Date: Fri, 15 Nov 1996 16:20:32 -0600
Reply-To: "Nichols, David" <nichols@SPSS.COM>
Sender: "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From: "Nichols, David" <nichols@SPSS.COM>
Subject: Re: 6.1 SPSS SUPERSEDED?
There are a lot of differences. Statistically, there are the new GLM
and VARCOMP procedures. Release 7.5 adds a COMPARE keyword to the
EMMEANS subcommand in GLM that will give all pairwise comparisons
among estimated marginal means for main effects, simple main effects
or averaged main effects. Finally, automatic pairwise comparisons in
FACTOR is enhanced: component analysis and a couple of types of
factor analysis on covariance matrices is a big one. It will now
give post-rotation sums of squared loadings (and percentages of
variance for orthogonal rotations). It will also save factor scores
for all cases when using mean substitution, and will create scores
for a new set of cases.
DISCRIMINANT adds leaving one out classification. Both of these are
now in redone output. In both cases, it's much better than the old
LOG REG has a Hosmer-Lemeshow goodness of fit test added, an option
to choose the cutpoint in the classification table, two pseudo-R^2
measures, confidence intervals for the odds ratios (added for 7.0),
and a better layout of model test statistics. That is, where current
releases have Model and Improvement, which give reduction in -2LL
compared with the previous block and step, 7.5 gives Model, Block
and Step values, which compare with the intercept only model, as
well as the previous block and step.
REGRESSION dialogs now allow you to check boxes for the change in
R^2 and the zero order, part and partial correlations, which have
to be specified in syntax in earlier releases. You also now have
the option of writing coefficients and their covariance or correlation
matrix to a file.
These are some of the nicer things I can think of off the top of
my head, related to the things I do.
Senior Support Statistician
>From: Tom Davis[SMTP:firstname.lastname@example.org]
>Sent: Thursday, November 14, 1996 2:41 PM
>To: Multiple recipients of list SPSSX-L
>Subject: 6.1 SPSS SUPERSEDED?
>Is the SPSS 6.1 for windows really superseded by the 7.5 for windows?
>What are the specific differences? I'm trying to decide whether to
>make the move to 7.5.