I’m engaged in a fairly major research effort right now. We’re trying to understand some business conditions a little better, and are using a variety of tools to try get a handle on it.
Oddly enough, I also happen to be taking a graduate course in research at the moment.
And I’m wondering about research methods, mostly because I’m investigating something that I don’t want to assume that I know anything about. And when you’re doing that, the standard survey will not work.
Think about it: a standard survey has questions and lists of potential answers. To create that, you need to know (or think you know) at least a large fraction of the possible universe of answers. In my case, I don’t want to assume any of the answers.
So I’ve constructed three one-page “surveys.” They’re basically short-answer questions. And I’m wondering if I can apply a sort of fuzzy analysis to the answers that I get.
My inspiration comes from tagging. Tagging is the opposite of taxonomy. Taxonomy is a science of classification: phylae, categories, rows, matrices. Slots that you create and slots that you fill. A place for everything and everything in its place.
To me, tagging is a much more organic beast. It grows exponentially. It accepts that fact that something things don’t fit into just one category. In fact, many things fit into many categories. It’s an inherently scalable way of dealing with complexity – because in the tagging world, you don’t have to manage that complexity. You don’t have to beat it into intellectual submission, understand it, categorize it, or make it all make sense.
You just do it … and “it” builds “itself.”
I’m wondering if we need to develop new ways of analyzing and modelling datasets that are (self)organized by tags. Maybe they already exist. Maybe a hundred postdocs are already hot on the case.
I hope so, because I’m going to be getting a lot of fuzzy data. And I’m planning on tagging it and putting it in a shaker and seeing what comes out.