This is the last installment of the Collected Wisdom on Job Talks series (there are 4 posts in all, an intro, a post focused on preparation, one focused on delivery, and this post). Thanks to all the folks who contributed to these ideas. Major contributors include Carlos Martinez del Rio, Julia Parrish, Stephanie Hampton, Doug Levey, Lars Brudvig, Ellen Damschen, Carl Bergstrom, Toby Bradshaw, Ray Huey, Billie Swalla, Janneke Hille Ris Lambers, Emily Carrington, Bill Moody, David Perkel, Richard Strathmann, and Robert Cleland. Feel free to add your two cents. I will update these posts as new material comes in. I am particularly interested in expanding the resources section, so use the comments section to let me know where you got your inspiration for your great talk.
A bundle of tips for building a clear slide
- Titles: Use a simple declarative/informative brief title at top of EACH slide. Not “Effects of A” but “A Blocks B” or “A is Inactive” or “A activates B.” This will help those who doze or looked down or jotted a note when you first said what the slide was.
- Size of lettering. 20 to 35 point in PowerPoint! Everyone makes them too small. 18 points is the lowest you can go for a label. The larger size is for titles. A convenient rule of thumb is that the lowercase letters of any label need to be 1/40 of the size of the picture at least–you can measure this during practice projection. This goes for the numbers labeling tic marks on graphs and the indication of their units too. Don’t be embarrassed to use large letters. Look at any billboard. No printed figure labeled for publication has letters big enough for a projected slide made from the same picture. Re-letter figures from the literature to meet this requirement. You can erase the old letters in Photoshop and re-label in PPT, or cover the old letters with new ones in a white box. To make a test, project something in the auditorium and go to the very BACK of the room and ask if it is clear. Can you really read the smallest letter? Similarly a transparency made with the 12-point type that we would consider generous for a printed document is not visible on projection.
- Lines: Often you will import graphs from another program. Either in PPT or in the original program, make the axis thickness and symbol sizes adequate. A 1 point line thickness does not show. If you import graphs and then shrink them on the PPT page, letters, lines, and symbols will become smaller.
- Colors: Colors are great but try also to have contrasting brightness. Objects differing in color but not brightness are hard to see. Contrast is paramount. Textured backgrounds just make it hard to see the stuff you are presenting. They obscure your message. Dark reds and blues are brilliant on the computer screen but disappear if put on a black background. Dark reds are fine against white. Yellow disappears against a white background
- Consistency and continuity: Movies have a continuity editor who makes sure that the cars stay the same color from scene to scene and people wear the same clothes coming out of the door as they did going into it. Similarly, try to keep the same color/symbol/thickness for control data versus that for test data. In diagrams, represent the same object the same way each time. Use decent graphics and layout and a consistent theme — so it looks and feels like the slides were made for the talk you are actually giving rather than a bunch of slides from other talks
- Transitions: use av to supplement your message, not provide the main entertainment. zooming slides, too many animations, lots of videos, etc. etc. the audience should remember YOU first, your message/science second, and your av third.
You know when you are ready when…
- You know what is on the next slide without having to think, and you can transition easily from one slide to the next
- You have eliminated all the text “crutches” from your talk
- You know what part you would cut if your talk is running too long and you are able to cut it on the fly without a pause
- You are comfortable explaining all of your material and working with the audience directly
- You know what points you need to make for every slide you show
- Watch Amy Cuddy’s ‘power poses’:
- Garr Reynolds’ blog Presentation Zen is pretty good if you can get past the self-promo and wise sage schtick.
- It is worth studying – though not necessarily emulating – the so-called Lessig method.
- Watch, and re-watch, some fantastic seminars. If you are associated with a strong academic department now, the best place to look is the seminar of the last couple of people you hired.
- For better or worse, Powerpoint is the universal medium of research seminars. There’s a growing realization that Powerpoint was adopted without hard thought about how humans process visual and written information. Although this seems like a boring debate, it’s important for everyone who uses Powerpoint to ponder a bit. Michael Alley (Penn State) has done audience-based research on this. Click here for some of his thoughts on this.
 It is excruciating when a speaker says ‘I should skip this since I don’t have time’, and then they end up talking a long time about it anyway (And this is where having less text on the slides useful because it’s way less obvious if you’re skipping something…)