Group Exercise in Selection: What Separates Leadership from Domination

Salle d’évaluation vue de haut, six chaises autour d’une table d’acier couverte de feuilles vierges

A candidate walks out of the group exercise convinced he owned the room: he opened the discussion, proposed the plan the group followed, and twice pulled a wandering participant back on track. The rejection arrives a few weeks later. The exercise did not measure what he thought it measured, and the observers’ grid was not wrong.

What the group exercise measures, and what it does not

The group exercise is neither a knowledge test nor an eloquence contest. It is a stand-in for the workplace: people who do not know each other are placed in front of a problem none of them can solve alone in the time allotted, and the exercise watches what the presence of others does to each of them. The material is deliberately low on technical content — a logistics trade-off, a priority ranking, a resource allocation — so that no candidate arrives with a domain advantage. Anyone looking for the right answer has already missed the point.

The recruiter’s question fits in one sentence: does the group’s decision end up above or below what its best individual member would have made alone? A crew exists for this reason. The added value of the collective does not come from adding up competencies but from the circulation of information between them: an objection that surfaces, a doubt that gets voiced, a decision that leaves a trace. The exercise looks for the behaviours that make this circulation possible, and those that cut it off.

The confusion arises because both show the same surface signal: someone speaks, often, early. Leadership and domination share initiative-taking, volume, and occupation of space. They diverge on a single, verifiable point — what they do with what others produce.

Matrix of speaking time and retained contributionTwo-axis matrix. On the horizontal axis, speaking time; on the vertical axis, the contribution retained by the group. Bottom left, the bystander; bottom right, the dominator; top left, the quiet supporter; top right, the leader.quiet supportlittle seen, very usefulleadermoves things forward and says sobystandercarries no weightdominatortakes up space, contributes nothingspeaking timeretained contribution
Speaking time and retained contribution are two independent axes. Two quadrants fail for opposite reasons, and the quiet supporter ranks above the dominator in most behavioural grids.

The exercise isn’t about who talks the most, but who makes the group’s decision better than the best individual could. A leader helps others contribute; a dominator silences them.

In a group of four pilots, one candidate (A) proposes a plan, ignores objections, and monopolizes the discussion. Another (B) asks for objections, reformulates them, and ensures the final decision is recorded. Candidate A’s idea is rejected; Candidate B’s is implemented. The observers score B higher because his contributions left a trace.

The markers the observer actually records

A trained observer does not note traits. « Charismatic, » « self-effacing, » « authoritarian » are not scoreable: they are conclusions, and two observers do not reach them at the same moment. What gets logged are time-stamped occurrences — sentences spoken, actions taken — tied to an indicator on the grid. The debrief will demand the example; an impression without an example does not survive the confrontation.

The indicators that recur from one grid to another are few and very concrete: initial framing, meaning stating the objective, the time constraint and the decision rule before diving into content; naming and calling on whoever has not spoken; reformulation, which takes up another’s idea and checks that it has been understood; how an objection is handled; traceability of the decision, with its rationale. Almost always added to this is an in-exercise injection — a constraint that changes, new data dropped in, reduced time. This moment is the real subject of the exercise: it separates the person who re-establishes the framing from the person who re-establishes control.

Same apparent behaviour, two readings on the grid
Observed moment Scored as leadership Scored as domination
Opening States the objective, time and decision rule, then falls silent Announces his solution before the problem has been laid out
Another’s idea Reformulates it, names it, records it Ignores it, then claims it as his own later
Objection Asks for the reason, decides and explains why Responds to the tone, form or status of the objector rather than the substance
Silent participant Calls on him by name on a specific point Comments on his silence or speaks in his place
Change in instructions Restates the objective, redistributes, replans Speeds up alone and stops listening
His own idea set aside Acknowledges it in front of the group and serves the decision reached Comes back to it, reformulates it, puts it back on the table
Debrief presentation Has the group validate the summary Presents the summary as his own

The two profiles look alike as long as nothing goes wrong

Plot elapsed time on one axis and share of speaking time on the other: the two curves start out together. The first minutes belong to those who dare, and in a group of strangers the first person to bring structure gets immediate credit, whatever his intention. That is why experienced observers note little during this phase: it discriminates nothing.

