Data Reduction: What the What?

Okay, so, with the ready and willing help of Luker, I have spent much of my time this semester dismissing quantitative social science work (or at least a large part of it) as trivial number-crunching.  You know, you can't really know what's up with whatever social group you're researching if you just send out a bunch of questionnaires and ask them to rate various aspects of their lives from 1 to 5 (knowing that most of the time, they'll pick 3).  You actually have to go out and be with these people.  Talk to them face-to-face.  Live in their world.  Take a bunch of pretty pictures and try to pass that off as your dissertation (didn't Luker talk about something like that at one point?  Or am I crazy?).  Then, take the greatest hits from your notes, string them together, and presto!  You have an article!

I'm exaggerating, of course.  I know that some quantitative analysis is required even in the squishiest of qualitative research projects.  Count up the number of people you talked to, and then tell us the percentage that have used a chat-based reference advisory service, or whatever.  But then all of a sudden I'm in the middle of chapter 10 of Luker's book, and I'm colour-coding my notes and cutting and pasting them and then colour-coding them again and then arranging them into neat little piles, or else I'm virtually chopping up my virtual notes and passing them through some magical machine that will arrange them in informative clusters.  And at some point I have to talk to a born-again Christian about sex education.  Or something.

My point here is that the methodologies described by Luker in this chapter seem awfully positivist to me.  I'm sure they are useful, but doesn't this stuff go against the salsa dancing agenda she has been trying to establish?  I know that she would said "no", and explain to me that the data that we gathered was 100% salsa certified - that is, we employed the more flexible approach to research that she has outlined.  But still, after all that, we're supposed to designate all these rigid, absolute categories?

When I really think about it, I get the feeling that Luker isn't the most postmodern of researchers (or "po-mo", as she calls it at one point).  All these ideas about moving up "leveling up" the relevance of your research, and finding specific answers to specific questions, and so forth, seem rather absolute.  Not that I'm saying her approach is wrong.  I just find it interesting.

- Matt

2 comments:

Devon said...

I agree! I felt like Luker was trying to make research into a messy process and then suddenly at the end we were suppose to clean it up into a neat, orderly filing system.
I'm glad it wasn't just me that was puzzled by these two conflicted ideas.

Stephanie Lauren said...

Matt - this post was really funny. I now want to include the term "100% salsa certified" in my final paper for research methods :)

While I agree with some of your points though, I think that there is a method to Luker's madness (and her large emphasis on quantitative methods for data reduction). Ultimately, she wants to help social science researchers be able to conduct their small, unique case studies, but she also wants us to incorporate reliability and validity into our findings. By using methods that show you've thought carefully about the structure of your research design and that you've approached the analysis of your data from a systematic way, you'll be less likely to have your research ripped apart by someone more invested in quantitative methods.

On pg. 203 Luker says, "We have to protect ourselves and our work, and... the integrity of the enterprise." I think one way of protecting ourselves is to use more structured methods when we analyze data. This helps us describe how we were able to sift through transcripts of 30 interviews and come up with key themes and findings.

In fact, doing the peer review made me realize how important it is to describe how you code your data and arise at your themes. The greatest problem I had with the paper I reviewed was that the author casually mentioned in one sentence how they had coded the data and then didn't touch on this issue anywhere else in the paper. It made it harder for me to accept the themes she/he presented, as I didn't really understand how specific block quotes from one or two interviews led to the author developing a particular theme. I needed more information in order to understand the study!

While most of us are collecting research through qualitative methods, I think you need to employ more structured quantitative methods such as coding themes (and even cutting out pieces of interviews and rearranging them) in order to really get down to the nitty gritty aspects of your research. Also, there's the more practical fact of how to handle so much information and data. I think using these structured methods in the data analysis and data reduction aspect of your study can really help you recognize themes, make larger connections, and most importantly, help you explain to others how you came to your conclusions.

Post a Comment