On Fri, 14 Mar 2003 10:39:10 -0600, Rodney Sparapani <[EMAIL PROTECTED]> wrote:
> Mike Lyon wrote: > > >We have come down to two different statistical packages and I wanted > >to solicit opinions of SPSS and SAS/STAT packages. We need this for > >beginners (me) as well as more experience people involved in > >biological research. The common tests that we use are: general linear > >model univariate, multivariate, repeated measures, regressions as well > >as data reduction and transformations. > > > >Any info would be appreciated. > > > >Thanks > > > > > I'd recommend SAS over SPSS. It's alot more expensive, but you get alot > more too. Especially if you need to do statistics and data processing. I had this discussion a few years ago with one SAS fan who insisted that SAS did a lot more statistics, and I matched him with SPSS, frequently line-for-line in the syntax. SAS historically had better file-reading: especially from IBM data tapes and EBCDIC files, but also from other complicated formats. SAS historically spend a lot of time and effort in their data- reporting. Do you satisfy your accountant? Do you lay everything in the right place on the page, for this report that you will be producing every week? - There is a SAS group on the Internet to ask, and an SPSS group. I contribute to the SPSS group. The SPSS group handles a lot of statistics questions, and it is moving in the direction of having more questions about 'scripting', etc. When I last looked at the SAS group, several years ago, almost all the questions were about "How do you force the printer to < ... > " and so on. And you need to have your own SAS-specialist statistician (at least one) on hand, to keep you from screwing up badly -- because the statistics output is more confusing than SPSS, and what they give you is relatively unselected - and unwise, if you want to think of it that way. -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
