Sunday, June 04, 2006

Results Week Is Finally Here

It has been a rough month. My students are working like crazy, and I have been, too. But starting on Monday, we will start to see some payoff! We can start crunching the numbers. To me, this is the thing that makes this whole lifestyle worthwhile.

I see two reasons to be an academic researcher. The first reason is that one has figured out a way to set one's own schedule, and one has come up with some algorithm for churning out papers. Numbers beget success in this business, and many people with various inferiority complexes learn how to produce numbers without ever producing a single interesting idea or approaching anything akin to advancing science. They learn the churn without having even a vague notion of what it really means to do science. Sadly, I know far, far too many people like this.

The other reason to be an academic researcher is that you are driven by a passion to answer questions. People of this latter ilk can hardly think of anything else when it comes close to time to analyze their data and find out the "answers." It has killed me to be away from the lab this weekend when I could have been helping my graduate students analyze data.

You see, I have this relentless, unyielding drive to know the answer. We spend months designing studies, negotiating IRB Hades, and running participants. In the end sits the proverbial pot of gold at the end of the rainbow: a bunch of raw data waiting for analysis.

Patience never having been a strong suit, I tear through this like a lion tears through a speed challenged zebra. Most of our (and by our, I mean the Indiana group) research involves within-subjects analysis. Without being overly technical, this means that each participant sees a variety of conditions, and therein serves as their "own" control.

It also means that inferential statistics cannot simply be "run" on the data. Lots of coding and recoding must take place before the "answers" can be had. And if nothing else, I am good at the single minded pursuit of this recoding. More times than I can recount, I have stared at an SPSS monitor typing things such as:

if (order eq 1) pa_a_1 = sam_a_1.

Because, in the end, answers come out. Tomorrow will mark a lot of this kind of coding. But by the time I go to bed tomorrow, I will have some answers. If time allows, I will mention them here.

This all started on a September 2000 evening in the office of my master's advisor, Dr. Robert Meeds. We were working on a study at breakneck pace to make a conference deadline. I was introduced to Unix, repeated-measures ANOVA, and SPSS in the same evening. That night launched something great. The next night, my colleague Manish Gupta and I left the bar early to go enter more data. That never happened (read that study here if you agree to follow copyright laws).

I do research because I cannot possible imagine spending my life any other way. I'm still downright giddy that I get paid to do this. I sit at my desk and try to explain how the human mind processes mediated messages. I derive hypotheses and design experiments to test those hypotheses. And in the end I get evidence about whether I was on the right track. Then I start with the next study.

Eventually these experiments are published in academic journals (read some here and here). We do this with some frequency, and this frequency provides good exposure for the universities that hire us. It makes them happy with us. But this frequency is never, ever the grail we seek. Instead, it is a byproduct of the process that we love. It is part of the process that drives us. But it is always the science and never the numbers in the driver's seat.

As my colleague and good friend Dr. Paul Bolls is often quoted as saying, "Trust us. We're scientists."


Post a Comment

<< Home