Self Marking Your MMI Practice Honestly

Most students practising alone assess themselves the same way: they finish a station, sit back, and decide it went pretty well or pretty badly. That verdict is almost entirely a report on how anxious they felt, not on what an assessor would have written down.

The core problem is that you have access to your intentions. You know you were being empathetic, you know what your half finished sentence was heading towards, you know which point you meant to make before the bell. The assessor has none of that. They have only what left your mouth.

Three reasons self marking goes wrong

• You mark your intentions. Everything you thought counts as something you said, and the gap between them is invisible from the inside.
• You mark your mood. A station that felt calm gets a good score. One where your voice shook gets a bad one, even when the content was better.
• You mark the last thirty seconds. Whatever happened at the end colours the whole station, which is why a fumbled ending feels like a disaster and a tidy one erases four minutes of drift.

None of these are fixable by trying harder to be objective. They are fixable by changing what you are looking at and what you are asking.

Rule one: mark from a recording, never from memory

Audio is enough, and audio is easier to be honest about than video, where you will spend the whole time looking at your face. Record every practice station on your phone. Then wait, ideally until the next day, before you mark it.

The delay is doing real work. Marking immediately means marking while your intentions are still loud. A day later, you hear what a stranger would hear, which is the only useful vantage point you have available.

Rule two: ask evidence questions, not score questions

Giving yourself six out of ten for communication is meaningless. You have no calibration, no comparison group, and next week your six will mean something different. Replace scores with yes or no questions where the evidence is a specific moment in the recording.

A workable sheet, and you can answer all of it in five minutes:

• Did I say a clear position out loud, and can I quote the sentence? Not implied, not gestured at. Quoted.
• Did I say why, using a reason rather than a restatement of the position?
• Did I name the strongest opposing argument in a form its holders would accept?
• Is there a moment where I changed direction because of something I heard? Timestamp it.
• Would someone who only heard this recording know what I would actually do on Monday morning?
• How many sentences did I start and abandon? Count them, do not estimate.
• Could this answer have been given, word for word, to a different prompt in the same category?

That last question is the sharpest one on the list. Generic answers are the most common reason capable candidates sit in the middle of the pack, and they are invisible to the person giving them.

Rule three: mark against domains, not vibes

Real assessors are not forming a general impression, they are working through domains: communication, reasoning, insight, and something like professional judgement, with the exact wording differing by school. Our breakdown of how MMI scoring works in Australia walks through the usual shape, and you should check the university's current admissions page for their published criteria where they share them.

Marking domain by domain forces a useful separation. A station can be warm and empty, or rigorous and cold, and a single overall feeling hides both. Sort your evidence questions under the domains and you will start seeing which one is consistently letting you down.

Marking a roleplay station without an actor's view

Scenes are the hardest to self assess, because half the performance belongs to someone else. If a friend played the other character, ask them exactly two questions afterwards, and ask them in role: how did that conversation feel from where you were sitting, and was there anything you were ready to tell me that I never gave you the chance to say? Their answers are far more useful than a score. Our guide to MMI roleplay stations explains why hidden information is usually the point of the scene.

Marking your own roleplay from audio, the key test is simple: count how much of the talking was yours. If you spoke for two thirds of a scene designed to draw information out of someone, you have your finding, and no rubric was required to locate it.

Track patterns, not scores

Keep one document. For each station, write the date, the type, and two lines: what happened, and what I will do differently. Nothing else. Numbers tempt you into tracking a score that is not real, whereas notes accumulate into a pattern you can act on.

After ten stations, read the whole document in one sitting. That is where the finding is. If seven entries mention running out of time before landing a position, your problem is not ethics knowledge, it is pacing, and no amount of extra reading will touch it.

The honest limits of marking yourself

Self marking catches everything you can recognise. It cannot catch the things you do not know are problems, and those are usually the expensive ones: a tone that reads as dismissive, a habit of finishing other people's sentences, a confidence that tips into arrogance in exactly one topic.

For those, you need something outside your own head. Friends and family are generous graders, which makes them poor calibration. MasterMed's live AI interviewer runs timed MMI stations and marks you against a rubric, so the read on your station comes from something other than your own memory of what you meant. The first speaking station is free on the trial, no card, and the trial does not convert by itself.

Use both. Self marking is what you do after every station because it costs five minutes and builds the habit of hearing yourself accurately. Outside assessment is what you do periodically to find out whether your five minutes are pointed at the right things. If you are still working out the format itself, start with our overview of what an MMI is in Australia and New Zealand.

The uncomfortable part of marking yourself properly is that your good stations turn out to be fewer than you thought. That is not a reason to stop. It is the first genuinely useful information you have had, and everyone who improves goes through it.