As much as I blather on about loving statistics, I realized that I’ve never really explained what I like about it. So I figured I’d give it a shot.
Surprisingly (considering how obsessively in love I am with stats now), I was hesitant to take the University of Idaho’s intro statistics course that was required for my psych major. But I had to, so I did, and though I had no issues with the class or concepts and did well, I still wasn’t all that enthusiastic about the material in the end. Mean, median, mode, z-tests, and chi-squares. Who cares, right?
I wanted to go to grad school for psychology at that time, and I’d heard from my advisor and several other people that psychology graduate programs really liked students who knew their statistics. So grudgingly I made up my mind to complete at minimum a statistics certificate (like 15 credits of stats courses), or at best get a full minor (which was only two or three more stats/math courses, I can’t remember now).
To help facilitate my weak stats understanding (and to have the class on my transcript), I took PSYC 456: Tests and Measurements in my third semester.
And that’s when things changed.
This class introduced me to the useful aspects of statistics in an applied setting—inter-item correlations and their ability to reveal good and poor test items, predicting student’s final exam grades by their previous assignment and test scores, assessing personality and determining correlations between proposed traits…holy freaking crap.
And it went from there. The next semester I took STAT 401 and STAT 422, the latter full of grad students whose mental asses I kicked on every exam. You all know me and know how insane I get about things once I decide I like them. I had gotten that way about statistics.
Every stats class after this made me love the subject more. Why? I guess because I love how statistical procedures are able to extract meaning from gallons of data that, at first glance, may just appear to be a jumble of meaningless numbers without any pattern. Human beings love to measure things. Statistics allow us to measure with meaning. Take factor analysis. This procedure allows you to take a set of multivariate data (data in which each subject has measurements on multiple components) and “reduce” it down to a smaller number of “factors,” or components responsible for the most variation amongst the subject’s measurements.
Despite the negative reputation statisticians and their methods have gotten thanks to crappy researchers and phrases like “there are three kinds of lies: lies, damned lies, and statistics,” statistics are incredibly useful, incredibly revealing, and interesting as hell. To me, it’s really exciting to be able to describe data in totally new ways by bringing the meaning of huge datasets to the surface for everyone to see and understand what the data are actually saying. Not to mention the power of statistical visualizations—when done appropriately, graphs and figures can speak volumes. Hundreds or even thousands of subjects and numbers can be meaningfully reduced to a couple lines and colors and yet make an incredible statement. THAT’s powerful.
It’s also fun, especially if you don’t quite know what the data will reveal. And regression’s like the psychic component of the math world. How cool?
I love stats. Love, love, love.