QCE Specialist Mathematics IA2 and IA3 examinations: how to get top marks
IA2 and IA3 are one instrument sat twice. IA2 draws on any three of the five Unit 3 topics, IA3 on any three of the five Unit 4 topics, and each carries 15% of your Specialist Mathematics result. Your teacher writes the paper, you sit it unseen with 5 minutes perusal and 90 minutes working, and there are no bands to reach: your raw score becomes a percentage and a fixed fifteen-step ladder turns that percentage into your mark. Every step is a strict inequality, which QCAA illustrates with a case of its own: exactly 80% is 12 marks, not 13.
This guide describes the Specialist Mathematics 2025 v1.0 General senior syllabus, which the 2025 subject report calls "first examinable in 2026". The three reports it draws on cover the 2023, 2024 and 2025 cohorts, all of whom sat Units 3 and 4 under the superseded 2019 syllabus, so every figure below is labelled with its cohort year. The fifteen cut-offs came through the version change identical, digit for digit, while the descriptor prose beside them was rewritten.
The specification your teacher writes to
Your school writes the paper and QCAA approves it before you sit it. The specification is a single sentence repeated for each instrument with the unit swapped, and the conditions are identical:
| Sampling | any three of the five topics in Unit 3 (IA2) or Unit 4 (IA3) |
| Time | 5 minutes perusal, 90 minutes working |
| Supervision | individual, supervised |
| Provided | the QCAA Specialist Mathematics formula book, which your teacher must supply |
| Technology | required, and the paper must carry both technology-free and technology-active questions |
| Notes | none permitted |
| Split sittings | allowed across two consecutive sessions only if each session's questions are unseen and no teaching or feedback happens between them |
Questions may ask you to interpret unseen stimulus, calculate using algorithms, and draw or label graphs, tables or diagrams. Both papers assess all six assessment objectives and both are marked on one criterion out of 15.
What each paper is allowed to assume you know
This is where Specialist Mathematics differs from the other two QCE maths subjects. The syllabus names Mathematical Methods as assumed knowledge alongside your own earlier units: Methods Units 1 and 2 stand behind IA2, and Methods Units 1, 2 and 3 stand behind IA3.
That gives a Specialist question a much longer reach than its topic label suggests. A Unit 3 vectors question can require Methods differentiation without announcing it, and a Unit 4 statistical inference question can lean on the Methods normal distribution work.
IA2: the Unit 3 examination
IA2 samples any three of these five:
- Further complex numbers
- Mathematical induction and trigonometric proofs
- Vectors in two and three dimensions
- Vector calculus
- Further matrices
Vector calculus was a sub-topic under 2019 and is now a topic in its own right, with its nominal hours nearly doubled. It is the largest single increase in Year 12 and it is now one of the five things IA2 can pick.
You cannot know which three your school will choose, so four of the five is the working minimum. Two topics you revised hard may not appear at all, and the one you skipped may carry a third of the paper.
IA3: the Unit 4 examination
IA3 samples any three of these five:
- Integration techniques
- Applications of integral calculus
- Rates of change and differential equations
- Modelling motion
- Statistical inference
Modelling motion was also promoted from sub-topic to topic for 2025, and a new differential equations sub-topic was carved out of rates of change, which lost most of its hours to it. Both changes push Unit 4 towards calculus applied to a physical situation.
What a pre-2025 revision list gets wrong
Only one piece of subject matter crosses a unit boundary in Year 12, and it leaves. Subsets of the complex plane determined by straight lines and circles moved from Unit 3 down into Unit 2, where the verb also strengthened from identify to identify and sketch. Under 2019 it was directly examinable in IA2. Now it is Year 11 work that IA2 can reach only as assumed knowledge.
Two things moved the other way, up from Unit 2 into Unit 3, so IA2 can now sample them: determinants, inverses and matrix equations for square matrices beyond 2 × 2, all with technology; and the binomial expansion used to prove multi-angle trigonometric identities, which now sits alongside De Moivre's theorem in the induction topic.
