Goals: Start by thinking about your goals for the presentation. What do you need from the group? Is there a decision you’re facing where you need to think through the options? Is there something that’s confusing you or defeating you? Don’t give us a blow-by-blow account of everything you’ve done. Tell us what we need to know to help you achieve your goal for this meeting.
Limit your content. Lab meeting should be a discussion, not a monologue. If you cram in too much material, you shut out discussion, which is a waste of an opportunity. Aim for 15 slides or fewer.
Describe your data: Whenever you show a data plot, you should explain every aspect of that plot. Read us the axis labels. Tell us what different colors or symbols represent. Then point out the feature(s) of the plot that you want us to notice. Give us time to see what you want us to see. Do this for every single plot. If you skip any step, your audience will be lost. If you feel self-conscious doing this, then you need to get over that self-consciousness. If you feel you don’t have time for this, then you need to be more selective in limiting your content. When you show a schematic or a model, you should also walk us through it carefully, just as if it were data.
Embrace messiness: Don’t discard loose ends and contradictory data. What doesn’t fit? What might lead to a distinct or contradictory story? The group might be able to help you see these things in a new light. If you hide your puzzles, you miss an opportunity to benefit from group puzzle-solving.
Tell a story: Whenever possible, try to tell the story you might tell in a manuscript, or at least a chapter in that story. This doesn’t mean showing all your data. What is means is that you try to sketch the arc of a story. Often you will want to skip some past “chapters” in a given lab meeting in order to focus on recent material: just devote a quick high-level slide to relevant past chapters, to remind us how those chapters advance the story. If your story ends abruptly (because you don’t know the ending), then that’s fine - just get us started.
Help the team: Consider points that might be useful to other projects. What will particularly interest your audience? Is there any news that they can use?
Tell us what to expect: Give us a quick visual roadmap at the beginning. If we know the sections of your talk in advance, we can pace our questions accordingly. If you tell us where you particularly want feedback, we can refrain from trivial questions in the rest of the presentation.
Go big: The presentation screen in WAB236 is small, and many folks are sitting far from the screen. So you need to make your display items big, with big fonts. Consider limiting each slide to just 1 or 2 large display items. Otherwise we can’t see your data and it’s hard to avoid tuning out.
General advice on research seminars (see below) applies to lab meeting presentations also.
Before your start outlining your talk, identify a specific person you want to imagine in the audience. Ideally, this person is a nonexpert, somebody you really respect, but also someone who is somewhat sympathetic. Try developing the entire talk with that person in mind. With each slide or major point, ask yourself, “What does this person know? What don’t they know? What will interest them? What will confuse them?” The more you can put yourself in their shoes, the better your talk will be.
Plan on 1 slide per ~2 minutes. If you have 20 minutes and 20 slides, you have far too many slides; 10 would be a better number.
Don’t write an elaborate written script for yourself before practicing your talk out loud. You should develop your spoken delivery by speaking, not writing. Speaking and writing are different. If you read a script you have written without much oral practice, you will find yourself stumbling over it, and your audience will stumble also. Oral practice will keep your delivery spare and powerful.
Stand up for your oral practice. It makes it feel more real. You should be on your feet as much as possible – not sitting at your desk fussing with slides and typing your script.
Aim to show relatively few data plots (often just 1 or 2 plots per slide), but the plots you show should be rich, and you should spend plenty of time on each plot.
Whenever you show a data plot, you should explain every aspect of that plot – including the axes, the symbols, the colors, etc. Walk the audience through the plot methodically. If you feel you don’t have time for this, then it’s a sign that the data plot is not all that essential, or you are trying to pack in too much material. If you walk them through the plot, then the audience has time to actually comprehend it. But if you just flash the plot up without methodically explaining it, there is really no time to extract any meaning from it.
Establish a clear, strong visual style, and stick to that style in the talk. For example, try to employ schematics that you can re-use at several points in the talk. When possible, try to build up your schematics throughout the talk. Try to use a consistent color-code for different types of data.
If you have text on your slide, you should speak that text aloud. For example, if you have a phrase on the slide, then you should speak that exact phrase. If you have a sentence, you should speak that exact sentence. Don’t paraphrase the text, because this forces your audience to read one text while listening to an alternative (similar but competing) phrasing. This is simply confusing.
Even if you are clear, some audience members will get lost. You can get them back on track by starting with a Roadmap/Outline/Agenda slide, and then revisiting that slide when you transition to a new major section of the talk.
Some points will be relevant for everybody, whereas a few slides will be mainly for experts in the field. It’s OK to have some points “for experts only”, but in this case, be sure to signal this to the audience. For example, you might say, “Now, if you’ve actually done these sorts of experiments, you might know that it’s critical to control for X, and so I just want to show you how we control for that….” Or: “This slide is mainly for the handful of experts in the room, but I need to show it, so please bear with me….”.
