PMean: Fighting SASism

Steve Simon


I told a story today in a webinar workshop that I thought I should get down in writing for my blog. It involves a prejudice unique among statisticians called SASism.

Somehow a talk on survival analysis mutated to a discussion of Tobit regression

I was working with some very nice (and very smart) Economists back then because I was teaching Business Statistics in the College of Business at Bowling Green State University. We were discussing some data set that had an interesting feature. It had a continuous variable representing something like charitable contributions on your IRS tax form. It was using some sort of independent variables like the taxpayer’s age, income

This is an example of data that can never be negative. I guess it could be negative if you stole money from the collection plate at church on Sunday. Don’t do this

The other interesting feature is that in addition to non-negative values

The Economists suggested that I use a Tobit model and I had to confess my ignorance. I had never heard of Tobit regression

For what it’s worth

Fast forward to the 1990s and I was attending a talk on post hoc tests for analysis of variance. There’s a whole slew of these

I’ve switched from SAS to R

When I teach these days