The bigger shift is in what you are asked to do with content that stayed put. Of the 41 reworded Year 12 dot points, 31 changed their leading verb:
| Content | 2019 asked you to | 2025 asks you to |
|---|---|---|
| Inductive proof | use an initial statement and an inductive step | use an initial statement, an assumption statement, an inductive step and a conclusion |
| De Moivre's theorem, integral powers | prove and use | use, with the proof obligation moved into the induction sub-topic. It has not left the course |
| Cartesian equation of a path from a vector equation | derive it | understand and use it, with the circle, ellipse and hyperbola forms now printed in the syllabus |
| Differential equations in context | formulate and use | model and solve problems using provided differential equations, so the equation is supplied |
| First-order differential equations | solve simple ones | determine general and particular solutions, so an initial condition is standard |
| Simpson's rule | apply it using technology | apply it with and without technology, which makes it examinable by hand |
| Partial fractions | use in simple cases | use where the denominator has two distinct linear factors, and no further |
The four-stage induction requirement is the clearest of these. QCAA has written down the stages it will mark, and the effective-response bullets in the 2024 report reward exactly those four.
Turning your raw score into a mark out of 15
One criterion, Foundational knowledge and problem-solving, worth 15 marks. The instrument-specific marking guide, or ISMG, is the marking guide QCAA prints inside the syllabus for each internal assessment, and this one gives your teacher nothing to match your response against question by question. The sequence is arithmetic:
- Before you sit the paper, your school builds a marking scheme: a per-question map of where every mark and half mark lives. QCAA treats it as its own artefact, distinct from the ISMG. It is uploaded for the endorsement review before you sit the paper, and it has to be available again at confirmation, the check run on a sample of each school's marking once the papers are marked.
- Your paper is marked against that scheme, question by question, and the raw marks are totalled.
- The raw total becomes a percentage of the paper total, computed without significant rounding. QCAA has asked schools to show that calculation in full in each of the last three rounds, and prints worked cases like
40.5/48 = 84.375%and48.5/60 = 80.833%. - The percentage is read off the ladder below.
| Your percentage | Criterion mark | What the descriptor says at this level |
|---|---|---|
| > 93% | 15 | consistently correct recall and use; authoritative and accurate communication; astute evaluation of the reasonableness of solutions; reasoning used to correctly justify procedures and decisions; fluent application across simple familiar, complex familiar and complex unfamiliar situations |
| > 87% | 14 | (same descriptor) |
| > 80% | 13 | correct recall and use; clear communication; considered evaluation; reasoning used to justify procedures and decisions; proficient application across all three situation types |
| > 73% | 12 | (same descriptor) |
| > 67% | 11 | thorough recall and use; communication; evaluation; reasoning used to justify; application in simple familiar and complex familiar situations only |
| > 60% | 10 | (same descriptor) |
| > 53% | 9 | recall and use; communication; evaluation of some solutions; some reasoning; some application to make progress in simple familiar situations |
| > 47% | 8 | (same descriptor) |
| > 40% | 7 | some recall and use; basic communication, and nothing else. Evaluation, reasoning and application have dropped out |
| > 33% | 6 | (same descriptor) |
| > 27% | 5 | infrequent recall and use; basic communication of some knowledge |
| > 20% | 4 | (same descriptor) |
| > 13% | 3 | isolated recall and use; partial communication of rudimentary knowledge |
| > 7% | 2 | (same descriptor) |
| > 0% | 1 | isolated and inaccurate recall and use; disjointed and unclear communication |
Eight descriptor blocks stretch across fifteen marks, so the prose can never separate two adjacent marks. Only the percentage can. And the top four marks all require the descriptor's third situation type, which means the complex unfamiliar questions have to be attempted.
Four conversions QCAA has printed from real Specialist papers, which also show how much the paper total varies from school to school:
| Raw score | Percentage | Criterion mark | Where |
|---|---|---|---|
| 46/63 | 73.02% | 12 | 2023, named as effective practice |
| 57.5 | 95.8% | 15 | 2023, marked up as two section totals of 33 and 24.5 |
| 54/56 | 96.4% | 15 | 2024, two sections of 28 marks each |
| 53/55 | 96.36% | 15 | 2025 |
Nothing fixes the paper total. Only the 60/20/20 split of the marks inside it is specified.
