Original post date: Oct 27, 2010
I may have acute impostor syndrome like 85% of the time, but there's one aspect of academic science in which I'm solidly confident in my talents, and that's the art of the seminar. Giving a seminar combines two things I very much enjoy: public speaking and making something pretty. First you get to make something pretty, and then you get to stand in front of your pretty thing and talk about it! Does life get any better? Well, perhaps.
Having just put together and given a talk for my interview on Monday, all of my PowerPoint skillz were recently put to the test. I was very pleased with the final product and got some lovely feedback, so while it's all fresh in my mind, I thought I'd share some of what I think are the most important considerations when getting ready for a talk. Please add your own in the comments!
1. Tell a story. This is huge, and so few people do it well. Trust me, nobody in your audience is there to witness a data barf-o-rama, and nobody is going to think you're hot shit because you have 150 figures. Instead, they will be confused and/or think you're annoying and probably tune out. Pick your data that genuinely make sense to present together, and build your story around that. Start out by telling them why they should care about your topic, and take them through your thought process for how you decided on the experiments you did. If your talk is a two-parter, tell them that right up front; that way when you switch gears it won't come as a surprise. People love it when the things they were told to expect actually happen.
2. Pick a clean, simple look for your slides. All of your slides. The same look. This means no gradient backgrounds, no patterned or photo backgrounds, and keep the font continuous throughout. Sans serif (but obvs. no Comic Sans). Dark-on-light is far easier to read than light-on-dark, and the inexplicably popular canary yellow on royal blue makes me want to rip my eyes out. Don't do it.
3. PowerPoint slides are free! In other words, there's no need to cram six graphs (or a ton of text, for that matter) into one slide. Don't you want people to be able to see your data? If the answer is no, don't f%*king present it. If yes, blow that shit up, click "Insert New Slide" and spread it out. It will help keep your talk nice and paced, and it will keep the flow going better, as you won't have to start each slide with the maddening "I know these graphs are kind of hard to read, but if you look at this tiny data point on the 3rd panel from the right..." The more talking about data and the less making apologies you do, the more your audience will like you.
4. Come up with a concrete intro. One of my biggest pet peeves is when people start their talk by saying, "the title of my talk is blah blah." People can read, dude! Your title is there in like 74 pt. font. Tell me why you're here!
5. Plan your transitions. Whenever I feel like I have my talk basically done, I go through slide by slide, and write down what I think I'll say to move from one to the next. Nothing says "amateur" like a person who looks surprised to see the next slide that pops up. You know what else looks amateur? Animated slide transitions. Every time I see a talk where someone decided to have each slide "fly in" or do "venetian blinds" or whatever, I want to punch them in the face. Why are you trying to distract me? I'm a scientist, not a newborn.
6. Summarize along the way. Whenever you've presented a good chunk of data, it's a nice idea to have a slide to recap, because people have probably forgotten already. State 2 or 3 main findings for people taking quick notes, and they might actually remember what they learned from your talk after cookie time is over.
7. Lay off the laser pointer. Another one of my pet peeves is people who feel compelled to point at every word on every slide, like a bouncing ball in a sing-a-long. Not necessary! As noted in No. 4, People can read, dude. The laser should be used only to point at something that actually needs pointing at.
8. Practice for timing, you egocentric asshole. We've all witnessed the following: "Just 2 minutes left? OK, um, I'll just fly through these next 20 slides then...oh, but I really just want to show you this great new data we just got..." By the time 15 minutes have gone by, the entire audience is shooting eye-daggers at the speaker. Do you want to be the target of eye-daggers? I didn't think so. Practice your talk out loud--in front of your lab, your roommate, your significant other, your dog--I don't care. Practice it and make sure that you end when you're supposed to. You're telling a story, remember? Finish. The fucking. Story.