The Anatomy of an Abstract: A Worked Example
About a hundred and ten words doing four jobs. This is the only piece of academic writing the whole subject ever takes apart in front of you, so it is worth taking apart slowly.
What the example is, and what it is not
The slide is titled "Example analysis of an abstract" and reproduces an image: a short abstract with four labels around it, colour-coded, each pointing at an underlined stretch of the text. The slide attributes the image to a university student-services writing-tips page. No author, title or date is given for it, and the abstract's own study is not cited.
It matters that this is the only one. Across fourteen weeks the subject names dozens of sections, structures and components — see Structuring a Research Report for the inventory — and shows you the inside of a finished piece of writing exactly once. Everything else is a list of what should be present. This is the single place you get to see what present looks like.
The abstract in full
Read it once more without the annotations and notice how little it contains. There is no literature, no theory, no researcher, no institution, no aim statement, no limitation and no future work. Five sentences, four jobs, and nothing that is not doing one of them.
The four labels, as printed
THE ANNOTATION, EXACTLY AS THE FIGURE GIVES IT
| Label as printed | Position | What it points at |
|---|---|---|
| "Why there was a need for some research" | Left, red | Sentence 1 — the vulnerability of the parts |
| "What was done" | Right, red | Sentence 2 — "It was investigated whether…" |
| "Key results plus how the investigation was done" | Left, blue | Sentences 3 and 4 — the two results sentences |
| "Significance of the results" | Right, green | Sentence 5 — cost effectiveness, serviceability, environmental benefits |
Four labels across five sentences. The results move is the one that runs to two, and the third label's second clause — "plus how the investigation was done" — is examined further down this page.
- Need
- What was done
- Results
- Significance
That sequence is the whole lesson of the figure. It is not a template the source offers you — the slide presents it as an analysis of someone else's writing, not as a form to complete. But it is a diagnostic you can run on your own abstract in about ninety seconds, and the section at the end of this page does exactly that.
Walking the five sentences
Sentence by sentence
Sentence 1 — the need
"Various parts in gas turbine and waste incinerator plants are vulnerable to high temperature oxidation and corrosion." One sentence, no citation, no hedge. It states a condition that exists whether or not anybody studies it. Notice what it does not do: it does not introduce the researcher, the discipline, the paper or the literature. It establishes that there is a problem in the world.
Sentence 2 — what was done
"It was investigated whether High Velocity Oxy-Fuel (HVOF) thermal spraying of intermetallic compounds (specifically nickel aluminide, NiAl) onto such parts could protect them from this." The whole subject of the study in one sentence, and framed as a proposition put to the evidence — whether a particular thing could achieve a particular result. That framing is the transferable part: it commits the study to something answerable rather than announcing a topic.
Sentence 3 — the first result
"Initial results have shown that the coating provides improved properties compared with more conventional materials, such as other nickel alloys and stainless steels." The result is comparative — improved against a stated alternative — which is what makes it a result rather than an observation. The word "initial" is doing real work: it marks the finding as provisional without withdrawing it.
Sentence 4 — the second result
"In addition, some results in erosion testing showed that good corrosion resistance is combined with excellent protection against this form of wear." A second and different result, explicitly flagged as additional so the reader does not mistake it for a restatement of the first. It is also the only place in the abstract where anything resembling a method appears, and it appears as two words.
Sentence 5 — the significance
"The feasibility of using lower cost, less resistant substrates with a high performance protective coating offers the possibility of enhanced cost effectiveness allied with increased serviceability and ongoing environmental benefits." The move from what was found to why anyone should care, and it is hedged twice — "feasibility" and "the possibility of". The claim is about what the result opens up, not about what has been established.
Three of those five observations are worth lifting straight out of the engineering context. A need sentence states a problem that exists independently of your research. A what-was-done sentence names a proposition rather than a subject area. A significance sentence hedges in proportion to how far it is reaching beyond the evidence — the same discipline that published work applies when it states a contribution, documented at Stating Limitations and Contribution.
That close reading is this library's. The figure supplies four labels and no commentary — it does not explain why any sentence is written as it is, and it draws no lesson from the wording.
What the length demonstrates
About 110 words against a guideline of "usually … 300 words or less". The example uses roughly a third of what the same week says an abstract typically runs to, and it still completes every move the annotation identifies.
This is the most practically useful number on the page, because word guidelines get read as targets. They are not. An abstract at 295 words is a choice you made, not a requirement you met, and the demonstration that four moves fit comfortably in 110 words is the strongest argument available for cutting yours. The guideline itself, and how firmly it binds, is set out at Writing an Abstract.
The distribution is worth noting too. One sentence each for need, what was done and significance; two for results. If you want a default allocation, that is a defensible one — the results move is the one that earns extra space, because it is the only move carrying something the reader cannot get anywhere else. The proportion is this library's observation of one example, not a rule, and certainly not a ratio the source states.
The element that is not there
What the specification requires
- The overall purpose of the study and the research problem(s) investigated
- The basic design of the study
- Major findings or trends found as a result of your analysis
- The list continues beyond the file's extractable text and cannot be recovered
What the annotated example does
- Why there was a need for some research
- What was done — the subject of the investigation, not its design
- Key results, across two sentences
- Significance of the results — a move the specification's recoverable part does not name
The two columns disagree in both directions, which is easy to miss. The specification asks for a design statement the example does not make. The example makes a significance move the recoverable part of the specification does not ask for. They are not two versions of the same list — and the same week's study notes give a third account of what an abstract contains, compared at Abstracts, Introductions and Conclusions Compared.