The split happens in two places. The first real disagreement — not a polite nuance, but a contradiction that challenges the plan. Then the injection. At both points, the curve of the leadership profile dips: he asks a question, leaves a silence, draws out the one who is contradicting him. The dominator’s curve rises, because there disagreement is treated as a threat and the available response is more volume. Over the whole exercise, the first speaks often less than average during production and more during framing and debrief; the second is flat and high, because he does not listen enough to have a reason to vary.

Share of speaking time of the two profiles over the course of the exerciseTwo curves of share of speaking time against time, with a horizontal line marking equal share and two vertical markers: the first real disagreement, then the injection of a new constraint. The dominator profile, dashed, stays high and flat from start to finish, always above equal share. The leadership profile, solid, rises during framing, dips below equal share at each of the two markers, then rises again for the debrief.equal sharedisagreementinjectiondominator profileleadership profileframingproductiondebriefduration of the exerciseshare of speaking time
The difference does not show at the start, but at the first disagreement. The dominator profile keeps a high, constant share of speaking time, never below equal share; the leadership profile drops below that share during production and at each moment of regaining control, because he puts others to work, and rises again for the debrief.

The classic mistake: confusing visibility with contribution

The most frequent mistake is not aggressiveness, it is a bookkeeping confusion. The candidate counts his own contributions to the discussion, while the grid counts retained contributions. These are two different units and they are not proportional: an idea stated once, picked up by someone else, amended and then written into the final decision carries more weight than a string of interventions that left no trace.

The shapes this confusion takes are recognisable. Repeating the same proposal louder and slower, instead of asking why it was not taken up. Grabbing the marker and never handing it back — writing on the board is a form of power, and monopolising the pen is scored as such. Correcting a participant on form when one has nothing to oppose on substance. Finally, producing the summary alone, out loud, without having it validated: the group nods along from fatigue, and the observer notes an unshared decision.

Path of a retained contribution and a lost contributionTwo horizontal tracks. At the top, solid line: proposal, picked up by another, objection integrated, written trace. At the bottom, dashed line: repetition of the same idea, rising volume, silence from the group, no trace.what remains in the gridproposalpicked upobjectionwritten tracewhat disappears at the debriefrepetitionvolumesilenceno trace
An idea exists in the grid only if it has left a trace. The top track produces scoreable occurrences; the bottom track produces noise, however much energy is spent pushing it.

The timekeeper trap. Appointing yourself timekeeper is the most tempting shortcut to existing without taking any risk on substance, and the easiest one for the observer to read: once the framing phase is over, someone who only ever announces the time remaining has produced an indicator of presence and no indicator of contribution. The role is only worth something if it serves the decision: announce the time remaining together with what is still left to decide, and propose dropping a point to meet the deadline.

Withdrawal, the other mode of failure

The symmetrical failure is talked about less and just as frequent. The competent candidate who stays silent loses the exercise, not because discretion is a flaw, but because a grid only scores what is observable. Sound reasoning that stays inside someone’s head does not exist for the exercise, and it will not exist in the line either: the objection kept to oneself is exactly the mechanism that crew resource management training tries to undo.

The correction is not to talk more, but to make visible what one is already producing. Three actions are enough and cost little: voicing disagreement the moment it arises rather than at the end, and stating it as a verifiable fact rather than an opinion; asking that one’s contribution be recorded when it is retained; naming others. A sentence that begins with someone’s first name and ends with a precise question produces more scoreable occurrences than a long solitary monologue.

Three interventions that count. Before going in, set yourself three moments rather than a volume target: a framing intervention right at the opening — objective, time, decision rule; an integration intervention at the first disagreement, reformulating the other’s objection before responding to it; a traceability intervention before the debrief, having the group re-read the decision and its rationale. Three well-placed interventions are worth more than continuously holding the floor.

What happens when the observers get together

The scoring does not lock in inside the room. After the exercise, the observers consolidate their notes, and this is the part candidates know least about. Each has followed one or two participants, each arrives with time-stamped notes. The rule is simple: a score is defended with an example, and a disagreement between observers is settled by comparing examples, never by the seniority of whoever is speaking. A candidate described as a dominator whom no one can cite a moment where he crushed a contribution sees that rating dropped.

This is also where the group exercise is put back in its place: one station among others. The same competencies — communication, leadership and teamwork, situational awareness, workload management — are re-scored in the structured interview and above all in the simulator session, under the constraint of real time. A profile that shines in the room and collapses in the simulator is not a contradiction: it is a sign that the observed behaviour was a role, and a role does not hold up when the machine sets the pace. Blow-by-blow accounts circulate on the forum, with the usual caveat: an account is not a grid.

