True North: A Confidence-Scored Translation of the Big Five Model for Type-Based Personality Reporting

Type-based personality instruments modeled on the Myers-Briggs Type Indicator (MBTI) remain widely used in coaching, hiring, and team-building contexts despite well-documented test-retest reliability problems. Independent psychometric review estimates that 35% to 50% or more of respondents receive a different four-letter type when retested within a few weeks (Pittenger, 2005), a pattern predicted by the statistical cost of dichotomizing continuous trait scores at an arbitrary midpoint (Cohen, 1983). This report describes the design of True North, a self-report personality tool that measures the empirically-derived Big Five domains using two public-domain instruments — the IPIP-50 (Goldberg, 1992) and the IPIP-NEO-120 (Johnson, 2014) — and translates continuous domain scores into a familiar four-letter label while explicitly flagging any domain within a defined indeterminate zone (45%–55%) as unstable rather than resolving it silently. We describe the scoring architecture, the domain-to-letter correspondence used, its documented limitations, and a proposed test-retest validation design for future work. This document is a design and methodology report; it does not present new empirical validation data for the True North instrument itself. Live implementation: https://lochness-paris.com/true-north/Companion article: https://lochness-paris.com/your-mbti-result-has-a-coin-flip-problem-heres-the-test-we-built-instead/