Critiquing a Journal Article
Three clauses of an activity brief define a deliverable most writers have never produced. A submitted example in the supplied material shows what the difficult part looks like when it is done well — and what costs marks even when the analysis underneath is sound.
Three clauses that define the deliverable
Four instructions are packed into that. Find five new articles that support your research questions; write a critique of each; cover three named things in every one; reference them consistently. The word budget is for the whole set.
- Relevance
- How the argument is presented
- Outcomes
WHAT EACH OF THE THREE PARTS ASKS FOR
| Part | The question it answers | What a weak entry does instead |
|---|---|---|
| Relevance | Why does this article matter to my research questions, in my context? | States the article's topic, which describes the paper rather than connecting it to anything |
| How the argument is presented | How is the case built — what design, what evidence, what comparison, what reasoning links them? | Summarises the method in one clause and moves on, or skips the part entirely |
| Outcomes | What did it find, what follows from that, and do I accept it? | Repeats the abstract's conclusion with no verdict of the writer's own attached |
Column one is the brief's wording. Columns two and three are this page's reading; the brief names the three parts and does not elaborate on them.
A third deliverable, not a variant of the other two
The three literature deliverables in this material look similar from a distance and are graded on different things. Confusing them is the commonest structural error in the whole exercise.
THREE DELIVERABLES BUILT FROM THE SAME READING
| Annotated bibliography | Article critique | Literature review | |
|---|---|---|---|
| Unit of organisation | One entry per source | One entry per source | One section per theme |
| What it is mainly doing | Describing and evaluating each source in turn | Taking a position on each source's argument and its use to you | Synthesising across sources into an argument of your own |
| Does the writer's voice appear? | In the usefulness, limitations and reflection elements | Throughout. A critique with no position is not a critique | In the framing and the synthesis, rarely in the first person |
| Is method commented on? | Aims and research methods are one element among eight | Yes — how the argument is presented is one of only three parts | Only where methodological difference explains a disagreement in findings |
| Hardest part | Compressing eight elements into a paragraph | Having a defensible view of how the case was built | Getting the sources to speak to each other rather than in sequence |
The annotated bibliography's eight-element structure is at Writing an Annotation: The Eight Elements; the review's job is at What Is a Literature Review?.
The critique is the most demanding of the three because of its middle clause. An annotation can be written well by someone who has read carefully and understood the findings. A critique cannot — it requires a judgement about whether the way the case was built supports the claim it carries.
The middle clause is the hard part
"How the argument is presented" is asking about construction. Not what the paper concluded, but how it got there and whether the route holds.
SEVEN THINGS YOU CAN SAY ABOUT HOW AN ARGUMENT IS BUILT
| Feature | What to look for | The sentence it produces |
|---|---|---|
| The comparison structure | What is being set against what — two groups, two periods, with and without a condition | "The case rests on a comparison between high- and low-effectiveness organisations, so everything depends on how those groups were defined." |
| Where the evidence comes from | Primary collection, existing records, or deduction from stated assumptions | "The argument is analytical rather than empirical; no data is collected, so the claim stands or falls on its assumptions." |
| The assumptions carrying the weight | Premises stated up front that the conclusion depends on | "Two assumptions are relaxed, and the welfare conclusion follows from those relaxations rather than from any observation." |
| The order of the case | Whether the problem is named before the method, and whether the method answers it | "The paper opens by naming the quantity it cannot observe, then builds the design around approximating it." |
| What is reported and what is emphasised | Whether unfavourable results appear, and where | "The insignificant average effect is reported in the abstract rather than left to the discussion." |
| Scope claimed against scope demonstrated | How far the paper says its contribution reaches, against what it actually shows | "The method is demonstrated once, in one organisation, and the paper is explicit that it does not claim completeness." |
| The structure imposed by the venue | House-style devices — abstract boxes, key concept lines, page limits | "The absence of a methodology section reflects a seven-page format, not an absence of method." |
Synthesis for this page. The brief names the requirement and supplies no list of what to comment on.
