How DecipherU Match works

What it does

You guess first. Then you answer questions about your interests, the kind of work you prefer, how you work, what you value and a few reasoning problems.

We score those answers on our servers and compare them with the O*NET profile of the occupation each license leads to. The result is a relative fit label with a range, not a single number.

If your answers carry too little information, for example the same answer all the way down, many answers given very fast, or a whole group left blank, Match says so and shows no cards. Nothing is saved in that case.

If your closest match is not high, Match says that plainly and shows the full ordered list of every profile before any card. The cards stay available, but they are not forced on you.

What it cannot tell you

It cannot tell you whether you will pass an exam, get a job or like the work. Fit says how your answers compare with the O*NET occupation profile behind each license, not with answers from people who hold the license. It is a soft signal. In large studies of hiring, interest scores relate to job performance with a correlation of about .24, and the spread across studies is wide (Sackett et al., 2022). That figure is about job performance, not about how well a person fits a license.

The fit bands are relative. Strong fit means near the top of the licenses compared. It does not mean a high level of fit in absolute terms. We will set absolute levels after the pilot.

Readiness is separate. It comes only from real practice questions, never from the quiz.

Licenses and certifications

A license is a legal permit from a state board: you may not do the work without it. A certification is a credential from a professional body or a company: no state requires it by law, though an employer may. Every card and every line in the full list says which one it is, and a certification is never called a license.

Both kinds are compared with the O*NET occupation profile they lead to, in the same way. Time, cost and record-policy facts for a certification come from the body that issues it, and where we have not checked them the card says so.

Who may not use it

Match is not a hiring, admission or licensing tool. Employers, schools and agencies may not use it to screen, rank or decide about anyone. There is no employer view and no feed of results, and we do not sell or share results.

How the result is built

1. Interests are scored on six areas and compared with O*NET occupation profiles.
2. Abilities are scored from short multiple-choice items. A gap counts only where a license needs the ability. Being over-qualified is not a miss.
3. Work style, values and the kind of work you prefer add smaller parts.
4. Your state, education, time and money set flags, never exclusions. The one hard rule is a minimum age that a licensing board publishes. A record question never removes a license. It adds a note to check with the board. Where we have not checked a time, fee or record policy yet, the card says so.
5. Each license is compared with how it scores for people who answer at random, so a license that fits almost everyone does not crowd out the rest. Licenses are then ranked by fit, then by how surprising they are, then by pay, openings and the time and cost to get licensed.
6. A driver permit test, a food handler card, a food protection manager certificate and Praxis Core are entry credentials, not careers. They are compared but never get a card.

How a surprise is chosen

A surprise is a license that fits you well and that you did not name. We compare your fit with how widely known and how large the license is, and with how often it appears when only interests are used. A license you ruled out gets a boost if your answers sit close to it.

A result may also name one license your answers line up with least. It appears only when the absolute fit is low and your answers were not rushed. It does not mean you could not do the work.

How we keep it honest

Each sentence in a result comes from facts we hold. A checker drops any sentence with a number or trait claim that is not in those facts. The default text is a fixed template. A language model writes it only when we switch that on, and we record the model and prompt on every result.

We do not infer traits from how fast you answer. Speed is used only to mark a result as low confidence.

Privacy and age

No account is needed. We store your answers and result under a random id, with no name or email. Your result opens only in the browser that took Match, or in your account if you sign in. A share card is separate and shows only the license name, one line, the fit label and the date. You can delete your result from its page or with the delete link. If you lost both, email privacy@decipheru.com with the date you took Match.

People under 13 cannot use Match and nothing is stored. People aged 13 to 17 can take it, and their answers and result are never saved. We do not track people under 18. Match sets no cookie and records no event until the age question is answered and you are 18 or older.

We never use face, voice, video or emotion data.

How long it takes

Match has 79 questions, one on each screen. The first 7, about your situation, are optional. We estimate about 14 minutes from a typical time for each question, and the screen shows time left from your own pace once you have answered enough. These are estimates. We have not yet measured how long real people take, and the beta will tell us.

There is no timer anywhere. A slow answer is never marked down. Speed is used only to mark a result as low confidence when many answers are far too fast to be real.

Access

Every question is a native form control, so a keyboard works without a mouse: arrow keys move inside a group of answers, Tab leaves it, and Enter presses Continue. Tapping or clicking an answer can move on by itself, and you can turn that off. Choosing an answer with the keyboard or a screen reader never moves on by itself. Progress and time left are text as well as a bar. Motion is off when your device asks for less motion, and the layout holds at 400 percent zoom.

We check the pages with the axe accessibility engine in a real browser and by keyboard. A recorded test with a screen reader is still to do, and this page will say when it is done.

The beta and the questions we ask you

This version is a beta. After your result we offer a few optional questions: how much your top result feels like you, from 1 to 5, and which of your cards you would consider looking into. You choose to answer, nothing is required, and you can keep your result without answering.

Your answers are saved with your result under the same random id, with no name, no email and no free text. They are removed when you delete the result. We use them in aggregate to check whether results feel right and whether people act on them. They do not change your result. Results for people under 18 are never saved, so those people are not asked.

A later step would ask people who come back whether they took the exam. That is planned and not built yet.

Versions

Every result records the three versions that produced it. A change to any of them is logged here.

Instrument

match-instrument-1.0.0-beta

Engine

match-engine-1.0.0-beta

Data release

O*NET 31.0 with BLS and state board facts; see each card for the date

Data sources and licenses

• O*NET 31.0 Database, USDOL/ETA, CC BY 4.0. We modified and combined it. USDOL/ETA has not approved or endorsed these changes.
• U.S. Bureau of Labor Statistics Employment Projections and Occupational Employment and Wage Statistics, public domain. Each number on a card names its table and year.
• State licensing boards, and for certifications the body that issues them, for time, fees, minimum age and record policy. Each line carries the date we checked it.
• IPIP, public domain, for the wording style of work-habit items.

Built on

Holland's interest theory and Prediger's two axes. The Theory of Work Adjustment for values. Klimov's classification of work by object and purpose, in our own words. Lent, Brown and Hackett's career theory for the try-it step. All items are original. None is copied from any commercial or proprietary test.

Validation

Beta, not validated. The weights are placeholders, the mapping from licenses and certifications to occupations was done by one coder and has not been double-checked, and many time, fee and record-policy facts are not checked yet. This beta is also our first study.

The plan, in order. Every stage is planned unless this page says it ran, and we will publish the results, including the bad ones.

1. Cognitive interviews with 15 to 20 people on wording and reading level. Pass: no item misread by more than 2 people.
2. Pilot A with 300 or more completers, to check structure, reliability and timing, with an established interest inventory as a second measure. We use such an inventory only with its publisher's written permission. Match itself uses only items we wrote, so it needs no one's permission to run.
3. Pilot B with about 100 retesters at 2 to 3 weeks, to check that results hold steady.
4. A swap and acceptance study with 40 to 60 people: can you pick your own report over a random other one at least 75 percent of the time, and does no statement get accepted by more than 85 percent of takers.
5. A predictive study with at least 190 completers per outcome, asking whether fit relates to practice score, completion and the exam result at 6 to 12 months, beyond what prior practice score and education already tell us, and whether it works equally across groups.
6. A known-groups check with existing license and certification holders: do they land closest to their own credential?
7. Calibration before any pass percentage appears: at least 200 passes and 200 fails per exam family. Until then Match shows no pass percentage at all.

The technical manual is at decipheru.com/validity/match. It lists what has not been measured yet, in the same order as this plan.