Presentations are a privilege

by Jane on June 3, 2009

Blogger Seth Godin reminds us that presentations are a precious opportunity to change minds. To translate his advice to a scientific context, I’d make the following two substitutions:

  • “emotional pictures”  —> “screen-filling, well-explained, compelling plots”
  • “to change minds”   —> “to educate”.

Godin partially echoes Edward Tufte’s argument that slideware is a terrible way to share information. Which is precisely why we publish papers, not powerpoint slides, in scientific journals.

However, while experts will read your papers, most astronomers never will.    They may, however, attend your 50 minute scientific talk.  Given that this is how we communicate with each other, how can each of us make our talks more informative, interesting, and understandable?  What resources are you using?  What strategies have you adopted?  What terrible techniques would you like to kvetch about?

{ 5 comments… read them below or add one }

1 Jane June 17, 2009 at 1:23 pm

Kurtis, I’ve been heavily influenced by a few experimental talks you gave at Arizona. When appropriate, I try to use images or animations to put the correct cartoon in the audiences’ brains. And I love the idea of title = take-away sentence. Even if it’s a more boring procedural slide like “Contamination is not important.”

What do you mean about “telling a story”? Examples? I’ve heard this advice before, and am still trying to figure out how to apply it.

I’m curious about the tablet idea -send examples when you’ve experimented more!

Reply

2 Kurtis W June 17, 2009 at 12:03 pm

Two books, Beyond Bullet Points by Cliff Atkinson and The Craft of Scientific Presentations by Michael Alley led me to make a few big changes in my presentation style.

First, I now make the title of each slide a one-sentence takeaway point for that slide. “Impact Probabilities” conveys very little information; “Asteroid B612 has a 91.2% chance of hitting the Earth in 2061” conveys just about everything that needs to be remembered.

Second, I have eliminated almost all bullet points from my slides. Some times it is not possible, but if the bullet points are just a listing of points I want to make about a given slide, I put them in the speaker tools notes. Audiences will read the bullet points and ignore what you are saying or trying to convey, and often draw the wrong conclusions. Also, eliminating bullet points allows me to enlarge graphs so that they are more easily seen.

Third, I try to make my presentation tell a story. Sometimes this means that the typical talk pattern of introduction, experimental design, results, and conclusions gets more mixed up; I may intersperse the background throughout the talk rather than have my audience try and remember an obscure fact 40 minutes after the time I said it.

My next goal is to try and annotate graphs during a talk using a writing tablet. I have the tablet, and I know it is possible to do, and I can see graphs where that would be useful. I just need to spend the time to get it all working and to make sure I can do this seamlessly.

Reply

3 Jane June 18, 2009 at 8:21 am

(I just ordered the books you recommended.) Thanks for this helpful discussion, Kurtis — this is really great stuff. I hope AstroBetter readers find it useful, too.

Reply

4 Kurtis W June 18, 2009 at 7:54 am

The Beyond Bullet Points book describes the “story” idea in detail, though I’m not sure his explanation fully works for scientific talks. But the basic idea is that the classical drama comes in three parts: an introduction, where the main characters are introduced, and the problem the characters will face is exposed; a development section, where the characters interact with the problem, searching for a solution; and a conclusion, where the plot is resolved and loose ends are wrapped up. In a scientific talk, the problem is the scientific question; the “characters” are more loosely defined. The astrophysical objects are likely some of the characters, but things like observations can be either characters or plot development. So, unlike what is suggested in the Beyond Bullet Points book, I don’t worry quite as much about defining the characters and plot development in my mind.

When I talk about the “story”, what I mean it is to do a little work like an author does in writing a book. I ask myself a few questions before typing a single slide:
* What is the purpose of my talk (to convey research, to get a job, to form new collaborations, to get telescope access, etc.)?
* Who is my audience (an astro department, a physics department, the public, my cats, etc.)?
* What are the conclusions I want to make?

This latter one has, to me, proven the most important. It ties into the other two points, but I let the conclusions guide the design of my talk. For example, if I want the audience to remember that 7 solar-mass stars may go supernova, but I don’t want to talk about Type Ia supernovae, then I just avoid talking about type Ias, even though I find that area interesting and often desperately want to slip something in.

The next step (for a new talk) is to write my slide titles. It is possible to take lists from Word and import them into PowerPoint as titles of slides (and presumably from Pages into Keynote, though I’m a recent migrant to Keynote), so I can write the titles all on the same page. I basically try and use the titles to outline the thought process I want the audience to go through.

After I import these titles into PowerPoint, I add the graphics. For some of my titles, I find that there is no appropriate graphic. Often, these tend to be conclusions I am drawing from combining previous data. Here I still often struggle to decide what to do. Sometimes I eliminate the slide and just say the point out loud; sometimes I make a bulletted list of previous slide titles and re-iterate those, pointing how it leads to a conclusion; sometimes I repeat a graphic to emphasize how those data lead to that conclusion.

After I’ve written the talk, I go and look at the slide sorter page. If I’ve made the take-away point of each slide the title, I can “read” the story of the talk by just reading the titles. A bad slide title will interrupt the flow and stick out. Then I can delete the slide, hide it in case of questions, or re-work the slide. For example, a AAS talk may read:
* Hot DQ white dwarfs are mysterious; we don’t know their origin.
* The structure of a hot DQ white dwarf may give us clues to their origin.
* Asteroseismology allows us to study the interior structure of pulsating white dwarfs; do hot DQs pulsate?
* Hot DQs could also by Type Ia progenitors, though we doubt it.
* Hot DQs do indeed vary!
* Hot DQ variations don’t look like normal white dwarf pulsations.
* We need more data on these mysterious hot DQ white dwarfs!
Here the Type Ia title sticks out as otherwise obstructing the flow of a story, so I’d take it out (Actually I’d probably combine it in my explanation about the first or second slide).

I guess my main point is that the slides are one of many tools used in a talk, they are not the talk itself. And, like any tool, they should be used when appropriate and useful, and not used when they aren’t appropriate. If a given point is better made by drawing on the white/black board, or by hand waving, or by singing an operatic aria, then do it that way.

One last tip that I try and use is the “blank screen” button. If you want to explain something and need the audience’s full attention, or if there is no appropriate slide for a topic, or when you are taking questions at the end of a talk, blank the screen. That forces your audience to pay attention to you, not to whatever the point on your last slide may have been.

Reply

5 Kurtis W June 18, 2009 at 7:57 am

Oh, one other point that I often forget. If your audience has a lot of non-native English speakers, then more words should be included. But, again, don’t write out your talk on your slides. Short, important phrases are very useful. Long sentences will slow down the non-native speakers as they try and digest it, and will distract the native speakers from what you are saying.

Reply

Leave a Comment

Previous post:

Next post: