There were a substantial number of statistical enhancements made for 7.5. These include enhancements to GLM that allow automatic computation of pairwise comparisons for any main or simple main effect (between or within), user assignment of error terms for some situations not previously available, and a new VARCOMP procedure. LOG REG was enhanced with two pseudo-R^2 measures, a Hosmer-Lemeshow goodness of fit test, a clearer set of model LR test statistics, and the ability to specify a cutpoint for classification other than the default .5. FACTOR, DISCRIMINANT, CLUSTER, QUICK CLUSTER, and PROXIMITIES were moved into the Base module (NLR, CNLR, and LOG REG were moved from the Advanced module to the Professional one), converted for pivot table output, and in the case of FACTOR and DISCRIMINANT, some new features were added. FACTOR will now do analyses on covariance matrices where appropriate, and PROMAX rotation was added. Cross-validation (leaving one out) was added to DISCRIMINANT. I'm probably leaving out a few things as well.

David Nichols Senior Support Statistician SPSS Inc. nichols@spss.com

>---------- >From: Joel Winter[SMTP:s994299@UMSLVMA.UMSL.EDU] >Sent: Wednesday, September 17, 1997 1:05 PM >To: SPSSX-L@UGA.CC.UGA.EDU >Subject: Diff between v7.0 & v7.5 > >I was wondering about the difference between 7.0 & 7.5. I've run both on >other peoples machines (I'm still happy with v6.1) and don't know whether to >recommend to a 7.0 user to upgrade. I searched the SPSS Web site and it >seems >that there is no cheap easy patch or upgrade. > >So, what's the cost? Benefit? > > >================================ >Joel Winter >s994299@umslvma.umsl.edu >