I spent the last six months hard at work preparing the bulk of my thesis. My thesis happened to be on a very interesting topic (size estimation for elusive populations) and I thoroughly enjoyed every moment of it *. As I neared the end, I'd accumulated so much esoteric knowledge over the ins and outs of the recurrent event model, I found that I was skimming articles on the topic and understanding them. Fantastic! I know enough about this to teach other people!
With my thesis done, it was time to do just that. So I made this lovely presentation with tons of formulas and tables and hey, a total braindump into LaTeX! Fantastic!
It wasn't until I was dutifully giving my practice defense in front of the other students that I realized just how useless it was in conveying the really important parts of my thesis, namely: why I did what I did, what was new, and what made it interesting. Instead I'd just managed to barf up a slightly smaller version of my thesis, and that wasn't supposed to be my goal. So back to the drawing board.
In my thesis, I have to convey the idea of a full population where only some people are known. So I opened up Photoshop and went to work making my population.
Aside from my rather horrible choice of colors, I think it's clear that this wasn't going to work. So I quickly made a list of the aspects I was going to need to convey:
- A big population
- Of which some are known
- And some of those people are captured more than once
- They have to have covariates
- And, well, there's also an element of time.
- Of course, the methods I developed for my thesis, I need a sensible representation of those, too.
So it was a lot of stuff that I wanted to get across. To distract myself from all the hard work I had ahead of me, I opened up Flash and made a Meeple wiggle around a little and avoid my cursor. "Ahh! Help! An event!" I dunno. It made sense at the time.
And then an ah-hah moment. I've got to use animation, because that's the only way I can really convey a timeline. So I set to work and spent something like 36 hours almost straight through and came up with something amazing! A meeple that wiggled, moved when moused over, and ALSO CHANGED COLORS.
At this point my (developer) husband got back from his conference, looked at what I'd done, and informed me that it was going to be impossible for me to produce what I wanted alone in time for my thesis defense the next week.
So, lucky me, we started in on it together, building up a graphical representation of my simulation study so that I could demonstrate the progression over time, the change in covariate values, and the way in which I built up the missing covariate histories compared to the old ways of doing things. It was a masterful success.
So this was my method, and it worked out okay because I happened to have access to a little extra information on the side. Obviously making a flash game for your thesis defense isn't the right solution for everybody, but maybe for some people it is.
But let me remind you what is probably not the best solution for your thesis defense:
Because you can do better. Take time to think about ways to creatively convey your point and do what it takes to make it happen because as wonderful as your thesis might be, if no one understands it, no one will use it.
* That's not true, there were a ton of setbacks and I'm pretty sure I considered throwing my laptop out the window on no fewer than two occasions.
Danielle has her Msc in Methodology and Statistics and works mainly in R to bring order to chaos.