Writing Workshop: Talks and Posters

One of the underestimated tasks in nonfiction writing is to impose narrative shape on an unwieldy mass of material.  -William Zinsser

[Note: Most of the material from these writing workshop blog posts, plus a lot more never-blogged material, is now available in my book, The Writing Workshop: Write More, Write Better, Be Happier in Academia.]

Researchers often present our work in the form of talks or posters. We like these formats because they are a relatively quick and painless way of learning about each other’s work. It’s much easier to take in a dozen talks and posters during one day of conferencing than to read a dozen papers.

At least, it should be easier. But not all academic talks and posters are easy to understand. Some difficulty is unavoidable because the material is highly technical and abstract, and it will always be hard for non-experts to understand. But some of the difficulty is unnecessary—it stems from bad writing and bad delivery. I know this because I’ve sat through a lot of talks (and stood through a lot of posters) that I found confusing, boring and frustrating even though I’m an expert in the same area of research. In most cases the research itself was fine; it was just the talk or poster that was badly written and delivered. This post explains my six principles for how to create academic talks and posters that are as easy as possible for people to understand and enjoy.

1. Start Early

If you wait until the last minute to write your presentation, you won’t have time to practice and revise it. The more important the talk is to your career, the more important it is that you prepare it early, so that you can practice giving it at your lab meeting, your roommate’s lab meeting, the lab meeting of that professor who asked you to take care of their dog last summer, etc. Each time you give the presentation, you get comments that help you improve it. Over time, it gets a lot better. But you can’t go through these cycles of practice and improvement if you write your talk on the plane on the way to the job interview where you have to give it. Why do that to yourself? Start early.

[Note from Editor, Prof. Lisa Pearl: Perhaps an example timeline would be helpful. I gave an hour-long invited talk on Nov. 15; I started putting it together on Oct. 1, based partially on material I had presented before. I finished a complete draft by Nov. 1 and practiced the hell out of it (i.e., some part of it out loud every day) until I was happy. As another example, for my 30-minute advancement presentation as a grad student, I started putting it together two months in advance, and practiced the entire talk out loud every day for three weeks prior.]

2. Speak at the Right Level for Your Audience

The first challenge of any research talk is to guess what your audience will already know, and what you will need to explain. Occasionally, speakers explain too much. I recently went to a job talk that was very clear, but seemed to be aimed at too broad an audience. The speaker explained absolutely everything, as if the audience of cognitive science professors and graduate students had no prior knowledge of the area. This was annoying and boring for the audience, and the candidate also spent so much time explaining basic-level concepts that they didn’t get to show what was new, exciting and important about their research. I heard one audience member remark afterward that it would have been a good lecture for an undergraduate-level psychology course in 1975. Ouch.

The opposite problem—assuming too much and explaining too little— is far more common. I’d estimate that over 50% of academic talks make this mistake, and with inexperienced speakers (such as graduate students) the rate is about 85%. Speakers usually don’t even know they’ve done this, because most audience members won’t raise their hand and say, “Wait, what? I don’t know what you’re talking about.” Instead, people in the audience will listen politely for a few minutes, then give up and start checking email on their phones. At the end of the talk there will be polite applause and few or no questions.

So the first key to a good talk is pitching it at the right level. To remind you, here are the levels from the earlier post on Defining Your Research Question.

Public level: At this level, you don’t assume that the listener has any specialized knowledge of your field. The rule of thumb is that you pretend to be writing for the New York Times or Washington Post, where the audience is smart and well educated, but knows nothing about your area of research.

Disciplinary level: At this level, the rule of thumb is to assume your audience knows everything that undergraduate majors in your discipline are taught. More specialized terms and concepts (the ones that most people learn in graduate school, or while doing their own research) should be explained the first time they are used.

Subfield level: At this level, the rule of thumb is to assume that your audience knows whatever you would teach in a graduate-level seminar on your research area. For example, if I speak to a group of people who study children’s understanding of numbers, I assume they are familiar with the research literature. But even so, I still define terms that might be ambiguous. For example, the term ‘number sense’ means one thing in the cognitive neuroscience literature and a completely different thing in the education literature. Because I talk to people from both those worlds, I always define the term when I use it.

Lab level: At this level, you assume that the listener shares all the same background knowledge and uses all the same labels for things as you do. You can really only assume this when talking with your collaborators and research assistants.

