How the assessment reaches a conclusion
Five stages, in order. Nothing reaches a ranking without passing the first three.
| # | Stage | What happens |
|---|---|---|
| 1 | Data-quality gate | Infrequency items, inconsistency pairs, identical-answer runs, response time, missing data and an impression-management scale produce a verdict. Not interpretable stops the engine. |
| 2 | Constraint filter | Budget, acceptable systems, fee status and language remove routes the family cannot access. Removed pathways are listed with the reason, so the cost of each constraint is visible. |
| 3 | Knockout screen | Two tiers. On a pathway's critical work-conditions dimensions, a gap of more than 2 points (on 1–5) resting on at least three pieces of evidence demotes it to contra-indicated, however well it scores elsewhere. A gap of 1.5–2, thinner evidence, or situational responses that contradict the workday choices produce a caution: the pathway stays ranked, its interval widens, and it goes to the top of the testing list. |
| 4 | Compensatory fit | Fit = Σ w·q·sim ÷ Σ w·q. A module with weak evidence loses influence rather than dragging the score down. |
| 5 | Confidence interval | Widened by lower evidence quality, disagreement between aptitude sources, uncoded human modules, high impression management, missing grades and the younger age band. |
Weights
| Module | Ages 16–18 | Ages 14–15 |
|---|---|---|
| B · Interests | 0.18 | 0.12 |
| C · Personality facets | 0.08 | 0.08 |
| D · Work conditions | 0.24 | 0.26 |
| E · Revealed behaviour | 0.22 | 0.26 |
| F · Aptitude | 0.18 | 0.16 |
| G · Motivation | 0.1 | 0.12 |
Interests carry 0.18 rather than being the whole instrument because interest–congruence relationships with satisfaction are real but modest in the literature. Revealed behaviour carries 0.22 because what a person has voluntarily done predicts better than what they say they enjoy. Personality sits at 0.08 because it is largely expressed through the work-conditions module and would otherwise be double-counted.
The twelve dimensions
| Dimension | Low | High | Source |
|---|---|---|---|
| Interruption tolerance | Needs long unbroken stretches | Energised by being pulled into others' problems | Bexley — authored |
| Autonomy requirement | Comfortable with clear direction | Needs wide latitude over what and how | O*NET Work Context 4.C.3.a.4 · 4.C.3.b.8 |
| Deferred-reward tolerance | Needs visible results quickly | Can work for years before seeing a result | Bexley — authored |
| Ambiguity tolerance | Prefers defined problems | Comfortable with no right answer and no spec | O*NET Work Context 4.C.3.b.8 (partial) |
| Social load | Prefers mostly solitary work | Wants people contact most of the day | O*NET Work Context 4.C.1.a.4 · 4.C.1.a.2.l · 4.C.2.a.3 · 4.C.1.b.1.e |
| Emotional labour | Prefers problems to feelings | Steady with upset, frightened or angry people | O*NET Work Context 4.C.1.d.1 · 4.C.1.d.2 |
| Procedural load | Resists protocol and paperwork | Comfortable with procedure, documentation, exactness | O*NET Work Context 4.C.3.b.4 · 4.C.3.b.7 · 4.C.3.b.2 |
| Hierarchy tolerance | Needs a flat structure | Comfortable in a long, steep apprenticeship | Bexley — authored |
| Pace & deadline pressure | Needs a steady, self-set rhythm | Sharpened by constant deadlines | O*NET Work Context 4.C.3.d.1 · 4.C.3.d.8 · 4.C.3.d.4 |
| Visibility & accountability | Prefers shared, behind-the-scenes responsibility | Comfortable being the named person | O*NET Work Context 4.C.3.a.1 · 4.C.3.a.2.a · 4.C.3.a.2.b · 4.C.1.c.1 |
| Physicality | Desk and screen | On your feet, hands-on, on site | O*NET Work Context 4.C.2.d.1 · 4.C.2.b · 4.C.2.c |
| Depth versus variety | Many different problems | One problem, for years | O*NET Work Context 4.C.3.b.7 · 4.C.3.c.1 |
36 career pathways
Grouped by family. Each is mapped to an O*NET occupation and a Holland code.
New & emerging: Social media & the creator economy · Game design & development · Applied AI & automation · Product management · Sustainability & ESG · Sport, e-sports & events business · Biomedical engineering & health technology · Fintech & digital finance
Technology: Software engineering · Data science & AI · Cybersecurity
Business: Consulting & strategy · Entrepreneurship & general management · Marketing, brand & digital media
Finance: Investment banking & corporate finance · Accounting & audit
Media: Journalism, writing & communications · Film, media & performing-arts production
Design: Architecture & urban design · Design — UX, product & graphic
Health: Clinical medicine · Nursing & advanced practice · Dentistry · Physiotherapy & sport science · Psychology & counselling · Public health & epidemiology · Veterinary medicine
Science: Biomedical & life-science research · Environmental & climate science
Engineering: Mechanical & aerospace engineering · Civil & environmental engineering · Electrical & electronic engineering
Law & policy: Law · Government, policy & international affairs · Economics
Education: Education & teaching
What this instrument is not
This is a structured decision aid built on established frameworks in vocational and personality psychology. It organises evidence a family already has, and makes the reasoning behind a recommendation visible and challengeable.
- It is not a validated psychometric instrument. It is assembled from validated frameworks; it has not itself been independently validated or normed against a representative population. All scoring is rationally derived until the practice's own norming programme reaches sufficient volume.
- It is not a measure of intelligence, and it makes no prediction of success in any occupation.
- It is not diagnostic. Nothing here identifies a clinical or learning condition.
- Interest inventories have modest predictive validity — real, replicated, and weaker than this industry usually admits. That is why interests are one input among six rather than the whole instrument.
- Occupational profiles describe averages. Software engineering at a two-person start-up and at a bank are not the same job. Treat every profile as a starting point for a conversation, not a description of a specific employer.
- It does not replace judgement — the student's, the family's or the counsellor's.