Read on Twitter
I'm currently reviewing abstracts for a national conference, on a topic slightly outside of my research area. I'm noticing a recurring set of writing issues that detract from a lot of abstracts. (1/n)
I was putting together a “helpful hints” list for my own trainees, and I thought I would share (this is my opinion only - I'd love to hear others add their opinions). IMO, main abstract pitfalls include: (2/n)
(1) Why should we be excited about your study? Rather than focusing only on the 10,000-foot view (XYZ disease is a major cause of death but hasn’t been cured), instead focus on the specific gap you are testing (i.e. current Rx don’t target a key disease mechanism). (3/n)
So many abstracts start with a nearly identical statement about severity of a particular disease, or its unknown mechanism. Some different approaches: Has anyone ever tested the hypothesis in your model or tissue? Has a clinical population been ignored in previous studies? (4/n)
Does your hypothesis test an existing dogma? Are you confirming a conclusion of another group? Can you use the Liam Neeson approach: You have a particular set of skills…(that no one has applied yet)? Are you following up on serendipitous data? There is no wrong answer! (5/n)
(2)Give the reader confidence in the rigor of your study. Include a clear explanation of exp numbers/design, with attention to rigor and reproducibility. For an abstract, sometimes you’ve only got a small experimental number. Sometimes that’s ok, but be upfront and clear. (6/n)
Alternatively, if the prose is more than ~10% numerical, it’s exhausting to read (again, IMO). Include only the most important numerical result (and point out that it has strong statistical underpinning). This strengthens the impact of your major conclusion. (7/n)
If you clearly articulate your n, statistical test, and p value, there's no need to include means + SEM, which are basically impossible to read. You could save space by leaving out the numbers, and instead include info about statistical approach and reproducibility. (8/n)
(3)Tell us early about your experimental model ( #justsaysinmice, humans, cells, in vitro) and how you’ve thought about which is best for stringent hypothesis testing. You need to explain clearly why the model you are using adds to the gap in knowledge. (9/n)
(4)Help us judge the potential impact of your work. Immortalized cells can be just as impactful as human tissue if it's the most rigorous model for testing all hypotheses (including null). Let your reader know that it’s impossible to overexpress, KO, etc… in other models. (10/n)
(5)Use less abbreviations. So many less. Less than you are even remotely considering. Like almost none. Maybe zero. (11/n)
(6)When you carefully use some “lay” language, you actually sound like the expert you are. I constantly have to work to avoid jargon, and I often fall back on it when I feel nervous that I’m not enough of an expert (or when I’m experiencing #impostersyndrome). (12/n)
(7)“Taken together, these findings suggest” is now my least favorite clause. It is in at least 75% of all abstracts I’ve read. I am so, so, so guilty of using this cliche as well, and I don’t even know where I learned it! (13/n)
Can we all agree to just write, “We conclude…” or “These studies support the conclusion that…” or “Our data confirms the clinical strategy of…” or “Our data refutes the current dogma that…”. Any of these would be better. (14/n)
(8)Finally, for the many trainees who struggle with imposter syndrome, or are part of groups who have been traditionally left off “the podium”, or who are minoritized by our scientific spaces, I want to affirm your ways of writing and speaking. (15/n)
This set of opinions is meant to affirm that there is more than one right way to present your data and ideas. (16/n)
Advisors and reviewers: we need to do support trainees from traditionally minoritized groups, and to intentionally and supportively advise and sponsor them as they write these abstract drafts, revisions, and submissions. (17/n) @sbarolo, @BlackInImmuno, @sacnas
These are just my opinions (n=1), and also carefully follow the guidelines required by wherever you are submitting. Also -consider following some great scientists and writers, too: @zugenia, @shrewshrew, @KYT_ThatsME, @iGrrrl, @Afro_Herper, @ravenscimaven (18/end)