The 60/20/20 blueprint
IA2, IA3 and both external papers are built to one question specification, within ± 2%:
| Degree of difficulty | Share of marks | Objectives it typically targets |
|---|---|---|
| Simple familiar | 60% | 1, 2 and 3: recall knowledge, use knowledge, communicate |
| Complex familiar | 20% | any of the six |
| Complex unfamiliar | 20% | 4, 5 and 6: evaluate reasonableness, justify procedures and decisions, solve problems |
Complex unfamiliar has a definition the syllabus states in full and both versions state identically: all the information needed to solve the problem is not immediately identifiable, the required procedure is not clear from the way the problem is posed, and the context is one you have had limited prior experience of.
Read the cue, then use the method it names
Several Specialist problems have a second solution route that sits outside the syllabus, most often a physics formula for a motion or circular-motion problem. QCAA has told schools in each of the last three rounds to cue the expected approach inside the question, using phrases like "using a vector calculus approach" or "definite integration using technology", and to build the marking scheme around the approach cued.
Take those phrases literally. A correct final answer reached by a route the scheme was not written for has fewer places to attach marks, and on a multi-mark question that is most of the marks.
The same logic runs through the technology split. The syllabus requires each internal paper to provide opportunities for both technology-free and technology-active responses, so one 90-minute paper carries both, and the mode changes where the marks sit. On the technology-free half they are on the algebra. On the technology-active half they are on setting the problem up and interpreting what comes back, so write the expression you evaluated as well as the value it returned. Find out before the day whether your school marks the modes by labelling questions, by cueing them in the wording, or by splitting the paper into sections.
What the marked excerpts reward
There is no prose artefact here to compare against a model answer. The excerpts QCAA reproduces are photographs of handwritten algebra covered in ticks, half marks and marker annotations, and three mechanisms show up in all three years.
All three mechanisms are generous and all three are conditional on the same thing: legible working for the credit to attach to. QCAA even asks schools to annotate a mark awarded for a step you implied and never wrote, which only happens when the lines around it make the step unambiguous.
Where the marks actually went missing
These are the errors QCAA's own markers annotated on Specialist internal papers, quoted from the reports:
| What went wrong | The marker's note |
|---|---|
| A sign error in a discriminant, in a particle-collision problem, where two mistakes happened to cancel | "mistake in discriminant (two mistakes actually cancel)" (2023, IA2) |
| Substituting into the wrong time value and carrying it forward | "but 6² − 7×6 + 10 = 4" (2023, IA2) |
| Solving a three-dimensional vector problem as though it were two-dimensional | "3rd dimension? they could be in different planes here!" (2023, IA2) |
| Choosing a unit vector that does not match the orientation the response set up | "this would actually be k, given the orientation of the system!" (2023, IA2) |
| Losing a constant factor part-way through an integration | "forgot the 6" (2024, IA3) |
| Drawing a boundary on an Argand-diagram region without marking whether it is included | "convention is a dashed line, however, a written statement will suffice" (2024, IA2) |
Two of the six are three-dimensional vector bookkeeping and one is a boundary convention. None of them is a gap in understanding, and all three cost marks on a paper where 6 to 7 percentage points is a whole ISMG mark.
What happens to your paper before and after you sit it
Two QCAA processes bracket the exam, and they behave very differently in this subject.
Endorsement is the approval QCAA gives your school's paper before anyone sits it. Across the 2023, 2024 and 2025 rounds it was granted at first application to 43%, 50% and 30% of Specialist IA2 papers, and to 47%, 44% and 61% of IA3 papers. The 2025 IA2 figure is the lowest number anywhere in this subject's reporting, and it came with 231 alignment flags against 122 for IA3 in the same round. Alignment to the syllabus specifications is the dominant reason a Specialist internal exam is refused, in every year and on both instruments, and the recurring faults are the ones this guide has already named: items labelled complex unfamiliar that are not, questions resting on assumed knowledge alone, and schemes that do not match the approach the question cued.
