49

Many times I've attended and will attend international conferences presenting a poster in my life. I used to think that this poster works have been considered less important than scientific papers or oral communications, but important anyway for scientific divulgation.

So, when presenting a poster, I've always been thinking: "Okay, it's not as important as a oral communication, but, if they accepted it, this means they found it interesting".

Then I read the "Ten Simple Rules for a Good Poster Presentation" article by Thomas C. Erren and Philip E. Bourne, and I was pretty disappointed after reading point 4:

Rule 4: Poster Acceptance Means Nothing

Do not take the acceptance of a poster as an endorsement of your work. Conferences need attendees to be financially viable. Many attendees who are there on grants cannot justify attending a conference unless they present. There are a small number of speaking slots compared with attendees. How to solve the dilemma? Enter posters; this way everyone can present. In other words, your poster has not been endorsed, just accepted. To get endorsement from your peers, do good science and present it well on the poster.

Do you agree with this statement?

What is the overall importance of a poster session during an international conference?

Is is just a way to collect subscriptions and money, or is it an important occasion for science divulgation?

Thanks

Edit: I'm speaking about poster session vs. oral session, I do know that a poster with presence is better than nothing. As Woody Allen once said: "The 80% of success is just showing up" ;-)

  • 19
    Acceptance of a talk isn't an endorsement either (except in some disciplines like CS where giving a talk is conditioned on successful peer review of a paper). – David Ketcheson Jul 26 '12 at 8:10
  • This is very boread: what subject do you speak about? I've never attended a conference with poster sessions (and I gave many contributed talks) because of my subject. My colleague-next-door has never given a talk (and has made a lot of posters instead) because of his subject. – yo' Jan 8 '14 at 20:36
  • It's both, they get money and subscriptions, poster presenters get the chance to get some visibility, which may mean citations (not to their poster, but related papers), which means a better h-index, because it's all about impact. Anyway, the visibility pays-off when there is something to make visible, and a poster is not that thing, but an advertisement of that thing. – Trylks Feb 17 '14 at 11:12
  • @DavidKetcheson: Poster papers in some CS subfields are usually peer-reviewded, as well. It's a brief, more superficial review, and the expectations the paper has to live up to are lower, but some submitted poster papers nonetheless get rejected, so the rest could be said to be endorsed to some extent. – O. R. Mapper Oct 16 '15 at 21:37
  • 1
    @O.R.Mapper: It's worth noting that at some CS conferences, you don't submit a "poster paper" per se, you just submit a paper. You might then be asked post-review to present it as a poster rather than a talk, but the level of peer review involved is no different than for any other paper. – Stuart Golodetz Aug 28 '16 at 16:30

11 Answers 11

29

A poster session is a good way to disseminate your work: but it has nothing to do with how you get credit for your work.

Some computer science venues have moved to a model where papers are reviewed and accepted, but the vast majority of papers only get a poster presentation at the conference. Another model is where all accepted papers are invited to a poster session. So it really depends on your area and venue.

  • 5
    Is poster session a good way to disseminate one's work? A talk means (say) 15min x 100people, a poster means 3min x 10people (optimistically). And in the later case, they are distracted by many things. – Piotr Migdal Jul 27 '12 at 0:19
  • 14
    The right comparison here is not poster vs talk, but poster vs nothing. – JeffE Jul 27 '12 at 3:17
  • 9
    Um, have you carefully observed those 100 people - probably 5 are NOT checking email, surfing, making their own slides, or ... And in cs conferences, 100 people for a talk is a large factor over estimate – Suresh Jul 27 '12 at 14:47
  • 1
    @Suresh I'm aware of it. But it's easier to estimate effective attention going the other way: for an each day of a conference, I effectively pay attention to 5h of talks, and say - 20min of posters. Of course, it's a very rough estimation, but still there is one order of magnitude of difference. – Piotr Migdal Jul 28 '12 at 12:11
  • 2
    In a good 6 hour poster session, you can get up to 50 or more people. Busy posters have 10+ people listening at all times. – Memming Feb 17 '14 at 20:59
22

It might be better if things change in the direction of making "posters" and "poster sessions" have greater weight, but, at the moment, in "research" mathematics, I fear this is not so. (This does not deny that undergrad "research" is typically showcased in exactly this way.)

