Should I learn R instead of SAS (created 2010-04-05).

This page is moving to a new website.

I got a question from a statistician beginning her career asking whether she should learn SAS or R. That's a very personal question and there is no perfect answer. Here is what I wrote.

SAS is a more polished product which shows in broad and diverse range of documentation, training, and user groups. Not that R is without these, but rather that R has less of these.

R is object oriented and SAS is procedure oriented. Some people, including me, see this as a big advantage for R, but other people don't really care. Certain areas, such as drug development, have adopted SAS as a standard, and if you do not know SAS, you will be swimming against the tide. It's not as if SAS is the only product used, but it has such a strong penetration that it would be hard on most team projects without knowing SAS. Other areas, like genetics, have adopted R as a standard. Again, this is not an exclusive thing, but a strong trend nonetheless.

I stopped using SAS in 1998 and I haven't missed it all that much. So I think you would be fine learning just R. Learn both if you can, but if you only have time to learn one well, I would certainly encourage you to learn R.