Two examiners attend the same simulator session: same crew, same failure, same exchanges. On the grid, one rates the action plan at the minimum acceptable level, the other finds it comfortable. Neither is acting in bad faith—it’s their calibration that has diverged. The entire system for maintaining instructor and examiner competence addresses this precise issue: maintaining not the action, but the judgment, and keeping it aligned with that of others.
The evaluator’s paradox: a judgment that wears out silently
A pilot’s competence is almost self-maintaining: each line stage confronts them with the machine, weather, and periodic checks, and the gap is quickly visible. An instructor’s competence is already more sheltered—transmitting, diagnosing an emerging error, and adjusting the difficulty of a scenario are actions that no one verifies at every session. An examiner’s competence, finally, receives practically no natural feedback: a too-generous grade does not cause an incident on the spot, and a too-strict grade does not trigger any alarm. The drifting evaluator produces perfectly plausible evaluations.
The regulatory framework takes this problem seriously: judgment is a perishable skill, and its degradation is invisible to the one experiencing it. An examiner does not feel themselves becoming lenient; they see candidates as “generally better than before.” An instructor does not feel themselves lowering their standards; they find their trainees “at level.” No self-assessment-based system can correct this.
The system’s response is structural: each level of the training pyramid is observed by the level above. The crew is trained by the instructor, the instructor is monitored by the examiner, the examiner is observed by a senior examiner or inspector, and the authority closes the loop by descending back to all levels through standardization. The value of a control—what a recruiter, insurer, or foreign authority grants to a qualification—rests entirely on the integrity of this chain.
Judgment is like a muscle that weakens over time without exercise. Since examiners don’t get immediate feedback, the system uses higher-level oversight to keep everyone aligned. It’s like having a coach check your technique regularly to prevent bad habits from forming.
An A320 instructor in Paris evaluates 100 candidates per year. After 5 years, their grading becomes slightly more lenient without them noticing. A senior examiner observes their sessions and finds their “satisfactory” ratings are now equivalent to what used to be “good.” The instructor then attends a standardization seminar to recalibrate their judgments.
Three paths, one cycle: the mechanics of revalidation
In the European system, an instructor certificate typically follows a three-year cycle, and its revalidation is not played out at the deadline but over the entire duration of the cycle. The mechanism relies on three paths: attested instructional activity volume, a refresher seminar, and a competence evaluation conducted by a qualified examiner. Depending on the certificate category, two of these three paths suffice—and this is a design choice, not a convenience.

The logic is worth examining closely. The active instructor proves their maintenance through what they already do: documented activity and a seminar that recalibrates them. The less active instructor, however, cannot rely on a practice they did not have: the combinations push them mechanically toward in-situ evaluation, i.e., direct observation. The system directs each profile toward the control they truly need. And to prevent a very active instructor from indefinitely escaping external scrutiny, the competence evaluation periodically returns by right.
The deadline, however, changes the nature of the requirement. As long as the certificate is valid, extension is a matter of proof: activity logs, seminar attestation, evaluation report. Once the deadline is passed, the burden shifts—the renewal requires a refresher whose content is determined by a training organization, followed by an evaluation that is no longer optional. The same privilege costs far more to recover than to maintain.
The examiner adds another level: their authorization, most often based on an instructor certificate that it presupposes, follows its own schedule. Its maintenance combines a minimum activity of conducted tests, periodic standardization with the authority—or recognized by it—and, central to the system, a test conducted under the direct observation of an inspector or senior examiner. And the authorization does not survive the underlying qualifications: an examiner whose underlying qualification lapses loses, in the same motion, the right to examine.
| Aspect | Instructor certificate | Examiner authorization |
|---|---|---|
| Typical cycle | Three years | Three years, on a separate schedule |
| Maintenance paths | Two out of three paths: instructional volume, refresher seminar, competence evaluation | Minimum activity of conducted tests, standardization with the authority, test conducted under observation |
| Who observes | A qualified examiner, during the competence evaluation | An authority inspector or senior examiner |
| After the deadline | Refresher defined by an organization, then evaluation | Complete re-standardization before any new test |
| What the cycle protects | The quality of transmission | The consistency and comparability of judgment |
The trap: conducting a test with an expired authorization. The examiner’s schedule is separate from that of their flight qualifications and instructor certificate, and nothing materially blocks the conduct of a control outside validity. The test is then without value: the candidate believes they have an extension they do not, the discovery occurs months later during an audit, and regularization involves the candidate, operator, and authority.
The right reflex: maintaining a separate schedule for evaluator functions, separate from that of qualifications, and consuming the windows early. Seminar placed at the start of the window, observed test scheduled as soon as the last period opens, activity log kept as it goes: each element consumed early turns a deadline into a simple formality.
Standardization: manufacturing inter-judge fidelity
The refresher seminar is the least understood component of the system. From afar, it’s a day of regulatory reminders; up close, it’s a calibration bench. The heart of the session, in organizations that take it seriously, looks like this: a filmed or replayed session, individually graded by each participant on the current grid, followed by a comparison of the grades—and it’s the gap between evaluators, not the observed crew’s performance, that becomes the focus of the work.
