QCE Psychology IA3 research investigation: how to get top marks

Your IA3 is worth 20% and is marked out of 20, five marks each across four criteria: Forming and Finding, Analysing, Interpreting and Evaluating. You collect no data. You evaluate a claim relevant to Unit 4 using secondary evidence you research, and the marks go to how well you narrow that claim into an answerable question, compare evidence across sources, and then say what your findings do and do not prove about the claim you started with.

This guide describes the Psychology 2025 v1.0 syllabus, which applies to Year 11 from 2025 and Year 12 from 2026. Any advice you find mentioning a Communication criterion, or criteria named Research and planning, Analysis and interpretation, or Conclusion and evaluation, was written for the superseded 2019 syllabus and those names no longer exist. The student excerpts quoted here come from the 2022 to 2025 subject reports, all marked under those old criteria, but they were chosen for what the students did with evidence, which the renaming does not touch.

What the task actually asks

You evaluate a claim relevant to Unit 4, chosen from a list your teacher provides. Unit 4 is social psychology, interpersonal processes, attitudes and cross-cultural psychology, so your claim will sit in one of those. Your job is to turn it into a research question you can actually answer with published evidence, then reach a defensible position on what that evidence says about the claim.

The syllabus also flags this instrument as the one that assesses science as a human endeavour subject matter alongside inquiry skills.

Conditions. Around 10 hours of class time, up to 2000 words written or up to 11 minutes multimodal. All six assessment objectives are assessed.

What can be done as a group under the 2025 syllabus: selecting the claim, identifying the relevant scientific concepts, and conducting the research. This is new. Under the 2019 syllabus students worked individually throughout, and the 2024 report told schools to write authentication strategies for individual work only "as group work is not an aspect of IA3". That advice is now superseded. Everything after the research, the analysis, the interpretation and the write-up, is still yours alone.

How it's marked: the four criteria

Communication is not a criterion here. Referencing and scientific language are marked inside Forming and Finding, and genre conventions are marked inside Evaluating. Those two sit in different criteria, which is easy to miss if you are scanning for a single place where writing quality is judged.

IA2 and IA3 group the same five inquiry stages in opposite directions. IA2 splits Forming and Finding into two criteria and combines Interpreting and Evaluating into one. IA3 does the reverse. If you learned the IA2 arrangement first, see the IA2 student experiment guide, because the instinct it gives you points the wrong way here.

The claim, and narrowing it into a question

All four reports held ask for the same thing about claims: a claim is a single assertion. The 2025 report puts the tightest frame on it, asking for one broad independent variable and one broad dependent variable. QCAA's before-and-after, from the 2023 report, is the clearest version. A workable claim: "the presence of others affects the way we behave". The version to avoid packs several assertions into one sentence by saying the presence of others affects the way we think, feel and behave.

Claims QCAA has printed as good examples: "Social media is changing the nature of relationships", "Conformity can increase prosocial behaviour", "Aggression is a learned behaviour", and "socialisation influences antisocial behaviour". Claims it has flagged to avoid are value statements and colloquial sayings, one named example being "Do unto others as you would have them do to you".

Your school writes the list, so the part you control is the narrowing. That narrowing is the first thing the Forming and Finding descriptor asks about.

The 2024 report puts the target in one line: research questions should be "clearly developed to be distinct from but still relevant to one aspect of the claim" and specific enough to include "one clear independent and one clear dependent variable".

What counts as a credible source

The 2025 syllabus is more permissive than most students assume. Evidence must come from scientifically credible sources "such as: 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 websites and magazines are 2025 additions; the 2019 list had four items and included neither. QCAA describes the change as clarifying the nature of student-accessible credible sources. The 2022 report had already said the same thing under the old rules, that sources need to be scientifically credible "although not necessarily drawn exclusively from peer-reviewed journals".

Analysing: sufficient means across sources

The 2024 report is explicit that sufficient and relevant evidence "was selected from more than one credible source", and that thorough identification of limitations was "specific to the research question and/or claim". The 2022 report adds how sufficiency is judged: across the whole response, by how far the sources together address the research question.

So a single strong study cannot carry this criterion. The top band asks for sufficient and relevant evidence drawn from more than one credible source, with sufficiency judged across the whole response.

Interpreting: quality of evidence, and the conclusion

The conclusion in this criterion answers your research question. The claim is handled separately, in Evaluating.

Discussion of the quality of evidence is where responses most often go thin. The 2022 report asks for an insightful discussion using "specific features" and examining limitations "with respect to addressing the research question or the claim". The 2024 report asks for "discussion of the effect of limitations on the ability of the evidence to respond to the research question and the claim". Naming a limitation is not the discussion. What it does to the weight the evidence can carry is the discussion.

Strong responses also report the source that disagrees with them. A 2024 excerpt reads: "Both Eduardo et al., 2022 (p=0.016, p=0.048) and Hagal et al., 2015 (p<0.0001) indicate that HH compliance increases under overt human observation because of the Hawthorne effect with statistical significance. However, Wu et al., 2018 (p<0.001 and p=0.13) indicates that HH compliance increases under human observation due to the Hawthorne effect with partial statistical significance."

Evaluating: extrapolating to the claim

This is the message the reports return to every year, and the top band turns on a single verb.

Application uses your findings to answer your research question. Extrapolation says what those findings license you to conclude about the broader claim, and where that licence runs out. Four consecutive reports say some version of the same sentence. The 2024 wording is the bluntest: credible findings are "specifically extrapolated to the claim, beyond simply addressing the research question". The 2025 wording asks that findings address "not only the research question but also the claim".

Where schools and QCAA disagree most

Across all four cohorts held, the least reliably marked of the four IA3 criteria was the one covering conclusions and evaluation, called Conclusion and evaluation under the 2019 marking guide. School judgments on it agreed with QCAA on 87.27% of samples in 2022, 90.32% in 2023, 88.43% in 2024 and 89.87% in 2025, and the disagreement was almost entirely downward: between 9% and 12.12% of marks were reduced each year, against 0% to 0.84% increased. Overall IA3 agreement across those years ran between 83.03% and 85.65%.

All four cohorts were assessed under the superseded marking guide, and QCAA has published no criterion-by-criterion mapping to the 2025 set. Read the finding as being about conclusions and evaluation as work, and leave the 2025 criterion names out of it. What it tells you is where the generous marking happens, and it is the same part of the response the reports have written advice about four years running.

The Psychology study guide covers how confirmation moves marks across all three internal assessments.

Common mistakes

Concluding on the claim instead of the research question. Your conclusion in Interpreting addresses the question you set. The claim belongs in Evaluating, reached by extrapolation.

Stopping at application. Answering your research question well and going no further is the 2 to 3 band in Evaluating, however good the answer is.

Extrapolating without limits. The reverse error. A response that jumps from one narrow finding to a full endorsement of the claim has not evaluated anything. The strongest exemplars in the set all say how far the evidence reaches and then stop.

Building on one source. Sufficiency is judged across the whole response, and a conclusion resting on a single study has nothing to corroborate it.

Discussing source credentials instead of evidence quality. That an article is peer-reviewed is not a discussion of what its limitations do to your question.

Treating self-report as a throwaway limitation. Naming it costs nothing. The mark is in explaining the bias it invites and what would measure the same thing better.

Offering only an improvement or only an extension. Both are needed for the top band, and they are different moves.

Treating the whole task as collaborative. Selecting the claim, identifying the concepts and researching can be shared. The analysis, the interpretation and the write-up cannot.

Weak versus strong wording

The right-hand column is student work QCAA reproduced and annotated in its subject reports.

Before you submit

• Does your rationale show the research question growing out of the claim, with both variables narrowed?
• Does your research question name one clear independent and one clear dependent variable?
• Have you drawn evidence from more than one credible source, and do those sources together cover the question?
• Does your conclusion answer the research question, leaving the claim for Evaluating?
• Does every limitation say what it does to the evidence's ability to address your question, not just that it exists?
• Have you extrapolated to the claim, and said where that extrapolation stops?
• Have you given both an improvement and an extension, with the extension reaching an aspect of the claim your question did not?
• Is your referencing consistent, with every source acknowledged?

For the other 20% of your internal marks, the IA2 student experiment guide covers working with data you collect yourself. The Psychology study guide covers how all four assessments fit together.