How the reference ranges are built

Provenance: three sources, and they do not carry equal weight

The library rests on three sources. They are of different kinds, and that is worth knowing before you start using it.

The list of occupations comes from the Russian Ministry of Labour's occupational register and professional standards: 287 occupations drawn from 34 official fields of activity. Role titles follow the register and the professional standards rather than our own invention. The practical consequence is that you will usually find your vacancy under its formal job title, not under the internal label your company uses for it.

The stage model is Ichak Adizes's organisational lifecycle — eight stages, from Infancy to Bureaucracy. It is an external, published model that has been debated for decades. We did not invent it, and we have not quietly reworked it to suit ourselves.

The psychometrics — the eight profile parameters and the norms attached to them — rest on NeuroFrame's own validation research: a comparison sample of 10,000+ executives and entrepreneurs, with predictive value tested on 3,000+ employees against real KPIs and manager ratings.

This calls for candour. Two of the three provenance lines have an external anchor: the occupational list can be checked against the register, the stage model against Adizes's own work. The third line, "NeuroFrame validation research", has no external anchor. It is our data, obtained on our sample, and you cannot independently verify it today. What we do about that: we disclose the full evidentiary logic to a client's methodologists under a confidentiality agreement, and we offer calibration of the ranges against the client's own high-performing employees where the stakes justify it. What we will not do is present our own validation as if it were independent.

Nine facets: how a role is broken down

A range is not assigned to an occupation wholesale. The role is first broken down along nine dimensions — and these describe the work itself, not the person doing it.

A pipeline then does the rest. The facets set the baseline centres for the eight profile parameters: a role with a high cost of error raises the centre for conscientiousness, a role with long work cycles raises it for persistence, a role with high social intensity raises it for teamwork. The lifecycle stage then shifts those centres and alters the width of the corridor, narrowing it in some places and widening it in others. Dampers then come into play, preventing the stage from overriding the nature of the role altogether. All boundaries are rounded to the nearest five: figures such as "65–90" are not false precision but deliberate coarsening.

Ranges are usually 20–30 points wide. A wide range is not carelessness — it is a message. On that parameter the role makes no real demand, and it should not be used as a selection filter.

The nine facets of role analysis

• Cognitive complexity — how demanding the problems the role has to solve are
• Knowledge novelty — how often the role requires learning something new
• Cost of error — what a mistake leads to, from inconvenience to danger to life
• Degree of regulation — how far the work is prescribed by procedure
• Autonomy — how independently the person has to act
• Social intensity — how much of the role consists of dealing with people
• Level of leadership — whether the role answers for other people's results
• Required risk appetite — whether the role calls for acting under uncertainty
• Length of work cycles — how long the tasks are that must be carried through to the end

Six principles that move the ranges

The lifecycle stage does not move the requirements arbitrarily. Six substantive principles sit behind the shift, and every one of them can be contested on its merits — which is the right way to argue with the library.

The third principle carries a caveat we would rather state ourselves. The rise in demands on progress monitoring and conscientiousness holds reliably from department-head level upwards. At lead-specialist level the picture is less clean: among individual specialists there are many roles with a very high cost of error — doctors, for instance — and they pull the averages up. That is a feature of how the occupational list is composed, not a refutation of the principle, but it is worth knowing about.

What moves the boundaries

• Company age and the cognitive domain. The younger the company, the more mental efficiency and learnability matter: there are no processes and no mentors yet, so thinking and learning fall to the individual. At mature stages these demands ease for routine roles — by then the process does much of the thinking.
• Technical content of the role. The more technical competence a post requires, the higher the demands on the cognitive domain and the lower the mandatory floor for teamwork: such a role deals chiefly with equipment and data rather than with people. The upper boundary for teamwork is not lowered — a personable engineer is no worse, it simply is not a selection criterion.
• Seniority. The higher the position, the more progress monitoring and conscientiousness matter: a manager answers for other people's results, and their inattention scales across the whole structure.
• Innovation intensity and company youth. The closer a company sits to the innovative edge, and the younger it is, the higher the demands on openness across every role without exception: the environment changes weekly. At later stages the bar for openness falls — but not to zero.
• Cost of error ignores stages. The first damper: where a mistake is expensive, demands on conscientiousness do not soften at any stage.
• Openness against stagnation. The second damper: for roles that live on novelty, the lower bound on openness is held even at late stages.

