Miller's Pyramid
Miller's Pyramid is a four-level framework for assessing clinical competence that moves from knowledge at the base, up through the application of knowledge, then demonstration of a skill under observation, and finally what a clinician does in real practice. George Miller named the levels knows, knows how, shows how, and does, and because each level calls for a different assessment method, the framework is a quick way to check whether a program's assessments reach the level it claims to measure.
George Miller published the framework in Academic Medicine in 1990, in a five-page supplement paper titled The assessment of clinical skills/competence/performance. He was making an argument rather than drawing a diagram: assessment in medical education had drifted toward what is convenient to test, and no volume of written examination tells you whether a physician can actually do the work.
Each level has its own methods. Knows and knows how are reachable with written and oral tests, from multiple choice questions through clinical problem-solving exercises. Shows how needs observed performance, which in practice means an OSCE or another station-based examination with a standardized patient. Does can only be reached by watching a clinician work with real patients over time, which is what workplace-based assessment and entrustment ratings are reaching for.
The framework earns its keep as a check on your own claims. A program that assesses communication with a knowledge quiz is sitting at the bottom of the pyramid while reporting on a competency near the top, and a competency committee looking only at in-training exam scores and rotation evaluations has no shows how evidence at all.
Later authors writing in Academic Medicine have proposed extending the pyramid upward, toward entrustment decisions about how much supervision a learner still needs and toward professional identity formation, meaning who the learner is becoming rather than what they can perform. The four original levels have held up well enough that most competency-based curricula still map onto them, which is part of why a 1990 supplement paper is still cited weekly.
What this looks like in a program
- Map every assessment in your curriculum to a level, then look hard at which competencies have no evidence above knows how.
- Assess communication with station-based observation rather than a written test, since shows how is the lowest level at which a conversation can be judged at all.
- Treat repeated practice as the bridge between knows how and shows how, because learners rarely make that jump from reading and lectures alone.
Related terms
Last updated August 2026