How MMI scoring works in Australia

You will not be shown your MMI mark. Australian schools do not email a station-by-station spreadsheet. Anyone offering last year's "cut-off of 5.6" from Reddit is not a source.

What you can know is the design. An MMI is independent samples. Each station has its own assessor and its own score. Those scores are combined, then combined again with academic and admissions-test results in a way that school publishes, or does not.

What an MMI is is the format. This page is the mark, as far as public pages go.

Independent stations, not one vibe

The point of a circuit is that station three has not heard station two. A clumsy opening does not follow you down the corridor. A charming ethics speech does not carry the role-play. That is why recovery is built in, and why performing the same speech six times is a waste.

Schools that are not MMIs do not work this way. UNSW is two interviewers in one room; they have heard everything you have said, and they have read your MAP form. Otago Alternative is a Zoom panel. Sydney's standard metropolitan MD currently has no interview to score.

If your invite is a panel, stop thinking in stations.

Typical circuits in this country run somewhere between six and ten stations of about five to ten minutes. Those ranges are typical, not a promise. Monash currently publishes six stations. Adelaide currently says a minimum of six of ten minutes. UQ currently briefs eight. JMP currently describes eight of eight minutes. Auckland currently assesses seven of eight Kira stations. Do not average those into a fake national circuit.

What public pages actually say about combining scores

Only quote a weighting if that university published it for your intake.

Undergraduate examples:

• Monash: ATAR aggregate, UCAT ANZ total, and MMI are equally weighted (one-third each) for offers. December interview selection can be UCAT-only; January puts ATAR back in.
• Adelaide: offers 20% UCAT / 40% interview / 40% academic. Interview invites are UCAT-only (sections 1 to 3; SJT as a tie-break).
• UNSW: academic rank, UCAT ANZ overall (SJT not used), and interview. Not averaged. A high mark in one area will not compensate for a low mark in another. Algorithm not disclosed. No interview feedback.
• UQ: each MMI assessor gives a score that contributes to the admissions rank. They do not publish a public 20/40/40 split on the interviews page. Treat forum "weightings" as unverified.
• Auckland MBChB (2027 limitations): GPA 60% / MMI 25% / UCAT-ANZ 15%, with a Dean-set MMI minimum. An older PDF described a "fair score" adjustment for scenario difficulty and interviewer harshness; the 2026 MMI webpage does not repeat that paragraph. Do not freeze the PDF.

GEMSAS graduate examples, from the 2027 school entries (confirm there):

• Melbourne: course offers GAMSAT 25% / GPA 25% / interview 50%. Interview ranking is equal GPA and unweighted GAMSAT.
• ANU: offers 50% interview / 50% the composite GPA/GAMSAT used for interview ranking.
• Deakin general stream: 50% GPA + GAMSAT + adjustments, 50% MMI.
• Griffith: 50:50 interview-selections rank (GPA/GAMSAT) and overall interview score.
• Macquarie: offers 50% GPA / 50% MMI. GAMSAT is used to select for interview, not in the offer formula as published.
• Notre Dame (Sydney and Fremantle): offers 50% GAMSAT/GPA/Casper/bonuses and 50% interview.
• UWA (standard): GPA 30% / GAMSAT 20% / interview 50% (rural column uses a rural rating and lower percentages on the other three).
• Wollongong: offers 70% interview / 30% admissions bonuses. GPA, GAMSAT and Casper are hurdles or interview-ranking tools, not offer arithmetic, on the 2027 entry.

If a school has not published a split, we will not invent one. JMP describes a circuit without handing you the arithmetic.

GEMSAS also says interview scores are converted to a common standardised scale (for example a z-score) before they may travel to a lower-preference school. That is consortium policy, not a claim that every Australian undergraduate MMI uses the same transformation.

What the interviewer is probably scoring

They are not scoring your UCAT. Verbal Reasoning, Decision Making, Quantitative Reasoning, and Situational Judgement are finished. Overlap with SJT is values, not a section score.

They are not scoring an OSCE checklist. You are not being marked on a cannula.

Public competency lists cluster: communication, ethical reasoning, empathy, teamwork, motivation, cultural safety, critical thinking. Monash names advocacy, collaboration, critical thinking, empathy, ethical reasoning, motivation and resilience. UQ names kindness and compassion, integrity, adaptability, verbal communication. Adelaide names problem solving, communication, teamwork, empathy, integrity, preparedness, cultural awareness, motivation. Wollongong names communication, empathetic and ethical approach, reflective manner, teamwork, decision-making, rural and Aboriginal and Torres Strait Islander health.

Those lists are not secret rubrics. The sheet in the room is unpublished. Treat third-party "1 to 7 Likert with anchors" as generic MMI research language, not as Monash's form.

Follow-ups are where a lot of the mark sits. A rehearsed opening that dies when they change one fact is a written answer delivered out loud. Ethics and why medicine fail the same way: slogans, then silence.

AHPRA's Good medical practice is the values frame in Australia. Not the GMC. In New Zealand, MCNZ and Te Tiriti.

What you cannot know, and should stop hunting

You cannot know your station scores, whether interviewer three is a "harsh" rater, a numerical cut-off for an offer, or whether finishing thirty seconds early cost you a point (UQ says quality, not length). You can know whether you answered the stem, took a position, named the people affected, and handled a push-back. That is enough to practise.

How this should change prep

Practise independent starts. After a bad station, walk it off. Do not apologise to the next interviewer. Virtual versus in-person does not, on public pages, come with a different marking scheme: same competencies, different room.

Do not try to game an unpublished rubric. If a commercial course sells "this is exactly how they mark," they are guessing.

After UCAT, the Sep to Dec timeline is the calendar. One marked mock circuit teaches you more about scoring than a leaked table, because you will hear which stations actually collapse.

MasterMed's first speaking station is free on /interview. No card. The trial never auto-converts. Use it to get a mark against a rubric you can see, then remember the university's sheet is still theirs.

Re-read the current admissions guide the week you are invited. Weightings move. This page is not a contract.