Why your GAMSAT score may have plateaued, and what to check next

The second attempt score barely moves, despite months of extra study. More questions, more videos, another course. The same result. The score tells you marks were lost. What it does not tell you is whether those marks are being lost to the same recurring reasoning step, across different questions, in a pattern you cannot see from a percentage.

Topic versus pattern

When you review a wrong answer, the first thing you see is the topic: graphical data interpretation, argument structure, stoichiometry. What you need to find is the step in your reasoning where the error occurred. Those are not the same thing.

The topic is what the question tested. The pattern is what you did wrong. Graphical data interpretation is a topic. Assumes a linear relationship without checking the axis scale is a pattern. The first describes the question. The second describes the error. You can study more graphical interpretation questions and still lose marks to the axis-checking step if that step is the one that keeps breaking.

A flat score has many possible causes. One of them is that you have addressed the topic variable by covering more material, while a recurring reasoning pattern that produces the same kind of error across different subjects remains unaddressed. This article gives you the checks to run to find out whether that is what is happening.

Checks you can run yourself

Take your last ten to fifteen wrong answers from recent practice. For each one, write down what you actually did. Not what the question was about. The step. Did you eliminate an option because it used a word the passage did not use, without checking whether the passage supported the idea in different words? Did you choose an option because it sounded precise, without checking whether the passage constrained that precision? Did you abandon a multi-step problem at the second step because the path was not obvious, when working the first step would have made the second visible?

If the same step appears three or more times, you have found a pattern worth investigating. If the steps do not repeat, extend the sample or look elsewhere: the explanation might be a knowledge gap, a time-management issue, or simply that fifteen attempts is not enough evidence yet.

Check whether similar errors appear across different subject areas. If you are assuming linearity on graph questions in chemistry and also in economics passages, that suggests a reasoning pattern rather than a subject-specific content gap. If the errors stay within one subject, check whether the recurring step is about that subject's content, or whether it is a reasoning behaviour that has only appeared there so far.

Check whether some correct answers came from weak or uncertain reasoning. A low-confidence correct answer where you were guessing between two survivors is evidence of the same kind as a wrong answer: the reasoning that got you to those two was not strong enough to separate them. If those near-miss correct answers show the same reasoning step breaking as your actual errors, the pattern is broader than the percentage suggests.

What to do with what you find

If the steps do repeat, you have a lead worth following. A recurring reasoning pattern is correctable once identified, and correcting it is a structured sequence: recall what you did, retry the question without the explanation in front of you, identify where your reasoning diverged from the correct pathway, compare the two, and correct the specific step. Not a re-read of the explanation. A process that makes the correction something you did, not something you were told.

Then you check whether it transferred. A different question, later, where the same reasoning step is required. Not to confirm mastery. To find out whether the correction generalised or whether you memorised one specific case. If the error recurs, the correction did not transfer, and you return to the step that failed. If it does not recur across several opportunities, that is evidence the correction is holding. That sequence is what a guided mistake review is built to enforce.

If the steps do not repeat, say so plainly to yourself, and look elsewhere. The plateau might be a coverage gap you have not yet identified, it might be timing, it might be confidence affecting option elimination, or it might be that you need a larger sample before a pattern becomes visible. A negative result from this check does not mean the plateau is unsolvable. It means this particular explanation is not the one.

The evidence requirement

Running this check by hand is possible, but it requires revisiting each wrong answer and reconstructing the reasoning you used at the time. For an attempt from three months ago, that reconstruction is harder than for one from last week, and memory of the reasoning process is not the same thing as a record of what you selected.

Verantrix records which option you selected, how confident you were, how long you took, whether you were correct, and whether you changed your answer, on every practice question. It uses that evidence to identify recurring reasoning behaviours while the attempt is still fresh, and it designs your study plan around correcting them rather than around covering more topics. If you have sat the GAMSAT before and the score has not moved in proportion to the preparation you put in, one thing worth assessing is whether the marks are going to a recurring reasoning pattern that more content coverage will not fix.