Glossary

GBA (game-based assessment) evaluates a person through a game simulation: they solve tasks while the system reads their behaviour, instead of asking them to describe themselves. A conventional questionnaire relies on self-report, which is easy to tailor to whatever an employer appears to want. A game records the actions themselves — what someone did, where they paused, what they went back and redid.

See also: Stealth assessment, Digital trace, Profile

A digital trace is the full set of recorded actions a person takes during a game session: clicks, pauses, changes of decision, checks of the instructions. It is the raw material of the assessment — conclusions rest on what someone did, not on what they said about themselves. A single action means nothing; meaning appears in recurring combinations across the session.

See also: GBA (Game-Based Assessment), Stealth assessment, Profile

A profile is a set of eight values describing how a person thinks and behaves at work. Three parameters belong to the cognitive domain, five to the personality domain. A profile describes working behaviour and nothing beyond it: it is not a portrait of the person in general, not a measure of intelligence, and not a judgement about who they are outside their tasks.

See also: Target profile (role range), Percentile, Digital trace

The reference base is the sample of profiles against which an individual result is compared. At NeuroFrame it comprises more than 10,000 executives and entrepreneurs from over 500 companies. Without such a base a percentile means nothing: "higher than 70%" only answers the question once you know higher than whom. The closer the base is to a person's real working environment, the more meaningful the comparison.

See also: Percentile, Calibration, Profile

A percentile places a person relative to a comparison sample rather than counting points scored. A value of 70 reads as: this result is higher than that of 70% of people in the reference group. It is not "70 out of 100" and not a school-style mark. The scale is always tied to whom the person is compared with — change the reference group and the number changes.

See also: Reference base, Range (min–max), Profile

A target profile is the desired range of values on each of the eight parameters for a specific role at a specific company stage. It is not a portrait of an ideal person but the zone in which the role is usually performed steadily. The same post at the go-go stage and at the stability stage calls for noticeably different ranges.

See also: Range (min–max), Organisational lifecycle, Profile

A range is the lower and upper boundary of desired values on a parameter for a given role. Below the minimum, the risk is that the person will need support to meet the demands of the post. Above the maximum, the risk is mismatch in the other direction: the work does not provide the load they are used to, which over time turns into dissatisfaction.

See also: Target profile (role range), Gap (deviation), Overqualification

A gap is the difference between an actual parameter value and the nearest boundary of the desired range. If the value falls inside the range, the gap is zero. A gap is a prompt for a conversation, not a verdict: a deviation on one or two parameters is worth exploring at interview and understanding in context, rather than treating as grounds for refusal.

See also: Range (min–max), Target profile (role range), Percentile

Calibration is the adjustment of role ranges to a particular company, based on the profiles of its own effective employees. The general library describes typical demands of a profession; calibration accounts for what makes this company different — its market, its culture, the way its teams are built. It requires enough profiles for the result to be statistically stable.

See also: Target profile (role range), Reference base, Range (min–max)

Mental efficiency is the quality and speed of thinking under load: holding the terms of a task in mind, switching between them, and deciding when time and information are short. It belongs to the cognitive domain. Higher values suit roles with a dense flow of decisions; lower values indicate that a person works better at a calmer pace and with fewer parallel tasks.

See also: Progress Monitoring, Learning Agility, Profile

Learning agility is the speed with which a person draws a lesson from new experience and changes how they act. It is measured not by the volume of knowledge held but by how quickly behaviour shifts after feedback. High values matter where the rules change often; with low values, learning takes considerable effort and calls for more time, repetition and mentoring.

See also: Mental Efficiency, Openness to New, Progress Monitoring

Progress monitoring is the ability to notice in good time that things are going off course and to correct them. The person checks interim results against the goal and does not carry an unworkable approach through to the end out of stubbornness. With low values, deviations are noticed late, and regular check-in points with a manager are useful.

See also: Mental Efficiency, Result Focus, Conscientiousness

Openness to New is curiosity combined with a readiness to put new things into practice. The second half matters most: interest without application stays a conversation. High values help where the approach has to change and unfamiliar ground has to be covered; with low values a person prefers proven methods, so change is best introduced in small steps with the reasoning explained.

