How is IA3 different from IA2?

IA2 uses primary data you collect yourself in an experiment. IA3 uses secondary evidence you research about a claim your teacher gives you, and no data collection is involved. The criteria are grouped differently too: IA2 keeps Forming and Finding separate but combines Interpreting and Evaluating into one criterion, while IA3 does the opposite, combining Forming and Finding but keeping Interpreting and Evaluating as two separate criteria.

Do my IA3 sources have to be peer-reviewed journal articles?

No. The 2025 syllabus also accepts books and podcasts by well-credentialed scientists, "popular" science websites or magazines, and websites of governments, universities, independent research bodies or science and technology manufacturers. Podcasts and popular science sources are new additions in the 2025 syllabus; the 2019 list was narrower.

What's the most commonly missed requirement in IA3?

Extrapolating your findings back to the claim, not just answering your research question. QCAA's own top-band description for Evaluating asks for findings extrapolated to address the research question and the claim. A conclusion that stops at the research question, however well argued, has not done what that criterion is asking for.

QCE Physics IA3 research investigation: how to get top marks

Q: What is the QCE Physics IA3 research investigation?

It's a research investigation worth 20% of your Physics result, marked out of 20 across four criteria of five marks each: Forming and Finding (merged into one criterion), Analysing, Interpreting, and Evaluating. You evaluate a claim relevant to Unit 4 using secondary evidence you research yourself rather than data you collect, and it assesses science inquiry skills together with Science as a Human Endeavour subject matter. It takes roughly 10 hours of class time.

Q: Can I work in a group for IA3?

For three steps only, under the 2025 syllabus: selecting the claim, identifying the relevant concepts, and conducting the research. Everything after that, including your analysis, interpretation, evaluation and write-up, has to be your own individual work.

Your IA3 is worth 20% and is marked out of 20, five marks each across four criteria: Forming and Finding (one merged criterion), Analysing, Interpreting, and Evaluating. Interpreting and Evaluating are separate here, which is the opposite of IA2. You do not run an experiment. You evaluate a claim relevant to Unit 4 using secondary evidence you research yourself, and the marks reward how well you weigh that evidence against the claim, not how many sources you list.

This guide describes the Physics 2025 v1.2 syllabus, published October 2024, which applies to Year 12 students from 2026. Advice that mentions a Communication criterion, or criteria called Research and planning, Analysis and interpretation, or Conclusion and evaluation, was written for the superseded 2019 syllabus and those names no longer exist for your assessment.

What the task actually asks

You start from a claim your teacher provides from a list relevant to Unit 4, pose a research question addressing an aspect of it, then gather secondary evidence from scientifically credible sources, analyse and interpret that evidence, and evaluate the claim against it. It takes roughly 10 hours of class time. Your response is written up to 2000 words, or multimodal (at least two modes delivered together) up to 11 minutes.

IA3 assesses science inquiry skills and Science as a Human Endeavour subject matter. That is worth knowing even though SHE content never appears on your exam: it means the historical or societal context around your claim is not a detour from the task, it is part of what Forming and Finding is marked against.

Your claim comes from Unit 4, so it will sit inside special relativity, quantum theory or the Standard Model. Reading around the topic your claim belongs to before you start researching makes it far easier to tell which sources are actually relevant.

What can be done as a group, under the 2025 syllabus: selecting the claim, identifying the relevant concepts, and conducting the research. Everything after that, the analysis, interpretation, evaluation and write-up, is yours alone. If you and a classmate researched together but hand in near-identical evaluation sections, that is exactly the problem the individual-work requirement exists to catch.

How it's marked

IA2 and IA3 group the same five inquiry stages differently. For IA3 specifically, the 2025 ISMG runs:

Criterion	Objectives	Top band (4-5)
Forming and Finding	1, 2, 6	a considered rationale identifying clear development of the research question from the claim; a specific and relevant research question; selection of sufficient and relevant sources; appropriate use of genre conventions; acknowledgment of sources through appropriate referencing
Analysing	3	identification of sufficient and relevant evidence; thorough identification of relevant trends, patterns and relationships in the evidence; thorough and appropriate identification of the limitations of evidence
Interpreting	1, 4	justified scientific argument(s); a justified conclusion linked to the research question; fluent and concise use of scientific language and representations
Evaluating	5	justified discussion of the quality of evidence; extrapolation of credible findings of the research to the claim; suggested improvements and extensions that are considered and relevant to the claim

Notice Analysing asks only for the limitations of evidence, not uncertainty. There is no primary data in IA3, so uncertainty calculations, error bars and trendlines, the vocabulary that decides IA2's Analysing criterion, do not apply here. Writing about "uncertainty in the data" in an IA3 report is answering the wrong assessment.

If you have already done IA2, one contrast is worth holding onto: IA2 keeps Forming and Finding as two separate criteria and merges Interpreting with Evaluating. IA3 does the reverse: it merges Forming with Finding and keeps Interpreting and Evaluating separate. It is easy to apply backwards if IA2 is the one you learned first.

What counts as a credible source

The syllabus is more permissive than most students assume. Credible sources include:

books and podcasts by well-credentialed scientists
"popular" science websites or magazines
websites of governments, universities, independent research bodies or science and technology manufacturers
scientific journals

Podcasts and popular science sources are 2025 additions. QCAA says it widened the list deliberately, "to direct students to a wider variety of student-accessible sources."

