, , , ,

While I’m busy in the lab most days this week, I thought it would be a good time to talk about things that are supposed to save us time when reading papers – the abstract.  If you recall in my last Friday Scribbles, I usually write the abstract last.  This is exactly the opposite way my PhD supervisor suggested, but I think I have some good reasons for waiting until the end.

In “ye olden days of yore” before electronic Table of Contents (eTOC) alerts, Web of Knowledge, or Google Scholar (and also when I started my undergrad degree, so just over 10 years ago), the main way academics found new papers in their field was by 1) subscribing to relevant journals personally or browsing new issues in the Library, or 2) consulting Biological Abstracts.  Biological Abstracts is (it still exists!) a regular publication that, as the title may suggest, prints the abstracts of indexed biological journals.  These abstracts were categorized, so I might check out the ecology or zoology sections.  And essentially, I would browse the tome (hundreds or thousands of pages published each month) looking for any papers of interest, or use an annual index for certain topics.  Biological Abstracts is online now, and integrated into Web of Knowledge.

But in this pre-digital age, the abstract was incredibly important since it could make or break whether an academic tracked down and read the whole article.  Nowadays, the abstract isn’t even sent in eTOC alerts for many journals (e.g., those published online at BioOne, and some titles published by Wiley or Springer) – just the title and authors (and sometimes the page range).  But abstracts are still an important part of papers.  So how do I write them?

In the 3rd year of my undergrad degree, I took a course called “Theoretical and Evolutionary Ecology”.  This was my first experience writing an abstract.  Our prof gave us a photocopy of a complete paper, but the abstract was removed.  We had one week to write an abstract for the paper (on egg size vs. number in trout, if I recall).  We then turned these in, and with our graded abstract, she provided the actual abstract of the paper.  I think this is a great way to get students writing abstracts, and to introduce them to critical reading of the primary literature.


One of my papers with the abstract removed.

Abstracts at most journals have one thing in common: they have a maximum length.  This can range from 100-300 words depending in the journal or contribution type (short communication vs. full paper vs. review).  One of the reasons I like to write the abstract last is because by that time I have a pretty good idea of where I want to submit the paper.  Some journals (e.g., those like Journal of Animal Ecology published by the British Ecological Society) also have specific formats and requirements for their abstracts.  I’ll focus on the more common “free-form” type of abstract between 250-300 words.

I generally break the abstract into the same sections as my paper:

  • Introduction: 2-3 sentences
  • Methods: 1-2 sentences
  • Results: 2-3 sentences
  • Discussion: 2-4 sentences

I usually start my papers’ introduction with a broad idea, theory, or concept, so this usually forms the first sentence of my abstract.  The next two focus in on my problem, and state my actual objective.  I don’t think I can emphasize this last point enough.

The description of methods is brief, but sufficient to tell the reader what I did: “I measured floor tiles in 23 academic and 25 government buildings in April 2012”.  Unless my analysis is novel, or atypical, I generally leave out anything about the statistics since these details will interest only a small number of readers.

It’s nice to include some actual numbers in the abstract: “Government floor tiles (0.04 ± 0.01 m2) were significantly smaller than academic floor tiles (0.06 ± 0.01 m2)”.  Notice that I didn’t put in any p-values, F-ratios, AIC weights, etc.  Remember – this is supposed to be the “take home message”.

The last part of the abstract puts my results in the context of the introduction, and I usually wrap up with a concluding sentence.

Lastly, some journals also request between 4-10 key words.  These should be chosen to maximize the searchable terms so that others will find your paper.  If you include a species name in the title, there’s no need to include it in the key words since most searches are for terms in the title and key words – don’t use up a valuable key word repeating what’s already in the title.  Use higher taxonomic groups or English common names (if not in the title), geographic areas/study sites, or the major concept in the first part of the introduction.  They key (get it?) is to have your paper appear in search results for people writing papers that would cite your work.