See also: Learning Agility, Risk Appetite, Conscientiousness

Risk appetite is the readiness to act under uncertainty without waiting for the full picture. It rests on the balance between sensitivity to reward and sensitivity to possible loss, as described by Reinforcement Sensitivity Theory. High values are needed where the cost of delay exceeds the cost of error; low values mean caution and a need for fuller data before deciding.

See also: RST (Reinforcement Sensitivity Theory), Openness to New, Result Focus

Result focus is the drive to see something through when the quick result fails to arrive. High values are valuable in long projects with delayed returns. A very high result focus paired with weak progress monitoring carries the risk of holding on to an approach that is not working; low values mean a person needs short cycles and visible interim results.

See also: Progress Monitoring, Conscientiousness, Mental Efficiency

Agreeableness is steadiness in social situations together with a willingness to help colleagues. It shows how comfortably a person works through other people rather than alone. High values matter in roles where the result is assembled out of agreements; with low values a person is more distant in communication and task-focused — a working style, not a shortcoming. The name is shared with the Big Five trait, but this parameter is read from observed behaviour rather than from a self-report questionnaire, so validity figures published for questionnaire-measured agreeableness describe a different construct and do not carry over.

See also: Conscientiousness, Openness to New, Profile

Conscientiousness is attention to detail, method and dependability: the person finishes what they start and does what was agreed. It is one of the five Big Five factors and one of the steadiest predictors of work performance in occupational psychology. With low values an employee needs clear instructions and regular support and review.

See also: Big Five, Result Focus, Progress Monitoring

The Big Five (Five-Factor Model, FFM) is an empirical model of personality structure built on five factors, developed since the 1930s and now the mainstream account in psychology. The factors were not derived from theory but from analysis of the language people use to describe one another, and have been replicated across dozens of cultures. Five NeuroFrame parameters rest on this model.

See also: CB5T (Cybernetic Big Five Theory), PEN model, Conscientiousness

RST (Reinforcement Sensitivity Theory) is Jeffrey Gray's neuropsychological model. It describes the balance between an approach system (BAS, sensitivity to reward) and an inhibition system (BIS, sensitivity to punishment and uncertainty). At NeuroFrame it underpins the risk appetite parameter: caution and decisiveness are treated as different balances of these systems, not as a virtue and a fault.

See also: Risk Appetite, Big Five, PEN model

PEN (Psychoticism — Extraversion — Neuroticism) is Hans Eysenck's model of temperament, among the oldest biologically grounded accounts of personality. Eysenck tied its three dimensions to differences in how the nervous system works, not to observable behaviour alone. Historically it precedes the Big Five and was partly absorbed into it.

See also: Big Five, RST (Reinforcement Sensitivity Theory), CB5T (Cybernetic Big Five Theory)

CB5T (Cybernetic Big Five Theory) treats personality traits as parameters of a behavioural control system: how a person sets goals, judges the distance still to cover, and adjusts their actions. It explains why stable traits show up in particular acts at all, which makes it a natural fit for assessment built on observed behaviour.

See also: Big Five, ECD (Evidence-Centered Design), Progress Monitoring

ECD (Evidence-Centered Design) is an engineering methodology describing how a specific observable action becomes evidence of an unobservable psychological construct. It answers not "what are we measuring" but "how are we measuring it and why does this count as evidence". It sets out an explicit chain: construct, observable behaviour, rule of inference.

See also: Stealth assessment, Digital trace, Construct validity

Stealth assessment is measurement embedded invisibly in gameplay and built on ECD principles (Shute, 2011). The player sees no questions and cannot tell which action speaks to what, so tailoring responses to expectations is far harder than with a questionnaire. Taking part remains voluntary and informed: the method is unobtrusive, not covert.

