By day I'm a scientist. Really, I'm a facilitator/force multiplier for the scientists around me, but I have a scientific background and I do enjoy the field. Most of the time my job is to make sure people know what they're doing, make sure all the necessary materials are available, and to help out where I can, especially on some of the more onerous, time-consuming tasks. Sometimes as part of my job I get to go to lectures/talks where people present their research, which is usually a bit of a treat and usually something that gets me excited about the science I do.
Yesterday I went to a talk that failed to excite me and I've spent a bit of time thinking about what was wrong and how it could have been better. Not that I'm likely to ever give a talk again, but I do think there are some basic storytelling ideas that are applicable to other projects I'm working on. I'll talk about those at the end.
First, only present questions you're going to answer. A talk should consist of the presentation of a problem (big question) followed by how you're going to address that problem (data you're going to collect). Basic, huh? I think it's very important to draw a connection between the question you're asking and the measurements you're making. Otherwise it just feels like you're throwing a bunch of measurements at the problem and seeing what sticks. There should be some mechanism that connects the problem with the proxy.
It's also a good idea to limit the number of questions you present in a given talk. Setting up a problem and setting up an interpretation take a bit of time. If you try to cram too many ideas into the talk your audience isn't going to be able to follow because, more than likely, you won't have time to tell us what we need to know. That doesn't mean you can't have questions come up through your data--realistically you should have some number of unanswered questions at the end that are suggested by the data. But the big question at the beginning should have been addressed in some way and those additional questions should be "tantalizing questions we hope to address in the future."
Second, the data should be presented in an easy to read way. Plots should be legible to the audience--fonts big enough, colors neither garish or so subtle they can't be read, everything labeled so the audience can understand the figure without the presenter guiding us through.
Also, talk about all the data you put up! As a scientist I am well aware that we measure far more stuff than is ever seen by anyone outside the lab. I know how much work goes into some of these measurements that may never actually see the light of publication. Still, if it doesn't help you tell your story it's a distraction. I start wondering why it's there and what it tells me.
Anyway, I know this is supposed to be my fiction blog so I'll try to make a connection to fiction here.
The first connection is, I think, pretty obvious. By the end of a story the major plot line should be resolved. That's one I've had a few issues with myself. It's been much more interesting to me to create new problems than to resolve any of them, and that's lead to some pretty unsatisfying stories. My most successful stories haven't actually been stories as much as they've been character sketches of a moment of decision, which is a resolution of a sort, but it's kind a limiting plot-wise.
The second connection I'd make is between making nice graphs and making nice prose. The prose needs to be clear and correct enough for people to understand what's going on. There needs to be enough detail that a reader can follow what's happening and can picture it in their mind.
People always talk about the narrative of a talk, or how you tell a story in a talk and, as a reader of fiction, that's never quite made sense to me. I'm glad that only like half-way through my not so illustrious career I've finally figured out what's meant by that!