I attended the school of the Reverend Maclean. He taught nothing but reading and writing. And being a Scot, he believed that the art of writing lay in thrift.
Half as long.
– A River Runs Through It
Most science powerpoints could use a good dose of Reverend Maclean.
A follow up to my talk rant from a few days ago. Over the years, I’ve lost count of the number of talks I’ve heard. Some were truly memorable (in a good way!), and now and then, even a less-than-great talk will still let brilliant science shine through. But most of the talks I’ve seen had powerpoints that detracted from – rather than complemented and enhanced – the core messages. Way too often, you see slides like this:
Here’s the thing. Though everyone learns differently, most research shows that audiences don’t do well at balancing complex slides with what you’re saying. Busy slides are a trap scientists fall into all the time, I suspect because we are so drilled in being complete about the methods we use and the data we produce. Leave details out of a manuscript or grant proposal and prepare to get hammered. So it’s understandable.
But it’s an approach that absolutely KILLS effective storytelling, which in the end is what almost any talk needs to be. I know that’s an uncomfortable feeling for many scientists. For the most part, it’s not really the way we are taught to share information. (Story? What do you mean? This is science dammit! I need to show all my methods and my data!)
No you don’t. You need to show just enough of them to deliver the take home messages in a memorable and credible way. Worried about leaving detail out? Keep the talk on the short side and have the details ready in reserve for any questions. Typically, even the most engaged and up-to-speed audience will only take 2-3 key points home from a talk….unless you cram that talk full of 27 different points. Then they’ll take nothing away.
Look, I was as guilty as anyone else for a long time. Then I heard the late George McGovern talk at a predominantly science meeting. Later that month, former Senator Tim Wirth, in a similar setting. In each case, the politicians blew away the scientists for the power and clarity of their messages. And neither one of them used a single visual aid.
Wirth and McGovern were there to inspire, not to present data. So they had a built in advantage. But the contrast was startling and made me rethink my entire approach to talk preparation. Now, the first thing I do is ask myself: What story do I want to tell, and to do it best, do I need slides at all?
Typically, it being science, the slide answer is yes. We usually want to see some data, and we like to see images that put those data in context. But the answer is not always yes – there are times where scrapping all visual aids makes for a more effective talk. And even for those forums where showing data is clearly preferable, think about the occasional computer failure you’ve seen at a professional meeting – every once in awhile, the person so afflicted turns that fubar into a memorable, beautifully summarized presentation of the main message.
Where slides are needed, most of us could use a crash course in design. Here, as with other aspects of communication, help is out there. Todd Reubold’s advice and presentations are a great resource. So too are sites like slidesthatrock.com, and hosts of others. All tend to converge around some pretty simple rules. Lose the bullet points. In fact, lose most of the text period. That which remains should be large and visually evocative. Images should be sharp, memorable and complementary to the point. Graphs and tables should be as simple, clear and full of contrast as possible.
Here’s an example of how my slide design has changed since paying attention to these kinds of resources. First, an image from a talk a few years ago:
Not so terrible right? Or so I thought. But still, this one slide has ten different bits of information on it, and it was only one of 38 slides in a 50 minute talk. Eleven other slides also had bullet points of some kind or another. Sure, it’s information that might help some people who tune out for a bit, but for the most part, data show that slides like the one above create divided attention streams – a killer to effective information transfer. Better to just say the points in a way that keeps audience attention on you, not your slide. Good stories are not told with bullet points.
Here’s a slide from this month, for a talk with similar overall goals:
Here too, I had a few pieces of a complex story I wanted to summarize as part of the talk lead in. But I did so in speech only, and my sense is that it was much more effective than using the bullet point crutch above. (Though I don’t know, because formal talk assessments are rare – unlike the other products of our business which go through peer review all the time…).
And what about showing data? Often hard to avoid some complexity there right? For the most part, no, if the visuals are planned well. Here’s a slide from a talk I witnessed a while back (blurred on purpose):
Tons of text, bunches of graphs with hard to read axes, etc. The same meeting had several other talks where full page data tables were shown, any cell of which was nearly impossible to decipher. We all see slides like this all the time – from professional meetings to job talks to department seminars. For a paper – maybe ok. Sometimes you just need to stick it all in there. But for a talk, much better to follow Todd Reubold‘s rule: Restrain, Reduce, Emphasize. It’s just Reverend Maclean again, repackaged.
And again, it’s not that hard. Sure, people like Hans Rosling and David McCandless (among others) have made an art out of cool data visualization. If you can pull that kind of thing off, awesome. But high end graphics aren’t essential. Here’s a before and after example from Todd’s presentation, for a mock situation in which the speaker wants to show data on country-level differences in how much biking people do. The first slide is a pretty fair representation of the kinds of data slides we often see (many of my past ones included…):
But what do you REALLY want the audience to take home from your data? In Todd’s example, it’s a comparison of Danish and U.S. cycling rates. So boil it down, and highlight the main point. Easy to do even with basic graphics packages.
Above all…every bit of a powerpoint should be viewed as spice on the main meal: the story you are telling. If you get hit with computer failure, it shouldn’t matter much. That’s where other forms of communication training can really help. Try improv — either formally, or just within lab / student /etc groups. Use video and critique each other. Have fun with it. Watch great storytellers and speakers analytically to pay attention to what works. Notice, for example, that a great speaker will do just what a great book does: vary the pace and flow in ways that further emphasize the key points. Powerpoint design can help here too – for long talks, seek ways to break things up. Toss a short video in there. If you can do funny, do so – not everywhere, but just as an occasional interlude. Consider a completely blank, black slide that forces all attention on you for a particular interactive moment. Use plot twists and surprises. And so on.
Do enough of this and you’ll realize that while truly gifted orators are just that…most people are capable of giving perfectly solid, if not excellent presentations. (Don’t believe me? How many parents do you know who can’t read or tell their kids a compelling story? Same rules apply.) Like anything else we do, it just takes time and practice.
The more I pay attention to two things: Reverend Maclean’s Rule (one I’m clearly not following in this voluminous post…), and a commitment to storytelling, the better my presentations seem to be. A work in progress for sure, but a goal worth chasing — for all of us. Because as I wrote last week, poor presentations slow or even stop the flow of information and ideas. And that ain’t good.