Most academic talks and posters should be pitched at the disciplinary level. This is the right level when you are speaking at a conference or to a university department where the audience is mainly other researchers in your field. Presentations to broader audiences, such as to faculty from across the university, or to a non-university audience, should be pitched at the public level.

Occasionally you can pitch a talk or poster at the subfield level, but that’s pretty rare. The last time I gave a presentation like that was when the NIH had funded my lab and a few others under the same program, to work on the same narrowly specified set of problems. Part of the deal was that we all had to go to NIH headquarters each year and make a presentation about our progress. So that was a meeting of a small group of people (maybe 10 labs total) working on closely related topics. Those presentations were pitched at the subfield level.

No talk or poster should be pitched at the lab level, but early-career scientists (particularly graduate students) often make the mistake of doing this. I think they get used to talking about their work in the lab, and they don’t realize that the rest of the world outside their lab doesn’t have the same background knowledge. These are the talks where the audience spends the time checking email and no questions are asked at the end.

3. Tell a Story

A lot of academic and scientific talks are boring. Not just boring to people who don’t care about the research, but boring to other researchers in the same area. When you can’t hold the attention of people who have chosen to spend their careers thinking about this stuff, you’re doing something wrong.

Because bad academic talks are boring, one of the most common pieces of advice you hear is that you must grab the audience’s attention and hold it. But how are you supposed to do that? The answer is simple: By giving your talk the structure of a story.

Popular story types like romances, murder mysteries, police procedurals, and even jokes all grab and hold people’s attention in the same way: They create some kind of tension and then relieve it. The clearest example of this is the joke, which in its classic form consists of a setup that creates tension and a punchline that relieves it.

Romances thrive on romantic or sexual tension, which arises when the lovers want each other but are kept apart. The tension is resolved when the lovers finally get together. Adventure stories and thrillers create tension by putting characters in danger. The tension is resolved when the danger is gone and the characters are safe. Murder mysteries and police procedurals create tension through curiosity and unanswered questions: Who did the crime? How will they be caught? How high up does the conspiracy go?

All of us who grew up with popular fiction, TV and movies expect these rules to be followed. We expect that the joke will have a punchline, the lovers will get together, and most important for the present post, the mystery will be solved. We know that the questions raised at the beginning of the story will be answered by the end, and waiting for those answers is what keeps us paying attention. The way to structure a research presentation like a story is by raising some problem or question at the beginning, and then resolving it over the course of the presentation.

Aside: Are Scientific Stories Helpful or Harmful?
As you may know, I’m a fan of open science in general and of registered reports in particular (see here and here). And you may be asking whether there’s a contradiction between the believing in registered reports and encouraging people to make their scientific talks more story-like. The whole point of a registered report, after all, is to take the storytelling element out of science. For too long, the scientific process has been skewed by publication bias and (boring findings and null findings don’t get published), which creates pressure on scientists to ‘find’ surprising and exciting things. This has made a small group of academic parasites publishers very rich (read my rant here), but it has been terrible for science.

To understand why it’s been terrible, imagine what would happen if the only way crime data were published was in the form of true crime stories in the news. Police and public officials could collect crime data but could not make them public, except as part of a good story.

Think about how that would distort our vision of reality. We would hugely overestimate the rate of splashy, attention-grabbing crimes such as terrorist attacks and child abductions (which are in fact incredibly rare). We would also have no way of knowing the actual frequency of any crimes, because only the interesting ones would be reported.

That’s the situation we have in science, because publishers who want to maximize their sales and citations have prevented scientists from publishing most of our data. Most of the data we collect are not particularly interesting. It’s rare that we find something truly surprising (that’s what ‘surprising’ means, after all), but there are plenty of boring data that are nonetheless worth collecting. And if they are worth collecting, they are worth publishing—if only so that multiple labs don't have to spend time and resources collecting the same data over and over again, unaware that some other lab has already done this work.

If academic publishing had not been hijacked by a bunch of greedy jerks during the twentieth century (here’s my favorite article explaining how it happened), we wouldn’t be in this situation. Our best hope for improving things is to adopt open science practices (e.g., registered reports) and support fair open access publishing.

So here’s what I think about scientific stories: When we are doing research (designing studies, collecting and analyzing data, drawing conclusions and reporting findings), we should not have to worry about telling stories. Studies should be published if they ask interesting questions, use rigorous methods, and draw reasonable conclusions from the data. Publication should not depend on the ‘novelty’ or ‘impact’ of the findings at all. (This is a great benefit of registered reports.)