Confirmation is the check QCAA runs on a sample of the marking afterwards, and it barely finds anything. Agreement between school and confirmed marks ran 99.01%, 100.00% and 100.00% on IA2 and 99.67%, 100.00% and 100.00% on IA3 across those three cohorts. Two consecutive perfect years on both instruments.
Read together, those two say something practical. Once your raw total is settled the mark out of 15 is close to mechanical, so there is no band judgment to argue with and no descriptor to write towards. The whole contest is over individual raw marks, question by question.
Where our own data lines up with QCAA's advice
QCAA publishes nothing at the level of a single dot point. We do, and two caveats come with it. Our questions are external examination questions, not school-written internal ones, and all of those papers predate the current syllabus. So the figures below say something about the content your paper will sample and nothing about this instrument.
On the Unit 3 side that IA2 draws from, the weakest outcome in our sample is modelling displacement, force, velocity and relative velocity at 30.9% across 427 marked AusGrader attempts, followed by using De Moivre's theorem for integral powers at 47.0% across 546. The first of those corroborates QCAA's 2024 advice about extending vector concepts from two dimensions to three, which is also what one of the marker annotations above is about. Proving complex number identities involving modulus and argument sits at 53.8% across 876 attempts, against QCAA's 2025 advice to practise algebraic manipulation of complex numbers in both forms.
On the Unit 4 side that IA3 draws from, implicit differentiation is one of our two weakest outcomes in the subject at 32.6% across 263 attempts, and motion of a body in non-equilibrium situations under concurrent forces is 39.5% across 409. Statistical inference is the topic that costs the most marks anywhere in the subject, carrying more of the paper than any other topic at an average of 53.9%, which lines up with QCAA naming the comparison of the distributions of and , and the notation for their means and standard deviations, as a practice to strengthen in both 2024 and 2025. The full per-topic and per-outcome picture is on the Specialist Mathematics study guide.
Common mistakes
Preparing for a band instead of for marks. Two students whose papers match the same descriptor word for word can still finish a mark apart. Nothing in the ISMG prose is reachable; only the raw total is.
Solving a cued question your own way. If the wording names an approach, the scheme was written to it.
Treating a three-dimensional problem as flat. Two of the six annotated errors above are exactly this, and both are recoverable in seconds by sketching the axes before starting.
Answering a technology-active question with a bare number. The marks on that half of the paper are for setting up and interpreting, and neither is visible in a result on its own.
Writing a calculator method on the technology-free half. Nothing in the scheme can pay for a route the mode does not allow.
Revising three topics and hoping. The paper samples three of five and you find out which three on the day.
Assuming subsets of the complex plane are still IA2 material. They moved to Unit 2 for 2025, so they now reach your internal exam only as assumed knowledge.
In the room
There is no draft to check in advance, so run this as a per-question routine:
- Has the question named an approach, and am I using the one it named?
- Which technology mode is this, and does my working match it?
- If this is a vector problem, is it two-dimensional or three, and do my unit vectors match the orientation I set up?
- If this is an induction proof, have I written all four stages, including the assumption statement and the conclusion?
- Have I carried every constant through the integration, including the ones outside the integral?
- On an Argand region, have I marked the boundary as included or excluded, by convention or in words?
- If a value two lines up is wrong, is the working after it legible enough for a marker to carry the error through?
- How many marks is this question worth, and has my response given the scheme that many separate things to tick?
Three topics of five is an argument for breadth, and the cheapest way to find the two you are thinnest on is to attempt questions you have not seen from all five, which is what the Specialist Mathematics question bank holds. Read the IA1 problem-solving and modelling task next: it runs on the opposite principle, with four criteria and readable descriptors deciding 20 marks, and it is the only internal assessment you can plan against in advance. The rest of the subject is on the Specialist Mathematics study guide, and the QCE hub has the other subjects.
Frequently asked questions
Can a Specialist Mathematics IA2 question test Mathematical Methods content?