Dampers: why the stage does not get the last word

If the lifecycle stage were allowed to move requirements without constraint, some cells of the library would produce dangerous advice. A young company lives in chaos, and it follows formally that discipline matters less there. For a doctor, a pilot, an operator of critical equipment or a chief accountant that conclusion is inadmissible. The chaos of a young company does not excuse a mistake that costs health, money or a licence. So for roles with a high cost of error, demands on conscientiousness do not fall at any stage. On the same logic, progress monitoring is held up for legal roles: the price of a detail missed in a document, a risk or a deadline does not depend on how old the company is.

The second damper works in the opposite direction. At late stages — aristocracy, early bureaucracy, bureaucracy — the environment already selects for the agreeable and the predictable, and formal logic would suggest lowering the demand for openness. For roles that live on novelty — engineering, creative work, R&D — the lower bound on openness is held even there. A company that fills such roles with closed-minded people accelerates its own ageing, and the library should not help it do so.

A third constraint is of a different kind. We deliberately do not use the Agreeableness scale as a filter when selecting managers. Agreeableness as measured by questionnaires is the weakest of the five traits as a predictor of leadership, and the link does not replicate consistently across samples (Judge, Bono, Ilies & Gerhardt, 2002; ρ = 0.08). Our parameter shares that name, which is exactly why the caveat has to be stated: our value is read from observed behaviour rather than from a self-report, and the two constructs have not been shown to be equivalent, so the coefficient does not transfer. But it justifies caution: the scale describes a manager's style, not their suitability.

What the library cannot do

This is the most important part of the page, and we would rather you read it before you start making decisions on the back of the numbers.

The central limitation is simple: a range describes the typical requirements of an occupation, not the requirements of your company. It is built from official role descriptions and a stage model, not from observation of your team. If you have a non-standard role shaped around a particular person, an unusual distribution of authority, or a strong culture that changes what the job actually is, the library will give you a bearing but not a finished answer.

The ranges are a reference point, not a verdict: a gap on one or two parameters is something to explore at interview, not grounds for rejection.

Limitations we state up front

• The library gives typical bearings; it does not replace a full interpretation of an individual NeuroFrame report.
• The matrix is populated for every occupation-by-stage combination, including ones that are rare in practice — a reactor unit operator in a company at the Infancy stage, for example. Such cells are derived from the model's general logic rather than from observation, and serve as a bearing for atypical business configurations.
• A range is a reference point, not a filter. A gap on one or two parameters means the person may need support in that post, not that they should not be hired.
• Non-standard roles and distinctive cultures call for calibration of the ranges against the client's own high performers. That is separate work, and we say so openly rather than hiding behind the word "universal".

Why the operational key stays closed

We disclose the evidentiary logic of the method: what is being measured, through which behaviour it is captured, and how the model of evidence is constructed. A client's methodologists and decision-makers receive that in full under a confidentiality agreement.

What stays closed is the operational key — the scoring key and the reference walkthrough. This is not information withheld for the sake of mystique; it is the condition under which the method continues to measure what it is meant to measure.

The reason lies in the nature of implicit assessment. We observe behaviour rather than self-report, which is precisely why the system is close to impossible to game. But the moment the key becomes known to those being assessed, an implicit metric turns into an explicit one, open to deliberate optimisation: the person stops solving the problem and starts reproducing the model answer. Predictive value falls, and the normative base is devalued — not for one participant, but for every subsequent sample. This is the classic coaching effect, and it cannot be undone.

Transparency towards the client and security of the key against test-takers are two different requirements, and they do not conflict. A client is entitled to understand the basis on which decisions about its people are made, down to the methodological core. A test-taker should not be handed instructions for displaying a result they do not have. Both requirements are met at once.

This is not a private position of ours. Securing scoring keys is a standard requirement of professional assessment practice: the Standards for Educational and Psychological Testing (AERA/APA/NCME) and the ITC Guidelines on Security of Tests, Examinations and Other Assessments.

For the same reason, our public materials give neither per-facet values for individual occupations nor the size of the stage adjustments: a handful of examples is enough to reconstruct the whole lattice. The pipeline can and should be described — that is what this page is for. The numbers inside it cannot.