An article critique is a short paper that summarizes a published article and then judges how well it makes its case: whether the methods fit the question, whether the evidence supports the conclusions, and what the work adds. Your instructor wants to see that you can read a source as a careful, fair skeptic, not just report what it says.
Below you'll find the standard format, critical questions for each kind of article, APA and MLA setup, and a complete article critique example built on a real open-access study you can read yourself.
An article critique is an evaluation of one article that explains its argument, weighs its strengths and weaknesses, and reaches a reasoned verdict. The University of Arizona Global Campus Writing Center's guide to article critiques says the task is to read a piece of research critically and "identify and evaluate the strengths and weaknesses of the article." A critique can include a brief summary, but your evaluation is the main event.
Massey University's article critique guide breaks the job into three verbs: identify the article's purpose and main argument, evaluate how convincing it is, and respond with your own assessment. It also asks you to do this "in an objective and balanced way." So critique does not mean criticize. You can conclude that an article is excellent, as long as every judgment rests on evidence from the text.
The easiest way to go wrong is to write a summary instead of a critique. The table shows the difference, with sample sentences about the study critiqued later on.
| Summary | Article Critique | Literature Review | |
|---|---|---|---|
| Question it answers | What does the article say? | How good is this article, and why? | What do many studies show when read together? |
| Your judgment | None | Central: your thesis is a verdict | Present in how you group and weigh studies |
| Sample sentence | Kross et al. (2013) found that Facebook use predicted later dips in mood. | Because Facebook use was measured with one self-report slider, the study cannot separate passive scrolling from active messaging. | Early studies of Facebook and well-being relied on one-time surveys; later work added daily sampling and randomized experiments. |
If your course also asks for an annotated bibliography, its evaluative annotations use the same skill in compressed form.
To write an article critique, read the article twice, test it against a set of critical questions, keep the few judgments that matter most, and build the paper around one clear verdict. Here is the process step by step.
Before you open the article, find out exactly what your instructor is grading. Check:
If you choose the article yourself, pick one that meets the stated criteria and has a methods section you can follow.
On the first pass, read like an ordinary reader. Focus on the abstract, the end of the introduction, where the research question usually sits, and the start of the discussion. Then close the article and write two sentences: what the authors asked and what they concluded. If you can't, read it again before you start judging it.
On the second pass, read like a reviewer. Annotate as you go:
Use the question bank in the next section. Keep a separate "can't tell" list for things the article never makes clear, such as how participants were recruited. The Critical Appraisal Skills Programme (CASP) checklists advise reviewers to "never make assumptions about what the researchers have done." If you can't tell, the article wasn't explicit, and that gap is itself a fair point to raise.
You'll end up with more notes than space. Keep the two to four points that most affect whether the conclusions hold, and include at least one genuine strength. A problem with the sample or the main measure outranks an awkward sentence or a typo in a table, because it changes what the study can prove.
Your thesis should give your overall judgment and the main reasons for it in one or two sentences. A reliable pattern is: [Article] [succeeds or partly succeeds] in [aim] because of [strength], but [weakness] limits [what the reader can conclude].
For example: "Kross et al. (2013) use a strong real-time design to show that Facebook use comes before small dips in mood, but a small local sample and a one-item measure of Facebook use limit how far the results generalize." For more models, see this guide on how to write a thesis statement.
Build every critique paragraph from three moves:
Weak: "The sample was too small." Strong: "Because all 82 participants were recruited through flyers in one college town, the findings may not hold for young adults outside a college town." The second version names the detail and says why it matters.
On your last pass, check that the summary takes up no more than about a quarter of the paper, that every judgment is tied to evidence, and that your verbs match the study design (more on that below). Recheck each quotation against the article word for word, and cite it in your required style every time you quote or paraphrase.
A journal article critique should judge the article by the standards of its own type, because an experiment, an interview study and an opinion essay succeed or fail in different ways. Name the type first, then use the matching questions.
