A critique of an article, which many courses call a critical review, does two jobs: it tells the reader briefly what a journal article did, then judges how well it did it. The judgement earns the marks. Every guide we read puts a short summary before the evaluation; they disagree on how long it should be.
EdCitation publishes this guide. On 26 September 2026 we read seven universities' guides, from Australia, Britain, Canada and the United States, and the 2024 checklist for randomised trials from the Critical Appraisal Skills Programme (CASP). The template and the worked example, a critique of a real open-access trial read in full, are ours.
EdCitation never writes, rewrites or edits any part of a critique: the judgement stays yours, which is the point of the exercise. It does the checking around it. Cite a source builds the reference for the article you critique from the publisher's record and screens it for retraction, and Verify references checks that article and every other source you cite, free.
What is a critique of an article?
A critique is a reasoned judgement of an article's strengths and weaknesses, backed by evidence from the article and from other work. University of Hull Library (2026) frames the task as deciding whether the article achieved what it set out to do and whether it can be relied on. University of New South Wales (n.d.) adds that being critical means questioning, not criticising, and that a fair review draws on related texts.
The terms overlap. Federation University Australia (n.d.) says a critical review is sometimes called a "summary and critique". The University of North Carolina at Chapel Hill Writing Center (n.d.) says a class critique is written for your lecturer, to show close reading and a grasp of the field; books have their own guide.
Summary and critique are different jobs
The summary reports; the critique judges. University of Kent (2025) gives a review two purposes: a descriptive one, telling the reader what the text covers and concludes, and a critical one, judging its quality, meaning and significance. Federation wants the summary dispassionate, clear to someone who has not read the article, with the author's ideas in the present tense and findings in the past.
Critical writing vs descriptive writing covers the difference at sentence level.
How is an article critique structured?
Most guides give five parts: the full reference, an introduction, a summary, the critique and a conclusion, then a reference list if you cited other work.
| Part of the critique | What goes in it | Source |
|---|---|---|
| Heading | The article's full reference, in your course's style; build it with Cite a source | Queen's; Federation (usually APA 7); Guelph |
| Introduction | The topic, the article's aim and main findings, and your overall verdict: positive, negative or mixed | UNSW; Federation |
| Summary | The key claims, most to least significant, reported without judgement | Federation; UNSW |
| Critique | Strengths and weaknesses judged against named criteria, with other sources where they help | UNSW; Guelph; Kent |
| Suggestions | How the research could have been done better | UNSW; Federation; UNC |
| Conclusion | Your overall judgement restated, and the article's significance | UNSW; Guelph; Federation |
| References | Every other source you cited; check them with Verify references | UNSW; Guelph |
Where the universities disagree
They differ on five points; your brief outranks them all.
- How much summary. University of New South Wales (n.d.) says about a third of the review, and in another paragraph of the same page a quarter to a third. Federation makes summary and critique roughly equal. Kent asks only for a short summary.
- Whether to summarise separately at all. The Queen's University Library (2026) guide, prepared by the University of Toronto Mississauga library, builds the review around a thesis about the article, with no summary section. One University of Guelph Library (2026) approach sets the article in its research context instead of summarising it.
- Quotations. Queen's tells you to quote the article to illustrate your points; Federation asks for no quotations and no headings at all. UNSW calls headings optional.
- The order of the critique. UNSW lets you order points by importance, by balance, or one criterion to a paragraph. Guelph follows the article's own sections: introduction, methods, results, discussion. Kent puts strengths then weaknesses in a short review, and goes section by section in a long one.
- Mostly positive or negative. UNSW and Federation agree: with a mostly positive verdict, give the negatives first and end on the positives; reverse it for a mostly negative one.
How long should each part of a critique be?
Few guides give lengths, and none a word count for each part. Read your brief first; Check your paper reads its word limits and required headings into a checklist, each shown with the sentence it came from.
- The whole critique. Federation says usually around 800 to 1,000 words or more. Kent's simple structure is for a short review of about 500 to 1,000 words, and changes above 1,000. UNSW's pattern covers anything from one page to more than four.
- The introduction. One paragraph for a journal article, two or three for a book (UNSW).
- A one-page critique. One paragraph of strengths and one of limitations (UNSW).
- The conclusion. As short as possible; Federation says a review of 800 words may not need one.
Our rule: take the summary share from your brief; if it is silent, a third sits in the middle of what the guides say.