The questioner's potentially-cynical-sounding quoted "market analysis" is I think perfectly correct, if slightly exaggerated. That is, every enterprise "needs to" support itself, and this is typically by giving larger numbers of people the impression of participation.

I have no experience whatsoever of anyone in graduate admissions or post-doc hiring or... caring at all about any poster-session contribution. A "contributed talk" is also of essentially zero weight, exactly (as other answers mention, and the question anticipates) there's no filter (a.k.a. "peer review").

On the other hand, "doing things", as opposed to "not doing things", is a plus. Passivity and quietude are vastly less valued than activity, whether peer-reviewed or not. Ok, the market drives certain conceits, so don't be fooled by these, but, still, "being seen" is a good thing. Maybe not a resume-padder, but better than not being seen, by far.

17

My personal experience (atmospheric science / remote sensing) is that poster sessions have little significance. You can present your work, maybe you will have some interesting discussions with scientists or get interesting ideas from others. In my field, posters are not peer-reviewed and virtually always accepted, if not clearly off-topic or rejected for political reasons. For me, the main point of going to a small (<300 people) conference where I have only a poster is to be able to speak with famous scientists in my field (who of course have talks), do networking, etc.

My advice: do your best on your poster, make it informative and attractive (not too much text please!), but don't expect a huge attention or exposure. The conference consists of more, much more, than just the poster session where you present your work. Use the opportunity to be with senior scientists in your field!

  • 1
    This is extremely off-topic, but I'm always surprised at how big some fields' conferences are (I'm in math). I've only been a handful of conferences with >300 attendees and have thought they were uniformly awful. <50 seems quite normal to me. – Ben Webster Feb 17 '14 at 14:37
  • The American and European Geophysical Union annual meetings have >10,000 attendees, but at the particular specialised session you still meet the people from the narrow field of specialisation. – gerrit Feb 18 '14 at 0:31
16

I will speak toward my experience, as a poster presenter, and a poster reviewer. My experience is primarily in the domains of Epidemiology, Public Health or Medicine so take that for what you will.

  • "Does this mean they liked my work?". It means they don't hate your work. While conferences do need people, and their only rate-limiting factor is the size of whatever room they're having the poster session in, they don't accept all posters. I've reviewed posters, and some of the posters I marked down heavily didn't end up getting accepted. Admittedly, you'll see reject rates of like 20-25%, but that's still not nothing.
  • "Is it as good as a talk?" Probably not. But keep in mind an abstract gets accepted as a talk because it fits two criteria: Quality and Theme. The talk has to be good. But as importantly, the talk has to be on a subject that fits well within a session, so an organizer will grab it. It's possible this doesn't happen, even if your abstract is amazing - sometimes, it just doesn't fit in somewhere.
  • "Is it important?" Yes! It's not the most important thing you've ever done, and it's not going to change your career. But from standing next to posters I've gotten complements - and critiques - of my work that have made it better. I've made contacts that have turned into collaborations. I've talked to important people, and had my work in front of them. Someone I know got a "Here's my card, why don't you shoot me an email after the conference..." Admittedly, I've gotten all that from talks as well, but a poster isn't just some sad way to justify your registration fee.
9

I would guess that this depends on your domain. In computer science, talks at conferences present peer reviewed papers. Posters are also reviewed (e.g., ICST 2011, SEFM 2010). However, in my experience, there are not that many candidates submitting posters, i.e., the reviewing may be less stringent. At some conferences, posters are combined with, e.g., tool exhibitions and/or doctoral symposiums. Advanced PhD students are sometimes encouraged to prepare a poster and to give an overview of their work so far.

8

Sometimes a poster session is better than an oral presentation due to time constraints in the latter. During a typical oral presentation you will have 10-15 minutes for questions and comments while a poster session lasts for few hours, allowing for more in-depth discussions.

