Skip to main content
All resources

Use OSCE Room · 4 minute read

Read your feedback—and turn it into your next step

How to read your AI report: what each result means, how to check the evidence, what coverage means, and how to turn it into your next step.

Published by OSCE Room · Updated · Guide version 2

The idea to keep in mind

Start with one result and its evidence, not the total score. Check what the report could and could not assess. Then choose one skill to practise.

What the AI report is based on

When you submit, an AI examiner compares your saved work with the station's marking guide, called a rubric. Each point in the rubric is a criterion: something you can be seen to do. The evidence can include your recorded speech, the actions you took and your record entries.

The AI checks whether you did each thing, and it accepts different ways of doing it. It does not compare you with one ‘perfect’ answer.

The report is written by AI. The stations and suggested approaches have not been checked by a clinical expert, and the AI can make mistakes. Use the report to understand what was captured. Clinical accuracy, hands-on skill and your official exam result are separate questions.

Start with one result, not the total

Pick one criterion and read its result and explanation. Then open the evidence link and look at the words, action or record entry it points to.

Ask yourself: Was I talking to the right person? Does the quote keep my meaning? Was the entry made at the right time? If your memory and the saved evidence are different, look into it before you make it your next study goal.

What the result labels mean

  • Demonstrated: your saved evidence showed this criterion.
  • Partly demonstrated: some parts were shown, but not all.
  • Not demonstrated: the saved evidence did not show it.
  • Not assessable: there was not enough evidence to judge, for example because speech was not captured.

Coverage: how much could be assessed

Coverage tells you how much of the rubric the AI could assess from your evidence. It is not your score. For example, a report may assess your conversation and records, but not your hands-on skills, which a screen cannot see.

If coverage is low, the report may not show a total, or may show a range. That tells you what the system could check. It is not a hidden fail. Do not compare two totals as if they were the same when one attempt used help, or one report assessed much less.

Is it a skill gap or a technical gap?

A missing word, a dropped connection or talking to the wrong listener can limit what the report can judge. ‘Not assessable’ is different from a mistake you made. If important evidence was not captured, the AI cannot rebuild what you said.

Check the report's limitations and the transcript. OSCE Room cannot currently replay your audio or re-transcribe it. If a number or the word ‘not’ was captured wrongly, treat it as a technical problem: test your setup before repeating the skill. Do not change your clinical study based only on an uncertain transcript.

Safety findings and reference approaches

A safety finding points to a possible safety problem in your evidence, based on the station's draft rules. Look at the evidence and the station's references. A good result in other areas does not cancel a safety issue.

A reference approach is an example of one way to handle the fictional task. It is not a script to copy, and it is not an approved treatment plan. Use it to understand structure and reasoning. For clinical learning, use trusted official sources and hands-on support.

Choose your next practice

Turn one gap into something you can practise. If your handover had no clear request, practise just the request. If units were missing, practise a few record entries. If you misread the task, do a reading-only drill.

Also keep one strength, so you remember what to keep doing. After your next attempt, compare the evidence for that skill, not just the total.

A three-line reflection

  • What I did well: one strength, and the evidence that shows it.
  • What needs work: one gap, or one evidence problem to sort out.
  • What I will do next: one specific practice or setup check.

Go back to the same attempt to review progress

Progress & feedback opens your saved attempts and reports. Each attempt keeps the station version you practised and your saved evidence. If the station is updated later, your old attempt does not change.

If the site offers an updated report, it keeps your original evidence and earlier reports. A new report is a new reading of the same attempt, not a new performance. When tracking progress, keep in mind the mode, the quality of the evidence and the report version. No practice result predicts an official pass.

Not clinically reviewed. These are study explanations and platform guidance, not patient-specific clinical advice or an official marking scheme. OSCE Room is independent of NMBA and Ahpra; practice feedback does not certify competence or predict a pass.

For exam rules, check the current official handbook and your own Ahpra correspondence. Digital practice supports preparation alongside supervised hands-on learning.