Some of your results will be unsurprising, whereas other results may be surprising. Your audience will be more interested – and better able to follow the talk – if you make a clear distinction between surprising and unsurprising results. For example, you can say some results are “as expected”, or “as you might have predicted”, or some experiments were “confirming”. When you talk about something being surprising, you can say “We were surprised”, or “We didn’t expect”, etc.
The Introduction and Conclusion sections are arguably the most important sections. The job of the Introduction section is to tell the audience what gap you are going to fill, and why they should care. The job of the Conclusion section is to summarize the high-level take-home messages and why they should care. If you are tempted to skimp on these sections, or if you feel nervous about them, it’s a sign that you should spend some time brainstorming about your Introduction and Conclusion with a colleague or mentor, to try to refine your big-picture pitch. It’s common to feel nervous about these sections, but the solution is to spend time thinking about them, and get help early and often.
some suggestions from Rachel (specifically re: “job interview chalk talks”)
Your goal in the chalk talk is to communicate your future research Aims and to fosterdiscussion/excitementin the audience. The audience will not get excited unless you let them talk and interact with you. So you must regard this as a managed conversation and not a monologue.
If you have a 1 hr time slot, then you need to plan for no more than 20 minutes of “presenter content”. In other words, when you practice in a room alone, your content should not last >20 min.
Start with a statement of your overarching question, problem, or hypothesis. Try to deliver a bold and confident statement that they will remember. Good eye contact and body language are crucial at this particular moment.
Then do a quick summary of your Aims (1-2 minutes of presenter-content per Aim, ~5 min presenter content total). If you are lucky, nobody will interrupt you during this quick summary. You can try to forestall interruptions by saying at the outset that you are going to begin with a quick summary of each Aim that will last ~5 min. “After I finish that quick summary would be a great time to start asking big-picture questions. At that point I’ll loop back around and discuss each Aim in more detail.”
After you finish your summary, say “This is a great time to ask big-picture questions, before I go into detail.” The questions that will make you shine are the “big-picture” questions… assuming you can answer them. So encouraging big-picture questions is a strategic move. Piddly little questions are dangerous because they tend to pull you down into tiny details that will be boring to non-experts. (Here are some good big-picture questions: “These seem like fundamental questions - can you educate us about why they are still open questions in your field?” “Can you tell us about your long-term vision?” “How might your work change paradigms in your field?” “What is the most exciting thing you think you might discover in the next 5 or 10 years?” “Can you draw our attention to the most innovative part of your research program?” “What’s your approach to balancing risk-taking and exploration with productivity?” “What would you say is the most innovative part of your research program?”)
After your quick summary, plan to go through each Aim again in more detail (~5 min per Aim, no more than 15 min of presenter-content in total). You will be interrupted constantly, so this will last forever.
Try to respond to each question as if you really wanted the feedback – i.e., you must suppress your irritation and impatience. On the other hand, fakeness is easy to detect. So the challenge is to get yourself into a mindset where the valuable thing is the discussion, not the content you have planned. Remember that each audience member will want to talk, and letting them talk makes them happy.
If you are getting dragged down with piddly little questions too early, it’s OK to try to defer one of them, in order to send a message to the audience and to pick up your pace. E.g., “If it’s OK with you, I’d like to defer that question for a few minutes until we get to the part of the talk where we discuss Aim 2 in more detail.” But then you must remember to go back to that question… and that imposes a cognitive load on you.
As you go, it’s extremely helpful to telegraph which Aims or sub-Aims are the “safer” ones, and which are the “riskier” ones. A good research program has a balanced portfolio. Don’t make them guess.
If it looks like you’re not going to make it through your Aims, you can ask them to let you jump ahead. At this point, feel free to break the rules if you need to. E.g., if Aim 2 is sort of standard stuff (slightly ho-hum) and you really want to get to Aim 3, then it would be OK to skip to Aim 3. That said, it’s fine if you never get to Aim 3 if it’s not your most exciting Aim. Do whatever you need to do in order to get them excited about your research. Your goal is not toget throughyour content. Your goal is to get them excited. Use a time-crisis as an opportunity to focus on your best stuff.
If at all possible, prepare the whiteboard / zoom-board ahead of time. You will want to establish a space for each Aim. Draw in any relevant circuit diagrams or other crucial visuals. But there is a chance you won’t get any time in advance to prepare the board (and you will be pretty rattled anyhow), so practice your quick draw skills in advance. If color is important to you, then bring your own whiteboard markers; there is also a slight chance you will be given a blackboard, so make sure you have a plan in mind for monochrome.
Adjust your plans for writing/drawing to your skill level. Don’t make them watch you slowly write/draw for a long time. But don’t write/draw so fast that it’s illegible. If you are not good at writing/drawing on a whiteboard, then plan to use the marker very sparingly, and then do it carefully. It’s better to have a few neatly written things, versus many sloppily written things.
Sometimes people will ask how your Aims might map onto your first R01 application. There is no right or wrong answer here, as long as your answer is coherent. Just be ready for it.