1983/01/01 by Manfred Thaller, Thaller, Manfred
Mathematics · #Probability and Statistical Research
paper · pdf · doi:10.12759/hsr.8.1983.4.88-95
SAS is reviewed quite favorable by most people; indeed the only reason why it is not recommended more generally is, that is has that far been restricted to IBM systems - and IBM is definitely not very popular among academic computing centres. SAS is a very large system. In our opinion one should describe it as a core of sophisticated statistical programs embedded in a consistent software environment. As a result of this design SAS programming falls more or less apart into two categories: on the one hand the statistical procedures can be used without any technical expertise; on the other hand everything that goes beyond that, particularly data handling, can relatively soon reach the limits of the canned solutions which are provided by appropriate options of the relevant commands.