But doing your research is one thing, and giving a talk about it is another. When I give a talk or a poster, I want to make things as easy and pleasant for my audience as possible, so I look for the story in the work I’m presenting. I don’t mean I invent stories that aren’t there; I mean that I try to organize my work in the form of a story. That means I start by introducing a problem or question (or more than one), and then I use the rest of the presentation to answer the question(s) or resolve the problem(s).

Finding the story in translational or applied research

Easy peasy. You’re already working on something people care about; you just have to show them how your particular research connects to the big problem you are trying to solve. For example, if I were doing the research described in the (100% fictional) chemistry example in another post, I might start a talk like this. (Here and throughout the post, the rectangles on the left are slides and the text on the right is what the speaker says while each slide is visible.)

This opening connects a big problem the audience cares about (curing cancer) to the specific question your research addresses (how can we optimize the process of making Talinexatol?) If they want to learn the answer to the second question, they will keep listening to your talk. 

Finding the story in basic research

If you do basic research, you have to work a little harder to help your audience understand and care about the question you are trying to answer. But don’t worry, you can do it. After all, there’s some question in there that interested you enough to make you choose this problem to work on, right? So you just need to show that to your audience.

The easiest way to raise a question in your audience’s mind is by presenting them with a puzzle— something surprising or counterintuitive that piques their curiosity. It could be some surprising facts about the world, or just an apparent contradiction.

For example, I recently gave a talk about how number concepts are related to language. I started by saying, “Here are three things that I think are true about numbers and language: (1) I think that number is unrelated to language; (2) I think that numbers are related to particular languages; and (3) I think numbers are related to language in general. In this talk, I’m going to explain how all three of these things can be true, depending on what you mean by number(s) and what you mean by language.”

Here’s another example, from the Philosophy elevator pitch, which starts like this:

We standardly take ourselves to know a great many things, but there are some apparently compelling philosophical arguments which purport to show that knowledge is impossible. I’m interested in working out how these arguments go awry, and in the process discovering something important about the nature of knowledge (and related notions, like reasons, evidence, and so on).

This is definitely basic research, and the philosopher introduces it by way of an apparent contradiction: As human beings, we think we know stuff. But some philosophers say we can’t know anything. What's up with that?

Of course, when you start your talk with an unanswered question, puzzle, or contradiction, you implicitly promise that you will answer the question, solve the puzzle, or resolve the contradiction by the end of the talk. Of course if it’s a really big question, you probably can’t answer it completely. But you should at least be able to show how your work is contributing to answering it .

Use brief stories to make points within the talk

Even if you can’t figure out a way to structure your whole presentation like a story, you can use stories make smaller points within the presentation. This will still make for a better talk than if you didn’t have any stories, because stories engage an audience’s attention and hold their interest.

My former student Ashley studies how people think about social hierarchies, which includes how they feel about winners and losers. In order to introduce the idea that adults like winners, Ashley sometimes shows a photo of her father, wearing what appears to be a baseball cap with two brims facing in opposite directions. She puts the picture up on screen and says something like, “This is my dad. Can you see what’s unusual about the hat he’s wearing? Yes, it’s actually two hats sewn together. It’s a UC-Berkeley hat on one side and a UCLA hat on the other. You see my brother went to UCLA, and I went to Berkeley. You can see in this picture that he’s watching a football game. It’s actually the UCLA-Berkeley football game. And here’s the key question: Can you guess which team is winning?” Ashley points out that the UCLA side of the hat is facing forward in the picture, and she explains that her dad supports whichever team is ahead, turning his hat around to show his changes of allegiance. Ashley then goes on to present other examples and show some hard data establishing that adults like winners. But she introduces the idea with a story and an image that are relatable and fun.

As another example, my student Emily studies a classic decision-making problem called the explore/exploit problem. She often introduces this problem by give audiences a hypothetical explore/exploit task. For example she might say, “Imagine that after this talk, you decide to go out for dinner. Do you go to your favorite restaurant, or try a new one? To go to your favorite restaurant is to exploit a known resource; to try a new place is to explore. Exploring is considered riskier than exploiting, because you might not like the new restaurant. But it also has potentially greater rewards because you might like it even better than your old favorite place.”