Because the grid is not enough. Well-written behavioral markers reduce dispersion, but they do not eliminate it: two honest readers of the same reference produce different grades, each anchored in their own history of what a “good” crew is. Inter-judge fidelity—the property that ensures a grade means the same thing regardless of the grader—is not decreed in a manual: it is manufactured through repeated confrontation of judgments on common cases.
The discipline that follows is simple to state: a grade is justified by observable facts—what was said, done, or omitted, and when—never by a general impression. A standardized evaluator can answer, for each graded competence, the question “What did you see that justifies this level?” This is also what makes the grade defensible: in front of the contesting candidate, the organization, and if necessary, the authority.
Cross-observation: the observed test
The most singular mechanism in the system remains the observed test: periodically, an examiner conducts a real test—real candidate, real stakes—while an authority inspector or senior examiner, seated in the background, evaluates them. The observer does not grade the candidate; they grade the conduct of the test: completeness of the briefing, conformity of the scenario to the prescription, neutrality during the session, quality of the debriefing, solidity of the documentation.
The most scrutinized moment is the one that occurs rarely: managing an insufficient performance. Deciding a failure, announcing it, justifying it with facts, documenting it so it holds—this is the part of the job that wears out the most and is seen the least. A profile with no failures is in fact a signal, just as an abnormally high failure rate: both extremes betray a personal calibration that has replaced the standard.
Cross-observation also exists, in less formal ways, within organizations: instructor seated in the back of a colleague’s session, double reading of decisive sessions, concordance meetings where borderline files are reviewed by several. And it is doubled by a rule of independence: the one who provided the training for a candidate for a test is, except in cases provided for by the text, not the one who administers it—and existing ties are declared. Judgment is also protected by distance.
What the data reveals: drift can be measured
Competency-based training programs have added a level to the system that classic regulation did not know: measurement. As soon as each session produces structured grades—by competence, by phase, by scenario—the organization has statistical material: distribution of grades by evaluator, actual use of the scale, compared failure rate, correlation of each grader with the collective on comparable sessions.
These data make visible drifts that no introspection detects. Indulgence that sets in over two years is read in a sliding average; the halo effect is read in abnormally correlated grades between supposedly independent competencies; the central tendency is read in a scale whose extremes are never used. Each of these signatures has its countermeasure, and it’s precisely this that structures effective standardization sessions.
| Drift | Manifestation | Countermeasure |
|---|---|---|
| Indulgence or severity | Entire distribution shifted up or down on the scale | Statistical comparison with peers, blind-graded filmed cases |
| Halo effect | A brilliant competence pulls all other grades upward | Competence-by-competence grading, based on observed facts |
| Central tendency | Rejection of extreme grades, everyone “in the middle” | Levels anchored on described behaviors, justification of median grades |
| First impression | The initial briefing contaminates the grading of the entire session | Phase-by-phase grading, review before debriefing |
| Temporal drift | Calibration that slowly shifts between two seminars | Periodic recalibration, longitudinal tracking of grades by evaluator |
A point of doctrine is worth stating: the data does not sanction, it triggers. An atypical grading profile is not a fault—it’s the starting point of a targeted standardization conversation, additional observation, or a deliberately chosen calibration case. The worst system is the one that collects without feedback: evaluators know they are measured, never see anything come back, and the measurement becomes an object of mistrust rather than a professional tool.
What you should check in your own file
For those holding—or aiming for—a instructor certificate or examiner authorization, this mechanism translates into concrete checks. Schedules first: that of the certificate, that of the authorization, that of the underlying qualifications, which almost never coincide. Windows next: a well-placed competence evaluation can, in cases provided for by the text, serve multiple purposes at once—it’s the regulation that defines the allowed combinations, not local practice. Finally, check who is qualified to conduct your evaluation: the required seniority of the examiner depends on what is being evaluated, and an evaluation conducted by the wrong level of examiner is worthless.
The gray area of the system deserves to be named: the independent instructor, who operates outside any structure, benefits from no internal loop—no concordance meeting, no colleague in the back seat, no data. For them, the regulatory cycle is not one floor among others: it’s their only calibration. The text is designed as a safety net, not as a complete maintenance regime; nothing prevents doing better, by seeking voluntary observation or joining the standardization sessions of an organization.
For those preparing an instructor qualification or building a path toward examination, finally, this system says one useful thing: the competence that will be controlled is not only that of the pilot, but that of the future judge—and it is worked on as such, with grading cases in support, from the preparation of the qualification.
A final methodological reminder: this article describes mechanisms, not your situation. The cycles, windows, and allowed combinations vary according to the certificate category and issuing authority; the applicable regulation, decisions of your authority, and approved documentation of your organization prevail in all circumstances over what precedes.

Leave a Reply
You must be logged in to post a comment.