A summative OSCE decides whether a candidate passes. A formative OSCE tells a candidate what to work on. The same circuit can serve either purpose, but not both at once. Feedback given during a summative exam changes the performance it was meant to measure. This page sets out what changes when the purpose is learning. It covers how mock OSCEs fit and what the evidence says about their effect.
Why the purpose has to be decided first
Purpose drives every other design choice. A summative OSCE needs enough stations to make a reliable pass-or-fail decision. It needs examiners trained to score and not to teach. It needs a defensible standard and an appeals route. A formative OSCE needs stations that expose common weak spots. It needs examiners trained to give feedback. It needs a follow-up so the feedback is used.
A program that builds one and uses it as the other gets the worst of both. A summative exam with feedback in the station is neither reliable nor fair. A formative exam scored as if it were summative teaches candidates to perform rather than to learn.
The AMEE guide on OSCE organization treats the split as part of examiner training. Its learning outcomes for examiners include written feedback for every exam. Verbal feedback appears for formative exams only. That is the line. Verbal feedback belongs in the formative station, and only there.
What is the difference between a formative and a summative OSCE?
Feature | Formative OSCE | Summative OSCE |
|---|---|---|
Purpose | Show the candidate and the program where the gaps are | Decide whether the candidate meets the standard |
Stakes | Low; no progression decision rests on it | High; progression, graduation, or licensing |
Station count | Fewer, chosen to sample common weak spots | Enough for reliability, set by the blueprint and the evidence |
Examiner role | Observe, score, then give feedback | Observe and score only |
Feedback | In the station, or right after the circuit | After results, in writing, per station and overall |
Standard setting | Optional; a threshold for guidance | Required; a defensible method chosen in advance |
Candidate preparation | Part of the teaching | Published handbook, blueprint and mock stations |
Analysis | Which skills the cohort is weak on | Reliability, station metrics, examiner comparison |
What changes when the OSCE is formative?
Four things change. Each one is a design decision, not a relaxation.
The stations change. A formative circuit samples the skills the program most wants to build. It adds the ones the last summative exam showed the cohort was weak on. It can afford a station most candidates will struggle with. The point is to find the struggle.
The examiner's job changes. After scoring, the examiner tells the candidate what went well and what to change. Doing that well takes training. Feedback that names a behavior and a next step is useful. Feedback that restates the score is not. The candidate section of this guide covers feedback models in detail.
The timing of feedback changes. In the station, feedback is immediate and specific. It costs station time. It can also affect the candidate's next station. After the circuit, feedback can be fuller and cover the whole performance. One surgical program added structured written feedback to its OSCE. It sent the feedback by email after the whole exam. That way, feedback on one station could not sway the next.
The consequences change. A formative result leads to a plan: practice, a repeat of the weak stations, a talk with a tutor. It does not lead to a decision on progression. Candidates need to know that in advance. Otherwise they treat the exam as summative and the feedback as a verdict.
How do mock OSCEs fit?
A mock OSCE is a formative OSCE run shortly before a summative one, in the same format. Its purpose is narrower. It builds familiarity with the format, timing, and pressure. Programs run mock exams with faculty examiners or senior students as examiners. Some are peer-led events that the students organize themselves.
The evidence on mocks agrees on one point and splits on another. Mocks reduce anxiety and raise confidence. One peer-led mock OSCE study measured state anxiety. The mock group scored 54.8 and the control group 63.6. The authors found the gap significant. A second peer-led mock study reported higher confidence. The effect on the later summative score is small or absent. The same anxiety study found summative scores of 8.60 against 8.30. That gap did not reach significance. A Canadian study found no effect on later scores.
The reading for a program is simple. A mock is worth running for the candidates' sake. It should not be expected to raise the pass rate. It cuts the share of the summative result that reflects an unfamiliar format rather than skill. That is a gain in validity even when the mean score does not change.
Two rules keep a mock honest. The live station bank stays out of it. And the mock's scoring scheme should be the same kind as the summative one. Candidates then learn how they will be judged, not just what the room looks like.
Can one OSCE be both?
Programs sometimes try to make a summative OSCE formative by adding feedback afterward. That works, and it is good practice. Written feedback per station after results turns a verdict into a learning event. It does not touch the measurement. The candidate section of this guide covers how to do it.
The reverse does not work. Feedback within a summative station or between stations changes what later stations measure. It also creates unfairness. Different candidates receive different feedback from different examiners. An evaluation of OSCE implementation in a low-income setting in Tunisia reached a related conclusion from the other side. It advised programs new to the format to start with formative OSCEs. It advised several stations per skill. It advised building educator skills before raising the stakes. A program that begins with summative OSCEs measures unfamiliarity as much as skill. Its examiners are untrained, and its candidates are new to the format.
Designing a formative OSCE
The list below is the minimum required for a formative circuit to produce usable feedback.
Limits of the evidence
The mock OSCE studies are mostly single-site, with one year group each. Several are peer-led events. Their organizers also wrote the evaluation. The anxiety effect is consistent across them. The score effect is not. The studies were too small to detect a small effect with confidence. The formative-first advice rests on one setting's experience. Its barriers were tradition, educator skills, and student familiarity. Those may weigh differently elsewhere. Treat both as good practice with moderate evidence, not as settled findings.
Frequently asked questions
What is a formative OSCE?
A formative OSCE is conducted to provide candidates with feedback rather than to make a progression decision. It uses the same station format. The examiner gives feedback after scoring. The stations are chosen to expose common weak spots. The result leads to a learning plan rather than a pass-or-fail outcome.
Can examiners give feedback during a summative OSCE?
No. Feedback at a summative station affects performance at later stations. It also creates unfairness between candidates who receive different feedback. Summative feedback should be written and sent after results, per station and overall. Verbal feedback in the station belongs in formative OSCEs only.
Do mock OSCEs improve exam results?
They reliably reduce anxiety and raise confidence. Their effect on the later summative score is small or absent in the published studies. Run a mock to remove an unfamiliar format from the result. Do not run it to raise the pass rate.
Should a new OSCE program start formative or summative?
Start formative. Implementation studies advise a formative first cycle. Examiners learn to score. Candidates learn the format. The program learns its logistics before a result carries consequences. A summative exam run with untrained examiners and new candidates measures the unfamiliarity as much as the skill.
How many stations does a formative OSCE need?
Fewer than a summative one. No reliability threshold has to be met for a pass or fail decision. Choose stations for the skills the program wants to build. Give each one enough time for feedback. A formative circuit of six to eight stations with feedback is common.