Ask an Expert! Or, How Statistics, Facebook and Polychoric Correlation Matrices Made Me My Own Library User
Frustrated with some data and fed up with my own inability to locate an appropriate statistical technique, I finally posted to Facebook in the hopes that a friend would commiserate with me:
"Bending my brain around ILL stats and thinking about exploratory factor analysis with categorical variables, despite the issues with it. Desperately missing [my old group of Emory PoliSci nerdbuddies and profs who were excellent at stats] and brainstorming these sorts of things."
Five seconds later, the prof I had tagged in the post replied, "Three words: polychoric correlation matrix." And I had four distinct reactions in rapid succession. They were as follows:
First reaction: sarcasm. Well OF COURSE polychoric correlation matrix, duh. Who WOULDN'T know that? Certainly not I. Pshaw.
Second reaction: confirmatory exploration. A quick Google search of that conglomeration of words, a quick scan of the Wikipedia description, and yep, this is much closer to what I need for what I want to do than I've gotten scouring statistics textbooks and incomprehensible math journal articles for two weeks. Until my eyes felt like they were bleeding, and my brain was mushy. Until all I wanted to do was curl up and cry in a corner until someone brought me a puppy. (Interestingly, my husband just got me a puppy for my birthday.)
Third reaction: gratitude. Thank you, Jeebus (and Professor Chris Zorn) that I have a direction and didn't have to pray I'd trip over this technique on my own. I was already stretching my husband's patience and our booze budget due to this thing.
My fourth reaction, and the one that prompted the blog post: chagrin. We beg our students and researchers to come to us as librarians for good direction before they get mired in the research process. Why didn't I go to the experts in the first place, the way I beg my students and faculty to do? The way the lit review section on expertise in my own darn dissertation says folks should do?
I know my reasons, and they likely echo those of my researchers. First, I thought I should be able to find the answer myself. Why didn't I ask my local methodologist professor buddies? Well, they're all on my dissertation committee, and I haven't touched my dissertation in forever, so I'm doing some guilt-hermiting (in which I crawl into a dark space and don't contact folks until I have something productive and useful to show them). I didn't want to look stupid for not knowing something, even though that something is admittedly quite far outside my wheelhouse.
Sigh. A lesson learned for myself: if even I fall into these traps, I need to make sure I continue to let me researchers know that it is okay to ask questions, if only so they're not making their lives harder than they have to be. My new mantra: Don't let guilt or ignorance waste your time. Ping an expert. Do it from the beginning.