What the submitted example does well
Four observable habits in the sample
Every entry connects to the writer's own context
Rather than summarising neutrally, each entry says what the paper means for the writer's own situation — relating a subsidy paper to how innovation is part-funded in the writer's own national context, and a portfolio-uncertainty paper to running a portfolio of roughly forty projects a year. Relevance is demonstrated, not asserted.
Every entry registers disagreement or doubt
Not one of the five is accepted at face value. One questions how investors would see profit in a risk-reducing approach that "will definitely eat up most profits"; one states "I am not 100% confident on this methodology as applicable"; one calls a paper's welfare conclusion "just one face of the coin".
Two entries name a limit rather than a fault
One observes that a method could create conflict where projects sit under different managers; another that a general article does not explain how non-innovative management could achieve excellence. Neither is an accusation of error — both bound where the paper applies.
One entry comments on the argument's structure
It notes explicitly that the paper compares high-effectiveness against low-effectiveness organisations. That is a comment on how the case is built rather than on what it found, and it is the entry that most directly satisfies the middle clause of the brief.
Registering doubt is the whole exercise
The single strongest feature of the sample is that a position is taken every time. It is worth being precise about why that matters, because the instinct running the other way is strong — published, peer-reviewed work feels like something to be reported rather than judged.
A critique with no position is a summary. If five entries all conclude that the paper is useful and well argued, nothing has been critiqued: the reader learns what the papers say, which they could have got from the abstracts.
Positions that are defensible
- The conclusion does not follow from the comparison the paper actually made
- The finding holds under conditions the paper does not have, in my setting
- The method assumes access, data or authority I could not obtain
- The claim is broader than the single case it is demonstrated on
- An alternative reading of the same result is available and is not addressed
- The measure and the grouping come from the same source, so the result may be about the respondents
Positions that are not
- The paper is old, with no argument about what has changed since
- The paper is hard to read, which is a complaint rather than a critique
- The finding is inconvenient for my argument
- The paper does not cover my topic — that is a selection error, not a flaw in the paper
- The sample is small, asserted without reference to what the design needed
- The journal is unfamiliar to me
What costs marks even when the analysis is sound
Every weakness observable in the submitted example is mechanical. None of them is a failure of judgement about the papers, and every one of them is assessed.
FOUR OBSERVABLE WEAKNESSES AND WHAT EACH ONE COSTS
| Weakness | What it looks like | Why it is marked |
|---|---|---|
| Citations inconsistent with the Harvard pattern | Author and year lead correctly, but page ranges and a database link sit in a non-standard order in the reference list | The brief names the Harvard system explicitly. Consistency is the whole point of a referencing system, so a near-miss is still a miss |
| Two citation systems in one document | Bracketed numbers used as reference markers alongside author-date citations | A reader cannot tell which system governs, and one set of markers usually points at a numbered list that does not exist |
| Industry-wide claims asserted without citation | A statement that participation in R&D rises every year while project success ratios fall, with no source attached | In a document whose subject is the evaluation of evidence, an unsupported empirical claim undermines the writer's standing to judge anyone else's |
| Inconsistent sentence-level grammar and article usage | Errors distributed through otherwise substantive analysis | Assessed independently of analytical quality, and cumulatively significant across a short piece |
Observations about one submitted document. They are described here because the failures are common and cheap to avoid, not as a judgement of the writer.
Writing one
A shape that satisfies all three parts in about a hundred words
Citation line
Full Harvard reference, on its own line, in the same pattern as every other entry. Settle the pattern once before you write entry one — see Harvard Reference Anatomy.
Relevance, in one sentence tied to your question
Name which of your research questions the article bears on and what it contributes to it. "This is relevant to my topic" is not this sentence.
How the argument is presented, in one or two sentences
Pick the single structural feature that most determines whether the claim holds — the comparison, the evidence source, the load-bearing assumption — and say what depends on it.
Outcomes, in one sentence
What it found, in your words rather than the abstract's, and what follows for your own study.
