing budget and B is over it. Can we test both against that line?" You have introduced evidence, challenged the conclusion and left the other candidate's standing intact, which is three things an assessor can classify.

The tail of this is just as important. If the group answers your objection, say so and move on. If it does not, say what your reservation is and let the group decide anyway. Agreeing to a weak option purely to end the friction is not collaboration, and it is marked as conflict avoidance rather than flexibility.

Checking the recommendation against the brief

Groups routinely agree something that breaks the budget, the timeline or the constraint they were given twenty minutes earlier. Being the person who reads the constraint back before the decision is finalised is a straightforward mark and takes one sentence.

What changes when the exercise is virtual

Video calls break the mechanism people normally use to take turns. Latency means the natural half-second gap you would step into has already been taken by the time you hear it, so discussions swing between talking over each other and awkward silence. The candidate who manages this deliberately reads as the most competent person in the room, and the method is unglamorous: finish by naming who you are handing to, then stop talking and wait longer than feels comfortable. Use the participant names on screen, which removes the excuse for not bringing people in by name. If there is a chat panel, treat anything you type there as observed evidence, because it is.

Role plays are not small group exercises

A role play puts you in a short one-to-one conversation with a trained role player: an unhappy client, an underperforming team member, a supplier in dispute, a colleague cutting a corner. It looks like a softer version of the group exercise and it is a different test with a different failure mode.

Three things follow from the role player being scripted. The escalation partway through is usually written into their brief, so it is not a reaction to you failing, and knowing that makes it far easier not to take personally. The information that unlocks the scenario is held by them and released in response to questions, so a candidate who talks for most of the conversation never finds it. And because the role player is instructed to repeat an unacknowledged concern, jumping straight to a solution buys you the same complaint a second time.

Scenarios built around a shortcut, a pressured figure or a favour for a friendly client are integrity probes rather than relationship tests. Handling the person warmly while letting the issue pass caps that competency regardless of how smooth the conversation was. In some formats the role player also completes a short rating on what the conversation felt like from their side, which is a reminder that the impression you leave is itself data.

What to prepare the night before

Three things, none of which involve the topic of the exercise, since you will not know it. Read the employer's published competency framework and write down the four or five behaviours it names, because that is the vocabulary your evidence will be sorted into. Rehearse the five or six sentences above until they are automatic: the invitation by name, the three-part summary, the halfway time check, the criterion-based disagreement, the constraint check. And for a role play, write yourself a two-line limits table: what you can offer, what you cannot, and who you would escalate to. That table prevents the single most costly error in the exercise.

CogniPrep's group exercise scenarios put you at decision points in a discussion and score the contribution you choose, and the role-play scenarios run turn by turn against a character with their own agenda, with feedback on what each response would have evidenced. If you have a full assessment day ahead, the exercise set also covers the e-tray and in-tray formats, which are usually the first thing you will sit on the morning.