To introduce the idea that different people follow different strategies of exploring or exploiting resources, Emily uses the example of her own parents. (I know, parents again—I swear I don't tell my grad students to mention their parents in their talks!) She describes how they follow a near-perfect exploitation strategy, eating dinner every Saturday night at the same Legal Sea Foods restaurant in Boston, sitting in the same booth and ordering the same meals. She shows pictures of her parents, the restaurant, the booth and the meals. It only takes a few seconds, but it’s totally charming; it makes everyone in the audience smile; and most importantly it clearly illustrates an exploitation strategy that they can understand.

If you happen to study psychology or any aspect of normal human perception, thinking or behavior, you’re in luck. Your audience is made up of (relatively) normal humans, so you can often demonstrate the phenomenon you study by having them do a few trials of your experiment for themselves.

For example, many of us who study number estimation do a demonstration where we flash an image (e.g., a flock of birds) up on screen for a second or so, and ask the audience to yell out how many birds they saw. In this way, can easily demonstrate that there is almost no variation for small numbers (if I show 2 birds, everyone yells out ‘two!’) and lots of variation for large numbers (if I show 20 birds, people yell out numbers ranging from about 15 to 25).

If participants in your studies heard a series of musical notes and then judged whether they were mostly ascending or descending, play the notes for your audience and ask them to make the same judgement. If your participants had to decide which of two witnesses was telling the truth, play the two videos for your audience and ask them to decide. Of course not all experiments with human participants can be demonstrated neatly in a talk. But if you can do this, it’s a great way to bring the research to life for the audience.

Caveat: Stories must be relevant.

If you do use something like a demonstration, an anecdote or an example, make sure it really does illustrate the phenomenon you want to talk about. The danger with stories and examples—because they engage people’s attention so effectively—is that audiences really think about them and remember them. So if your opening story or example implies that your talk will about one thing, and your turns out to actually be about something else, people feel annoyed and cheated.

For example, my student Emily (of the Legal Sea Food parents), way back when she was a new graduate student, was presenting a study of children’s propensity to take risks. ‘Risk’ was operationalized as a choice between two roulette-type wheels: One that paid out a single sticker on every spin, and another that paid out two stickers half the time, and no stickers half the time. Choosing the second wheel is considered a ‘risky’ strategy. (This is a child-friendly version of a task long used to studying decision-making in adults.)

Looking for a fun way to introduce the idea of kids taking risks, Emily started a talk with a picture of kids climbing a tree. She said something like, “Kids make decisions about risk and reward all the time. For example, these kids have decided that the fun of climbing this tree is worth the risk of falling.” Then she went on to present the study about the roulette wheels.

Afterward, some people in the audience complained that the gambling task wasn’t a good way to study why kids climb tall trees, because the types of risk involved seemed very different. Emily felt frustrated, because of course she hadn’t picked the gambling task to study why children climb trees—just the opposite. She had picked the tree example to get people thinking about kids and risk. The problem was that the tree-climbing example had been so engaging that some people in the audience really wanted to know how children decide which trees are too high to climb, and they were disappointed and irritated when Emily's work turned out to be about a different kind of risk.

My point is this: Examples, demonstrations and stories are are like flashing lights and sirens—powerful ways of grabbing and holding people’s attention. Use them carefully.

If it’s not a story, it’s probably a list.

Research presentations that don’t tell stories usually just present a bunch of information in some kind of logical order, which is essentially a list. The presentations may be very well organized, with main points and minor points—effectively sub-lists within the main list. But they aren’t stories, because they don’t raise a question at the beginning and answer it by the end. They don’t create tension and then relieve it, which means they don’t grab or hold the audience’s attention the way a story does.

One way you can tell that a presentation will be a list rather than a story is that it begins with an outline. I understand why speakers like to start with outlines. If you just want to convey a bunch of information, an outline can at least help your listeners structure the information in their heads so that they absorb it better and remember it better. For example, imagine that for some reason, I want to stand up and give a talk telling you my grocery shopping list. I have several options.