See also: GBA (Game-Based Assessment), ECD (Evidence-Centered Design), Digital trace

Validity is the property of an instrument measuring what it claims to measure. The question is not whether the instrument counts accurately but whether it counts the right thing. It is not a single number but a body of evidence: internal structure, links with external indicators, capacity to predict future performance. Hence validity is discussed by type — construct, criterion, predictive.

See also: Construct validity, Criterion validity, Predictive validity

Construct validity is the correspondence between an instrument's internal structure and the theoretical model it implements. It is tested by factor analysis: if the model posits two domains — cognition and personality — the data should decompose that way. Fit indices such as CFI and RMSEA show how well the observed data sit on the declared structure.

See also: CFI (Comparative Fit Index), RMSEA, Validity

Criterion validity is the link between an instrument's results and external indicators: other established tests, manager ratings, actual work outcomes. It answers whether the measurement agrees with what can be seen outside the procedure itself. It is expressed as a correlation coefficient r, whose magnitude must always be read alongside a description of the sample and the criterion used.

See also: r (validity coefficient), Predictive validity, Validity

Predictive validity is an instrument's capacity to forecast future effectiveness rather than describe what is already known. It is tested in sequence: measure first, then, after time has passed, compare against real work outcomes. It is the most demanding form of validity and the most valuable for hiring. It always concerns probability across a group, never a guarantee for one person.

See also: Criterion validity, R² (coefficient of determination), AUC (ROC AUC)

CFI (Comparative Fit Index) measures how well a statistical model fits the data in factor analysis. It shows how much better the declared structure explains the data than a model with no relationships at all. Values above 0.95 are conventionally regarded as good. It is read together with RMSEA: one index alone does not establish that a structure is correct.

See also: Construct validity, RMSEA, Cronbach's alpha (α)

RMSEA (Root Mean Square Error of Approximation) is a fit index that estimates the size of the approximation error in a model. Unlike CFI, lower is better: below 0.08 is acceptable and below 0.05 good. It takes model complexity into account, so it penalises extra parameters added merely to improve the fit.

See also: CFI (Comparative Fit Index), Construct validity, Validity

Cronbach's alpha is a measure of a scale's internal consistency: whether its individual items measure the same thing. The value rises as the items agree with one another. An alpha that is too high is also a signal — the items duplicate each other and the scale covers a narrower construct than intended. It is read alongside test–retest reliability.

See also: Test–retest reliability, Construct validity, CFI (Comparative Fit Index)

Test–retest reliability is the stability of a result when the procedure is repeated after an interval. High values mean the instrument captures stable traits rather than a mood on the day. For NeuroFrame the figure exceeds 0.83. Perfect agreement neither occurs nor should: people change, and every measurement carries a margin of error.

See also: Cronbach's alpha (α), Validity, Profile

r is the correlation coefficient showing the strength of the link between an instrument's result and an external criterion, on a scale from 0 to 1. The closer to one, the tighter the link. Coefficients from different studies cannot be set side by side directly: they depend on the sample, the criterion and the corrections applied, so any comparison needs qualification.

See also: R² (coefficient of determination), Criterion validity, Predictive validity

R² is the share of variation in work outcomes that an instrument accounts for; it equals r squared. If r is 0.5, then R² is 0.25 — a quarter of the variation is explained and three quarters depend on experience, tasks, the manager and circumstances. R² values from different studies are not directly comparable and must be read with their sample and criterion.

See also: r (validity coefficient), Predictive validity, AUC (ROC AUC)

AUC (ROC AUC) is a measure of classification accuracy: how confidently a model separates high performers from the rest. A value of 0.5 means guessing at random and 1.0 means flawless separation. Unlike r, it describes the quality of separation between groups rather than the strength of a relationship, so it is used where the decision is binary.

See also: Predictive validity, R² (coefficient of determination), r (validity coefficient)

The organisational lifecycle is Ichak Adizes's model: a company passes through a predictable sequence of stages, each with its own culture, its own tasks and its own demands on people. The practical consequence for assessment is that the same post calls for different ranges at different stages — strong leadership at go-go and in aristocracy means two different sets of requirements.

