About Us

We are members of the Andrew Gonzalez lab , in the Biology Department at McGill University.
Montréal, Québec, Canada

Wednesday, May 18, 2011

One-Talk Wonder

I just got back from CSEE 2011 in Banff, Alberta, where I was awarded a 3rd place prize for my student talk. I won 4th place last year (and a gift certificate at springer.com, where I picked up some great stats books). I seem to have developed a winning formula, and a few colleagues have asked for tips for giving a student conference talk, which I define as an oral presentation of 10-15 minutes (typically 12) with visual aids (PowerPoint or other ‘slides’) given to an academic audience of people in your discipline, but not necessarily your area of specialty.

I can sum it all up in a few short points, which I will explain in a bit more detail below:

I also have a few Other Notes about using presentation software (PowerPoint, KeyNote, Beamer, Presenter, etc.) and resources that I recommend checking out.

Why should I care?

This refers to two aspects of planning your talk: your personal goals (“I” refers to you), and what you want others to take away or remember from your talk (“I” refers to someone in your audience).

It might help to start by choosing your goals. What do you want to achieve with your talk?

  • Highlight an important recent result of your research
  • Articulate a conceptual argument
  • Raise important questions deserving of discussion
  • Take a deliberate position, thus inviting or inciting others to participate in a discussion
  • Provoke questions
  • Attract attention from certain people (by field or name)

The important thing to avoid at all costs is for your audience to ask themselves “Why should I care (about this talk)?” The answer should be obvious from the beginning to the end of your talk. Once you know what the answer is for your talk, it will help you plan the “story ark”, and the slides.

Less is More

You only have 10-12 minutes for a typical conference presentation. I often see people trying to cram too much information into too short a time. Even if you manage to speed-talk your way through it all, no one will remember it all. Choose one or two “take-home messages” that you want your audience to remember, and focus on those. Anything else will only distract them and you risk them not remembering anything, or running out of time.

I vividly remember nodding off in a talk about greenhouse gases emissions from lakes and wetlands, and having trouble interpreting the magnitude of measured values, until the presenter wisely showed a video of students breaking a hole in the ice on the surface of a frozen lake in northern Canada, and then lighting the escaping gas on fire. Fire shooting out of a frozen lake illustrates the issue far more elegantly and convincingly than a bunch of numbers. The numbers are absolutely important, and should be in a peer-reviewed publication, but the image of a flaming frozen lake is what the audience will remember, and will ensure that they actually remember the talk at all.

Keep it Simple

Avoid presenting too much novel information. Choose one or two salient points that you want your audience to remember (“take-home messages”). Place them into a context using general knowledge that is common to most people in your discipline (Biology, Chemistry, etc.). This could be “hot topics” that are being widely discussed in your discipline, such as climate change, speciation, global change, biodiversity, etc. Your audience is more likely to understand and remember new information when it follows from something they already know. The more specialized terms you have to explain, the bigger the risk that you will lose someone early on, and they won’t be paying attention to the important points later on. Nevertheless, you absolutely should define any technical terms or jargon you are using, but I would avoid it as much as possible. A general context also increases the relevance, thus addressing the most important question of any conference talk: “Why should I care?”

It’s ok to repeat yourself a bit. This holds true for concepts, terms, phrases, and especially visual themes. Consistent visual cues such as colours, shapes, icons, etc. will help to remind your audience of the important points you want to emphasize and help them to avoid getting lost or confused. Feel free to repeat your main findings / conclusions at least once throughout your talk. It may feel somewhat repetitive, but it will also increase the chance that your audience will actually remember it. You can mention your “take-home messages” more than once, in the context of different examples, different data or results.

As Brent Gurd once told me, while he was a PhD candidate at SFU, a talk should be structured as follows: Tell them what you’re going to tell them, Tell them, then Tell them what you told them.

I used to dislike “giving away the punchline” in the the intro or even the title of my talk, but I now realize that it isn’t necessarily a bad thing. Science writers often recommend having the main finding of your paper in your title. Rather than feeling like they don’t need to read your paper or listen to your talk, people will be interested to see what evidence you have to support your conclusions, or how you discovered your results, and you will have an audience.

Keep it Short

Nothing annoys me more than a talk that goes over time. Nobody likes it when a conference falls behind schedule, and it’s our own damn fault for letting it happen. My personal policy is that if I can’t impress you for 15 minutes straight, the least I can do is get out of your way quickly and give the next person enough time, and people in the audience to filter in & out between talks. If you’ve been to a few conferences, you will notice that people don’t tend to stay in one room for an entire session: they often move around between talks, and it is considerate to leave enough time at the end of your talk for this, as well as questions and discussion.

