Recently I’ve been dealing with data collected from an experiment where I record what people are doing with their eyes during an experiment. Every 0.002 seconds I sample X and Y co-ordinates of where someone is looking on a screen. Given that the experiment goes for about 20 minutes, I’m collecting over one million pieces of data per participant!

Obviously there’s not enough coffee in the world to make anyone go through each of those data points one by one, so I automate all of my analyses. But automation takes a lot of hours of programming, and therefore a lot of cups of coffee. When programming there is always a solution to any problem, error or bug that arises, and if you’re like me you have to keep working at it until you solve it.

So - I’ll share some data that I’ve managed to process and graph (more caffeine results in more output). Each line on the graphs below represents one eye movement. People were meant to be looking around the four corners of a box, while I flashed targets inside and outside the box. The graphs are actually the same, but the one on the top is the raw data, whereas the graph on the bottom has all the “cheat” eye movements (where the participant looked directly at the target) removed. For more detail, click the graphs.

These graphs represent roughly about two months of my life, or about 120 cups of coffee! The time it took to remove those few lines alone took about 20 hours of work, or four cups of coffee. But like I said before, the more coffee I drink, the more output I create - and now using the program I have written I can process the same amount of data as graphed above in just a few clicks of my mouse.

For anyone planning on doing research I recommend testing out my law, more caffeine equals more output. I guess it doesn’t matter if it doesn’t work for you - just come up with your own as long as it ends with “equals more output”!

**Are you studying Psychology@UQ and want to contribute to theuqpsycblog??**

Send Will an email to find out how: will.harrison@uqconnect.edu.au

Send Will an email to find out how: will.harrison@uqconnect.edu.au

Hey everyone, I've also launched my own website - check it out!

ReplyDeletehttp://www2.psy.uq.edu.au/~uqwharr1/index/Welcome.html

Will, your post doesn't actually say anything about how coffee affects the data output, all you're saying is that you drink alot of coffee while you're programming, and you get results. Where is the Null hypothesis?!

ReplyDeleteAlso, very pretty graphs :P

I would also like to point out that Will drinks instant coffee - imagine what he could achieve by swaping to espresso like the rest of us!

ReplyDelete