See also: Target profile (role range), Go-go, Prime

Infancy is the first stage of the Adizes lifecycle: the company is searching for a working business model and trying to survive. There are almost no processes, roles are blurred, and decisions are taken fast and revised often. What is asked of people is a readiness to act under uncertainty and to do work that falls outside any formal remit.

See also: Organisational lifecycle, Go-go, Risk Appetite

Go-go is the stage of rapid growth, where the task is to grow without losing control. There are more opportunities than hands; the company takes on many things at once and risks spreading itself thin. Speed, initiative and the ability to hold several directions at the same time are what count, while regular process still yields to pace.

See also: Organisational lifecycle, Infancy, Adolescence

Adolescence is the stage at which a company builds a system while trying not to kill its entrepreneurial spirit. Rules, division of function and professional management appear, and with them friction between the founders and the new managerial logic. People need the ability to negotiate and to build process without devaluing how the company worked before.

See also: Organisational lifecycle, Go-go, Prime

Prime is the stage of balance: the company is flexible and controlled at once, growing yet predictable. Processes exist but do not yet get in the way, and initiative no longer breaks the system. Holding this state is harder than reaching it — prime is precisely where the temptation arises to protect what has been achieved instead of trying something new.

See also: Organisational lifecycle, Adolescence, Stability

Stability is the first stage of ageing in the Adizes model: the company holds its quality but the pace of change slows. On paper everything is well — revenue is even, processes run smoothly — yet new initiatives become rarer, and "why change it" is heard more often than "what could we try". The task of the stage is to keep quality without beginning to age.

See also: Organisational lifecycle, Prime, Aristocracy

Aristocracy is the stage at which a company retains control and cash flow, but the focus shifts from results to form. Procedure, status and internal sign-off gain weight, and risk is read as a threat rather than an opportunity. What is valued is precision, dependability and the ability to work within established rules.

See also: Organisational lifecycle, Stability, Early bureaucracy

Early bureaucracy is the stage where the task is to avoid falling apart and to hold on to key functions and key people. Internal conflict and the search for someone to blame consume more energy than work with the market. Managers need the ability to steady a team and keep people able to reach agreement in an environment that discourages it.

See also: Organisational lifecycle, Aristocracy, Bureaucracy

Bureaucracy is the stage at which a company lives on the stability and dependability of its procedures. Work is organised around the rulebook: what matters is that everything is done by the rules and on time. Conscientiousness, precision and predictability are valued, while the room for initiative is narrow — worth bearing in mind when matching people to such roles.

See also: Organisational lifecycle, Early bureaucracy, Conscientiousness

A talent pool is a group of employees prepared to step into key positions. It is normally assembled in advance so that a manager's departure does not stall a unit's work. A profile helps to see what in a pool candidate already matches the demands of the future role and what is worth developing before appointment, so preparation has a focus.

See also: Succession planning, Target profile (role range), Gap (deviation)

Succession planning is the systematic preparation of replacements for critical posts before those posts fall vacant. It answers who would carry the work on if a key person left, and what has to happen today for such a person to be ready. It differs from a talent pool in being tied to named positions rather than to a general group.

See also: Talent pool, Target profile (role range), Organisational lifecycle

An assessment centre is a multi-day evaluation procedure with trained observers, business simulations and exercises — a recognised standard for in-depth assessment of managers. It is resource-intensive: participants' time, expert work and organisation. NeuroFrame addresses part of the same task in 30–60 minutes online, without replacing an assessment centre where live observation and discussion are what is needed.

See also: GBA (Game-Based Assessment), Profile, Predictive validity

Overqualification is a situation where a person's values sit substantially above the demands of the role. It is not "too good" but a mismatch: the work does not provide the load they are used to, interest fades, and the risk of boredom, dissatisfaction and departure grows. The practical conclusion is not rejection but a conversation about what this role can offer them.