What earns top marks, criterion by criterion

QCAA's own subject reports define its top-band adjectives more concretely than the syllabus does. All four are near-identical across the years held.

Rationale. A top-band rationale "clearly breaks the claim into its constituent parts and outlines the logical development from the claim to the research question." A rationale that summarises the general topic your claim sits in, without showing that working, stays in the middle band.
Research question. Specific and relevant research questions "include measurable independent and dependent variables that allow for a response related to the appropriate subject matter." A measurable threshold, a named unit or a stated time window is what turns a broad question into a specific one.
Analysis. Relevant evidence has to be drawn "from sufficient sources to enable thorough identification of the trends, patterns and relationships between them", and arguments are justified "by linking the relevant trends, patterns and relationships found within the evidence to support discussions." Comparison across sources is the mark, not the length of any one source's summary.
Quality of evidence. A justified discussion "includes verifying the credibility, bias, relevance, accuracy and recency of multiple sources", and should acknowledge the limitations in the methods and data behind each source, not just the source's reputation.
Evaluating the claim. This is the single most-missed requirement in the criterion: findings have to be "extrapolated to address not only the research question but also the claim, using credible evidence discussed throughout the investigation." Answering your research question is not the same as evaluating the claim.
Improvements and extensions. Considered improvements "address the identified limitations", while considered extensions "link to the experiment's validity." Both need to trace back to something specific your analysis found, not a general wish for more research.

Common mistakes

A research question that restates the claim. If your claim and your research question could be swapped without anyone noticing, Forming and Finding has nothing to show for the "development" it is marked on.

Treating limitations as uncertainty. IA3's Analysing criterion asks for limitations of evidence, not the error-bar and trendline language IA2 uses. There is no primary data to attach uncertainty to.

Stopping at the research question instead of the claim. See the callout above. This is the gap between a strong Interpreting response and a strong Evaluating one.

Listing source strengths with no weaknesses, or the reverse. QCAA's top-band discussions weigh credibility, bias, relevance, accuracy and recency together. A source list that only says what each source got right, or only what it got wrong, has not evaluated quality.

Letting your research drift outside Unit 4. The 2024 subject report notes the specific temptation directly: students researching nuclear physics claims can be pulled toward content that is "more closely aligned to Unit 2 subject matter." Your subject guide has the four-year trend on how often this costs schools their endorsement; for your own draft, the check is simpler: does every claim, concept and source you use trace back to Unit 4?

Doing more of the task as a group than the syllabus allows. Selecting the claim, identifying concepts and researching can be shared. Writing near-identical analysis or evaluation with a classmate undermines the individual-work requirement on everything after that.

Improvements or extensions that are vague wishes. "More research is needed" says nothing. A strong improvement names what it changes and why: sourcing more recent data because newer measurement methods carry smaller error margins, for instance, ties directly back to a limitation the analysis actually found.

Weak versus strong wording

These pairs are drawn from real IA3 responses QCAA reproduced and annotated in its subject reports, so the right-hand column is what a genuine top-band response actually did.

Weak	Strong, from a real marked response
A research question with no measurable threshold	"Does GPS technology compensate for the impact of time dilation due to special relativity on its accuracy, and achieve aircraft positioning accuracy within 10 metres?"
A research question with no units or timeframe	"How does special relativity affect the clock drift (microseconds) and vertical positioning accuracy (metres) of the global positioning system (GPS) over the course of 24 hours?"
A rationale that summarises the topic	A rationale that names the two postulates of special relativity, distinguishes it from general relativity as a competing cause of clock drift, sets a threshold, then narrows to aircraft GPS because "GPS error could compromise flight safety"
Evidence described but not computed	Taking a cited GPS mean velocity and substituting it into $t = \frac{t_0}{\sqrt{1 - v^2/c^2}}$ with $t_0 = 86400$ s to report an error of $-7.2138$ μs per day
"The source is reliable because it's a journal"	Naming the journal, its peer-review standard, the authors' standing, the publication date, and arguing why an older publication is still current because no major technological breakthroughs have occurred since
A strengths list with no weaknesses	A two-column quality-of-evidence table: strengths including that sources corroborate each other, weaknesses including a named margin of error that "majorly limits its accuracy" in one source
A conclusion that answers the research question and stops	Answering the research question, then extrapolating explicitly to the claim: naming the shared calculation method across other systems and stating the claim "can be extrapolated" and "accepted" on that basis
An improvement that is a wish	An improvement of "modern data", justified because sourcing from more recent reports means the equipment used likely carries smaller error margins

Before you submit

Does your research question develop clearly from the claim, with a measurable variable, unit or timeframe, not just a restatement of it?
Does your rationale show the logical steps from claim to research question, rather than summarising the general topic?
Have you drawn trends, patterns and relationships across multiple sources, not just summarised each one in turn?
Have you discussed the credibility, bias, relevance, accuracy and recency of your sources, rather than asserting one is reliable because of its type?
Does your conclusion extrapolate back to the original claim, not just answer your research question?
Does every improvement or extension trace to a specific limitation your analysis found?
Have you kept your claim, concepts and sources inside Unit 4, and kept the analysis, interpretation, evaluation and write-up as your own individual work?

For the assessment that comes before this one, see the IA2 student experiment guide. For how IA3 fits against the other three assessments, including how QCAA's confirmation of school marking has moved over four years, head back to the QCE Physics subject guide, or the QCE Physics subject page for real marked questions on Unit 4 topics.

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.