Psych/stats research problems

May 13, 2009 8:48 AM

I'm looking for fields of research involving emotional awareness and regulation which need a lot of statistics and machine learning.

I am a computational biologist. I like solving inference problems using Bayesian statistical models. Over the past couple of years, my interests have shifted. I've been reading a lot of psychology books, and started volunteering as a counselor on a suicide hotline. I like this kind of work, and I'm considering a career change. Some kind of work along these lines which also played to my strengths as a statistician/mathematician would be nice. The more I think about it, the more this seems llike asking how to mix oil and water, but I don't really know where to start in searching the literature, and was hoping that someone more familiar with it might have some ideas.

Here's an example of the kind of project I have in mind. I don't think it's feasible with current methods, but it gives a flavor of what I'm looking for: It would be interesting to find a way to automate the laborious analysis of video footage underlying Randall Collins's Violence: a micro-sociological theory in order to flag situations in which there seems to be a risk of escalation to violence.

posted by anonymous to science & nature (3 comments total)

Computational neuroscience is probably about as close as you're going to come. The wikipedia article is a decent start. I know grad students who do all sorts of interesting fMRI studies, which often require heavy duty computational power and stats to make sense of. I can almost guarantee that someone, somewhere is putting clinically depressed patients in the fMRI machines to determine what parts of the brain are "active" and how they differ from normal.
posted by chrisamiller at 9:18 AM on May 13

Yeah, any fMRI study requires major statistical analyses. As far as emotional awareness and regulation goes, you might be interested in the work of Kevin Ochsner. He's a social neuroscientist at Columbia whose work on emotion regulation is pretty compelling.
posted by solipsophistocracy at 10:39 AM on May 13

Bayesian inference is hot in cognitive psychology; this kind of higher-level approach might appeal to you more than neuroscience, if you're interested in stuff with clinical applications. Start by looking at Josh Tenenbaum's page and start following links from there. Psychology in general is very much in need of, and very much welcoming to, folks with strong math background and training.
posted by escabeche at 12:09 PM on May 13