Bringing causal models into the mainstream

Andrew Gelman points to John Johnson who has written a short piece about causal modeling in the mainstream. This blurb seems most important to me -

So, that leaves the last point, which may cause some controversy. "Do not try this at home." Causal analysis does not have a SAS proc or simple R routine (perhaps with the exception of two-stage least squares). This is going to have to come at the end of perhaps hours of data exploration, modeling, testing, rejecting, trying something else, and finally accepting. A causal model is not always going to be easy to write into a statistical analysis plan, and primary investigators may not want something so fluid in the plan.

Although Johnson works primarily in the pharmaceutical industry where the processes of randomization and experiment design are quite rigorous, I think this point is quite important for the social sciences as well. Throughout this year I've found that causal analysis is too often the goal of observational studies, when in fact they should perhaps be a lot more careful in making any strong causal inferences from (what is likely to quite flawed) data. I have taken some extra pains to do this in my MA thesis, and I'm happier with it as a result.