Empirical articles usually follow the introduction, methods, results and discussion layout. If that structure is new to you, this primer on empirical research explains each part. Use this section-by-section checklist:
| Section | Questions to Ask | Red Flags |
|---|---|---|
| Title and abstract | Does the title claim only what the design can show? Does the abstract state the sample, design and main result? | Causal verbs ("causes," "boosts") for a correlational study |
| Introduction | Is the research question clear? Is the gap in earlier research real and fairly described? | Only supportive studies cited; no clear question |
| Methods | Does the design fit the question? Who was sampled, how, and how many? How were key variables measured? Was ethics approval reported? | A convenience sample treated as representative; the main variable measured with one self-report item |
| Results | Do the analyses answer the stated question? Are effect sizes reported alongside p-values? Are dropouts and exclusions explained? | Unexplained exclusions; statistical significance treated as practical importance |
| Discussion | Do the conclusions stay within the data? Are limitations stated honestly? Are alternative explanations addressed? | Recommendations that reach far beyond the sample |
| Funding and references | Who funded the work? Are competing interests declared? Are the sources current and relevant? | Undeclared conflicts of interest; mostly outdated citations |
Interview, focus group and observation studies aim to explain experiences, not measure how common something is, so don't fault them for small samples or a lack of statistics. The CASP qualitative checklist groups its ten questions under three headings: "Are the results valid?", "What are the results?" and "Will the results help locally?" Its questions ask whether:
For a review, judge the search, not just the conclusions. Ask whether the authors named the databases, dates and search terms; set clear inclusion and exclusion criteria; assessed the quality of the studies they included; and explained how they combined results. A review that pools weak studies can only produce a weakly supported conclusion.
For non-research articles, focus on the argument. What is the central claim, and is the evidence for it current and sourced? Does the logic hold? Are counterarguments acknowledged? Do the author's expertise and the venue fit the claims? Judge the piece against its purpose: an op-ed is meant to persuade, so criticize weak evidence, not the fact that it takes a side.
For health research, the EQUATOR Network's library of reporting guidelines gives you a ready yardstick. It lists CONSORT for randomized trials, STROBE for observational studies, PRISMA for systematic reviews, and SRQR and COREQ for qualitative research. If an article leaves out an item its guideline expects, that omission is a specific, defensible point for your critique.
The standard article critique format has four parts: introduction, summary, critique and conclusion. That is the structure UAGC's Writing Center recommends, and it is the safe default unless your rubric sets a different one.
| Part | What to Include | Share of a 1,000-Word Critique |
|---|---|---|
| Introduction | Author(s), article title, journal and year; the article's purpose or main point; your thesis | About 100 to 150 words |
| Summary | Research question, method, sample and main findings or arguments, in your own words | About 150 to 250 words |
| Critique | Two to four paragraphs on strengths and weaknesses, each tied to specific evidence from the article | About 500 to 650 words |
| Conclusion | Your overall verdict, the article's significance, and what further research or revision would help | About 100 words |
Treat the word shares as a rule of thumb; the point is that evaluation fills at least half the paper. Inside the critique section, short papers can cover strengths and then weaknesses. For research articles, organizing by criterion (design, sample, measures, conclusions) works better, because each paragraph reaches its own mini-verdict.
According to the Purdue OWL's APA general format page, an APA 7 student paper uses 1-inch margins and double spacing throughout, including the title page. Recommended fonts include 12-point Times New Roman, 11-point Calibri, 11-point Arial and 11-point Georgia. A student paper's header holds only the page number, flush right.
The student title page lists the paper's title (bold, centered in the upper half of the page), your name, your institution, the course number and name, your instructor's name and the due date. Your reference list starts on a new page. Here is the reference entry for the article critiqued below:
Kross, E., Verduyn, P., Demiralp, E., Park, J., Lee, D. S., Lin, N., Shablack, H., Jonides, J., & Ybarra, O. (2013). Facebook use predicts declines in subjective well-being in young adults. PLOS ONE, 8(8), Article e69841. https://doi.org/10.1371/journal.pone.0069841
APA 7 lists up to 20 authors in a reference entry, and because this journal uses article numbers instead of page ranges, the word "Article" and the number take the place of pages. In your text, cite it as (Kross et al., 2013). Our APA style guide covers other source types.
The Purdue OWL's MLA general format page says to skip the title page unless your instructor asks for one. Put your name, your instructor's name, the course and the date in the upper left corner of the first page, then center your title. Add your last name and the page number in the upper right corner, half an inch from the top. Use 1-inch margins, double spacing and a legible 12-point font such as Times New Roman, and end with a Works Cited page. See our MLA style guide for Works Cited entries.
The article critique example below evaluates a real, open-access study, so you can read the original and check every point against it. The article is Kross et al. (2013), "Facebook Use Predicts Declines in Subjective Well-Being in Young Adults," published in PLOS ONE.
| Element | What the Study Reports |
|---|---|
| Question | Does Facebook use predict changes in how young adults feel moment to moment and how satisfied they are with their lives? |
| Participants | 82 people (mean age 19.52; 53 women) recruited through flyers in Ann Arbor, Michigan; each needed a Facebook account and a touch-screen smartphone and received $20 plus entry in an iPad raffle |
| Design | Experience sampling: five text messages a day between 10 a.m. and midnight for 14 days, plus the Satisfaction With Life Scale before and after |
| Measures | Five 0 to 100 sliders per text: current mood, worry, loneliness, Facebook use since the last text, and direct (face-to-face or phone) contact since the last text |
| Main findings | More Facebook use between texts predicted worse mood at the next text; mood did not predict later Facebook use; higher average use predicted lower end-of-study life satisfaction; direct contact predicted feeling better |
| Authors' caveats | Effects were "relatively 'small'"; mood was measured on a single bipolar scale; experiments are needed to establish causation; results may not generalize to other ages or networks |
Bold labels mark each part of the format. Notice that the summary is one paragraph and the evaluation takes up most of the paper.
