Friday, September 30, 2011

Intuition is fed by data.

This post is for my friend Rees, who's  FB wall was beginning to burst with our comments. I've moved the conversation here to save his friends from the excessive verbiage.

In order to save the poor people who originally made comments on that post, I'll respond here.

I think there are two side to the question of how intuition is fed by data. The side which you're focusing on deals with constantly pressing the data for more answers. You're taking information and looking for as many different possible questions which can be applied to it. For example, say you were looking at a list of suicide rates by nation. One way to look at the data is to ask yourself questions like, "Why does Lithuania place first?" "Does Japan rank high because of cultural views on suicide?" "Why do so many Muslim countries fall lower on the list than Christian countries?"

These sorts of questions are great for coming up with research questions. But that's about as far as you can take it. Notice, however, that these speculative questions all rely to some extent on the intuition you already have about social dynamics. In the same breath that you're asking the question you are also formulating a speculation as to the answer. That is one way that data feeds intuition.

Another side to my notion that intuition thrives from data comes from an experimental viewpoint. Whether you're doing longitudinal studies or controlled experiments, the whole point of science is to come up with a question and then choose the observational method that will best answer it. You use a great deal of specificity at this level. You give everything a definition, and you ask discrete questions. For example, you might ask the question, "Does changes in political leadership have an effect on suicide rates?" Then you look for not one data set, but two. You take the historical documentation of the two questions and begin looking for statistical correlations. If you find a result, maybe that every time a country goes through a significant political shift (note: you have to operationalize both the terms 'significant' and 'political shift') then the suicide rate increases, you have found a correlation. That is not the same thing as finding a cause. But you've narrowed your search a little bit. In that way you have fed your intuition at least by now begin able to say, "Well, I don't know what's causing it, but I see that these two things tend to happen at the same time." That's more than you knew before, and it guides your later questions.

I should make it clear that I made up those results. I have no idea if political upheaval has anything to do with suicide rates. My intuition tells me that the two are unrelated. That doesn't mean much, but it's ok to talk about what I think might happen or what might be behind something.

My guess is that your intuition is already well-developed, but that you're just out of practice noticing it. There's nothing wrong with looking at things from different angles. But your intuition shouldn't be something you hide from yourself, or from anyone else. There's a risk of becoming blind to your own bias.