Your position, in one sentence
A doubt, a disagreement, or a limit of applicability. This sentence is the difference between a critique and a summary, and the sample supplies one in every entry.
Before the set is submitted
- All five articles are new — none is drawn from a set you previously nominated
- Each one genuinely bears on a stated research question, not on the topic in general
- Every entry contains a sentence that could be disagreed with
- Every entry says something about how the case is built, not only what it concluded
- One citation system, applied identically in every entry and in the reference list
- Every factual claim of your own carries a citation, or is labelled as your own observation
- The total is inside the word budget without any entry having been left as a stub
What to carry forward
- Three parts, all compulsory: relevance to your questions, how the argument is presented, and the outcomes.
- This is a third deliverable, distinct from the annotated bibliography and the literature review, and harder than both.
- The middle part is the difficult one. Pick one structural feature per entry — the comparison, the evidence source, the load-bearing assumption — and say what depends on it.
- Every entry needs a position. Five entries that all approve of their papers have critiqued nothing.
- A limit of applicability is often more useful than a flaw, and requires only that you know your own context.
- Divide the word budget by the number of articles before writing, and design each entry to fit.
- Mechanical faults — mixed citation systems, uncited claims — are assessed separately from the analysis and are the cheapest marks to keep.
Frequently asked questions
How is a critique different from an annotated bibliography entry?
An annotation describes and evaluates a source across a set of elements — citation, aims, methods, scope, usefulness, limitations, conclusions and reflection. A critique has three parts only, and one of them requires you to comment on how the argument is built. The critique also demands a position: an annotation can be written well without ever disagreeing with anything, and a critique cannot.
Can I criticise a peer-reviewed paper if I am not an expert in its method?
Yes, and the more defensible route is often to identify a limit of applicability rather than a flaw. Saying a paper is wrong requires you to be right about its method. Saying a method assumes access, data or conditions that your setting does not have requires only that you know your own context — and it is more useful to your reader.
What does "how the argument is presented" actually mean?
It means the construction of the case rather than its conclusion: what is compared with what, where the evidence comes from, which assumptions the conclusion rests on, whether the problem is named before the method, and whether unfavourable results are reported or buried. One such observation per entry is enough at the lengths these briefs allow.
How long should each critique entry be?
The supplied brief allows 600 words for five articles, and the submitted example runs to roughly 90 to 130 words per entry. Those numbers belong to that brief and that submission — they are not a standard. The transferable move is to divide your own budget by the number of articles before you start writing, so no entry ends up as a stub.
Is it acceptable to write in the first person in a critique?
The submitted example does so throughout, and it is what allows every entry to register a position and connect the paper to the writer's own context. Check your own institution's expectations, because conventions differ and the surrounding material defers writing conventions to your supervisor. Whatever register you use, the position still has to be visible.
References and source attribution
- Five journal papers supplied as exemplars alongside the teaching material, published between 2000 and 2011, together with an assessed student critique of them. Observations on this page describe that one submission's structure and habits; it is anonymised throughout.
- Bryman, A. 2016, Social Research Methods, 5th ed., Oxford University Press, Oxford — cited in the supplied source; the standing reference for critical reading of research design.
- O'Leary, Z. 2017, The Essential Guide to Doing Your Research Project, 3rd ed., Sage Publications, London — one of the general research methods texts cited across the supplied source.
- Quinlan, C. 2011, Business Research Methods, 1st ed., Cengage Publishing — cited in the supplied source as the text behind its treatment of literature and evidence.
- The supplied teaching source: consolidated weekly teaching notes and slide material, including the activity brief requiring five further articles to be identified and critiqued for relevance, argument presentation and outcomes, referenced using the Harvard system.
Suggested questions for Ask KEVOS
- Rewrite my critique entry so it comments on how the argument is built, not just what the paper found.
- Is my objection to this paper a flaw or a limit of applicability?
- Check my five entries — do any of my position sentences work for more than one article?
- My critique is over the word budget. Show me which sentences are summary rather than critique.
- Find every factual claim in my critique that has no citation attached.