personality.naseebedu.com — a free career and personality test for Uzbek students aged 13–18, in their own language.
Most Uzbek teenagers choose a university direction with no structured information about themselves at all. The advice they do get is a relative's opinion, or a translated Western quiz that names a job and calls it destiny. TestMind is the alternative: fifteen minutes, no account, no fee, and a result that opens options instead of closing them.
A 75–95 question test, about 14 minutes, in four parts:
| Part | What it asks | Items |
|---|---|---|
| Personality | Big Five, adaptive | 25 or 35 |
| Interests | RIASEC, O*NET-style | from a 48-item bank |
| Work values | what makes work worth doing | 10 |
| School subjects | marks — optional | 11 |
Out comes one of ten archetypes with its own artwork, an interest profile, and ranked directions drawn from 96 careers, 55 majors and 16 fields. Ten take-away PDF guides, one per archetype, are there to download.
The personality section is genuinely adaptive: everyone answers a fixed core, and only students whose 2nd and 3rd traits land within 0.5 of each other get asked the extra ten. Simulation put that design at 58.3% exact-archetype accuracy against 56.1% for a flat 35-item test — better, and shorter, for nine students in ten.
These are not style preferences. They are why the tool can be handed to a fifteen-year-old.
- Personality can never block a career. It is weighted
0.08for careers and majors, and its term is floored at 0.5 — a quiet student still sees entrepreneur; a creative one still sees software engineer. Interests lead (0.50), because what someone enjoys doing predicts occupational choice better than what they are graded on at fifteen. - No ordered Holland code is ever shown. Nobody leaves believing they are an "SIA".
- No fake precision. Bands — 🟢 strong match, 🟡 worth exploring, 🔵 alternative — never "87% suitable".
- No salaries, no demand ratings, no employment forecasts. There is no Uzbek labour-market data behind them, and a fifteen-year-old would believe them.
- The result screen ranks careers once, from four signals, in one place. The archetype block above it is alphabetical and says in words that it is not a recommendation.
- Data minimisation. No full name, birthday, address, phone, ID or parent income. Email only when the student asks for their guide.
Static files on GitHub Pages. No framework, no build step, no backend. test.html is
the entire app — ES5 only, because it has to run on old Android phones over school Wi-Fi,
and that is a hard requirement rather than an aspiration.
/ 20 pages GitHub Pages serves as URLs
/ru/ /en/ the same 16 pages, translated
/assets/ css, js, fonts, character artwork
/guides/ ten take-away PDFs
/tools/ Python generators + puppeteer suites — never served
Content lives in Python and is compiled to JS, so the result screen and the archetype pages can never disagree. The site is trilingual; the test itself is Uzbek-only on purpose — a translated item bank is a different instrument, not the same one in another language.
tools/ holds the generators and the browser suites that drive a real Chrome to a real
result screen in all three languages — including when the content pack arrives late, or
never arrives at all. See tools/README.md.
Built for NaseebEdu schools. Free for every student who takes it.