Being Asked What You Would Do Differently

You have just told a story you prepared. It went well. Then the assessor says: looking back, what would you do differently? And a candidate who was fluent ten seconds ago says honestly, I think I handled it pretty well, and the temperature in the room drops.

That is the single most costly answer available to that question, and it is extremely common, because most preparation stops at the story and never runs one step past it.

Why it gets asked

Medicine runs on reviewing your own decisions. Case review, morbidity and mortality meetings, audit, supervision, reflective practice for registration: the whole professional structure assumes doctors will examine what they did and consider whether it was the best available option.

A candidate who cannot find a single thing to reconsider in their own story is showing an assessor exactly how they would behave in that structure. Not badly intentioned, just unable to look at their own work critically, which is a genuine problem for someone about to make consequential decisions for a living.

There is a second reason. The question is a live test. Your prepared material has been used up, so what comes next is unrehearsed thinking, and unrehearsed thinking is much more informative than a story you polished for a week.

The two failure modes

The first is the nothing answer. I would do the same thing again, or I cannot think of anything. It reads as defensiveness even when it is really just being caught off guard.

The second is the collapse, where a candidate hears the question as criticism and dismantles their own answer. Actually now that you say it, I probably should have done the whole thing differently, I did not really handle it well at all. This is worse than it feels. It suggests your position moves whenever someone applies pressure, and in a profession where you will sometimes have to hold a view against a confident senior, that matters.

The target sits between them: your decision stands, and one specific part of it could have been better.

A structure that works under pressure

Four short moves, in order:

• Name the part you would keep, in one clause, so the answer does not read as a retreat.
• Name the specific thing you would change, and make it specific: a timing, a person you did not tell, a question you did not ask.
• Say what that would probably have improved, so the change has a reason attached rather than being an apology.
• Where you can, point to a later occasion where you actually did it the better way. This is the part that turns a claim into evidence.

Assembled, that is about twenty five seconds. I would still have raised it with her rather than going straight to the supervisor. What I would change is when: I waited until the end of the shift because I wanted to be sure, and by then she had repeated it twice. Saying something the first time would have been slightly more awkward and considerably more useful.

Timing, information and inclusion

If you are stuck, almost any decision can be reconsidered along three axes. Timing: too early, too late, or over too long a period. Information: what you would want to know before deciding that you did not go and find out. Inclusion: who you would involve or tell that you handled alone, or the reverse.

Those three will generate a credible answer to almost any version of this question in a few seconds, which is exactly what you need when it arrives unannounced.

When it follows an ethical scenario

The question is also used after a scenario station, where it means something slightly different: what would change your approach, or what did you not consider. Here the honest move is to name the consideration you underweighted rather than to reverse your conclusion.

I would keep the same position, but I moved past the patient's own view too quickly and spent most of my time on the family. If I ran it again I would start by asking what she wanted, because that might have changed everything after it. You have conceded a real weakness without abandoning your reasoning, which is the exact behaviour the station is looking for.

It is also fine to say you would want more information and to name precisely which information and why it would matter. That is different from hedging, because you are specifying what would move you.

Prepare the second layer, not a second script

The practical preparation is simple and almost nobody does it. Take each story in your bank and write one line: the part of this I would change and why. Not a paragraph, one line. Then it exists when you need it. Our list of common MMI interview questions in Australia is a reasonable place to check which of your stories are likely to get used, and therefore which need the extra line.

Do not memorise the wording. A reflection delivered as smoothly as the story it follows sounds manufactured, and the small hesitation of genuinely thinking is part of what makes the answer land. Know the content, find the words on the day.

The same discipline applies to your motivation material. Assessors sometimes ask what you would do differently about your own preparation or your path here, and a candidate who has thought about the weaknesses in their own case answers that comfortably. Our guide to the why medicine answer covers how to build a position that can survive being questioned rather than one that only works unopposed.

Practise being interrupted

Follow ups are where circuits do most of their discriminating, and in some stations an interviewer will deliberately probe rather than let you run. Our overview of what an MMI is in Australia and New Zealand describes how that usually plays out, though station design differs by school and changes over time, so check the university's current admissions page for anything specific to your interview.

Rehearsing alone will not prepare you for this, because you never ambush yourself. MasterMed's live AI interviewer runs timed stations and marks you against a rubric, and the follow up arriving when you thought you were finished is the specific thing worth practising here. The first speaking station is free on the trial, no card, and the trial never converts by itself.

Every story you tell has a version you would run better. Find it before the day, and the question that ends other candidates becomes the part where you sound most like a doctor.