---
title: "OSCE glossary"
description: "Plain definitions of the OSCE terms an exam office uses: circuits, blueprints, cut scores, borderline methods, stringency, and the rest."
canonical: https://studydrome.com/docs/osce/toolkit/glossary/
updated: 2026-10-10
---

# OSCE glossary

Plain definitions of the OSCE terms an exam office uses: circuits, blueprints, cut scores, borderline methods, stringency, and the rest.

Twenty-two terms are defined the way an exam office uses them. These are the operational words: the ones that appear in a run sheet, a set of regulations, or a board minute. Each is defined in one or two sentences here and explained in full on the page that owns it.

Two groups of terms sit elsewhere on purpose. The measurement words, including validity, reliability, content specificity, and the standard error of measurement, are defined under [validity and reliability](/docs/osce/foundations/validity-and-reliability/) because they need a paragraph each rather than a line. The station statistics are explained in [reading station metrics](/docs/osce/quality/reading-station-metrics/).

## The terms

- **OSCE** — An objective structured clinical examination. Candidates rotate through timed stations. Each station assesses one task, on the same marking sheet for everyone.
- **Station** — One timed task with its own brief, examiner, marking sheet, and kit. It is the unit an OSCE is built from and scored on.
- **Circuit** — One full loop of stations that a group rotates through. A large cohort runs several circuits, side by side or one after another.
- **Rest station** — A timed slot with no marking. It balances a circuit when the number of candidates does not match the number of stations and gives candidates a break.
- **Blueprint** — The grid that maps stations to the domains the exam claims to cover. It is written first. A reviewer asks for it first, too.
- **Station bank** — The stored set of stations a school can draw on, with versions and usage history. A bank makes retirement a choice, not an accident.
- **Marking sheet** — The list of items an examiner scores at a station, with the scale for each. It is what makes two examiners score alike.
- **Global rating scale** — One overall judgment of the performance, from clear fail to excellent. The borderline standard-setting methods read this, not the item scores.
- **Critical item** — An item whose failure would cause the station or the exam to fail, regardless of the total. It covers acts that would harm a patient. The rule belongs in your regulations.
- **Cut score** — The mark at or above which a candidate passes. A method produces it, a board approves it, and it is stated before results go out.
- **Standard setting** — The process that turns scores into a pass mark. It is a judgment made against a set method. It is not a statistic taken from the cohort.
- **Borderline regression** — A method that fits a line from checklist totals to global ratings. The cut score is the total the line predicts at the borderline grade.
- **Borderline group method** — A method that takes the mean or median total of candidates graded borderline. It needs enough borderline candidates to be stable.
- **Angoff method** — A method where experts judge what a just-competent candidate would score. The judgments are made before the exam runs.
- **Conjunctive rule** — A second-pass condition on top of the total, usually a minimum number of stations passed. It stops one strong station from covering a failed one.
- **Simulated patient** — A trained person who plays a patient to a script, the same way for every candidate. Also called a standardized patient.
- **Examiner stringency** — How far one examiner marks above or below the average at the same station. It is measured by comparing examiners who shared a station.
- **Sequential OSCE** — A design where everyone sits in a screening set, and only those near the cut score sit more. It puts testing time where the decision is close.
- **Run sheet** — The exam-day document says who is where and when. It holds the circuit map, the rotation times, the signal plan, the staff posts, and the numbers to call.
- **Quarantine** — Keeping candidates who have sat the exam away from those who have not. It applies to any exam run in more than one session.
- **Station facility** — The mean score for an item or station as a share of the maximum. Higher means easier. It is the first number to check when a station looks odd.
- **Discrimination** — How well an item separates stronger candidates from weaker ones. It compares the top and bottom thirds. Near zero means the item tells you nothing.

## How to use it

Two uses. The first is internal: agree these definitions once and put them in your regulations, because most OSCE arguments are about words rather than evidence. The second is external: a candidate handbook that defines the cut score and the conjunctive rule in one line each prevents a certain kind of appeal.

Where a term has a definition your regulations already give it, your version wins. The purpose here is a common starting point, not a standard.

## Frequently asked questions

### Is a checklist the same as a marking sheet?

In everyday use, yes. The term "marking sheet" is the safer term in regulations because it covers the observable items, the scale, and the global rating together, whereas "checklist" is sometimes read as the item list alone. The guide uses marking sheet when the whole document is meant.

### Are simulated and standardized patients different?

The terms are used interchangeably in most of the literature. Where a distinction is drawn, standardized emphasizes consistency across candidates for an exam, and simulated emphasizes the portrayal itself. This guide uses simulated patient throughout.

### What is the difference between a cut score and a pass mark?

None in practice. Cut score is the term standard-setting papers use, pass mark is the term regulations use. Pick one and use it consistently in every document a candidate will read.

### Where do the measurement terms live?

On [validity and reliability](/docs/osce/foundations/validity-and-reliability/), which defines validity, reliability, Cronbach's alpha, the generalizability coefficient, content specificity and the standard error of measurement. They are kept together because each needs an explanation rather than a line.

> [!NOTE]
> **Where the definitions come from**
> Every term here is used in the sense that the cited sources use it. The reading list behind them is on the sources and further reading page in this section.