An OSCE generates more feedback material than any other exam a school runs: a score per station, a judgment per domain, and an examiner who watched the performance. Most of it is never sent. The published work agrees on three things. Written feedback after the whole exam is the practical default. Students rate the marking detail above the commentary. And a useful report can be built from data you already hold.
Why the material goes to waste
Three reasons, and all of them are fixable.
The exam was not designed for it. An OSCE is built to produce a defensible decision. So the marking sheet is shaped for scoring, not for telling someone what to do next.
The examiner has no time. A station runs to a bell, and the next candidate is at the door. Written comments compete with the score itself.
And nobody owns the send. Results release is a process with a date; feedback is often a good intention with no owner and no deadline. The fix is to give it both, on the same timetable as results release.
When should it go out?
After the whole exam, not between stations, and not on the day for a cohort still sitting.
The reason is contamination rather than kindness. A German surgical OSCE used structured written feedback. It went out by email after the whole exam had finished. The reason was precise. Feedback must not influence a candidate's performance at later stations. Any feedback that reaches a candidate mid-circuit changes the exam for them and for nobody else.
That leaves a window: after the last session, and before the cohort scatters. In practice, it means feedback is prepared alongside the analysis and released with the result, or shortly after.
Written or face to face?
Both work. Written scales; face-to-face does not.
A Malaysian medical school compared the two across two formative OSCEs with 116 fourth-year students. One semester used immediate face-to-face feedback. The other used an individual written report carrying domain scores, examiner comments, and the marking detail.
Measure | Face-to-face | Enhanced written |
|---|---|---|
Comfortable receiving it | 91.3% | 96.0% (p = 0.022) |
Felt it improved performance | 89.6% | 95.0% (p = 0.008) |
Distressed by it | 13.5% | 10.0% (no difference) |
The written report won on comfort and on perceived benefit. One finding complicates the picture usefully: students who preferred the written form had lower OSCE scores than those who preferred face-to-face, 43.8 against 47.2. Read that as a hint. The candidates who most need a conversation are the least likely to ask for one. The results letters page takes that up.
What students actually value
The marking detail. In the Malaysian study, students ranked the marking sheet the most valuable part of the report. They put it above the comments.
That is a design instruction. If the report shows only a total and a paragraph, the most useful element has been left out. Show the domain scores, show what the station was marking, and show where this candidate sat against it.
Free text still matters, but it is the part schools do worst. In the German study, 87.5 percent of students wanted the structured feedback continued. More than half still called it not concrete enough. They preferred the examiners' own comments to the pre-written options. Structured categories make feedback possible at scale; they do not make it specific.
A structure that examiners can actually complete
The German model is worth copying in shape. Six domains, each rated on a three-point scale with pre-written options, plus free text. They are structuring the encounter, time management, behavior toward the patient, manner of speaking, hygiene, and practical execution. Twelve examiners built it themselves through a structured group process. That is why they completed it.
Two lessons transfer. Examiners fill in what they helped design. So involve them in writing the categories. And pre-written options carry the load while free text carries the value. Leave room for both. The free text is the part candidates quote back.
A report that costs nothing extra
The cheapest useful feedback is already sitting in your score file. A proposed model from Western Australia builds an individual report from summative data alone. It calculates index scores expressing relative station difficulty and relative individual performance. Every student gets one report after the exam.
Its authors make a claim worth testing locally. Timely, individual, actionable feedback for every student is possible without changing the exam process or adding to examiner burden. Is your marking already captured per station and per domain? Then the report is a formatting problem, not a data one.
A note on automation. Language models are being tested for drafting feedback. The evidence so far is preliminary and mostly small pilots. Drafting under expert review is a defensible use; letting a model decide or alter a score is not.
Limits
Almost all of this evidence is self-reported satisfaction and perceived improvement, measured soon after. Whether OSCE feedback changes later performance is not settled. The remediation evidence on the results letters page is sobering, given that assumption.
The studies are also single-site and conducted in different cultures, with small sample sizes. Take the consistent direction and test the specifics on your own cohort.
Frequently asked questions
When should OSCE feedback be sent to students?
After the whole exam. Never between stations, and never while a cohort is still sitting. One published surgical OSCE sent its written feedback by email once the exam had finished. The reason was to keep feedback from changing how a candidate did at later stations.
Is written or face-to-face OSCE feedback better?
One study compared the two across 116 students. Written feedback scored higher on comfort, 96% against 91.3%. It also scored higher on perceived improvement, 95% against 89.6%. Distress was low for both. And written feedback scales, which face-to-face does not.
What part of OSCE feedback do students value most?
The marking detail. Students ranked the marking sheet the most valuable element of an enhanced written report, above the comments. A report showing only a total and a paragraph leaves out the part they use.
What should structured OSCE feedback cover?
One published model uses six domains on a three-point scale with pre-written options plus free text: structuring the encounter, time management, behavior toward the patient, manner of speaking, hygiene, and practical execution. Examiners helped write the categories, which is why they completed them.
Can OSCE feedback be produced without extra examiner work?
Largely, yes. One model builds an individual report from summative data alone. It uses index scores for relative station difficulty and relative performance. Its authors report that timely individual feedback for every student is possible without changing the exam or adding examiner burden.