Miller's pyramid provides a useful classification to understand the purposes of different types of assessment. This classification classifies all assessment methods into four main categories or levels: knows, knows how, shows how and does.

Knows: This level contains assessments that test pure facts; i.e. factual knowledge. So, assessments at this level cannot assess anything other than pure recall of knowledge. Simple multiple choice questions (MCQs), short answer questions and long essays are examples of the assessment methods that assess at this level.

Knows how: Assessment at this level involves assessing application of knowledge; i.e. whether the candidate can use the knowledge that they have learned to solve problems or to address practical issues. Any scenario-based written assessments (e.g. MCQ, short answer questions) that request the candidate to apply what they know to the scenario within the question are examples of assessments that satisfy this level.

Shows how: Not only the candidate should be able to apply the knowledge that they have learned to hypothetical scenarios, but they should also have the knowledge, skills and attitudes to carry out the necessary interventions to address the issue/problem within this scenario. Assessments that assess candidate skills under simulated conditions fall into this category/level; e.g. Objective Structured Clinical Examination (OSCE), long case, short case.

Does: At this level of assessment, the candidates need to demonstrate that the knowledge, skills and attitudes that they have can be transferred to real-life workplace situations. Hence, assessments at this level are called workplace-based assessments; e.g. Mini Clinical Evaluation Exercise (mini-CEX), Direct Observations of Procedural Skills (DOPS), Case-based Discussion (CBD), Multi Source Feedback (MSF), portfolio assessment.