Poster session strategies.

23 November, 2019

Poster sessions. I hate presenting at them, in general. I have had many good interactions, but the negative interactions often overshadow many of the good interactions. The subject of this post is how to get the most of poster sessions as a presenter or a visitor. I’ll also cover inappropriate behavior and some strategies on how to end uncomfortable situations that have worked for me.

I typically attend IEEE conferences in robotics and computer vision, and recently have gone to some plant biology conferences. I’m a white woman from the USA now in middle age. The IEEE conferences I attend are overwhelmingly attended by men. I am used to this, from grad school and work. I am also used to being in all-women’s environments from undergrad. These ideas are formulated from my own experience, so your mileage may vary.

Roadmap.

• General conference strategies.
  • Introductions.
  • Permissions.
  • Inappropriate questions and comments.
• Poster visitor strategies.
• Poster presenter strategies.
  • Badgering questions.
• Conclusion.
• In-text references.

Back to Tips and Tricks Table of Contents

Back to the academic conferences section

\(\quad\)

General conference strategies

First, I will go over some general strategies that are useful whether you are a presenter or a visitor.

Introductions.

I really like an introduction. I feel a little bit uneasy without one, but recognize not everyone feels the need to introduce themselves. Given that, that relic from the last century, the business card, still has a place. When I get a business card, I look at the card and the person, and sometimes the name’s spelling on the card can help me with the pronunciation of a name that is new to me. Graduate students in the U.S.A. – often the printing offices at the university will produce 1000 of these things for you at low cost or for free. And ask the PI to cover the cost, if there is one.

I have noticed some people take pictures of name badges as a substitute for business card collection. For them, that works. But for me, to have someone and their phone hovering around my chest is not my idea of fun, so I take my badge off in this situation.

Ask permission.

Ask for permission before taking a picture of someone and/or their badge. Some conferences mention this explicitly in their Code of Conduct and/or Social Media Policy, and others do not.

Inappropriate questions and comments.

For the record, the following questions and comments are inappropriate the first time you meet someone at a scientific conference. Some are inappropriate all the time in professional settings. Some of these I have observed in my own experience, others are from that treasure-trove called Twitter:

• ask someone on a date.
• ask/make comments about their age.
• ask/make comments about their romantic partner status (i.e. “are you married?”, “do you have a —friend?”, “shouldn’t you be married?”).
• ask if/when they are going to have children.
• ask/make any sexual comments (negative or positive, both are gross, ok?) or have discussions about sex.
• ask someone their race, sexual orientation, and/or make comments about their race, gender, sexual orientation, I am sure many others.
• make comments about their race, gender, and/or nationality’s ability to do certain kinds of research, positive or negative.
• make derogatory comments about a presenter’s institution, x100 if a graduate student.

In brief, the reason these questions and comments are inappropriate is because they are not related to the scientific content of the conference and presentation. A scientific conference is for …. science. As such, if you – as a presenter or as a conference attendee – gets asked this kind of garbage, you are under no obligation to answer or respond.

If you are the kind of person who finds that the entries in the list are part of their usual conversation, and does not understand why the entries are inappropriate or offensive, I suggest a whole book on this topic, Ijeoma Oluo: So you want to talk about race, reprint 2019, ISBN 978-1580058827. While this book deals with race specifically, many other topics are also treated well. The chapter on microagressions is excellent.

Responses I have found helpful include:

• “I have heard your question, but I am not going to answer it.”
• “This is none of your business.”
• Silence.
• Walking away.

One option is answering the question. My experience is that people who want personal information early on will often show a profound lack of boundaries later. And if you have the energy to explain the reason why these questions or comments are inappropriate, by all means do so.

Post-writing edit: I realized I did not talk about anger. Anger is an appropriate response to the types of things on the inappropriate question/comment list too. After all, anger may be an efficient way to get your point across.

Later, if you become professional acquaintances with someone at the conference, it is common to talk about shared personal experiences such as the two-body problem with partners, raising children, hobbies, etc. Sometimes, I volunteer this personal information and let the other person also volunteer information if they want to talk about it (and vice versa).

Anecdote.

At IROS 2015 in Hamburg, Zexiang Li from Hong Kong University of Science and Technology gave a plenary talk. I happened to like this talk quite a lot; in particular, he started the talk with theoretical topics and ended on the subject of optimal conditions for starting companies. In the process, he displayed care for his students (one of whom founded DJI) and pride at what they had been able to do independent of him.

During the question and answer period, instead of a question, someone made a racist comment. It fell into this category: “make comments about their race, gender, and/or nationality’s ability to do certain kinds of research, positive or negative.” Zexiang Li just said, “Thank you for your comment.” and the Q&A period continued.

