One of the goals I have identified for this blog is to chart some of my growth as a student over the course of my program. I feel like I need to state that from the beginning of this post, because I am going to proceed to explain my new love and obsession: methods of quantitative measurement.
I have the giddy feelings of zealot right now. You see, I am a storyteller, a writer. I am an ENGLISH teacher, for pete’s sake. I just assumed, upon entering school again with the number one goal of doing lots of research, that I would be doing Qualitative research. Numbers, I thought, don’t tell the whole story. There is so much missing. Maybe I’d throw in some quantitative pieces, just because people like numbers, but my heart would be in open-ended questions and interviews.
What I did not realize was that the numbers were, in fact, telling stories themselves. It just seems to me that the stories are buried in the language of statistics and probability. These are, granted, somewhat complicated stories, but they are explaining the world in their own way. I have given this some thought: it’s not that I am more confident in the results or I think quantitative methods are better. I just find them intensely interesting. I learn something new and I think it about it for days: learning about ANOVA post hoc tests right now, such great names! Tukey’s HSD (for Honestly Significant Difference. Something about the “honestly” tickles me.) and Bonferroni. There’s even one just called R-W-G-W-Q.
What I find most interesting is the rhetoric that surrounds numbers, the way this work lives out in the wild, without a statistician around to point out the nuances. Merit pay is a hot issue in Michigan right now for example. Michigan had changed its laws in order to apply for Race To the Top money, which it did not receive. Now, districts and law makers are scrambling to meet the requirements of the law, despite evidence that suggests that Merit Pay is actually not really a good way to improve student achievement. Yet here we have all of the biggest wheels turning for an expensive program that shows little or no results, despite the best data and analysis we can do. The numbers tell a story, but in this case, the listening isn’t going so well. Citing a statistic is a powerful way to put forth an argument, but I don’t think we as an audience for these arguments have a good way to question and evaluate those statistics.
In the end, I am dreaming about planned contrasts and regression. I think about it all the time. I want to talk to someone about it, but most people look at me weird and leave the room. I knew I had a problem when I got really taken with Holm’s test and looked up the original paper and read it ON MY PHONE. I just had to know.
Next up: learning R. Anyone want to help?
More fun with statistics for you:
- Andy Field’s Statistics Hell
- My new favorite social network, Cross Validated: http://stats.stackexchange.com/
- G* Power
- The R project for statistical computing