Date: Fri, 25 Feb 2000 10:59:22 -0600
Reply-To: "Nichols, David" <nichols@SPSS.COM>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: "Nichols, David" <nichols@SPSS.COM>
Subject: Re: differences between ordinal regression vs. catreg
The CATREG procedure performs a linear regression of a categorical dependent
variable on a set of categorical predictor variables, using optimally scaled
transformations of the variables if so desired. It's a special case of the
OVERALS algorithm. The PLUM procedure fits one of a series of nonlinear
cumulatiave response models for an ordinal response variable with
categorical and/or continuous predictors. In CATREG, all variables must be
numeric with positive integer values (or else the values get truncated). In
PLUM, the predictors can take on any values.
Principal Support Statistician and
Manager of Statistical Support
> -----Original Message-----
> From: Hans de Wit [mailto:firstname.lastname@example.org]
> Sent: Friday, February 25, 2000 4:02 AM
> To: SPSSX-L@LISTSERV.UGA.EDU
> Subject: [SPSSX-L] differences between ordinal regression vs. catreg
> Could somebody tell me about the differences between the
> ordinal regression and
> the regression using optimal scaling in Spss 10?
> The only thing i could think of is that Ordinal regression
> can take interval covariates
> and Catreg not really.
> Are there differences in the interpretation of the results?
> Hans de Wit