Calibration Conversation¶

Structured feedback conversation (ritual) — instantiates Competence Calibration Feedback

A structured two-way conversation that surfaces a person's own self-assessment, sets the evidence beside it without triggering shame, and lands on one concrete next step.

Calibration Conversation is the human container the whole loop happens inside. Where Benchmarked Feedback hands over a standard-anchored verdict, this mechanism is a two-way dialogue whose defining move is to ask first: elicit the person's own estimate of where they stand before any evidence is shown, then place the evidence beside it together. That ordering is what makes the calibration land — the person watches their own prediction meet the record, rather than being handed someone else's conclusion. Around that moment it does two things a memo cannot: it frames the gap as information about reliability rather than about worth, and it reads the room in real time so the message is heard instead of defended against. It ends not with a score but with one action the person has agreed to own.

Example¶

An on-call engineer is confident they are ready to run a Sev-1 incident solo and is chafing at still being shadowed. Their lead opens a calibration conversation. First the ask: "Walk me through how you'd handle a database failover under load — and how confident are you, one to five?" The engineer says five. Only then does the lead lay the evidence beside it: two recent shadowed incidents where the engineer reached for the right runbook but missed the step of declaring a comms channel, leaving stakeholders dark for twenty minutes. The gap is now visible to the engineer, not asserted at them. The lead frames it deliberately — "this isn't about whether you're good; it's about what has to be reliable at 3 a.m." — and watches for the flush of defensiveness, slowing down when it appears. They leave with one agreed step: run the next incident lead-with-safety-net, specifically drilling the comms-declaration reflex, and revisit in two weeks.

How it works¶

Its distinguishing structure is the sequence, not the content. Elicit before revealing creates the predicted-versus-actual moment inside the room. Frame as information, not verdict keeps the gap attached to future reliability rather than to identity. Monitor for threat — reading defensiveness, deflection, or withdrawal as signals the message is bouncing off, and adjusting pace or framing before it is rejected. Converge on one owned action so the conversation exits as commitment, not as a feeling. The mechanism carries no evidence and no standard of its own; it is the vehicle that lets evidence be received.

Tuning parameters¶

Elicitation order — self-assessment before or after the evidence is shown. Before produces the sharpest calibration moment; after is gentler but forfeits the contrast.
Directiveness vs. discovery — telling the person the gap versus guiding them to name it. Discovery sticks better and builds their own judgment, but takes longer.
Safety-to-candor balance — how much reassurance versus how much bluntness. Too safe and the message blurs; too blunt and it triggers the very threat that blocks hearing it.
Action scope — how large the agreed next step is, from a single nudge to a development plan. Match it to the size of the gap and the person's current load.
Cadence coupling — one-off versus a recurring slot. Recurring normalizes calibration so it isn't read as a special summons.

When it helps, and when it misleads¶

Its strength is that it makes confidence corrigible while keeping the relationship intact: the person co-produces the diagnosis and leaves with a step they chose, which is why it holds. It depends on — and builds — psychological safety, the shared sense that being candid about a gap won't be punished, which is the precondition for honest self-report in the first place.^[1]

Its failure modes are the ones any hard conversation invites. It can soften into a pleasant chat that never actually names the gap — the "feedback sandwich" that buries the message between reassurances. A fluent talker can narrate their way out of the evidence. And it is easily run backwards: a conversation staged to socialize a decision (a reassignment, a hold on autonomy) already made, using the language of "calibration" as cover. The discipline that keeps it honest is to keep real evidence in the room, name the gap explicitly even when it is uncomfortable, and record the action that was agreed.

How it implements the components¶

Calibration Conversation fills the human, safety, and action-commitment slice — the parts that make evidence receivable:

self_assessment — elicits and makes explicit the person's own estimate of readiness, as the conversation's starting point.
feedback_safety_frame — frames the gap as information for reliability, not a judgment of worth, so it can be heard rather than defended against.
learning_or_escalation_path — converts the exchange into one concrete, owned next action.
identity_threat_monitor — reads defensiveness and withdrawal in real time and adjusts framing before the message is rejected.

It does not generate the evidence it discusses (performance_evidence_set, Skills Assessment) or supply the standard the evidence is judged against (performance_benchmark, Competency Framework); it works with what those provide.

Instantiates: Competence Calibration Feedback — the container in which diagnosed gaps become receivable and turn into a chosen next step.
Consumes: evidence and standard-anchored content from Benchmarked Feedback and its evidence-generating siblings.
Sibling mechanisms: Benchmarked Feedback · Peer Review · Calibration Exercise · Exemplar Comparison · Confidence Rating Scale · Competency Framework · Skills Assessment · Reflective Error Log · Simulation or Case Test · Decision Rights by Competence · Supervised Practice

Notes¶

Without an evidence source in the room, a Calibration Conversation degenerates into opinion-trading — two impressions with nothing to arbitrate between them. It is the container, not the content; it needs a benchmark or an evidence set to be a calibration and not just a talk.

References¶

[1] Psychological safety, in Amy Edmondson's sense, is the shared belief that a team is safe for interpersonal risk-taking — that admitting a gap or a mistake won't be met with punishment or humiliation. It is a precondition for candid self-assessment: people only report where they actually stand when doing so feels safe. ↩