We checked whether BP reproduces its scoring results from 288 synthetic responses. This material explains archetype-condition and pattern-activation results. It does not contain a real-user validation study or AI report excerpts.
The synthetic-case JSON records input provenance, the internal replay check, public results and limitations. The academic-map JSON separately records reviewed links between concepts and literature. Item-level responses and internal calculations are not public.
We fixed the input and scoring code, then ran the calculation once in each of two separate processes. The complete results matched except for the execution timestamp. This is a software reproducibility check on one synthetic case, not reliability or validity testing with real users.
This example does not expose real user data, item-level responses, question text, internal axis codes, raw scores, scoring formulas, weights, thresholds or prompts. The input and complete output are retained as private reproducibility records.
No. Synthetic example 01 uses responses created by a verification tool. It does not copy a real user's answers or profile.
No. The public file describes the synthetic input's provenance and item count, but does not provide item-level responses or question text. It does not enable an external reader to rerun the internal calculation from this file alone.
Reproducing one synthetic input does not establish accurate measurement of real people's personalities. Test-retest reliability, validation against external instruments and peer review each require separate evidence.
The free MBTI assessment needs no payment. You get a type story and function strength cards.