It will be clear from the rest of my How to Write guide that the typical research paper is written in chunks, in no particular order, which are then stitched together to form what is hopefully a coherent whole. The abstract is the last thing to tackle and is one of the hardest components to get right. The importance of this little blurb at the head of the paper cannot be overstated as it has to serve many functions. Primarily it is a short summary of your research that acts as advertising. It will influence:
- Journal editors- does your research fit the scope of their journal? Is the quality of your writing up to scratch? Is your research design appropriate?
- Potential readers- do they want to go on to read the rest of your paper, or is it something they would rather not waste their time on?
And of course, we should never forget that the vast majority of readers these days may read ONLY your abstract and perhaps nothing else in your paper. The modern way of reading means that we like to find information instantly and with a minimum of critical effort.
So key message number 1, is to make sure your abstract really shines. Take your time getting it right.
What should be in the abstract?
There are several basic components that should be incorporated into your abstract:
- Problem statement: Why did you carry out your research? Was there a specific question that needed to be answered? What previous research were you building on?
- Methodological approach: What did you actually do to get your results?
- Results: As a result of completing the above procedure, what did you learn?
- Conclusion/implications: What are the larger implications of your findings, especially for the problem/question that was your initial motivation for the research?
Format and how to write the abstract
When you are writing the abstract for a scientific paper, you are attempting something quite different to the abstract for a conference. Your conference abstract is a way of saying to people at the meeting ‘Come and see me speak/present my poster. We can discuss in detail’. The conference abstract is more data rich and low on discussion. The paper abstract has to be more rounded and balanced in structure.
The length isn’t going to vary much. Journal abstracts are more or less unfailingly 250 words these days, which fits well with PubMed and Web of Science type search engines and what they display. There will be variation in structure as some journals like a formal structured abstract with discrete sections (Background, Methods, Results, Conclusion), but most do not.
Writing as if you have the formal sub-headings is a good idea as it will help you pull the whole piece together. If the journal doesn’t use them, simply take them out as a final step. The balance ought to be somewhere along the lines of Background- 10%; Method- 30%; Results- 50% and Conclusions 10%, but obviously can vary according to your needs. Bearing in mind we only have 250 words in total, your background and conclusions are going to have to be very focused. Essentially these are a couple of sentences each, so simple clear statements are key.
There is compelling evidence that maternal obesity during pregnancy exerts a programming effect upon health. This study evaluated the effect of maternal weight loss upon blood pressure in young adults.
This study has shown that an intervention which promoted a moderate weight loss in obese pregnant women, was associated with lower blood pressure in their adult offspring.
There are no frills in this kind of writing, I’m afraid. This is simple and functional.
The methods section of the abstract needs to be packed with information in order to grab the attention of editors more than anyone else. What/who was sampled? How many samples? How was data collected? What sort of research was it (qualitative, animal study, cell culture, epidemiological- case-control, cohort, RCT, cross-sectional)? Were there different treatment groups? Are there drug/supplement doses to report? It is all about facts, facts, facts.
The results will make up the bulk of the abstract. The emphasis is different to the conference abstract where you might insert tables or graphs and essentially try to put in detail of as many of your findings as you can cram into the space allowed. For the paper stick to headline findings and don’t feel obliged to give too much detail- just P values will do for example.
Blood pressure was lower at age 35 in men whose mothers had lost weight in pregnancy than in those who had not (P<0.01).
In women, but not men, there was evidence of a graded response to maternal weight loss such that blood pressure was 1.1 ± 0.5mmHg lower for every 1kg of lost weight.
You essentially string together a series of 3 or four sentences that give the main points from your paper.
Techniques for writing the abstract
I have tried several approaches over the years, all of which generate a decent first draft that can then be tweaked into a publishable form.
1. Make a list
The make a list technique is essentially a paint-it-by-numbers approach to abstract writing. The list is a predetermined checklist against which you insert one sentence at a time. Then put those sentences together and you have a first draft. The list would be:
What is the problem under study?
What was the hypothesis?
What was the study design?
What was sampled?
How many subjects?
What sub-groups were studied?
What was measured?
When was it measured?
What was compared?
Headline finding number 1
Headline finding number 2
Headline finding number 3
What do you think it means?
Why is this important?
2. Highlighting key phrases in the paper
With this technique I go through my paper and simply use the coloured highlighters in my word processor to pick out phrases and points that I think would be important for the abstract. A couple from the introduction, 4 or 5 from the methods, same from the results and the final statement from the discussion. Then pull those highlighted points out, stitch them together and you have a first draft.
This technique works OK- that first draft will be very clunky and will need a lot of polishing, but this enables you to produce an outline without putting too much thought and angst into the process.
3. ‘Write it for your mum’
I like this one a lot and it is my current modus operandi. What I mean is describe your research in very simple language that a well-educated lay person would understand. Stick to the Background/Methods/Results/Conclusion structure and imagine it as a conversation.
I am interested in this because of X. There was a question I really wanted to answer, so we carried out a study in which we did Y. The main things we observed were A, B and C. In the light of this we now think that Z is true.
This is a bit like writing your abstract from scratch without any reference to the paper, but I find it works well. The first draft may be a bit simple, but is enhanced by adding in small details here and there. The write it for your mum technique actually makes you engage brain to think about what you want to say and generates a well-balanced piece that is able to stand alone and be viewed separately from the rest of the work (as a good abstract should).