Speaking of questions, it can be strategic to leave some details out of your talk that are obvious to experts in the field, or which will invite obvious questions from them or others. This ensures that people will ask you questions and may help start a discussion, and lets you address those points during an informal period where other people are moving in & out of the room. It can take up to 30 seconds for someone to think of, or work up enough confidence to ask a question in a such a public setting. Having one or two obvious questions can give people a bit of time to think of more interesting ones while you engage with your audience about other things.

Be careful not to leave too much out, or people might get the mistaken impression that you have overlooked something important and they may not bother to ask if they question your competence. Feel free to ask your peers for help judging the difference.

For example, I work on “heterocystous filamentous nitrogen-fixing cyanobacteria”, which I identified using fluorescence microscopy with a special filter, and measured Nitrogen-fixation using the Acetylene Reduction Assay (ARA). The only thing I mentioned in my talk this year was “nitrogen-fixing cyanobacteria”, which was a big part of my main finding, but I didn’t mention any of the rest, or even which genera of cyanobacteria I found. Experts in the field will know what ARA is without having it explained, and non-experts are welcome to ask. I usually expect someone in the audience to ask how I measured Nitrogen-fixation, and I’m happy to explain the method. I did present the experimental design, which is important for interpreting the results. Nevertheless, details about ARA would be distracting and are not really important for the main point of my talk, which was about how rates of N-fixation by cyanobacteria respond to climate change and habitat fragmentation.

Timing my talk is still somewhat of a challenge for me. I find when I practice a talk, I generally take 1 or 2 minutes longer than when I present it in real life. In my case, I do things like repeat phrases I said wrong or forgot in a practice, whereas in the real talk, I just keep going because I’m so afraid of going over time. Other people have the opposite problem: they practice faster than they would when presenting. Having an audience can help keep you more honest about your time, but realistically, you will be practicing more on your own. Learn what your pattern tends to be, and when in doubt, make it shorter. Most conferences will flash you some warning when you have 5 or 2 minutes left. Decide beforehand where you need to be in your talk by that point, and what to skip if you realize you are behind. If you have a plan (and a backup plan), you won’t get as flustered when the warning comes up. Don’t be too disappointed about skipping things because of time. Remember, less is more, and it’s better to have a clear simple short talk, than a long rambling one.

The most important thing is to have a strong beginning and ending to make a good impression. I would recommend skipping a few things in the middle if you have to, in order to finish with a clear message at a relaxed pace.

The pace of introducing information is also important. Your audience will tune out or start checking their email if the pace is too slow, and you risk taking too long or not presenting enough information to keep people interested. If the pace is too fast, you may get flustered, or the audience will be so overwhelmed with information that they won’t be able to process it all. There is a rule of thumb that suggests “one slide per minute”, which I find is a useful guide, although I don’t follow it too strictly. I use it to judge the overall amount of material for a talk, although the ratio tends to fall as talks get longer (an hour lecture only needs about 45-50 slides in my experience). The definition of a “slide” might also vary, depending on how you do “animations” with your visual aids or presentation software. In general, I would define a new slide as anytime you replace everything on the screen with new information: moving or adding elements to the same background is the same “slide” for the most part. The one slide per minute guideline is still useful in this case, because it means not spending too much time with a static display, nor brushing past things too quickly. I also like to spend a little more time with graphs and results, to give people a chance to interpret it for themselves, and to ensure that I take enough time to explain things like axes and legends so that the audience can understand them and come to the same conclusions I did, or disagree and ask me about it at the end.

Practice, Practice, Practice and Rehearse

A talk is not simply a verbal presentation of a paper or idea. It is ultimately a performance, or maybe even a short one-person theatre. This doesn’t mean you need acrobatics and special effects to entertain your audience, but it doesn’t have to be boring, either. The best talks have personality, maybe a bit of humour, and something intellectually or emotionally compelling. If you express your feelings about your work, whether it is boredom, confusion, concern or excitement, your audience will likely feel the same way.

If you accept that you are putting on a bit of a show, then you realize that you must rehearse. I used to avoid practicing a presentation, because I didn’t want them to be too robotic, but I now realize that practicing a talk means more than just memorizing words. It can help to have a “script”: I don’t always write it down, but I meticulously plan the points I need to make, in which order, the appropriate visual aids, and in many cases the words I want to use. I don’t just memorize the words, however, I try to practice the delivery as well, including tone, pace, pitch, volume, etc. Rehearsing, even in my head, helps me to be more relaxed and enjoy the actual presentation, rather than allowing nerves to get the better of me. Yes, I still get nervous before a talk (I often carry a change of underwear in my bag .. just in case), and I have difficulty paying attention to the talks immediately before and after mine. Practice really does help, as long as you practice the important things.

