Jungian Typology (MBTI)

What it is

Four axes of preference — energy (E/I), perception (S/N), decision (T/F) and organization (J/P) — that combine into 16 types. Beneath the four-letter types lives the richer theory: the stack of cognitive functions (which process leads your mind, which supports it, which is the Achilles’ heel). It is a LANGUAGE of preferences, not a measurement: its value lies in the vocabulary it gives to what you always felt.

Where it comes from

What it does NOT measure

It isn’t a trait measure (the Big Five does that with far greater evidence), it doesn’t predict performance, and a type isn’t a box: the axes are continuous, and forcing someone in the middle into a letter is the model’s greatest weakness.

Evidence — limited evidence

The MBTI shows reasonable-to-good internal consistency (α = 0.845–0.921 aggregated across 193 studies, n = 57,170; Erford et al., 2025, Journal of Counseling & Development), but its central problem is the dichotomization of continuous dimensions into types. Test–retest studies indicate meaningful instability in classification: about 50% of respondents receive a different classification on at least one of the scales after only 5 weeks, and ~35% change their four-letter type after 4 weeks (Pittenger, 2005, Consulting Psychology Journal, citing McCarley & Carskadon, 1983, and Myers et al., 1998). McCrae & Costa (1989, Journal of Personality) showed that the instrument measures four relatively independent continuous dimensions that overlap with the Big Five (correlations around 0.7 for EI×Extraversion and SN×Openness) — not qualitatively distinct categories. The dichotomization itself entails a non-trivial loss of information (26%–32%; Harvey & Murry, 1994, Journal of Personality Assessment). Predictive validity for occupational outcomes is weak: R² ≈ 0.01 for leadership behaviors, with only 7 of 20 associations significant (Zárate-Torres & Correa, 2023, Frontiers in Psychology, n = 529). Technical reviews advise against routine use in selection and organizational contexts (Boyle, 1995, Australian Psychologist; Randall et al., 2017, who also report unsatisfactory reliability for the TF subscale = 0.61). The honest calibration is limited evidence: useful for self-exploration and reflection, inadequate for selection or development decisions that require classification stability and demonstrated predictive power.

The studies behind this

Real, verifiable references — click to check them yourself. That’s what separates a reading from a horoscope.

How it enters the crossing

Narrative tier (0.6; instrument 0.65). In the crossing, the lens reads the function stack in your episodes (it never asks "are you an introvert?") and the test result enters as an instrument — when the two agree, the type gains ballast; when they diverge, it becomes a tension.

The sixteen types