It can lean on it, but it cannot be built on it. The syllabus names Mathematical Methods as assumed knowledge for both internal examinations: Units 1 and 2 for IA2, Units 1, 2 and 3 for IA3. So a Methods technique can appear inside a Specialist question. QCAA has told schools in all three of the 2023, 2024 and 2025 rounds to avoid standalone questions that only test assumed knowledge, and its 2025 wording tells IA writers to avoid reliance on Physics contexts or Mathematical Methods subject matter.
What does it mean when a Specialist Mathematics IA question says to use vector calculus?
That the marking scheme is built around that route and no other. QCAA has told schools to cue the expected approach in every one of the last three rounds, using phrases like using a vector calculus approach or definite integration using technology, because several of these problems can also be solved with a physics formula that sits outside the Specialist Mathematics syllabus. Take the cue literally: the school's per-question mark breakdown has been written to the approach it names.
Does exactly 80% get 12 or 13 in Specialist Mathematics IA2?
Twelve. Every cut-off on the ladder is strictly greater than, and QCAA prints this case in the 2023 subject report as effective marking practice: 80% was awarded 12 marks out of 15, because 80% is not greater than 80%. Your percentage is also not rounded before it is read off the ladder.
Which topics can Specialist Mathematics IA2 cover in 2026?
Any three of the five Unit 3 topics: further complex numbers, mathematical induction and trigonometric proofs, vectors in two and three dimensions, vector calculus, and further matrices. Under the 2019 syllabus the unit held three topics and the paper had to sample all of them, so the change cuts both the predictability and the number of topics you can afford to skip.
Do I lose everything after an early error in a Specialist Mathematics IA3 question?
No, provided the marker can follow what you did next. Follow-through credit is expected practice and QCAA prints marked examples of it: a friction-on-an-incline response scored 3.5 out of 4 with the error labelled and the later work marked as carried, and a Newton's law of cooling response carries carried-forward ticks on its final lines. The credit attaches to visible working, so an answer with the middle steps missing has nothing to carry.
Sources
- Specialist Mathematics 2025 v1.0 General senior syllabus, QCAA, January 2024, pp. 12-13, 34-46. The IA2 and IA3 specifications and conditions, the question-specification table, the full 15-mark percentage cut-off ISMG, and the Unit 3 and Unit 4 topic lists.
- Specialist Mathematics 2019 v1.2 General Senior Syllabus, QCAA, August 2018, pp. 17-19, 31-39, 42-50. Superseded IA2 and IA3 conditions, the 120-minute working time, the all-topics sampling rule and the 2019 cut-off ladder, used for the before-and-after.
- Specialist Mathematics subject report, 2025 cohort, QCAA, January 2026, pp. 15-25. The 53/55 conversion, the endorsement and confirmation tables, the annotated IA3 excerpts showing half marks and carried-forward credit, and the rulings on cued approaches and misidentified complex unfamiliar items.
- Specialist Mathematics subject report, 2024 cohort, QCAA, January 2025, pp. 16-25. The alternative-solution and implied-mark rulings, the Argand-diagram boundary convention, the advice against splitting complex familiar questions into parts, and the 54/56 conversion.
- Specialist Mathematics subject report, 2023 cohort, QCAA, January 2024, pp. 20-29. The strictly-greater-than worked example, the annotated IA2 excerpts on particle collision and three-dimensional vectors, and two further percentage conversions.
- AusGrader marking data, QCE Specialist Mathematics, AusGrader. 14,641 marked attempts on questions from board external papers, by self-selected AusGrader users, used only where it corroborates a QCAA finding. Internal assessments and school-uploaded papers are excluded, and the pool carries SCSA and VCAA questions mapped to this syllabus alongside QCAA's own. Not a QCAA cohort.
Syllabus and assessment material referenced in this guide is © State of Queensland (Queensland Curriculum and Assessment Authority), licensed under CC BY 4.0. See our QCAA licensing notice. AusGrader is an independent study tool and is not affiliated with, endorsed by, or operated by the QCAA.
Keep reading
Practise what you just read
Work through real QCAA Specialist Mathematics questions and get your written answers marked against the official criteria, instantly.