1. I could just read you the list: almond flour, butter, eggs, cheddar cheese, salt, pepper, heavy cream, baking powder

2. I could organizing the list better, breaking it up by different sections of the store and ordering it from the section with the most items to the one with the fewest. Then I would have a list like this. DAIRY: Butter, cheddar cheese, eggs, heavy cream. BAKING AISLE: almond flour, SPICES: black peppercorns. If my list looked like that, I might feel the urge to start my talk with an outline, saying something like, I have to get things from three sections of the store: Dairy, Baking, and Spices

3. I could make it a story. I could start with a picture of delicious-looking scones, and say something like, A couple of weeks ago, my friend and former student Ashley texted me a picture of these gorgeous black pepper and cheddar cheese scones that her husband Rob made. I found this really annoying. [First question raised in audience’s mind: Why was Barbara annoyed?] You see, Ashley knows that I recently gave up eating flour and sugar. So it seemed like she was taunting me and my pitiable sconeless existence. But of course, she’s too good a friend for that. It turns out there’s no flour or sugar in these scones at all. [First question answered. Second question raised: How do you make scones without flour?] They’re made with almond flour—Yay! I decided to make them immediately. The scones required almond flour, butter, an egg, shredded cheddar cheese, salt, pepper, baking powder, and heavy cream. I already had the salt, pepper, baking powder and egg, so I went to the store to get the rest of the ingredients. 

Looking over these three options, you can see how No. 2, where the information is organized into an outline structure, is better than No. 1, where the same information is presented as a flat list. But No. 3, which has a story-like structure, is much more interesting than either of the first two. In fact, I’m pretty sure that reading No. 1 and No. 2 didn’t make anybody want to go to the store and buy that stuff. But No. 3 probably made at least some readers want to make the scones. (Here’s the recipe. You’re welcome.)

[Note from Editor, Prof. Sarah Lawsky: I would add a caveat here. On the one hand, the story is much more interesting. On the other hand, it also includes a lot of information I don't need. Depending on what my goal is, I might choose (2) over (3) in some cases.]

4. Use Text Sparingly

A very common mistake made by academic speakers is putting too much text on their slides and posters. In general, humans communicate find speaking easier than writing. That’s why when you are sitting in a cafe with a friend and they ask you about your research, you answer them with speech. You don’t write out a description of your work on a sheet of paper and slide it across the table for them to read.

I don’t mean that visual information is bad. Even sitting in that cafe, you might find some things hard to describe to your friend in words, and you might take out your phone to show a photo, or use pen and paper to make a sketch.

The same is true in talks. Let’s imagine that you are doing a research project where you build a robot that can crochet stuffed toys. In particular, you have designed this robot to be self-aware and to recognize representational art, and you hypothesize that it will be faster and make fewer errors when it crochets a toy robot than when it crochets an otherwise similar toy mummy. How do you describe the robot and mummy toys to your audience? One way is to put written descriptions up on screen.

But in this case, you are asking your audience to read the slide and listen to you at the same time. It’s like asking people to breathe underwater-- they can’t do it! So they’ll either read the slide and tune out what you are saying, or they will look away from the slide to listen to you. It’s a mess, but academic speakers do it a lot. (Even cognitive scientists, who should know better because we study attention.) A better idea is to put images on the slide. The audience can look at the images while they listen to you.

[Note from S.L.: I am even a little stricter. I have a slide only if it has a picture, graphic, graph or very rarely a phrase that I cannot write on a whiteboard. I do not have slides for purely entertainment purposes.]

Here’s a real-life example from a talk I gave last spring at the meeting of the Mathematical Cognition and Learning Society. The purpose of the talk was to tell people that registered reports are a new article type being offered by the society’s journal. Here’s the beginning of the talk, showing each step of the animations. (Notice how I use the W key to ‘go to white’ when I want the audience to just focus on me. More on this in No. 5, below.)

After the meeting was over, I created a separate, stand-alone version of the talk to post online. For the stand-alone slides, I added text to represent what I had said aloud at the meeting. So the stand-alone slides were much busier and more cluttered than the slides I used in the live talk. Here’s the first one.

Notice that the slides for the live talk don’t contain much text. Text is used only to emphasize key words and phrases and to cite sources. Academic speakers often make the mistake of preparing stand-alone slides for a live talk.

Maybe it’s because we’re used to writing papers, or maybe we’re afraid that once we get up in front of the audience, we’ll forget what we were going to say. But for some reason, instead of making slides that illustrate and emphasize our spoken presentation, many academic speakers make slides that compete with and distract from what they are saying. If you can avoid doing this, your talk will instantly be better than 75% of the talks at your next conference. A good rule of thumb is to read aloud everything that you want the audience to read, as soon as you show it to them. [Note from L.P.: I have this rule too!]

