Having just returned from a long trip where I gave three talks, one of the first things I saw when I started following social media closely again was this post on how to do better presentations. The advice is the usual stuff-- more images, less text, don't read your slides, and for God's sake, rehearse the talk before you give it-- and it's generally very good. Given the two very different types of presentation I gave over the last few weeks, though, I think it's important to add one note about the design of the visuals, which is this: when you're putting a talk together, keep the final medium and the secondary audience in mind.
I'll attempt to illustrate with slides from my big European excursion. The first talk I gave was a public lecture in Bristol for the Festival of Ideas. Those slides:
A few things to note, here: First, there are a LOT of slides-- 73 of them, for a talk that went about 45 minutes. Some of that is because I've spread out some animation into separate slides rather than a single slide with transition effects, but even if you factor that out, it's be over a slide per minute, which is my usual rule of thumb for these things.
The other thing to notice about these is that there are very few words-- it's almost entirely images, many of them self-explanatory, others very much not. In fact, if I weren't trying to make a point about slide design, I wouldn't've bothered uploading them to SlideShare, because they're so cryptic.
That's by design-- these slides are for use in a public lecture aimed at people who aren't scientists, and there it's much more important to be visually striking and entertaining. It demands a different kind of approach-- I usually refer to this as the TED style-- and a much greater level of memorization and rehearsal. I actually have the text for this entirely written out somewhere (though I departed significantly from it during the actual talk) because it's critical to have the ideas and timing down, as there are very few cues on the slides, and some of the jokes depend on timing. This is a single-audience talk, intended for the medium of oral presentation only.
Now, here are the slides for my two talks at the Nordita workshop:
These are also on the long side, though not as much as the counter at the bottom might make it seem-- both have some "extra" slides at the end; the blank slide marks where I intended to stop, and the slides after that are for extra reference in case anybody asked a question about one of those issues. They're each around 38 slides of actual talk, intended for a 45-minute block; some of these are quick images that weren't up for very long, so it's under the one-slide-per-minute rule of thumb.
The most striking difference between these and the Bristol slides, though, is that they're much more text-heavy. This is partly because of the timing-- I didn't have as many opportunities to rehearse these, and there isn't a full written-out text for either of these. It's also partly the different audience, which required a good deal more technical detail. But mostly it's because I knew I was going to do exactly this: share them on-line.
Well, okay, not exactly this, in terms of uploading them to SlideShare and embedding them in a blog post. But the slides will be posted on the workshop web site at some point, and that imposes an additional constraint on the design, namely that they ought to be comprehensible to someone who's reading them without hearing my prepared text at the same time. They're talks that have to work in two media, for two slightly different audiences. Which means they need to have some extra words-- not a whole bunch of complete sentences, or anything, but enough to give a reader a sense of the context and the overall narrative that you can't easily get from images alone. In the case of the Bristol talk, I suspect someone who wasn't there would have a difficult time figuring out what the hell some of those images are about; some people who were there might be a little hazy on a few of them. For the Nordita slides, I hope the general thrust is clear just from the slides.
That's a factor that sometimes gets omitted in discussions of how to do presentations-- not every presentation is just a presentation. A lot of PowerPoint (and equivalent) slide sets are doing double duty, as both a visual aid for a live oral presentation, and as a record of that presentation that will be studied later both by people who were there and need a refresher, and people who weren't, but still need some of the information that was in that talk.
(And, in fact, a number of images and explanations in the Nordita talks come from online versions of PowerPoint slides used by people whose research I was talking about...)
That's an important distinction to keep in mind when putting together a set of slides for a talk. While there is some advice that works well for any audience-- don't just read the slides, and for the love of God, rehearse the talk before you give it-- there are other rules that might need to be bent to account for a secondary audience who's reading the slides, rather than hearing the talk. In the case of a public lecture, where nobody is likely to look up your slides as a reference, a total lack of words may be appropriate, but if you can reasonably expect people to be reading your slides by themselves at some later time, it's a good idea to have some words.
This is also true of things like classes-- once or twice, I've made class notes available online, and I've had people critique the slides I use in class as "too much text." Which misses the secondary audience, namely students who weren't in class and need to catch up, or who were in class but need a refresher.
There's a whole spectrum here, of course. The sort of thing I was trying to do in Bristol is one extreme, and a class is probably at the opposite extreme, where it's useful to spell out as much as possible. Workshops like the one at Nordita are in the middle, shading toward the "class" end of the spectrum, while something like a talk at a conference would probably drift a little more toward the "all images" end-- while I might put the slides online, APS doesn't host an archive of every PowerPoint presentation used at the March Meeting or anything, so they're unlikely to be widely read by people who weren't there. (Been a while since I did a "regular" conference talk, though; feel free to invite me somewhere to do one, and we'll see what I come up with...)
This is a somewhat narrow point, of course, but it's a thing that often bugs me about online mockery of PowerPoint slides. While it's true that there are cases where people make awful text-heavy slides and read them near the audience, some sets of text-heavy slides are actually a well-done resource for a secondary audience of readers rather than listeners. If you're a good presenter, you'll be aware of both of those audiences, and prepare accordingly.
But for the love of all that's holy, practice your talks before you give them. That's good advice no matter what kind of slides you're using.
Yes, good advice. One size does rarely fit it all. Another thing that bugs me about the typical physicist's talk is that a lot of speakers put more thought into their slides than they put thought into what to say along with the slides. Ie, I sometimes wonder why they don't just send us their slides and be done, why bother with the audio? Sure, not everybody is a good speaker, but sometimes I would appreciate at least the attempt to get information across verbally. (Oh, yeah, I'm totally typing this as somebody is giving a talk ;) )
Certainly there are times when you expect somebody to read your presentation offline--sometimes there will be formal procedures for this, other times I have informally asked (or been asked) for a copy of a presentation, transferred by e-mail or USB stick. In a conference, if I'm citing a published paper, I'll usually include a text box with an abbreviated citation, in case somebody wants to know the source. But a presentation is no substitute for a technical report or a paper. I think it's reasonable to have your lecture notes online, for the reasons you state, but there is always the question of whether a presentation file is the right format for the audience in question. A class, oral conference presentation, or public lecture is almost always best in PowerPoint or something similar, but often a different format will work better for your non-live audience.
There is actually a good reason for limiting the amount of text on your slides, namely that IME one of the common ways for presentations to go wrong is to use a too-small text font. PowerPoint's default font sizes are what they are for a good reason; use smaller text at your own peril. The problem is that often, if you try to put too much text into a box, PowerPoint will automatically reduce the font size to make the text fit. You can usually undo this by making the box bigger, but that isn't always an option. After all, you also have to leave room for the figure, and having too-small text for the axis labels on your figure is another common way for presentations to go wrong.