How do you critique a research paper? The questions to ask
Work through a published appraisal checklist for the paper's design, then turn the answers into judgements. For a randomised controlled trial, the Critical Appraisal Skills Programme (2024) checklist has eleven questions in four sections: whether the basic design is valid, whether the study was methodologically sound, what the results are, and whether they will help locally. Each takes yes, no or can't tell. CASP warns never to assume what the researchers did: if the paper does not say, answer "can't tell", and read a paper with many of those with caution. CASP has checklists for other designs too; critical appraisal in our glossary names other tools.
Hull groups its questions under credibility (should you trust it?) and robustness (is it good research?), adapted from two 2007 British Journal of Nursing papers by Coughlan, Cronin and Ryan, which we have not read. Our questions below draw on CASP, Hull and UNSW.
Method and sample
- Is the question stated in terms of who was studied, what was done, what it was compared with and what was measured?
- Was allocation random, and was it concealed from those enrolling participants? Who was blinded?
- Who was recruited, and how? Would the people who refused or were excluded differ from those who took part? That is selection bias.
- Was there a sample size calculation, and did its assumptions hold?
Analysis and precision
- Was everyone who entered the study accounted for, and analysed in the group they were assigned to (intention to treat)?
- Are the numbers consistent between the abstract, the tables and the text?
- Is each effect reported with a confidence interval, or only a P value?
Claims against evidence
- Does the conclusion claim more than the results show?
- Would the results apply to other people or settings (external validity)? This is CASP's tenth question.
- Was the study registered in advance, and has the article since been corrected or retracted?
A structure template for a critique (ours)
Ours, built from the guides above; your brief overrides it.
- Heading. The article's full reference. Paste its DOI into Cite a source and choose your course's style.
- Introduction, one paragraph. The topic, what the article set out to do, what it found, and your verdict in one sentence.
- Summary. The design, the sample, the main results with their figures, and the authors' conclusion, in that order, with no judgement.
- Critique, one criterion to a paragraph. The judgement first, then the evidence from the article (a table, a number), then what it means for the conclusion.
- Suggestion. What would have made the study stronger; for a published article UNC advises the past conditional (it would have been stronger if the authors had...).
- Conclusion. Your verdict again, and what the article is still good for.
- References. Every source you cite, run through Verify references before you hand in.
Article critique example: the PARACHUTE trial
The article is Yeh et al. (2018), a randomised trial of parachutes in The BMJ's 2018 Christmas issue, open access under a CC BY-NC licence, read in full on PubMed Central. It is satire with a serious point, reported in the standard form, so it critiques well. The critique is ours; every figure is from the article.
Introduction. Yeh et al. (2018) set out to test whether parachutes prevent death or major injury when people jump from aircraft, and found no difference between a parachute and an empty backpack. This review argues that the trial's internal methods are largely sound, and that its result cannot be applied to any real jump, which the authors themselves make plain.
Summary. Between September 2017 and August 2018 the investigators screened 92 adults on aircraft; 23 were willing and eligible and were randomised, most of them members of the study team. The primary outcome, death or major traumatic injury within five minutes of landing, occurred in no participant in either group, and none had occurred at 30 days. Participants jumped from a mean altitude of 0.6 m at 0 km/h; those screened but not enrolled were on jetliners at a mean of 9,146 m. The authors conclude that only people at almost no risk could be enrolled, and that trials of treatments already believed to work may enrol those least likely to benefit.
Critique. The trial's structure is its strength: the question names its population, intervention, comparator and outcome; allocation used block randomisation stratified by site and sex; the analysis was by intention to treat; and a composite of death and major injury is hard to bias. Its weaknesses are reported by the authors themselves: allocation was not concealed from the investigator who assigned treatment, nobody was blinded, and the trial was not registered in advance. More serious is precision. The sample size assumed jumps from 4,000 m and a 99% event rate without a parachute; with no events in either group, no confidence interval could be calculated for the primary outcome, so the trial shows only that no difference was seen here, not that parachutes do not work. The reporting also needs care: Table 1 gives 11 participants with a parachute and 12 with a backpack, while the Results text gives 12 and 11. Neither version changes the outcome, and a reader cannot tell which is right. Finally, the trial fails CASP's tenth question outright: nobody at 0.6 m and 0 km/h resembles a person who needs a parachute. Yet the evidence for that verdict comes from the authors' own comparison of participants with those screened, a comparison they recommend other trials report.
Conclusion. The trial is sound in procedure and, by design, of no use for a real jump. It would have answered its question had participants been enrolled at altitude, which, the authors suggest, belief in parachutes would make hard to achieve. Its lasting value is the lesson its authors draw, that interpreting a trial takes "more than a cursory reading of the abstract".
The example states its verdict first, keeps the summary free of judgement, lets one checklist question decide, and reports an inconsistency without guessing which figure is right. A course critique would also set the trial beside the 2003 BMJ review by Smith and Pell that it answers; we could not retrieve that paper's full text, so this example leaves it out. For the reading before a critique, see how to read an academic paper fast.
How EdCitation checks the article you critique and your references
EdCitation is the right tool for the checking a critique needs, because it looks each reference up in the publisher's record and never writes one, and it flags a retracted paper. A critique of a withdrawn paper that never mentions the retraction fails however good its judgement; retracted papers in your reference list says what to do.
Cite a source: the heading reference, and its record
From the DOI 10.1136/bmj.k5094, Cite a source returned this APA 7 entry, from Crossref's record, on 26 September 2026:
Yeh, R. W., Valsdottir, L. R., Yeh, M. W., Shen, C., Kramer, D. B., Strom, J. B., Secemsky, E. A., Healy, J. L., Domeier, R. M., Kazi, D. S., & Nallamothu, B. K. (2018). Parachute use to prevent death and major trauma when jumping from aircraft: Randomized controlled trial. BMJ, 363, Article k5094. https://doi.org/10.1136/bmj.k5094
APA 7 lists up to 20 authors, so all eleven appear. The record showed no retraction and one correction, published in 2018 as 10.1136/bmj.k5343. We read it: it amends the Methods' last sentence, on the threshold for statistical significance. PubMed Central's copy already carries the corrected sentence.
Verify references: the whole list
We pasted this guide's nine references into Verify references. All nine came back verified: the trial with "The DOI resolves to this record and the title matches.", each university page with "The page is at this address, and its title matches.", and the CASP checklist, a PDF, with "The address opens a PDF whose first pages carry the entry's title." A source that cannot be checked is reported as such, never as not found.
Beyond the free tools
With Pro, $8 a month, References from a file checks every reference in the uploaded critique, matches each in-text citation to the list, and links published corrections. With Max, $24 a month, Theoretics QA judges the finished critique against your brief's own instructions, with quotations from your paper; it never rewrites anything. See pricing. No tool can promise a mark.
Quick questions
What is the difference between a summary and a critique?
A summary reports what the article did and found, without judgement. A critique judges how well it did it, against criteria such as method, sample and whether the conclusions follow from the results.
How long should the summary be in a critical review?
About a third of the review, says UNSW, or a quarter to a third; Federation University makes summary and critique roughly equal. Your assignment brief decides.
Which checklist should I use to critique a research paper?
Use the checklist for the study's design: CASP publishes one each for randomised trials, cohort, case-control, cross-sectional and qualitative studies, and others. Its 2024 trial checklist has eleven questions.
Can EdCitation write or improve my critique?
No. EdCitation never writes, rewrites or edits any part of a critique. It builds the article's reference in Cite a source, checks it and your other sources in Verify references, and reads your brief into a checklist in Check your paper, free.
References
- Critical Appraisal Skills Programme. (2024). CASP checklist: For randomised controlled trials (RCTs) [Checklist]. https://casp-uk.net/casp-checklists/CASP-checklist-randomised-controlled-trials-RCT-2024.pdf
- Federation University Australia. (n.d.). Critical reviews. Study skills. https://www.federation.edu.au/students/study-skills/study-resources/assessments/critical-reviews/
- Queen's University Library. (2026, August 5). Critical reviews. Introduction to research: Humanities and social sciences. https://guides.library.queensu.ca/introduction-research/writing/critical
- University of Guelph Library. (2026, June 8). Writing your critique. Write a critical review of a scientific journal article. https://guides.lib.uoguelph.ca/c.php?g=708045&p=5041273
- University of Hull Library. (2026, September 9). Journal critique. Other assessments. https://libguides.hull.ac.uk/other/critique
- University of Kent. (2025, September 29). Critically reviewing books and articles. https://student.kent.ac.uk/studies/literature-reviews/critical-book-and-article-reviews
- University of New South Wales. (n.d.). Critical review. UNSW Current Students. https://www.unsw.edu.au/student/managing-your-studies/academic-skills-support/toolkit/writing/critical-review
- University of North Carolina at Chapel Hill Writing Center. (n.d.). Writing critiques. https://writingcenter.unc.edu/esl/resources/writing-critiques/
- Yeh, R. W., Valsdottir, L. R., Yeh, M. W., Shen, C., Kramer, D. B., Strom, J. B., Secemsky, E. A., Healy, J. L., Domeier, R. M., Kazi, D. S., & Nallamothu, B. K. (2018). Parachute use to prevent death and major trauma when jumping from aircraft: Randomized controlled trial. BMJ, 363, Article k5094. https://doi.org/10.1136/bmj.k5094