A transparent method for different kinds of tests
Introsia publishes practical self-reflection tools and entertainment-first profiles. The method is not identical to a clinical instrument, aptitude examination, or academic research scale. Each product begins with a narrower question, defines what the answers can support, and states what the result must not claim.
1. Define the user question and the boundary
Before writing questions, we define who the test is for, why someone would take it, which concepts are in scope, and which promises are prohibited. Career Fit, for example, explores interests and work preferences without measuring ability. Inner Age uses a symbolic entertainment metaphor without claiming a real psychological age. Sensitive relationship and family products cannot diagnose people, infer another person’s motives, or make stay-or-leave decisions.
2. Build a mixed question set
Most current Introsia tests use 30 questions: 18 seven-point agreement items, 10 situations with exactly four mutually exclusive choices, and two written reflections. Career Fit uses 40 questions: 24 agreement items, 14 four-option tradeoffs, and two written reflections.
The agreement scale runs from strongly agree to strongly disagree. The middle position is treated as genuine uncertainty, not weak agreement. Selecting it opens a required explanation field and contributes a neutral numeric value. Four-option situations use explicit, reviewed score maps. Written reflections are separate from numeric scoring.
Sample questions on public pages are newly written illustrations. They are not the live bank and do not disclose weights, answer keys, or result thresholds.
3. Score the choices the product says it scores
Scoring is deterministic: the same completed scored answers under the same question and report version produce the same result. Different products use different rules. Some select a nearest editorial profile. Others identify a lower support area, rank interest patterns, or compare lifestyle priorities. East Asian Family Patterns separately tracks applicability and returns Mixed / Not Enough Applicable Context when there is not enough applicable input.
We do not convert private writing into hidden scores. The short analysis transition after the last question compares the scored answers and tradeoffs; it is not evidence of AI reading, psychological inference, or text interpretation.
4. Keep private writing local
Middle-answer explanations and the two written reflections stay in the browser on the device. They are not sent for analysis, do not affect the result, do not appear in the report, and are not included in analytics, advertising payloads, share links, or URLs. A person using a shared or monitored device should still consider local device privacy and clear browser data when appropriate.
No account is required. Browser storage also holds versioned assessment state so progress can resume. Question and report versions prevent one product or version from silently loading another product’s answers.
5. Create a deliberately limited free preview
Every test provides a free result preview. The exact fields depend on the product. Career Fit shows one archetype and one starting insight. Your Soul City shows one leading city/profile and one tradeoff. What Are You Mistaking for Love? shows the dominant pattern and its public core line. Other products show one primary profile or support question and a short public line. Scores, private text, secondary patterns, ranked alternatives, and paid chapters remain outside the free preview.
6. Make the paid report add real context
The optional report is a one-time $3.99 USD purchase with no subscription. Before payment, the page explains what the report adds. Depending on the product, that can include full dimension comparisons, alternatives, interaction patterns, tradeoffs, counterevidence, context checks, practical worksheets, and low-stakes experiments.
Payment buys the named digital report, not greater scientific certainty. It does not reveal private writing, turn Mixed into a hidden culture-specific result, or unlock a diagnosis. PayPal processes checkout and shows the final amount before approval. Product slug and report version are checked so one purchase cannot unlock a different report.
7. Review every result route
Before a product is approved, automated tests enumerate result types, rankings, and important combinations. They check minimum visible report length where required, repeated long paragraphs, grammar, wrong-dimension language, missing content, prohibited promises, privacy boundaries, and payment isolation. Human reviewers examine representative and edge-case reports, free results, mobile presentation, safety language, and whether a sentence could mislead despite passing a mechanical scan.
Entertainment reports may use a shorter narrative target when a 3,500-word report would create filler; the alternative standard must be defined before implementation. Long practical reports must meet their displayed-word requirement without hidden or duplicated text.
8. Separate facts, sources, and product inference
When a report uses external facts—such as historical episodes, occupational context, or changing city information—the editorial source and the product inference should not be presented as the same thing. Sources can support a fact. They do not validate an Introsia personality interpretation unless appropriate evidence actually exists.
High-stakes current information belongs with authoritative, up-to-date sources and qualified guidance. An Introsia report cannot replace medical, mental-health, legal, financial, safety, immigration, relocation, education, or credential advice.
9. Update visibly and accept correction
Questions and reports are versioned. Material changes to scoring, result meaning, paid content, or privacy behavior require a reviewed version change and compatibility rules for existing purchases. Editorial corrections can be made when a statement is unclear, unsupported, outdated, or inconsistent with product behavior.
To report an error, email support@introsia.com with the page URL, the exact sentence or behavior, and why it appears incorrect. Do not include private answers or full payment-card information. Read the Editorial Policy for the review standard and Privacy Policy for data handling details.