Conference abstract mistake

[bugbrained]

Hi guys,

There is an upcoming conference, at which I am presenting a poster for the first time. It was a super last minute lab decision where we sort of scraped together a project by reanalyzing videos collected by another grad student for a different project - anyway. We did a quick analysis, pumped out an abstract, and hit the submit button -- just as the clock struck midnight and the deadline was past. We got the abstract in anyway, because the PI knew a guy and was able to explain our situation. Now just recently, our stats adviser noticed that our Chi-squared/DF was too high for the poisson distribution to apply appropriately and had us run the analysis again using a negative binomial distribution. This has resulted in a change of significance (turning previously significant differences into insignificant ones) for several of our results, which were mentioned in the original abstract. Turning our original submitted abstract slightly incorrect. I feel like it is a bit too late to ask to resubmit a corrected abstract - it would be unfair to other presenters, looks unprofessional, and I don't want this reflect poorly on my PI. I'm also reluctant to suggest withdrawing from the presentation, since my PI pulled strings to make it happen and again, I don't want to reflect poorly. 

Any advice on how to address this??

To expand, out of the whole abstract, there is only one sentence that is incorrect: "For three threatened and endangered plant species, [the weevils] walked more than they rested." (I'm paraphrasing, obvs) The new results show that there is a visible difference between walking and resting for those three plant species, but it's not a significant difference. If we're being technical, I didn't use the word "significant" in that original sentence?? But it's still misleading??

I'd really appreciate some advice!!

Thanks!

[TakeruK]

In my field, conference abstracts are generally considered to be preliminary results and therefore, "mistakes" in abstracts are both expected and tolerated. Although I would not call your situation a "mistake" but instead, consider it as a preliminary result that was later shown to be incorrect after additional review. I would say that your situation is very common! A lot of people go even further before catching the mistake---sometimes I hear presentations that present one thing, but then later on, in the actual paper, it turns out that an earlier error meant that the results were slightly different.

This is OK. This is how science works! You present ideas, and then when you learn that you were wrong, you adapt them. For your situation, the right thing to do depends on your field. If your field is like mine, where conference abstracts aren't very meaningful, then I wouldn't do anything. In rare cases, someone might contact you because they found a result from your conference abstract intriguing and want to know if a peer-reviewed paper is ready yet, then you can let them know about the updated result. You should also talk to your advisor about the norms in your field.

[fuzzylogician]

Honestly, I would learn from this not to put together half-assed abstracts, but this can sometimes happen even if you are careful and take your time; mistakes happen, and sometimes it's just the case that between submitting the abstract and presenting the work you've learned new facts, run new analyses, or otherwise changed how you think about the data. That's fine. You just present whatever is the latest development, even if it's quite different from the original. That's all part of the natural development of science. At least in my field, you wouldn't change the abstract and you wouldn't withdraw it unless something went very wrong, but you'd just update the actual presentation. 

[GreenEyedTrombonist]

I once went to a conference where an abstract said someone would be discussing the architecture of Lord of the Rings and they decided that that was too obvious an analysis, so they presented on architecture in the real world (don't remember where) instead. Although this is an extreme case and outside the field, it shows that diverting from the original abstract isn't the end of the world. I once submitted an abstract and one week before the conference a study came out directly related to what I was talking about (it was exploratory research) so I rewrote my paper to include that study. That paper won an award at the conference, so the fact it didn't exactly match my original abstract was definitely not a big deal. I wouldn't worry about trying to resubmit or someone being upset the presentation doesn't exactly match the abstract. I'd just take this as a learning experience, as fuzzylogician said. 

[TakeruK]

For my field, it's common practice to write abstracts to present work that you have planned to do but have not started yet (however, you need to be sure that you will have it done by the conference, so typically only more experienced people do this and only when you already have all the data in hand etc.) Wording these abstracts is hard though, because you have to carefully balance saying things that will showcase the interesting results without committing yourself to a result that you might not find or might not have time to finish. So, for example, if you were in my field, in the future, for a result that is not yet peer-reviewed, I would have written something like, "We present our analysis on how often weavils walked compared to how often they rested" instead of a specific result one way or another. And, in the sentences leading up to this one, I would introduce the research question (i.e. why make this comparison?) and in the sentences afterwards, I might write about what new knowledge you would gain if it turns out they walked more and what other new knowledge you would gain if it turns out they rested more. 

For non-reviewed work, another reason to not be too specific in your abstract is that you don't want people to scoop you!

For context, my advice is in a field where for any given conference, a poster presentation is guaranteed and the oral presentations are competitive. Generally, a vague abstract like the one above will be hard-pressed to win an oral presentation slot, but it's a fine abstract for a poster presentation (like in this case). But since so many things depend on your field, this is something that would be good to discuss with your advisor.

[bugbrained]

Thank you so much for your replies, they've been super reassuring and insightful about the way things work at conferences!