P.Mean: Placing the mind of a statistician into software (created 2013-02-19). News: Sign up for "The Monthly Mean," the newsletter that dares to call itself average, www.pmean.com/news.

Someone asked about whether the American Statistical Association was going to "place the mind of a statistician into software." This means, I presume, creating a computerized system that could think like a statistician thinks. I was a bit skeptical and here is my dour reply.

Most people prefer to keep their brains inside their skulls, at least while they are alive. So what you are suggesting, I assume, is some type of process where the statistician talks to an interviewer asking questions and someone translates that into code. Hopefully not COBOL.

I suspect that the process of interviewing is error prone and the process of translating that interview into code is also error prone. So error prone that the process is unlikely to produce something of value when it is done.

There's a great debate over whether Rule Based Expert Systems can perform as well as or better than the average human. I, for one, am quite skeptical. I think these systems have value in helping to insure that an expert has considered all angles of the problem, as it is easy for us to overlook something. But there are too many subtleties that can't be coded, both because it takes too much time and because it is impossible to explicitly describe many of these subtleties. At best these systems might be able to serve as an intelligent assistant. But replace statisticians? Not likely for quite some time, I think.

This page was written by Steve Simon and is licensed under the Creative Commons Attribution 3.0 United States License. Need more information? I have a page with general help resources. You can also browse for pages similar to this one at Human Side of Statistics.