Pilot Psychometric Assessment Platform (v13d)

Multi-channel personality assessment for pilot and broader competency selection: eight main dimensions (six inspired by HEXACO, plus Resilience and Empathy) and 22 sub-dimensions, measured through forced-choice triplets (MFC), Likert items and a Situational Judgment Test, with further optional modules (ranking, adjective pairs, adjective checklist, point allocation, cognitive reflection). A single-file candidate form reads only a cleaned copy of the pool and never receives a key; access codes are locked by security rules with ownership and expiry checks and deleted after submission. A single-file admin panel builds the forced-choice blocks itself (54 by default), rescores every report from scratch and exports PDF and CSV. Content is generated from Python source files; the SJT has four cases of 18 scenarios with 288 options. Sixteen versions to date (v1 to v13d). An October 2026 scoring audit rebuilt consistency on a chance-corrected kappa, which lifted the simulated separation of honest answers from random, straight-line and fake-good ones to AUCs of 0.91, 0.95 and 0.93; the audit also added six validity indicators and a block generator balanced by sub-dimension. Next: a pilot of 100 to 200 people for norms.