In case you can't read....

Aug 04 2011 Published by under [Education&Careers]

I've had this conversation with a number of people over my career, but I have never been able to understand the culture of some fields when it comes to conference presentations. Let's say you open your conference program and see that Dr. Big Name is giving a talk in a session and you make a plan to go see her talk. You've read her lab's papers and you're excited to see the latest from the lab. Is there anything more disappointing than getting a summary of what the lab has published in the last 10 years?

Perhaps my field is unusual but talking about recent unpublished data seems to be the norm, not the exception. When I am putting a talk together it would never occur to me not to include a health dose of unpublished data. The only times in my career that I have talked about mostly published data have been when I first started as a postdoc and in the early days of being a PI, when I didn't have enough new data to even make a coherent story, but that accounts for maybe three professional talks out of many.

Is it a fear of being scooped or a penchant for keeping one's ideas close to the chest that promotes the Summary Talk? I don't know and even that I can't understand. I have had some of most productive conversations following a talk about unpublished data, when someone has approached me to discuss and idea or related data from their lab. Collaborations have even occurred via this mechanism. In short, only positive things have ever happened in my experience with presenting, and seeing talks that include, unpublished data.

So a short poll, but I would be really interested to hear from others in fields where unpublished data rarely grace the big screen. And if people do have accounts of scientists behaving badly when it comes to presented unpublished data, let's try and keep it to first hand experience rather than Lab Lore.

26 responses so far

  • qaz says:

    I think it's field dependent. Personally, I can rarely get enough information from a talk to know whether to believe a result or not. This means that unpublished data usually ends up with me thinking "maybe, maybe not".

    In my field, I find that the best talks are the ones that pull together decades of research to provide a big picture that is almost never available from all the little published papers. I once saw a talk by a senior grand-old-man in which each slide ended up covering one paper by one student from sometime in this person's super-long career. The contribution (and newness) of the talk was that I had never seen all the pieces put together before. (A good talk like this has enough of a citation on the slide that I can jot down where to go if I want to know details on any particular result.) Of course, some people can't put together a big picture, and their talk ends up being just a sequence of unrelated results which is a painful experience. But I find a good big-picture talk with all the data (X implies Y, so we went to check Y, and found it was really Z, which opened up A...) can be fantastic if done right.

  • proflikesubstance says:

    But I find a good big-picture talk with all the data (X implies Y, so we went to check Y, and found it was really Z, which opened up A...) can be fantastic if done right.

    And I agree, but these are typically hour long keynote-type talks where people expect this kind of thing. If the talk is a typical symposium talk, however, I don't want a rehash of your career in 15-30 (depending on the format) minutes.

  • hematophage says:

    I'm in the same field as you, so my vote isn't very interesting, but...yes.

  • Dr. Dad, PhD says:

    I'm in a highly competitive biomed field, and I was taught never to present something unless it was either submitted or ready to be submitted. I've loosened up as a post doc (and gotten some good returns from it), but I still consider the venue before I present unpublished work.

    At the huge society conference I tend to be conservative, but at a recent Gordon conference I gave a talk that was 80% unpublished. In the latter case there was definitely some strategy involved - I had come up with some "out there theories" and was writing a K99. I tested my theories on some of the most renowned experts and passed, so I used them in my K99.

  • pyrope says:

    I think my field is slightly less competitive because it's full of totally unique datasets from different locations. I don't really spend any time worrying about being scooped because I collect my own data. If I find something that has broad implications, I just get it out as quickly as possible. Still, the long time lag between submission and publication often means that multiple papers along similar themes will come out in the same year.
    I've only once presented work at a conference that was already published, and it had just come out the month before.

  • Genomic Repairman says:

    Why look at a poster or talk of 100% published work, I've already seen the stuff in a journal to start with.

  • Alex says:

    Although most short conference talks in my field are mostly on unpublished or very recently published data, one group is infamous for talks that are 90% "here is the big picture of what our lab does" and 10% new stuff by the speaker. Even a 15 minute talk by a student follows that format. Yes, we get it: the whole concept being studied in that lab is revolutionary. We know. That is why the student's abstract was accepted for a 15 minute platform talk rather than being sent to the poster hall with the masses. Now, could we get more than 2 minutes on your piece of the revolution?

  • Dr. Dad, PhD says:

    Just to clarify, I don't present published data. I was just taught to make it as damn close as possible....

    And yes, I think my adviser was overly sensitive, but for reasons I won't go into here.

  • Suze says:

    Final year materials chemist = keeping cards close to my chest. Once bitten, never again.

  • Dr Becca says:

    In neuro, I'd say that at smaller conferences and less high-profile talks at big conferences (i.e. not keynotes or featured lectures), the bulk of what you're hearing is unpublished. ALL posters are unpublished--in fact, I think (?) it's a rule at SfN that the content of posters can't be published already.

    I think it makes sense for the big-wigs to present published data at large conferences, as their talks are often attended by people not in their subfield, who likely haven't read their papers but are taking an opportunity to see someone who's made a significant impact on the field of neuroscience in general.

  • Girlpostdoc says:

    I have to agree with qaz - it seems field, organism, and academic stage dependent. Although I was trained as a population and evolutionary biologist, lately my postdoc research is more mechanistic. From the non-evolution conference (N=1) that I attended in this field it appears that published or close to being published research presentations are the norm. This was not true of the many evolution/ecology conferences that I have attended (N>10). That being said, I think even within the field of ecology/evolution there is a fear of scoopage, especially in experimental evolution where a very cool experiment can be performed in a matter of months.

