#rds2013 Managing Research Data: How I put talks together and thanks to CDRS

The Center for Digital Research and Scholarship (CDRS) at Columbia have done a truly magnificent job of capturing the Managing Research Data event. As a result we have access to videos, aggregated tweets etc. http://www.ustream.tv/recorded/29603442 contains my presentation (mins 31-50). Others have also contributed tweets analysis (see graph), tagExplorer (awesome), aggregation etc.

The reason it matters is that in my presentations I never know what I am going to say in detail. I work hard beforehand to get the most likely material into my head and turn it over and over. Normally I have probably 1,000++ “slides” to choose from, organised in directories, and make a list of the most likely directories. I often blog my thoughts beforehand and this helps in several ways: listing the items, possibly getting comments from the world and refining, and also something to fall back on if I can’t present from my own machine. (This happened on my farewell to CSIRO – because there were remote attendees the connection had to be central and I couldn’t use my laptop. But all of what I wanted to say was already on the web or the blog).

I am often nervous before talking. This is a good thing. It means I am taking it seriously. Indeed a very good touchstone is that if I take a talk casually (“I’ve done that before so I don’t have to prepare”) I may give a bad one. Audiences deserve commitment. I have – on 2-3 occasions – been truly terrified (one just a month ago). What matters is getting the right story for the right audience. And because I need to have a feel for the audience it can be very difficult to get it right until very close to the meeting. The “right” talk for the wrong audience can be a poor talk.

Since I use my own laptop (and insist on it) and because I agree that “Power corrupts; Powerpoint corrupts absolutely” (Edward Tufte) I use HTML. HTML has many virtues – it scales to the window, it wraps, and when I download a web resource I can just use it (although font size can be a problem). The disadvantage is that it is difficult to add multimedia without significant editing and it’s almost impossible to distribute the presentation (Powerpoint is a good container format and I don’t understand why the W3C has failed to generate good container approaches for multiple pages – which would then spawn editing tools). BTW I sometimes do PPT when I am forced to as part of a larger presentation – they want my “bit”.

I also edit my presentation constantly and – if I have the chance – I may be editing it during the speaker in front of me. This isn’t lack of preparation, so much as adding little details that reflect the makeup of the audience. If I don’t have this chance I find that I am always working after midnight the night before.

So when I give a presentation I know roughly what I want to say but I have far too much material to cover in the time. Because the slides are organised through links I don’t have to “skip over some” as so many Powerpoint presenters have to do. I note those slides which are essential and mark them so that I make sure I don’t forget them. I then ask the host to signal when I have 3 minutes left and make sure that I have wrapped up OK – I try never to overrun. I usually put the thanks first since I might forget them.

This works well for 30+ min presentations. The experience is slightly like my skiing – just out of control and having to think ahead while talking. That’s not the same as being rushed – it’s that I have to constantly make decisions about directions. (Linear Powerpoint is just click-click-click).

However one problem is that I can’t easily “mount my slides”. That’s also because there are interactive demos. So whenever I get the chance I ask to be recorded. And yesterday has turned out wonderfully – thanks again CDRS.

Yesterday however I had to fit a plenary lecture into 15mins. That’s tight, especially as I was going to be controversial. I agonized about how to do this. I knew that if I did my normal process I would seriously overrun. I therefore thought hard about using Powerpoint with timed transitions. But I just couldn’t feel happy about that. I had reserved a whole day beforehand to prepare. I was still exhausted from travelling back from Australia and not sleeping very well. So I spent the day writing blog posts. I wrote 8 posts in a day-and-a-half. They have the added advantage that people not at the symposium can read them. (Many slide presentations often don’t contain explanatory detail). At this stage I was very worried that the presentation would be woolly and unfocused. But during the blog posts I discovered the story to hang the presentation on. The messages that I wanted people to take away (see next post).

I created a linear list of topics. In 15 minutes you have to be close to linear but there were linkouts. I certainly needed to show some of those (e.g. Aaron Swartz). But a list of 20-odd links isn’t exciting as the main stream. So I interspersed those with images from the linkouts. (Generally my images are meaningful – If I show flowers, then the flowers have a clear message. This time I showed a cow on a common – because I was talking about commons).

So the mainstream is a mixture of images, links and short phrases or sentences that I scroll through. I haven’t done this before but feedback was positive. I start at the top and scroll down manually –there is no “slide” but often a concept fits on one screen. I don’t know how meaningful the final result is to people who weren’t there, but at least it links to the blog posts.

The next post shows the tweet analysis of what I said.