What you do about it is a decision, and the honest position is that the material does not make it for you. If your work will be judged against the prescribed sequence, the design sentence goes in. If you are modelling your abstract on the example, it will not. Whichever you choose, the useful discipline is to know which of the two you followed and to be able to say so.
Two further mismatches, and these are this library's reading
The source flags nothing about the example's own prose. Two things are visible in the transcribed text when you hold it against the same week's style slide, and both are worth recording — as observations about a supplied text, not as claims the source makes.
THE EXAMPLE AGAINST THE WEEK'S OWN STYLE RULES
| Rule, as the week states it | What the abstract does | Reading |
|---|---|---|
| "always use the past tense" | Mixes four: "It was investigated" (past), "Initial results have shown" (present perfect), "showed" (past), "offers" (present) | The example does not hold one tense. The rule is stated with always. |
| "Use the active voice when possible" | The central move — "It was investigated whether…" — is passive, and no agent appears anywhere in the abstract | The rule is hedged with when possible, so this is not a flat breach. It is still notable that the one sentence naming what was done is the passive one. |
| "concise, but complete, sentences" | All five sentences are complete; none is a fragment or a note | Met. |
| "one paragraph" | One paragraph, no headings or bullets | Met. |
This comparison is this library's, made by reading the transcribed abstract against the style slide two slides earlier. The source sets the two slides side by side nowhere and draws no comparison between them.
Running the four moves on your own abstract
THE MOVES AS QUESTIONS — THIS LIBRARY'S CONSTRUCTION
| Move | The question that sentence answers | How the example handles it |
|---|---|---|
| Need | What condition or problem makes this work worth doing, whether or not anyone studies it? | One sentence. No citation, no hedge, no reference to the researcher |
| What was done | What proposition was put to the evidence? | One sentence, framed as whether a particular thing could achieve a particular result |
| Design | How was it done — method, setting, sample, measure? | Absent. The week's sequence requires it; the week's example omits it. Decide which you are following and be able to say so |
| Results | What was found, and against what comparison? | Two sentences, the second explicitly flagged as additional so it is not read as a restatement |
| Significance | What does the finding open up, and how firmly can that be claimed? | One sentence, hedged twice — "feasibility" and "the possibility of" |
Five rows for four labels: the design row is the specification's requirement, not the figure's. It is set out separately rather than folded in, because folding it in would hide the disagreement between the two.
What to carry forward
- Four moves — need, what was done, results, significance — fit in five sentences and about 110 words. The 300-word guideline is an upper bound, not a target to fill.
- The need sentence states a problem that exists independently of your study. The what-was-done sentence names a proposition, not a subject area.
- The results move is the one that earns two sentences, because it carries what a reader can get nowhere else.
- The significance sentence hedges in proportion to its reach. Two hedges in the example's final sentence are doing real work.
- The abstract states no design, and the week's own prescribed sequence requires one. The material does not reconcile them, so make the choice deliberately and know which you made.
- It is a materials-engineering abstract borrowed for its shape. Take the structure; take nothing else from it.
Frequently asked questions
Is this abstract from a project management study?
No. It is a materials-engineering abstract about protective coatings, reproduced in a project management research methods subject as a structural illustration. Its technical content is not taught anywhere in the material and should not be read as engineering guidance. Only the shape is being demonstrated.
Should my abstract be five sentences?
The source makes no such claim. Five sentences is what this one example runs to, and it is an illustrative property of a single abstract rather than a rule or a norm. What the example does show is that the four moves fit comfortably inside a fraction of the 300-word guideline.
Why does the example not say how the study was done?
The source does not say. It requires "the basic design of the study" as the second element of its prescribed sequence and then annotates an example that contains no method, sample or procedure at all. This library records both statements and does not choose between them.
Does the four-move shape work for a qualitative study?
The source presents one example from one field and makes no claim about its range, so nothing in the material tests this. Nothing in the four moves is specific to engineering — need, what was done, results and significance are generic — but that observation is this library's and the source does not support it.
How long should each move be?
No per-move lengths are given anywhere. In this one example the split is one sentence each for need, what was done and significance, and two for results. That is a property of a single abstract, not an allocation the source prescribes.
Can I just copy the sentence patterns?
The patterns are worth borrowing — a comparative result, an additional result flagged as additional, a hedged significance claim. The wording is not, and the technical content certainly is not. Take the moves and write your own sentences into them.
References and source attribution
- The supplied teaching source: weekly study notes, slide decks and assessment activities for a master's-level research methods subject in project management. Author, institution and year not stated in the supplied files.
- Week 13 slide deck, 'The Challenge of Writing Up': slide 21, 'Example analysis of an abstract'. The annotated image is attributed by the source to a university student-services writing-tips page; no author, title or date is given, and the study whose abstract is reproduced is not cited.
- Week 13 slide deck, slides 19-20, carrying the abstract specification and four style rules against which the worked example is compared on this page.
Suggested questions for Ask KEVOS
- Label each sentence of my abstract with one of the four moves.
- My abstract has no design statement — draft one that fits in a single sentence.
- Which sentence in my abstract is the need sentence, and would it be true without my study?
- Rewrite my significance sentence so its hedging matches the strength of my evidence.
- Cut my abstract to 110 words while keeping all four moves.
- What does the supplied source actually say about the worked example on slide 21?