  • neuromusic says:

    In neuro, Doc Becca's comments are precisely why I avoid Dr Big Wig's Keynote, unless I'm trying to familiarize myself with a new area.

  • Bashir says:

    In my field I'd guess that most talks include data that is in press or at some close to publication stage. Though this definitely varies from area to area. I think the more populated and competitive areas are less likely to present data that isn't close to publication.

    I thought this post was going to be about some fields that read their presentations.

  • Unfortunately those most interested in unpublished data are usually those working on similar or even identical problems. You have only to look at who is asking the best questions to see this - at least in plant molecular research this is the case! If a particular experiment would take a lot of work to reproduce it can be worth the risk of presenting, but if a member of the audience already has similar data it can spurn them to get theirs out sooner rather than later. Unless it is already under review or, part of a manuscript almost ready for submission, you have to weigh the pros against the cons. A large lab or already established collaborative network might not be interested in bringing aboard a new member, especially if they have findings identical to those being presented. A new lab needs to be cautious in this respect.

    A big name should be more generous, but then again they do have to save guard the career of the student/postdoc who generated the data. Also the star or keynote speaker is expected to address a wider audience, and make their talk relevant to the overall theme of the conference. If it is details I want I usually find the smaller sectional talks, which are often given by student or postdocs from top labs, to be most informative.

  • proflikesubstance says:

    Dr. G - I hear this argument all the time, but I rarely, if ever, actually see this problem come up. I don't know how much of this paranoia is justified or just people seeing monsters in shadows.

  • ianqui says:

    In my (experimental) social science, most conferences explicitly say that you cannot submit to present already published or even accepted work. I supposed people squeak in, but generally, the large majority of what gets presented, even by bigwigs, is new work. Of course, I think that in my field people don't publish 100% of what they're working on (sometimes it's just a very small project, for example), so conferences can be an opportunity to get the work "out there" without necessarily needing to publish it.

  • Alyssa says:

    In my field (Astronomy), I'd say 95% of the talks are about unpublished data.

  • I know a lot of labs in my field (structural biology) that will only give talks/posters on papers that are at least submitted if not accepted. It's particularly the case in bigger labs that are working on really hot topics ā€” it's not uncommon for a competing lab to be working on the same protein structure, and they'll try to either get the crystallization conditions or quickly finish working on the structure and submit the paper.

  • Dr JK says:

    At least in Materials Science, the competition between the top groups is so tough that people don't often even share results with their collaborators, which sometimes is a real hindrance in the field. Almost everyone with a PhD knows someone who has been scooped somehow, and this makes everyone vary regarding presenting unpublished results. I have sometimes also witnessed talks where the speaker has changed the presentation to another one after noticing who are listening.

    Regarding the Big Guys presenting recent unpublished work, I suspect that they are sometimes unwilling to do this since they haven't been personally intimately involved in the research. Hence, it's better for them to let the Postdocs/Gradstudents present the work they are doing, and concentrate on the already published works themselves. Granted, this makes talks boring at times, but it is also frustrating to follow a talk when a part of the audience know better the presented work than the presenter -- I've also seen this a few times šŸ™‚

  • The Gordon Research Conferences I attend stress that you ONLY present unpublished data.

  • Sarah says:

    I find it varies a lot by professor even within my small subfield. The worst are talks that start with data collected before I was born.

  • Confounding says:

    I generally try to present stuff that's close to publication, but not yet in press - though admittedly, its hard to predict. When I don't do that, its one of two things:

    1. Further musings on a topic. When I said "more study was needed" or there were clear extensions of the work that didn't make press, lets talk about those.

    2. Side projects where, to be frank, this is publication. Things where I don't have funding or time to continue to pursue them, I doubt I can get them into a manuscript that won't slide off journals, but there's a couple interesting figures...

  • Karen says:

    The conferences in my field publish abstracts of the usual 15-minute presentations... and oftentimes this IS effective publication for several years. Right now I'm a student co-author on a paper that's been in press for three years, being held up by other, key papers in a Special Volume (these are a common means of publishing in my field). In fact, I might be able to summarize my MS thesis (due 10/31) in a paper for the volume! So... a lot of other people's abstracts from conference presentations go into a References Cited section.

  • TheGrinch says:

    Materials Physics/Materials Science.

    Top conferences in my field almost always have talks based on the recently published work. More so if the speaker is a PI. Unpublished data can wiggle in once in a while, usually from grad students or postdocs.

    In fact, to get a talk in the top conferences in my field you need to submit an abstract based on your recently published work (or in a very hot field), otherwise it is pretty much relegated to the poster.

  • chemprof says:

    Unless you're on the conference committee and automatically invited, you need to present new data in my field to be accepted for an oral presentation. Unfortunately many of the invited speakers (often the older guys) recycle old talks, but this is now being tolerated less. So much so that some of them just weren't invited to the biggest meeting this year - and when they complained they were told why.

  • HFM says:

    I started in a subfield of BE where conferences were about bleeding-edge data, so much so that people would leave holes in their poster and print the last figure or two on the hotel printers.

    For the last two years, I've been in cancer research. Virtually everything at conferences is published - people will put a picture of the journal cover in the corner of their poster or slides. On the rare instances that someone mentions unpublished data, it's either a dump of a failed project, or is done to mark territory...big lab X has sequenced 200 tumors already, you've sequenced three, and they want you to shove off.

    I miss my old field sometimes, but you can probably guess where the money is...

Leave a Reply