GAMSAT preparation resources: audit each by the job it does

Choosing between a question bank, a course, essay marking and official material gets easier once you stop comparing products and ask which job in your preparation is currently unfilled. This article is not a ranking and recommends no provider. It gives you a way to audit what you already own, decide what (if anything) to add, and notice when adding more is the wrong move.

The catalogue is built to overlap. One provider's course page lists video tutorials, practice questions, full-length exams and progress analytics in a single offer, with essay marking presented as a separate product. Another provider's page describes printed study guides, notes, simulated questions, mock exams and teaching videos. A bundle like that covers several jobs at once. That is convenient if you need all of them and redundant if you need one, and neither page can tell you which applies to you.

Six jobs a resource can do

This classification is ours. It is not ACER's and it is not any provider's. Its use is practical: write the job next to each resource you own, and see which jobs have two resources and which have none.

• Diagnosis: finding out where marks are being lost, using questions you have not seen before.
• Content repair: closing a specific knowledge gap that diagnosis has identified.
• Timed rehearsal: testing whether your approach holds up against the clock.
• Essay feedback: a reader's response to your Section 2 writing, which you cannot supply from the writer's chair.
• Benchmarking: measuring yourself on material you have not used to learn.
• Mistake review: turning a wrong answer, or a right answer you could not defend, into a changed approach.

Notice what is missing from the list: more questions is not a job. A question library can serve diagnosis, timed rehearsal or benchmarking depending on how you use it, but one set of questions cannot do all three. A question you have reviewed and remember cannot show you how you handle an unfamiliar one.

Providers' own guidance already treats these as stages. One provider's preparation guide opens with a diagnostic, moves to building knowledge, then practice questions and essays, then feedback and progress tracking, then realistic time pressure, and finishes with simulation. Another provider's tips recommend working through questions untimed first and timed afterwards, then comparing how your answers differ between the two. Both are provider advice rather than ACER guidance, and both fit the idea that each job comes with its own purpose and its own moment.

Four rules for running the audit

Name the unfilled job before you buy. Write down which of the six jobs the new item would do and which resource does that job now. If the honest answer is something you own and have not finished, the gap may lie in how you are using what you have rather than in what you have.

Cap what runs at once. Your hours are fixed, and every active resource generates review work: a full mock leaves a heap of answers to go back through, and a bank leaves a queue. One provider's tips caution against packing several full mocks into a short span, which is advice about workload and not about the examination. Our reading is that questions you lack the time to review are the first thing to cut. We would begin with a cap of three resources worked through at once, leaving the benchmark out of the count. That number is a judgement and not a finding. We have no data on the right number, and your situation may call for two or four.

Hold the benchmark back. A benchmark only measures something if the material is unfamiliar. Once you have worked through and reviewed a set, a later score on it may partly reflect memory of the set. One provider's guide places official ACER practice material late in its sequence, after a free online practice test, which fits that logic. We would reserve official practice material for the benchmarking job, since ACER administers the examination. What ACER currently publishes, and on what terms, is for ACER's own pages to settle. This article does not establish it.

Unbundle before you pay. List the parts of a package you would actually use. Essay marking is sold apart from the course on at least one provider's site, so a candidate who needs a reader for their writing is not obliged to pay for lessons as well. Several provider sites also list free practice material, and it is worth checking whether a free option already fills the job before paying for it. We have not assessed the quality of any of it. The audit applies to things that are not products too. A study group fills the feedback job or it does not, and a protocol for comparing your reasoning on your own questions is a way to find out which.

When the limit is review, not supply

Take the last block of questions you finished from any bank. For each wrong answer, try to write one sentence naming the step in your reasoning that went astray. For each right answer, ask whether you could defend it with the explanation closed. One provider's tips include a version of that second check. The sentences are the test: they show you something you can inspect, which a block score does not.

If you cannot write them, another bank would add questions without adding the thing that is missing, and a structured review process is the job to fill first. If you can write them and the same sentence keeps appearing, the pattern is observable but its cause is not yet established. That is the point where adding something becomes reasonable: content repair if the sentence names a knowledge gap, or practice aimed at that specific pattern if it names a way of reasoning.

Where this audit stops

It does not compare providers, and nothing here says any course or bank is unnecessary. A course may fill content repair, structure and essay feedback in one purchase, and for a candidate whose diagnosis shows several of those jobs empty, one bundle could involve less overlap than three separate buys. The cap of three is our judgement. We also have no evidence that any particular resource changes a score, so the audit tells you how to avoid duplicated effort and not how much any resource will help. Confirm anything about official materials, and the examination itself, with ACER.

What an audit cannot show you is whether anything has changed. A block score gives you a total, and it does not say whether the reason you keep losing marks is becoming rarer. Verantrix can show whether a specific reasoning pattern is occurring less often than it was, by recording how you reasoned on your practice attempts and naming the pattern those attempts share. Before you add anything this week, write the one-sentence reason for each of your next ten wrong answers and see whether a single reason keeps returning.