Should I resit the GAMSAT?

A four-point mean is commercially inconvenient in this market. Almost every page you will find on this question is published by a company that sells preparation courses, and a four-point expected gain is a difficult thing to sell against. We do not sell courses, so we can print it.

It is not an argument against resitting. It is an argument against resitting on the assumption of a large jump, and against paying for a course on the strength of an improvement figure nobody has measured.

The four-point finding is the best evidence available, and it is weak evidence. Anyone quoting it without these caveats, including us, would be overstating it.

• It is a mean with no distribution behind it. The paper gives no standard deviation, no sample size for the repeat subgroup and no confidence interval, so it cannot tell you what share of candidates improved, stayed flat or scored lower. A mean cannot tell an individual what to expect.
• It does not say what the four points refers to. Overall score or a section score is not specified, and there is no breakdown by section, by starting score or by preparation undertaken.
• The data is Australian, March sittings only, and covers 2005–2014. The test has since shortened from 75 questions to 62 in Section 1 and from 110 to 75 in Section 3.
• The authors administer the test. Three of the five sit or sat on the GAMSAT Policy Committee and two work for ACER. The paper contains no limitations section.

We publish it because a weak measured figure beats a confident invented one. If a better study appears, this page changes.

At most Australian schools, GPA carries equal or greater weight. ANU, Deakin, Griffith, Macquarie and UQ all shortlist on a 50:50 split; UWA shortlists 60:40 in favour of GPA. At Macquarie, GAMSAT drops out of the offer calculation entirely — the final ranking is 50% GPA and 50% interview University-published.

If your GAMSAT is already around the interviewee averages and your GPA is not, four more GAMSAT points will not change the outcome. Work out which half of the file is failing before you pay to re-sit the half that is not.

Wollongong treats GAMSAT as a pass/fail hurdle and never ranks it: a 51 and a 75 are worth exactly the same there. Sydney uses no overall score at all and ranks on section scores. Melbourne, Queensland and both Notre Dame campuses use the unweighted mean, so a resit that lifts Section 3 is worth substantially less at those schools than at one using ACER’s weighting. Check which method applies before you decide what to study.

Results last 4 years in Australia and 2 in the UK and Ireland, with Ulster a named exception at four ACER-official. Scores from different sittings are directly comparable, so an older result is not a weaker one. You may have more application cycles left on your existing score than you assumed — see how validity works.

Section 3 is the highest-leverage target under ACER’s weighting: it counts double and has the widest score spread of the three sections, so the same improvement moves your overall further there than anywhere else. In Ireland, where the overall score alone ranks candidates with no GPA component, Section 3 is half of the number that decides the offer.

Section 2 is the hardest to plan around. It is marked by three raters against a scale of one to ten with adjustments for rater severity, and it correlates only weakly with the other two sections — essay ability and science reasoning are close to independent Peer-reviewed. A strong Section 3 alongside a weak Section 2 is an ordinary pattern, not a contradiction.

11 pathways across 9 institutions admit without a GAMSAT score at all. Almost no preparation site lists them, because they are the one route that makes the product irrelevant.

Registration is A$568, £296 or €378 ACER-official. The larger cost is the calendar: GAMSAT runs across two sittings roughly three weeks apart, results arrive about two months after the multiple-choice papers, and there are no remarks and no appeals if the result disappoints again.

One structural point in favour of resitting: a lower subsequent score does not erase a higher earlier one while that earlier result is still valid. In Ireland, ACER forwards only your best overall score to the CAO. Confirm the position with each institution, since not every school states it.

Reasonable grounds, based on what the evidence supports:

• Your score is genuinely below the published cutoffs where you want to apply, rather than a few points below an interviewee average — those are different distances. Check the real numbers first.
• Something identifiable went wrong. Illness, a missed section, running out of time in Section 3, or sitting with no timed practice at all. A four-point mean is drawn from candidates who mostly did not have a specific fixable problem.
• You are below a hard section floor. Most Australian schools require 50 in every section, and several UK schools require 50 in each. A single section below the floor invalidates an otherwise competitive application, and that is a targeted problem rather than a general one.
• Your score is about to expire in a two-year market and you intend to keep applying in the UK or Ireland.

Weaker grounds, honestly labelled:

• You are already at or above the interviewee averages and want to be safer. The published evidence does not support expecting a large jump.
• This would be your third or later sitting. That is precisely where the evidence finds little further upward trend.
• Your GPA, Casper or interview performance is the weaker half of the application.

Nothing on this page tells you how much you personally will improve, because nobody can. What the data does support is where to aim: Section 3 carries the most weight and the most spread, timed practice addresses the failure mode most commonly reported, and the sections test close to independent abilities, so improving one does not lift the others.

Changing the result usually means changing the method rather than repeating it, and the materials are the cheapest variable to change. Our independent review of GAMSAT books sets out which titles are worth buying for Section 3 content and which duplicate what you already have.

Our free timed practice test reports a raw percentage on our own original questions. It deliberately does not show a predicted GAMSAT score: no public conversion from raw marks to scaled scores exists, and a live sitting also contains unscored trial questions. Any tool offering you a predicted score from a practice test is inventing it.