Data analysis and life

And ye, verily did the hordes of students return to the hallowed grounds of UQ, and mightily did we curse as lines at the Refectory and Merlos did lengthen accordingly.

This week, I find myself preparing for data analysis, and thus I shall write about it as well.

There are many different words and phrases for data analysis in the sciences. For instance: 'A love affair with statistics packages', 'Listening to music while attempting to break your keyboard', or more simply 'being a COLOSSAL nerd'. Any and all of the above are appropriate. The funny thing about it is that should you have any passion for science, you'll find yourself strangely immersed in not only entering the data, but then running each successive analysis, reviewing the results, interpreting said results and so forth.

It can feel like an anticlimax; you've worked and experimented (for weeks, months or perhaps even years) and now you have this great pile of data that you have to compile and work into something understandable. Not for yourself, as you should hopefully know already what you're doing. But going from having that data to creating a story that is understandable and interesting to other people is a skill, and the entire process is both a privilege and a pleasure. That's not to say it's not stressful and difficult as hell, as it certainly can be. Sometimes things won't go according to plan and you will wish to create a fist-sized and -shaped hole in your computer screen and the wall behind it (much to the shock of the inhabitants of the next room). You will almost certainly have a few late nights in the office. In spite of (and perhaps because of) this, it's totally worth it.

I personally doubt that many of us do this because it's a challenge. I certainly don't. It is absolutely a challenge, and I am expected to run experiments, to publish, etc. But as one of the senior academics said earlier this year, academia is/should be as much a vocation as anything else. So I do it because I love it, and the aforementioned necessities just serve as to legitimise the whole affair as a job.

Interestingly, people look at me funny from time to time and say 'but PhD's make no money' or even more amusingly 'that sounds so boring!'. Regarding money, it's cliche but I'd vastly prefer to do a job I enjoy rather than one that paid well, assuming I had to choose. I once tried working in an office job that paid well. I lasted 4 weeks before my brain melted out one ear from boredom. As for a PhD being boring, each to their own.

I have gone from talking about data analysis to a discussion of my perspectives on PhD's. Rather amusing shift, but I hope it proves enjoyable/interesting/informative nonetheless. I am off to finalise an Ethics proposal, so I shall see you all next week!

-Harrison

Thinking about talking

It is the holidays, apparently. For many postgrads this reads as 'work time where one does not teach'. For a PhD student, the concept of holidays is largely a self-directed process. It is perfectly legitimate to wake at 11am on a Monday and simply declare 'to hell with my work today' and spend the entire day in one's pyjamas . . . possibly writing a blog post or two (Note: I didn't actually do this, but the idea is a logical one).

Of course, provided your work ethic is not a withered broken husk that feeds upon guilt and shame (not to get too self-referential or anything) then you probably will come and work. For myself, this means experiments. Speaking of experiments, I mentioned that I would be doing so this blog around, and thus I shall avoid linguistically opening fire on the injustice of working during the holidays, and proceed with something more interesting. Potentially more dry, but just as interesting. At least, as far as I'm concerned it is, and I'm the important one in this relationship :)

The basic experimental design I use at the moment is called the Picture-Word Interference Paradigm, or PWI. Very simply, it is the presentation of a picture, with a word. The task is to identify the picture as quickly as possible. It's my kind of thing; gloriously simple and yet, when properly manipulated, can give you a true world of information to sink one's metaphorical chompers into.

We're using this design to work out how people select words to put into speech. True to form in the sciences, this is not nearly as simple as it sounds. It turns out there's quite a few stages in the speech process. You have a semantic/conceptual stage, where concepts are abstract representations without words or meaning attached (There is research to suggest that concepts are actually complex sensory representations). You then have a lexical stage, where the concept is attached to a kind of grammatical code, called a lemma. Then it gets actual sounds attached to it at the phonological stage, at which point we think it gets forwarded to motor areas to be turned into processes for your vocal cords etc. to deal with. And of course all this happens after your low-level visual processes have gone through and dealt with the basic properties of what you perceive.

Essentially the PWI throws a cognitive spanner in the works; by varying the relationship between the picture and word (sometimes the word is related to the picture, sometimes not), we can determine the process/structure by which we select words and meanings. Happily, in recent years there have been a few papers that generated an almighty stink amongst the researchers of speech production and thus new research is most welcome to try and resolve matters to some extent, which is what I'm aiming to do.

Speaking of which, my work ethic awakens with a roar and a snarl, and so I should get back into it. See you next week!