Moreover, not all people will manage to make it to your oral presentation because there will probably be many panels running in parallel. Poster session will typically run after all oral presentations have been finished.

8

One thing that seems to be missing from this discussion is the maturation process of a research effort.

A poster session is ideal for the early stages of the research, when you may not have much more than an idea, and you stand to benefit a lot from discussing your idea with other researchers from the same field. You may get some valuable pointers, and perhaps other conferencegoers will find your idea interesting. I've had some productive conversations with presenters at poster sessions.

A conference talk is better after you have some findings, formed some conclusions, and your are ready to present your results to other experts.

If you have a chance to collect even more data to bolster the validity of your research, you could eventually get it published as a journal article.

So, if you've been working on something for six or seven years, then, indeed, a poster session wouldn't be anything to write home about.

I agree with rule's overall sentiment: Poster Acceptance Means Nothing. In other words, a dozen poster acceptances won't put you on the fast track for tenure. However, that doesn't mean the activity has no value. Poster sessions can be especially beneficial for budding researchers, getting them used to talking about their research with experts from outside their university – maybe for the first time. Plus, some conference attendees will be interested in what they might see talked about at next year's conference.

So, are poster sessions a crowning acheivement? Hardly. But they still can be a productive stepping-stone, which is why I think it's sad when they are looked down upon.

6

Recently, several AI/CS conferences, such as AAMAS, or IJCAI put quite a bit of weight on poster sessions. Either each paper is required to be presented as a poster as well (AAMAS), or only selected papers get oral-only presentation and the rest poster-only presentation without making any difference in the conference proceedings (IJCAI). Poster sessions serve for dissemination of good quality work, being either promising, but still relatively preliminary, or targets a narrow audience, instead of general community.

Personally, when planned and executed well, I find poster sessions very useful and at times even more useful than plain oral presentations. Interactions by posters allow for more involved discussions with the authors. Of course this requires a good thought at logistics of poster sessions from conference organizers.

4

I basically agree with Rule 4, and analogously feel that giving a contributed paper doesn't count for much (and carries almost no evidence that your work has been vetted).

However, I would encourage you to see a poster presentation (or giving a contributed paper) as an opportunity to practice your presentation skills.
One advantage of a poster session is that you can try explaining a concept in different ways, and see which explanation is most well received. Although you're obviously honing your ability to communicate verbally, you should also be honing your ability to present information visually in a simple way. Many of the same principles that make for a good poster also make for good slides. And these principles are far from obvious to most people when they start out.

1

I'd say that there is a big difference in breadth vs depth. If you have a paper on a broad subject that is interesting to 50%+ of the researchers in that conference, then obviously a speaking slot is needed, but if you have some advances that are important but in a niche that really interests 5% of the attendants... then a poster would be more efficient in dissemination than a speaking slot.

I've been to conferences where I'm interested in something like 5 out of 50 posters, but those 5 topics were very important to me - it's impractical to have everyone speak, but it would be bad if those poster presenters wouldn't be there.

0

It's very field-dependent and conference-dependent. I can only answer based on the field in which I work (computer vision).

Generally speaking, the way big computer vision conferences like CVPR work is that you submit a paper, which is peer-reviewed, with the reviewers indicating whether they feel your paper deserves an oral presentation, a poster or rejection. In terms of prestige, it's seen as being distinctly better to get an oral presentation than a poster, but getting a poster is a non-negligible success: the important thing is that your peer-reviewed paper got accepted, and will be published in the conference proceedings. Indeed, most of the papers at conferences like CVPR are presented as posters rather than talks, but they're still very much seen as high-quality work.

Unlike in fields with different conference cultures (e.g. that mentioned by @J.R.), you won't get to present a poster at a conference like CVPR if your work is at an early stage: you'll just get your paper rejected. It's certainly not a medium for disseminating incipient work. Moreover, if you have twelve CVPR papers, most people would think you're doing pretty well, regardless of whether they were presented as talks or posters (primarily because in this context, poster acceptance => paper acceptance, and papers at major conferences are valuable).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.