Another aspect worth practicing and preparing for is the technical part: visual aids or presentation software. Make sure you know what platform and format will be used at your conference. Will you be allowed to use your own computer? Unless you are invited as part of a symposium, the answer is usually No. Typically, you will need to provide a PowerPoint or Adobe pdf file that will run on a Windows platform. Even if that is what you normally use, be careful if the conference computers have a different version of the software than what you used. Compatibility problems are always possible, so make sure you give yourself enough time check everything works as expected and to fix them if they arise, to avoid surprises during your talk. I use a MacIntosh computer most of the time, so I have gotten into the habit of using pdf as my preferred format. It means no smooth transitions, but also means fewer conversion or compatibility problems, and a certain degree of platform-independence. I learned last year, however that even this strategy is not problem-free. The “export to pdf” feature of Keynote sometimes creates files that will display incorrectly on a PC computer (apparently Apple has a different view of the pdf “standard” than Adobe). I was able to find a workaround, but it would have caused me a lot of stress if I hadn’t given myself enough time to check and address the issue the day before my talk was scheduled. I have used videos in talks before, which can be risky. If you do want to use a video, I suggest using a widely-supported format such as .avi, and make sure you have an alternative visual aid in case it doesn’t work.

Get Feedback

Get feedback before AND after your talk.

Part of your preparation should include giving a practice talk at a lab meeting, or for a few colleagues in your department before the conference (or at the conference if you’re last-minute like me). Our lab regularly spends an hour dissecting a 15-minute talk, and I find the feedback invaluable for judging which slides are working, and what parts make more sense than others.

If your talk is being judged for a student competition, chances are there are evaluation forms. Ask your conference organizers if you can get a copy of the judges forms for your talk, whether you win an award or not. This will also encourage judges to provide better feedback if this gets to be more common. Not all conferences will be willing to do this, and judges may not be comfortable having their comments read by the people they are evaluating, but it can't hurt to ask, and more requests might encourage conference organizers to make arrangements for speakers to get more feedback.

Other Notes

Bring some water with you in a spill-resistant container. It helps to have something to drink when your mouth & throat get dry, but you don’t want to spill it all over the computer, either. Feel free to take a few seconds between points, especially if there is a results graph up or something that the audience can look at. You may feel uncomfortably quiet, but your audience is not likely to even notice a pause of a second or two, especially if you’ve just given them something to think about.

Regarding acknowledgements, I often skip this part if I feel pressed for time. I still put up a slide with a list of names, but I don’t like to bore my audience by thanking people they’ve never met, and anyone in the audience will appreciate just seeing their names. I’ve actually gotten into the habit of putting important acknowledgements at the beginning, which makes for an easy introduction and helps warm up your audience with information they don’t need to worry about for most of the talk. This is a great place to acknowledge funding sources, important collaborators, big names, etc. On the other hand, having a few seconds at the end to spend with acknowledgements also gives people that time to come up with questions, or relax if their attention is starting to lapse. I’m not sure there are any hard “rules” about acknowledgements, and the choice is really up to you, so I would do whatever you feel most comfortable with, or whatever the people you are acknowledging actually prefer.

You may notice I didn’t discuss what to actually put on your slides. This really depends on your personal style and is less important to me than the other points I discussed. There is plenty of advice out there about PowerPoint-type slides and there are even formal principles to presenting visual information clearly, whether it be graphs, flowcharts, images, colour combinations, or text. Nevertheless, it is important to distinguish these from the most important aspect of a talk: the ideas themselves. Some people spend too much time worrying about their slides and not enough on what they are saying, the pace, and content from the audience’s perspective. If you are wondering how many bullet points to put on a slide, for example, I would say you are asking the wrong question and are probably better off with a different visual aid. I sometimes make a slide of bulleted text when I’m creating slides for a new talk, but they are almost always replaced or at least accompanied by a more visual illustration of the ideas I’m trying to convey.

Your visual aids should complement your presentation, not necessarily be the focus. Do not forget or underestimate the power of oratory. To see how relying on PowerPoint can ruin a good idea, look up “Gettysburg Address” and “PowerPoint” in your favourite internet search engine. It is truly horrifying.

That said, a compelling illustration can make a bigger impact that the most meticulous science. See my point above about showing fire shooting out of a burning lake to make a more memorable point than a bunch of big numbers.

Neil Dodgson has produced a video with some excellent advice for preparing a presentation. I won my first student presentation prize after watching his video, and it got me very excited about my talk again, after I started getting bored and anxious about it.