I don’t have any takeaways, other than “Thank you for your comment.” is another response, and this @#&$&* is persistent. This was during a plenary talk! The moderator did not censure the questioner or make any statement about how such comments were inappropriate. A lost opportunity.

Poster visitor strategies.

As a poster visitor, you may be interested in having more interaction with the presenter, but are unsure how to move the conversation forward. Here’s a selection of topics:

Introductions. Here, you can introduce yourself (Introductions) as well as the types of problems you have been working on. This gives the presenter some idea of why you are interested in the work.

Code. If you are interested in the code, ask about its availability. It may not be listed on the poster. Also, from my own experience, code releases take a lot of time, and are somewhat dependent on community demand. So if you have interest, express it.

Context and Domain.

• What were the circumstances that prompted this work? Why is it important?
• What features of the domain are usually unknown by those outside of it?
• How could I start working on projects in this domain? Is there a community to join?

Project Management.

• How many people were on the team, and of what types?
• What was key to implementing this project?
• If relevant to you, ask about grant programs.

Hardware.

• What tools did the team use?
• Did the tools work well, or not well? What would they change if they did the experiments again?
• What software packages would they recommend for similar work, and why?

At the robotics conferences, I ask people with UAS projects what control packages they are using and if they like them (and why), to help me decide on products for my own research program.

Your projects. Do you want to apply this work to your project? Explain your project’s context and constraints and ask the presenter if their work could be applied to the project. You will save yourself a lot of time.

Impact of this work to other domains. Do you have some insight as to how this work could be applied to other domains? This could be a good conversation. Start it.

Compliments and thank you. If you have a genuine compliment to give the presenter about the work or conversation you had, by all means, give the compliment. These words mean a lot, particularly against the backdrop of the regular disappointments of working in science and other logistical barriers to work that have nothing to do with science (see also: visas, funding, etc.).

More than one presenter. If there is more than presenter, be aware of the underlying dynamics if one of the presenters is the first author and the other is the advisor and/or is more senior. In the majority of those cases, the first author devoted most of the technical effort to the work, so ensure that questions are directed to the first author. In the topic list above, you may want to direct the project management questions to the more senior author and other questions to the first author.

Poster presenter strategies.

At your poster, you may have a mix of people with a large technical range: experts, people with a casual curiosity, undergrads, senior people, etc. The result is that you will have repeated conversations, but also new conversations with the individuals who have different levels of understanding of your area of expertise. This is an opportunity to practice your communication and teaching skills. Senior people can offer you some high-level ideas on your work, while you can practice explaining your work in a big picture way to those who are new to your area.

I encourage you to present to everyone at your poster, and not just to the most senior or famous person. At my last poster session, the people at my poster started having conversations in between each other about the poster and their own work, which I think is a great result.

Introductions. I often ask what the visitor is working on or interested in, and then I can customize the presentation to those topics.

Brief overview. The typical formula for poster presentations is to give an overview of the poster to visitors, and then answer questions and go into detail. This overview talk does not need to be long, maybe 3-5 minutes. While you are giving this talk, other visitors will be joining the audience, so you will have to repeat the overview, or potentially talk about other aspects depending on what people are interested in. For the overview talk:

• start with the motivation for the research. This motivation could be an application, or a problem formulation, depending on the field. The description should be understandable to people outside of your specific research area.
• situate your work within the larger research community and/or literature. i.e. what kind of method is this?
• and in layers go into greater detail,
• support the talk with figures on your poster,
• and end with the impact this work has for your program, the research community, or others.

Within this overview talk, it is important to note that while you may have been living and breathing this problem, visitors have not. Problem definitions are good.

Getting people to your poster. It is difficult to present your work without an audience, yes? Ways to publicize the time and location of your poster include telling people you meet at the conference, and publicizing it on Twitter. Concerning Twitter, I look at poster announcements for conferences I am not attending (and retweet them), so try to make your announcement descriptive to people who are and are not at the conference.

The following formula works well:

• conference hashtag, time and location,
• poster number,
• poster title,
• arXiv link, if there is one,
• picture of the poster, you with poster, or your favorite figure from the paper,
• if space, why do we care about this work? (impact),
• reply tweet for impact if there was not space,
• reply tweet with code link, if there is code.

At the plant biology conferences, speakers promote lab members’ posters during their talks as well.

Given all of this, many times the people who come to your poster are just passing through. You may have to invite them to your poster and indicate that you are ready to give an overview of the work.

Post your hours. Some conferences are now so pressed for space that the posters are only up for the time that they are presented, and then down they go. At other conferences, the posters are displayed for the entire conference, but only presented at certain times. In these situations, it is helpful to list your presentation times at your poster so interested visitors can come when you will be there.