Introduction. In "Facebook Use Predicts Declines in Subjective Well-Being in Young Adults," published in PLOS ONE in 2013, Ethan Kross and eight colleagues ask whether using Facebook changes how young adults feel from moment to moment and how satisfied they are with their lives. They conclude that it does: the more participants used Facebook, the worse they later felt. The study's repeated, real-time measurement is a clear advance over the one-time surveys that preceded it. However, a small local sample, a single self-report measure of Facebook use and causal language that runs ahead of a correlational design mean the findings support a warning rather than a firm conclusion.
Summary. The authors text-messaged 82 young adults, recruited through flyers in Ann Arbor, Michigan, five times a day for 14 days. Each text asked how participants felt, how worried and lonely they were, and how much they had used Facebook and interacted with people directly since the previous text. Participants also completed the Satisfaction With Life Scale before and after the two weeks. Using lagged multilevel analyses, Kross et al. (2013) found that Facebook use between two texts predicted worse mood at the second text, controlling for mood at the first, while mood did not predict later Facebook use. Higher average Facebook use over the two weeks predicted lower life satisfaction at the end, and direct social contact predicted feeling better.
Strength: design and rival explanations. The study's main strength is its design. Earlier research, the authors note, was cross-sectional, so it could not show which came first, Facebook use or low mood. By sampling experience five times a day and analyzing each report against the one before it, the researchers could test both directions, and only one appeared: Facebook use came before dips in mood, not the reverse. The authors also tested rival explanations. Direct social contact did not lower well-being, which argues against the idea that any socializing is draining, and the Facebook effect held when loneliness, which did predict more Facebook use, was controlled. Reporting is also transparent: the paper gives the average response rate (83.6% of texts), explains why two low-responding participants were dropped, and states that results did not change substantively under other cutoffs.
Weakness: the sample. The sample limits how far the findings travel. All 82 participants were recruited in one college town, their mean age was 19.5, and 53 were women. Volunteers for a study about Facebook may also differ from typical users. The authors acknowledge the age limit and call for research with other age groups and other networks, yet their closing claim about young adults in general rests on a convenience sample from a single city.
Weakness: measuring Facebook use. The study's central variable is measured with one slider: how much participants had used Facebook since the last text, from "not at all" to "a lot." That item cannot separate scrolling through other people's posts from messaging a close friend, two activities that could plausibly affect mood in opposite ways. The authors raise a related limit themselves, noting that a single good-to-bad mood scale can hide whether Facebook made people less positive, more negative or both.
Weakness: causal language. The article's language sometimes outruns its design. The title says Facebook use "predicts" declines, which fits correlational data, yet the abstract describes the aim as examining how Facebook use "influences" well-being and closes by suggesting that Facebook "may undermine it." Lagged data establish order, not cause. An unmeasured factor such as boredom, which the authors mention but did not measure, could raise Facebook use and lower mood at the same time. The authors concede that experiments are needed "to establish definitive causal relations" and that their effects were "relatively 'small.'" Later research has partly answered that call: in a randomized experiment, Allcott et al. (2020) found that deactivating Facebook for the four weeks before the 2018 U.S. midterm election increased subjective well-being.
Conclusion. Kross et al. (2013) offer a careful, transparent first look at how Facebook use and well-being unfold across a day, and their design answers a question that earlier surveys could not. Their evidence fits the claim that heavier Facebook use comes before small dips in mood among these participants. It does not yet show that Facebook causes those dips in young adults generally. The study is best read as a strong starting point whose own limitations section maps out the research that should follow.
References
Allcott, H., Braghieri, L., Eichmeyer, S., & Gentzkow, M. (2020). The welfare effects of social media. American Economic Review, 110(3), 629–676. https://doi.org/10.1257/aer.20190658
Kross, E., Verduyn, P., Demiralp, E., Park, J., Lee, D. S., Lin, N., Shablack, H., Jonides, J., & Ybarra, O. (2013). Facebook use predicts declines in subjective well-being in young adults. PLOS ONE, 8(8), Article e69841. https://doi.org/10.1371/journal.pone.0069841
You can check the outside source too: Allcott et al.'s randomized Facebook deactivation study in the American Economic Review.