Key words and phrases

In the talk above, I wrote out a few important words and phrases that I wanted the audience to learn or remember: ‘Garden of forking paths,’ ‘researcher degrees of freedom,’ ‘H.A.R.K’, and ‘Registered Reports.’ I revealed these words and phrases one at a time when I was ready to talk about them, and I read them aloud as soon as they appeared onscreen.

Quotations

If you quote someone, it's important to get their words exactly right. If it’s a long quotation, animate it to appear one clause at a time, and read each clause as soon as it appears. 

Text in figures and tables

Sometimes convention requires that you include some text in a figure or table that you don’t expect the audience to focus on. For example, some figures require error bars and you should include a label saying what the error bars represent (standard error, 95% confidence intervals, or whatever.) You don’t have to read that label aloud.

When you present a table, you should read it aloud in as much detail as you want the audience to read it. If the table has a title, show that first, followed by the row or column headers so that your audience knows what to expect. Finally, fill in the data themselves. 

How much data should you reveal at a time? That depends on how much you have to say about it. At one extreme, if you are going to focus on one cell in the table at a time, that’s the way you should reveal them.

If you want to talk about the table in less detail, you can reveal one row or column at a time. This might lead to a little distraction as your audience scans the table, but hopefully it will give them  sense of the variables and range of values without you having to talk them through each cell. 

If you only care about a few of the values or comparisons in a table, try to avoid putting the whole table up onscreen. It forces the audience to look at a lot of information that will distract from your message. Instead, figure out how to extract just the relevant information and omit the rest. For example, you might be able to turn the relevant information into a figure.

If you do for some reason decide to show a whole table when you want the audience to focus on just a few values or comparisons, use highlighting to direct their attention. This is not as good as eliminating the irrelevant information altogether, but it’s better than showing the table without highlighting, which leaves people wondering what they are supposed to be looking at. 

Citations and sources

If you talk about your own or someone else’s published research, it’s customary to cite the work by author and year of publication (e.g., Pearl & Lawsky, 2018). I usually put these citations in a lighter color (e.g., gray instead of black) in the bottom-right corner of the slide. Although citations are text, I don’t read them aloud because I don’t expect the audience to do more than glance at them. The same is true for other text indicating sources, like the designer’s name and website on the robot/mummy image slide above, or the URLs listing the sources of the jellyfish and lab images.

5. Give the Audience ONE Thing to Pay Attention To

The reason it’s a bad idea to put text on screen and then talk over it (as in the slide with robot/mummy bullet points above), is because it asks your audience to attend to two conflicting streams of information at once. Audiences are happy and comfortable when you give them a clear signal; they are unhappy and uncomfortable when you send multiple signals that distract from each other.

It’s okay to present visual and auditory information simultaneously, as long as they don’t conflict. That’s why it works well to put the occasional word or phrase on your slides, as long as you read them along with the audience. And it’s fine to put images on slides as long as they support and emphasize what you are saying, rather than compete with it. 

Figures

Figures can pack a huge amount of information into a small space. That’s good for papers, but often bad for talks. For simple figures, you can slow the flow of information to a manageable rate by presenting the figure one element at a time. 

For more complicated figures, often the best thing to do is create simplified version of the figure just for use in talks, as in this excellent example from Rougier, Droettboom and Bourne, 2014. The two figures represent the same simulation, but the figure on the left was designed to appear in a journal article, and the one on the right was designed for use in a talk. Many details have been removed from the second figure, and the yellow box and red dashed line have been added to make it easier for the speaker to draw the audience’s attention to specific details of the figure (i.e., the speaker can say, “Here in the yellow box, you can see . . .” or “The red dashed line represents . . .”)

Go to black, go to white, or mute video

As long as there is something to look at on the screen, people will look at it. When you want them to stop looking at the screen and look at you, it’s time to make the screen go blank. 

If you are using PowerPoint or Keynote, you can press the B key to make the screen go black, or press the W key to make the screen go white. If you are using some other presentation software, you may have a video mute button on your projector controls, which will let you do the same thing.

Going to black or white gives your audience a break from looking at the screen, and helps them devote their full attention to what you are saying. For example, you could go to a blank screen while you introduce a new idea, and then use images to show examples of it. A blank screen is also useful during the question-and-answer session of a talk, when people would otherwise be left staring at your last slide.

6. Practice, Practice, Practice

The surest way to make any talk better is to practice it ahead of time. The more important the talk is, the more you should practice. Start by practicing alone, and move on to practicing in front of anyone who is willing to listen and give you feedback. If you belong to a lab, ask if you can give your talk at a lab meeting. If you belong to a writing group, ask if you can give your talk to them. Here is a list of six things to check for when you practice your talk.