More than one presenter. I dealt with the more than one presenter situation in the following way. Since thankfully, I am past the Ph.D. era, this happened with a student and I was one of the senior authors. The student was directing the presentation and I was in a supporting role. I stood in the audience. If there was a question that the student felt he could not answer, he passed it to me. For context, the student was presenting a very multidisciplinary project at a robotics conference. I answered the entomology and project management questions.

Code.

• Tell people if the code has been released, and/or its time line for release.
• If you don’t currently have a code release time line, and people request a release, start a list and record names and email addresses. You can then balance interest in the code versus the work required.
• Dmytro Mishkin has an article about best practices for transferring your work to others.
• Need a way to log Github views and clones for grant reports and performance reviews? Use Noah Falhgren’s tool to log using the Github API.

Poster content. This post will not cover poster design, which is a big topic unto itself. I will mention that it is helpful to have some examples on the poster that are not in the paper, and that the poster content should be selected for presentation utility. In other words, the best figures for a paper are not always the best figures for someone to understand on a poster, or there may be other simpler figures you can add to the poster, such as flowcharts, that would aid in your presentation of the work.

Props and videos. Many people bring laptops or tablets to display video that supplements the poster. At the robotics conferences, they have had small shelves or tables at each poster that have worked really well for this purpose. Otherwise, sometimes the drinks tables are used. Pack a light device. If you don’t have a device, ask your institution or your PI to borrow one for the conference. I use a very limited laptop for travel (I don’t do programming while I travel), so it was cheap.

Meet your neighbors. When the poster session gets slow, meet your neighbors and ask them about their work. I had a great time at WACV 2018 within our poster corner, in which I met one of the organizers and someone who works on digital forensics. Would I have learned about the field of digital forensics, or gotten some insight about the computer vision community otherwise? Probably not.

See the rest of the poster session! Sometimes, the poster session gets very slow. When this happens, yes, you can sneak off and check out some of the other posters in your area and/or expand the definition of what a poster neighborhood is.

Responding to concerns about your work. Sometimes a visitor at your poster will want to discuss a point of your work in more detail because they do not interpret it as you do. In the early days of giving talks and going to conferences, I found responding to these types of concerns was uncomfortable. The natural response is to react defensively, which will not move the conversation forward. If I do not understand the other person’s position, I will ask clarifying questions. Often, the context the other person imagines, and mine, are totally different. Sometimes, I get a greater understanding of the research problem through these conversations.

There is a difference between genuine discussions about research and badgering, incessant questions, though.

Badgering questions.

Sometimes, there are persistent questioners at posters. I try to show a benefit of the doubt that confusion is related to language use, differences in technical background, and/or cultural differences, and I am very tolerant up to a point.

When it becomes very obvious that the person asking me questions has as their object to engage in a power play, appoint themselves my supervisor, or some other strange goal not related to communicating information, I am not tolerant. My prime objective then is to get this person to stop the behavior, or to go away. Here are some of my strategies:

• “I have already answered that question. What is it that you do not understand about my answer?”
• “If you want a different answer, you need to ask a different question.”
• “It seems that you are not listening to my answers. Why is that?”
• Refer the person to the paper for more technical details, if they are persistent and you are busy with other questioners. Have business cards ready to pass out so they can contact you via email. With this approach, it is great if you already have the paper up on arXiv. Some conferences take 2-3 months now to publish.
• Silence and ignoring the questioner if the badgering continues. Sometimes they feel the need to express their [negative] emotions and any response one offers will be wrong; I was in a situation where I could not fight back because of a massive power imbalance. I let them spend their energy without any response except for shrugs. It ended - finally.
• For a complex 5-part question, agree to answer the first part, and then switch to another person’s question (or then take the next part of the question). You are under no obligation to remember a 5-part question and take 10 minutes to answer one person’s question if more than one person is at your poster.
• Physical strategies to communicate that you are finished with this person: moving closer and/or increasing the volume. Note that I am a tall woman in a field with not very many women, most maintain a large radius around me. You’ll have to judge if this move starts a fight or ends it.

Conclusion.

This has been my perspective on poster session strategies. Many conferences are adopting Codes of Conduct; I see this as a positive development concerning community norms. In cases of inappropriate behavior, you may want to report it to the organization running the conference.

Suggestions? Catch up with me on Twitter @amy_tabb.

In-text references.

• Ijeoma Oluo: So you want to talk about race, reprint 2019, ISBN 978-1580058827.
• Dmytro Mishkin: Machine learning research as a product. How easy is it to use your work?
• Noah Fahlgren: nfahlgren/github_stats (Github statistics logging).

Many thanks to the early readers of this post for your feedback!