Distracted by abstracts: Tips for writing a good abstract for a scientific conference

ParsonsDr. Chris Parsons has been involved in whale and dolphin research for over two decades and has been involved in research projects in every continent except Antarctica. Dr. Parsons is an Associate Professor at George Mason University as well as the undergraduate coordinator for their environmental science program. He’s a member of the scientific committee of the International Whaling Commission (IWC), has been involved in organizing the International Marine Conservation Congress (IMCC) (the world’s  largest academic marine conservation conference) and is currently the Conference Chair and a Governor of the Society for Conservation Biology. In addition, Dr. Parsons has published over 100 scientific papers and book chapters and has written a textbook on marine mammal biology & conservation.

I have just read and reviewed through close to 100 scientific abstracts for a conference, and my main conclusion is that ” ‘abstract’ – this does not mean what you think it means!”

An abstract is supposed to be a concise summary of your entire paper or study. Basically a written version of the 30 second “elevator pitch”. In these days of information overload there is so much emphasis on publishing, and so many journals willing to accommodate, the number of articles in scientific fields has increased rapidly. As a result, academics are increasingly reading no further than the abstract, and often only reading the title. To test this I looked at some of my papers where the website they were hosted by kindly provided statistics on abstract page views and actual download rates. The download rates were approximately only 10% that of the abstract views across the papers (and I am naively hopeful that at least some of a downloaded paper will be read). The figures were similar for other articles, so it wasn’t just my papers. So 90% of people who see your work probably won’t go beyond your abstract. This makes it vitally important that all the information you want to convey about your work is in the abstract.

However, in a frighteningly high proportion of abstracts the key results and conclusions of studies are not even mentioned. One of the abstracts I read in this latest batch noted that the methods, results  and the conclusions of the study “would be discussed”. As an abstract this is useless. Too frequently place holder abstracts are submitted to conferences, with the assumption that results will magically appear before the meeting. But if you don’t manage to get that analysis done, you’ll be giving a presentation that will be lacking, will embarrassing you and damage your career. Moreover, a lot more people will see your abstract than will actually get to your presentation, so professional opinions may be made on you by the quality of your abstract rather than the final presentation. Plus abstract books are physical entities, whether electronic or hard copy, and will be around a lot longer than your 10 minute presentation.   So for your professional image and also for the sake of communicating your study it is in your interests to produce a good abstract.

With that in mind, here are some suggestions on abstract structure.

The opening 2-3 sentences should describe the issue your are dealing with. For example if your study is on by-catch rates of dolphins in Panama, describe the threat by-catch poses, or why we urgently need the data. These sentences establish the context of your study and it’s importance. You need to grab the readers’ attention.

Next 1-2 sentences should very briefly describe the method you used. This can and should be very simple as readers really want to know what you found rather than how you found it, unless the method is especially unusual or novel.  But the reader does need to see that your method (and sample size) is appropriate. Sample size is important to state a study based on 20,000 data points is potentially going to be much more reliable or conclusive compared to 10 data points.

Then write 3-5 sentences on your results. Your most important results should be noted clearly and concisely, emphasizing those that were statistically significant.

The most important sentences are the final 2-3 sentences. This is the conclusion or take home message of your study. This is the “so what?” or “what does this really mean?” part of your work. For example, for the dolphin by-catch case study above, the reader will want to know what is the impact of the level of by-catch reported- is it sustainable, will it lead to extinction of the local population of Panamanian dolphins? Is there any hope? What management measures are recommended?

Here’s a fictitious example:

“Recent fictionalized television documentaries have led many members if the public to believe that mermaids have been discovered.  Therefore there has been significant controversy over whether mermaids actually exist in various forms of social media. An inter-disciplinary meta analysis was conducted reviewing government and non-government data sets for records of mermaid sightings. The results of the analysis were that in over 120 years of government and academic oceanic surveys there was only one reported sighting of a “mermaid”. This sighting was further investigated and subsequent documentation revealed that the sighting was unreliable as the observer admitted that they “were very, very drunk”. Therefore, there is currently no scientific  evidence whatsoever that there is an extant population of mermaids in the oceans”.

Anyone who does not have a conclusion, or concludes “more research is needed” should be soundly smacked about the head with the conference abstract book or journal. Let’s assume all scientists will continue doing research in their specified area and let’s take the latter conclusion as a given. Moreover, anyone whose final statement is that their conclusions “will be discussed”, should be forced to eat the journal/ abstract book…

The abstract is potentially the most important part of your paper or conference presentation – treat it as such !

  1. Hi Dr Parsons,

    You present an excellent guide. I do wonder how much of the variation in abstract quality is driven by different views of what a conference presentation should be though. From my (limited) experience, what you have outlined speaks to a traditional style of presentation in which you are essentially walking through and advertising a published or in press paper. I generally find that type of presentation pretty pointless – unless I’m well out of my field – as I could just as easily read the paper if I haven’t already. The most interesting presentations, for my money, are those that tell me what someone is currently doing, rather than what they’ve done & published 12 months ago.

    You’re right of course that placeholder abstracts aren’t very informative, but they’re inevitable if you want to present bleeding-edge results and float new ideas yet are forced to submit an abstract months in advance of the conference. Yes, there is some risk that your analyses etc. won’t all come together in time. But I would much rather see a presentation in which not all the analyses are quite done, or the conclusions haven’t completely crystallised, than the same presentation that I’ve seen at the last three conferences (however polished it may be). I’m sure the time-lag between abstract submission and presentation plays a significant role in presentation recycling. The best solution I’ve come across, and which I know has been used in a few places (Evolution 2013, I think?), is to do away with abstracts altogether and only require a presentation title.

    None of this speaks to the quality of your advice of course, it just got me thinking!


  2. Eek! I just submitted an abstract where conclusions “will be discussed” at my presentation! I have the conclusions for part 1 of the talk. Part 2 is still being analyzed. Excuses, excuses, but I’d hope there’s more forgiveness for conference abstracts than published paper abstracts.

    Maybe it’s study-area specific? I’ve had very senior mentors who admit to going to talks based only on title, not abstract.

  3. I agree with Tom’s point on placeholder abstracts. This post is great advice for paper abstracts and for conference presentations on research that’s already done, but I personally tend to present preliminary results that still need some work before being ready to publish. Combine that with the 3-6 month lead time most conferences require for submitting abstracts, and a lot of the time I’m submitting a conference abstract on something that I don’t even have conclusions for yet. This actually helps me by forcing me to get my thoughts together before presenting, and really speeds up the process of getting all the analyses done and getting the results ready to publish. It also allows me to incorporate feedback from the conference when I’m writing the final paper. From what I’ve seen others present at the conferences I’ve attended (mostly fisheries and/or shark-related) this seems to be a more common approach than presenting on research that’s totally finished and published. Of course, it could also be that those of us studying fisheries and sharks are less organized than marine mammal researchers.

  4. I would actually argue that, for a conference, the abstract is not that essential to your talk. Word of mouth and the general appeal of the overall session are what draw people in. People are likely to skim the titles, but I can count on one hand the number of times I’ve referred to an abstract at a conference.

    So here’s my question: Of those 100 abstracts you reviewed, how many were actually rejected for being badly written (as opposed to the science not being good)?

    I’m not saying you shouldn’t take the time to craft a good, well-written abstract, but rather, if you’re a busy academic or grad student juggling numerous responsibilities, crafting a beautiful conference abstract that few will ever read is not necessarily the most effective use of your time. It takes 15 minutes to write a perfectly serviceable conference abstract. It takes hours to make the prose really shine.

    Manuscript abstracts, on the other hand, should be the best thing you’ve ever written. Every time.