See also: Range (min–max), Gap (deviation), Target profile (role range)

Human capital risk is the risk that the people in an acquired company cannot deliver what the deal model assumes they will deliver — either because their capabilities do not match what their roles require, or because the integration context prevents them from performing. It is one of the most documented causes of lost deal value, and the one for which standard valuation practice holds no line.

See also: NDC (NeuroDelta Coefficient), Human delta, Friction, Human capital due diligence

The NeuroDelta Coefficient expresses human capital risk as a multiplier centred on 1.0, in the convention of beta: NDC = 1 − human delta ÷ 50. Above 1.0 the deal carries priced human risk; below 1.0, a documented premium. It enters valuation through the channel already built for deal-specific risk — the company-specific risk premium in a build-up valuation, or a direct WACC adjustment — where (NDC − 1.0) multiplied by the equity risk premium gives the human capital risk premium in basis points.

See also: Human delta, Deal score, Company-specific risk premium (CSRP)

The human delta is the distance between the deal score and 50, the market-average team. Above the line the people add value; below it they cost money, and one point is worth roughly one per cent of the price. It is the same result as the NeuroDelta Coefficient, stated in points rather than as a multiplier — points for the discussion, the coefficient for the valuation model.

See also: NDC (NeuroDelta Coefficient), Deal score, Team score

The team score is one number describing how well the people in a company match what their roles require at that company’s stage of corporate life. Individual fits are blended in proportion to how much each seat can move the outcome. The scale is set so that 50 is the market-average team; a team fully inside the required bands on every parameter lands around 60. It says nothing about a transaction on its own — friction is subtracted separately.

See also: Deal score, Friction, Range (min–max)

The deal score is the team score minus friction: what the people are capable of, adjusted for what this particular transaction does to them. Two deals for the same target can have the same team score and very different deal scores, because integration depth, pace and governance are properties of the deal, not of the people.

See also: Team score, Friction, Human delta

Friction is the part of the deal score that comes from the transaction rather than from the people: culture distance, integration depth, governance and pace change, the growth engine after close, the integration timeline and key-person retention. A capable team can still be rejected by the receiving organisation — that is what friction measures. Past a certain depth the honest conclusion is not that the team is risky but that the deal is badly designed.

See also: Culture distance, Integration depth, Key-person retention, Deal score

Culture distance is how far apart the operating logic of the buyer and the target are — what one company rewards and the other punishes. It is the largest single item of friction, because it is one of the most thoroughly documented causes of lost deal value: cultural misalignment delays, reprices or terminates a substantial share of transactions. Distance is not a verdict on either culture; it is a cost to be priced or designed around.

See also: Friction, Integration depth

Integration depth is how much of the target changes after close: from staying standalone with its own systems and rhythm, through shared reporting, to full absorption where everything changes at once. Depth costs friction directly, and it also raises the weight of openness to new, agreeableness and learning agility in the team score — the deeper the integration, the more the outcome depends on capabilities of adaptation rather than of execution.

See also: Friction, Culture distance

Key-person retention is what holds the people the buyer is paying for: multi-year earnouts, retention pools and enforceable non-competes on critical roles. It is the one friction condition that works in the buyer’s favour — it subtracts risk rather than adding it. With nothing in place the general acquisition base rate applies: roughly a third of acquired employees leave within the first year, against 12 per cent of comparable regular hires (Kim, MIT Sloan, 2019).

See also: Friction, Human capital due diligence

The company-specific risk premium is the component of a build-up discount rate that carries risks unique to one company rather than to its market or its size. It is where country risk, size and illiquidity already enter a valuation — and therefore the natural channel for a measured human capital risk premium, computed as (NDC − 1.0) multiplied by the equity risk premium and expressed in basis points.

See also: NDC (NeuroDelta Coefficient), Human capital risk

Human capital due diligence is the part of pre-close diligence that examines whether the people in the target can deliver what the deal model assumes. Unlike financial and legal diligence, which verify what has already happened, it is the only strand that makes a claim about the future — which is exactly why it needs a stated method, a stated sample and stated limits rather than a set of interview impressions.

See also: Human capital risk, NDC (NeuroDelta Coefficient), Key-person retention