Timing

Do not exceed your time. To do so is rude to the audience, to the next speaker, and to the organizers of your talk. It makes you look unprepared and unprofessional. In their helpful book Public Speaking for Psychologists, Feldman and Silvia (2010) suggest using no more than 80% of your allotted time for the talk itself, leaving 20% to answer questions. When you practice your talk, keep track of your time. We’ve all been to presentations where the speaker gets the 5-minute warning when they still have 20 slides left. So they break into a sweat and start speed-talking like an auctioneer, racing to cram in everything they wanted to say. This is completely avoidable if you practice your talk and time yourself.

Pace

Ask your audience to tell you where you are going through the material too quickly, or too slowly.

Clarity

Ask your practice audience to tell you where the talk is confusing or unclear.

Interest

Ask your practice audience to tell you where the talk is boring or fails to hold their attention.

Check the tech

• Make sure the images show up, the animations work, the videos play, and the audio is audible. 

• If you will travel to give the talk, build in backup systems. When I travel, I have my slides in 3 places: (1) On my laptop; (2) online as a Google Slides presentation, and (3) on a USB drive, in case I have to transfer them to someone else’s computer. 

• If you will be running the talk from your own laptop, make sure to bring all the adapters you will need; don’t count on the conference organizers to provide them. If you do have to transfer your slides to another computer, click through them beforehand to make sure that everything—particularly audio and video—plays properly.

Practice answering questions

Most academic talks have a question-and-answer session at the end. Make sure you include this as part of your practice talk. Many inexperienced speakers live in fear of the question-and-answer session. They’re afraid that the audience will stump them with hard questions that expose weaknesses in their work. But that rarely happens. By the time you give a talk about your work, you’ve been thinking about it for a year or two at least. The audience has only been thinking about it for a few minutes (since the beginning of your talk), so you always know a lot more about it than they do.

Most questions fall into one of three categories. (1) The person asks you to clarify some aspect of your work; (2) they ask how your work relates to something else, which usually turns out to be their work; (3) they ask something totally bizarre that doesn’t make sense to you or the rest of the audience. No matter what kind of question it is, you follow the same guidelines for responding.

First, smile and nod. Try not to look defensive or angry, even if that’s how you feel. Treat every question as reasonable, and every questioner as well-intentioned.

Next, repeat the question. This serves several functions. First, it’s very likely that not everyone heard the question, so by repeating it into the microphone (or loud enough for everyone to hear), you are making sure that everyone can follow the conversation. Second, you want to make sure that you heard the person clearly and that you’re answering the right question. Third, if the question was unclear or didn’t make sense to you, this gives you a natural opportunity to reframe it.

For example, let’s say you’ve just finished giving a talk about your work training dogs of different breeds to find people who are trapped under rubble after earthquakes. Someone raises a hand and asks, “How is this related to corruption in the concrete industry in China and the developing world?”

Your first thought may be something like, It’s not related and that’s a stupid question, but try to find even a grain of a good question in there somewhere. For example, you could say, “Yes, you raise an important point—bad concrete makes the damage from earthquakes much worse than it would otherwise be. So, while we’re working on improving search and rescue with dogs, hopefully governments are working on building safer structures.” Then move on to the next question.

It’s also okay to ask a questioner to elaborate, or to invite them to say more. Often, questions that seem to come out of left field are from people who actually have something they want to say. So, another way to respond to the question about the Chinese concrete industry would be to say, “That’s really interesting. I don’t actually know much about the concrete industry in China, but do you think there could be a connection?”

Often, people will ask you to speculate about something that’s really outside the scope of the work. In this case, you have two options: Speculate (but be clear that you are speculating), or refuse to speculate, but talk about what information could be used to answer the question.

For example, when I give talks about my research looking at changes in social norms around children being allowed to spend time unsupervised by adults, people often ask what I think the effect of constant surveillance will be on the long-term development of this generation of children.

If I’m willing to speculate I might say, “Well to be clear, we didn’t measure effects on children. We just measured adults’ reasoning. But if I were to speculate about the long-term effects on children’s development, I guess I would say . . .”

If I don’t want to speculate, I might say something like, “Well no one knows, because no previous generation of children has been raised like this. What we really need, to answer that question, are some large-scale longitudinal studies that follow these kids for decades.”

