Giving and Receiving Feedback in Practice Sessions

Practice groups fail for one reason more than any other. Nobody wants to be the person who tells a friend that the answer was vague, so everyone says it was good, and four weeks later a whole group of people walk into a circuit having never once been told the truth about how they sound.

The opposite failure is just as common and does more damage. One person in the group has read a lot, decides they are the assessor, and delivers a running critique that leaves someone unable to speak by station four. Both versions waste the same evening.

Good interview practice feedback is a skill in its own right, and it is learnable in about twenty minutes. What follows is the set of rules that make a practice session worth the two hours it costs.

Why most feedback is unusable

The default feedback in a study group sounds like a verdict. That was strong. That felt a bit weak. You seemed nervous. None of it tells the person what to do differently on Tuesday, because a verdict is a summary of the listener's feeling, not a description of what happened.

The second problem is volume. If you hand someone nine observations, they will remember the harshest one and act on none of them. Feedback is only useful at the rate a person can absorb it, which is roughly one or two changes per session.

The third problem is authority. Nobody in a peer group is an assessor. The moment someone starts marking with a confident number, the group is optimising for that person's taste rather than for the criteria a real panel would use.

The rules that make it work

Agree these before the first station, out loud, so that nobody has to invent a policy in the moment when a friend has just spoken badly.

• Describe, do not diagnose. Say what you heard: you spent about ninety seconds setting the scene and thirty on what you decided. Do not say you are bad at structure.
• Two things maximum. One that landed, one to change. If you have five observations, choose the one that would improve the most other answers.
• Quote them. Repeat one actual sentence they said. People argue with characterisations and accept quotes.
• Make it actionable in the next repeat. Not be more confident, but start with your position and give the reasons after.
• Feedback goes to the answer, never to the person. The answer missed the patient's fear. Not you are not very empathetic.
• No verdict words. Ban strong, weak, good and bad from the room for the evening. It forces everyone to say something real.

That last rule sounds gimmicky and it is the one that changes the session most. Take away the easy words and people are left with the observation underneath them, which is what the speaker actually needed.

What is worth commenting on

Aim your attention at the things assessors are broadly looking at rather than the things that are easy to notice. Our explainer on how MMI scoring works in Australia sets out the kinds of domains that typically appear on a station sheet, and formats vary by school so check the university's current admissions page for what they publish. The useful territory is reasoning, whether the person engaged with the hard part of the question, and whether a listener could follow the shape of the answer.

There is a long list of things that feel like feedback and are not. Filler words below a certain frequency. Where someone's hands were. Whether they said doctor or clinician. If you find your group spending ten minutes on any of these, someone is avoiding the harder conversation about content.

Appearance is the clearest example. Sort it once, well before the season, using something like our guide on what to wear to a medical interview, and then never spend practice time on it again. It is a solved problem that keeps getting resolved because it is comfortable to discuss.

Feedback in roleplay is a different job

When you are the actor rather than the observer, your feedback is worth more than anybody else's in the room, because you were the only person receiving the candidate. Our piece on how MMI roleplay stations work covers the format itself, but the feedback rule is simple: report your experience as the character, not your judgement as a peer.

Say when you felt heard and when you stopped feeling heard. Say the moment you would have opened up if they had waited two more seconds. Say when an apology sounded like a procedure. That is information nobody watching can give, and it is usually the single most useful thirty seconds of the whole session.

Play the role straight while it is running. An actor who softens because they like the candidate has removed the entire point of the exercise, and an actor who invents extra hostility for sport has tested something no station will ask for.

Receiving it without going under

The hard part is not hearing the criticism. It is the twenty minutes afterwards, when your attention is on the comment rather than the next station, and every answer for the rest of the night is worse because you are monitoring yourself.

Three habits that help. Write it down and stop thinking about it, because the writing is what gives your brain permission to release it. Do not explain yourself, since defending the answer in the room converts feedback into an argument and you lose the rest of the comment. And do not act on a single voice: if one person says you rush and nobody else has, note it and wait for a second sighting.

Then, and this is the part almost everyone skips, repeat the station immediately with the one change. Feedback that is not tested in the next ten minutes is a note, and notes do not alter how you speak.

A session format worth copying

Three people. One speaks, one observes against a written prompt, one keeps time. Station runs to the clock. Sixty seconds of feedback, two things only. Immediate repeat of the same station with the change applied. Rotate.

Ninety minutes of that beats a four hour session where six people take turns and everybody gently praises everybody else. The repeat is the mechanism. Everything else is scaffolding around it.

Rotate the observing role deliberately, too. The people who improve fastest in a group are usually the ones who spend the most time watching, because you can hear a missing structure in someone else's answer long before you can hear it in your own.

When you have nobody to practise with

Plenty of applicants do not have three reliable people, and the honest answer is that solo practice with a recorder gets you part of the way. Listening back catches pace, filler and structure, and it catches nothing about how you responded to a person. MasterMed's live AI interviewer sits in between: it runs timed stations and marks against a rubric, so you get consistent criteria without waiting for a group to be free. The first speaking station is free on the trial, no credit card, and the trial never converts by itself.

The thing to protect, whichever way you practise, is the honesty. A group that has quietly agreed to be kind is more dangerous than no group at all, because it produces the feeling of preparation without any of the correction. If nobody in your sessions has told you anything uncomfortable in a fortnight, the sessions have stopped working.