Sentence starters help you sound precise without overstating your case. Adapt each one to your article.
| Purpose | Sentence Starters |
|---|---|
| Naming a strength | "A key strength of the study is..."; "The authors strengthen their case by..."; "This design allows the authors to..." |
| Naming a weakness | "The evidence does not fully support..."; "It is unclear from the methods whether..."; "This limits the extent to which..." |
| Weighing a point | "Although the authors acknowledge..., they underestimate..."; "This limitation matters less than it seems because..." |
| Concluding | "Overall, the article makes a [convincing / partly convincing] case that..."; "Future research should..." |
Notice when an article's verbs claim more than its design allows, and make sure yours don't.
| Study Design | Verbs That Fit | Verbs to Avoid |
|---|---|---|
| Correlational or observational | is associated with, predicts, is linked to | causes, leads to, proves |
| Randomized experiment | caused, increased, reduced (in this sample and setting) | proves for everyone, settles |
| Qualitative | participants described, the authors identify themes of | most people feel, the majority of the population |
| Review or meta-analysis | the pooled evidence suggests, studies consistently find | definitively shows |
Weak critiques tend to fail in predictable ways, and each one has a simple fix.
| Mistake | Fix |
|---|---|
| The summary takes over the paper | Cap it at about a quarter of your word count and start critiquing by the third paragraph. |
| Opinions without evidence ("confusing," "biased") | Point to the exact passage, number or method, then explain the effect on the conclusions. |
| Faulting the article for something it never tried to do | Judge it against its stated aim, then suggest what a different study could add. |
| Disagreeing with the topic instead of the reasoning | Critique the evidence and logic, even if you agree with the conclusion. |
| Presenting the authors' own limitations as your discovery | Say the authors acknowledge it, then add your analysis of why it matters. |
| New points in the conclusion | Move them into the body or cut them. The conclusion delivers the verdict. |
Run through this list before you hand in your critique:
If you'd like a second opinion before you submit, you can get feedback on your draft or work through the article with a tutor who will explain the methods and statistics and help you sharpen your own critique.
Follow your prompt, because length varies by course and level. As one example, a Portland State University graduate course guide asks for no more than five double-spaced pages and calls about three a good length. Whatever your limit, keep the summary to about a quarter of the paper and give at least half of your words to evaluation, since judging the article is the point of the assignment.
Use four parts unless your rubric says otherwise. The introduction names the author, title, journal and year and ends with your thesis. A short summary covers the research question, method and findings. The critique section discusses strengths and weaknesses with specific evidence. The conclusion gives your overall verdict and what future research should address. Format the paper in APA or MLA as assigned.
It depends on your instructor, so check the prompt. Many instructors expect a formal third-person voice, such as "the study does not explain." APA Style does allow first person when you describe your own steps, and the Purdue OWL notes that it is better than making the work itself the actor. If first person is allowed, use it sparingly and keep the focus on evidence rather than feelings.
Only if the assignment asks for them. Some critiques are closed-book and judge the article on its own terms. Others ask you to compare it with course readings or other studies. When outside sources are allowed, one or two strong ones that test a key claim, such as a later experiment or a larger study, are worth more than a long list of loosely related citations.
The two terms overlap, and some instructors use them interchangeably. When they differ, a review may lean toward describing the article's content and contribution, while a critique puts more weight on judging its methods, evidence and reasoning. Read the verbs in your prompt. "Summarize" and "describe" call for more summary, while "evaluate," "assess" and "critique" call for a verdict backed by evidence.
Yes. A critique is a balanced evaluation, not an attack, and Massey University's guide asks for an objective and balanced response. If an article is well designed and its conclusions match its data, your thesis can say so. You still need to explain why it is strong, with specific evidence, and you should still note any limitations, even minor ones, so your praise reads as earned.
Focus on what you can judge: whether the design fits the question, who was sampled and how, how the key variables were measured, and whether the conclusions stay within the data. Read the limitations section closely, and check whether the authors call their effects small. You can also ask a librarian, your instructor or a tutor to explain an analysis before you write about it.
Get affordable and top-notch help for your essays and homework services from our expert tutors. Ace your homework, boost your grades, and shine in online classes—all with just a click away!
Fast, secure, and handled by vetted experts.
Learn the 7 steps in the research process, from research question and literature review to data collection and write-up, with one worked student example.
Read More
Learn how to write a position paper step by step, with a class format, an annotated example and Model UN rules from NMUN and other conferences compared.
Read More
Learn how to write an advertisement analysis essay: choose an ad, read it with ethos, pathos and logos, draft a thesis and outline, and study a full example.
Read More