Finally, if you get a truly bizarre question and you have absolutely no idea how to respond, you can just look thoughtful and say, “Hmm, I guess I need to think about that some more. Let’s talk later.” But that’s like a get-out-of-jail-free card. You can only use it once per talk.

Principles of Posters

Just like with a talk, you make a good poster by starting early and practicing ahead of time. If possible, practice your poster before it is printed so that you can make changes based on the feedback you get. In my lab, we use a projector to show posters that haven't been printed yet.

The advice to use text sparingly is just as important for posters as it is for talks. Many conference posters are covered with text. I think that because we are used to writing journal articles, our first impulse is to make a poster by cutting and pasting parts of the article onto a poster board. We paste the abstract into one rectangle; we edit down the introduction and literature review to make it fit into another rectangle; and we just work our way through the paper like that. It’s the same impulse that makes people put way too much text on the slides for their talk.

You can see why this is a bad way to make a poster the next time you are walking around a poster session. Notice how you feel when you see a poster that is densely covered with text. It’s repellent—by which I mean it actually repels you. You don’t want to approach it or look at it, or try to read all those long sentences. Those blocks of text make you feel like running in the other direction.

But notice how you feel when you see a poster that is less cluttered, with more images and more empty space. That kind of poster draws you in. It gives your eyes a rest from the chaotic overactivity that poster sessions always have. (Incidentally, if you are interested in designing beautiful posters and slides, I highly recommend The Non-Designer’s Design Book (Williams, 2015). We love this book in my lab and consult it very often.

Once you have designed a beautiful poster, you will have to present it. The key is to make it a conversation, not a speech. One advantage that you have in a poster over a talk is that you are usually speaking to just a few people at a time. So don’t prepare a whole long speech. Instead, prepare an opening statement of a few sentences, and then let the conversation be guided by your visitor’s questions. Here are some examples of opening statements.

In this project, we built a robot that can crochet stuffed toys. We actually gave it self-awareness, and we wanted to know if it would be better at crocheting a toy robot than a toy mummy [point to pictures of the toys on the poster]. But it wasn’t. Here are the results [point to the results section of the poster, which should have figures or data tables with the key findings highlighted]. You can see here that actually the robot crocheted both toys equally fast. So now we think it’s either not actually self-aware, or it doesn’t understand what the toys represent. Or it loves mummies.

We looked at what people remember about interactions with other people. The basic idea is captured in this quote by Maya Angelou [point to quotation and either read it, or pause to let the person read it themselves]. So we brought participants into the lab and we interacted with them for a few minutes, and then we followed up five years later to see what they remembered about the interaction. We found that, just like the Angelou quote, people had a much stronger memory of how the interaction made them feel than of what the experimenter actually said or did.

This was an intervention study to see what motivates kids to practice the violin [point to picture of violin, or of kid playing violin]. One kid was rewarded with video games, [point to picture], one was rewarded with ice cream [point to picture], and one were just told that if she didn’t practice, her parents would be really disappointed [point to picture]. We followed the kids for five weeks, and we found that the video games worked best, followed by the ice cream. Parental disappointment was the least effective motivator.

After your opening statement, the other person can ask you whatever they want to know. This is the big advantage of posters—you don’t have to worry so much about pitching it at the right level for your audience, because your listener can guide the conversation to the right level by asking questions. You can talk about big ideas with some people and technical details with others. The poster is there to provide the visual information you need, such as pictures of your stimuli and figures or tables showing your results.

Your Work in Your Voice

Academic talks and posters should be a relatively easy and pleasant way for us to learn about each other’s work. After all, we all use language to communicate with the people around us every day. But when it comes to talking about our work in a professional setting, we get nervous. Maybe we’re afraid of being too informal, of not being taken seriously, or we’re afraid that we will forget what we wanted to say or say it wrong. So we try to write everything down and lock it in.

I’ve argued in this post that writing everything you want to say on your slides, or on your poster, will suck the life out of your presentation. If you want to write out the whole talk in your notes and memorize it, go ahead. But don’t turn your spoken presentation into a written one, where your audience is reading your slides instead of listening to you. You’ve done some research, and that’s a wonderful thing. You’ve added to the sum of human knowledge. Even if it’s a small contribution, it’s yours—and the world wants to hear about it from you, in your own voice.

Go to List of Writing Workshop Posts

Go to next post