Preparing without building a character

No posture wins the day, and the grids are built to detect it. Three competencies, however, can genuinely be trained and later serve on the line. Reformulation first: taking up another’s statement in one’s own words and asking for confirmation is a technical move, not a courtesy; it checks that the message got through and makes the idea available to the group. Announcing intent next: saying what one is going to do and why before doing it, which gives the group a handle to object — this is the very structure of an approach briefing, and the group exercise is its stripped-down version. Explicit time management last: announcing a milestone, keeping it, giving up a point to meet the deadline. This gets repeated in daily briefings as much as in a human factors module.

Two habits, conversely, cost dearly and are imported in good faith. Treating the exercise as an individual interview before witnesses, addressing the observers rather than the group: the grid scores the relationship to peers, not a performance for an audience. And transposing a field authority gradient into a group of peers — an experienced captain who carries the reflexes of the left seat into the room gets a fast decision and a silent group, which is precisely what the exercise is designed to spot.

What remains grey, and what needs checking

Two areas remain grey, and it is better to know them than to deny them. The first is linguistic. Many exercises are held in English; a non-native candidate sees his rate of speech drop, his reformulations shorten, and his interventions arrive with a time lag. A poorly used grid reads this as withdrawal. A well-used grid separates linguistic fluency, which belongs to a different assessment, from cooperative behaviour, and checks the same indicators over shorter units. If you are preparing for selection in a language that is not your own, work first on framing and reformulation phrases: they are what carries the score, and they are short sentences.

A man in a suit stands aggressively over a conference table, gesturing emphatically at seated colleagues in a tense business meeting.
Effective leadership requires collaboration, not intimidation.

The second is cultural. The threshold beyond which speaking up is perceived as an intrusion is not the same everywhere, and an international group shifts it for everyone. Here again, the remedy is the observable: an objection stated and recorded is legible across all conventions, an impression of dominance is not.

For anyone preparing for a type rating or a change of operator, there remains a concrete check, and it is not found in preparation manuals: the competency framework the operator actually applies, published in its training documentation, and the way it defines its behavioural indicators. Two operators can display the same competency title and score different things. Asking for the framework rather than for tips remains the best preparation available — and it is the same document one will later need to invoke against an examiner justifying a score.

In all circumstances, the applicable regulations, the competency framework and the operator’s approved training documentation take precedence over the descriptions given here, which are for educational purposes only.

Questions fréquentes

Should you talk as much as possible during a group exercise?

No. Behavioural grids score contributions retained by the group, not volume of speech. Three well-placed interventions — initial framing, integrating an objection, traceability of the decision — produce more scoreable occurrences than continuous verbal presence. A high speaking time with no trace in the final decision is even an unfavourable marker.

Is it disqualifying not to be the group's leader?

No, the leader role is not expected of everyone, and a group with several self-appointed leaders scores badly. What is disqualifying is the absence of observable contribution: sound analysis that stays inside your head does not exist for the exercise. So you must state your disagreement the moment it arises and ask for it to be recorded.

What should you do when another candidate monopolises the floor?

Competing with him on volume is the worst choice: the observers will then see two dominator profiles instead of one. The behaviour that scores well is reformulating his proposal to make it discussable, then explicitly opening the floor to a participant who has stayed silent. That produces two markers of cooperation where he produced none.

Does the group exercise weigh more than the simulator session?

They are two stations that shed light on the same competencies under different constraints, and their relative weight depends on each operator’s process. The simulator adds the constraint of real time and workload, which reveals behaviours that were being played. A good result in the room followed by a collapse in the simulator is read as a sign of a role, not a contradiction.

How should you prepare when the exercise takes place in English?

Focus first on short framing, reformulation and decision-proposal phrases, since they carry the score. A reduced rate of speech costs little if interventions come at the right moments and stay traceable. A properly used grid separates linguistic fluency, assessed elsewhere, from cooperative behaviour.

Do observers score personality traits?

No, they log time-stamped occurrences: a sentence spoken, a gesture, a specific moment, tied to an indicator on the grid. During consolidation, each score must be defended with an example, and a disagreement between observers is settled by comparing examples. An assessment without an example does not survive the debrief.

Commentaires

Leave a Reply

Découvrez l’écosystème FL330

FL330, c’est toute l’aviation professionnelle — explorez nos autres espaces.