Date:         Wed, 22 Dec 1999 16:03:39 -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: likelihood test

See pp. 46-47 of the SPSS Regression Models 9.0 (or 10.0) manual (or
corresponding pages of earlier editions) for discussion of the three chi^2
values. The significance (which many people call the p-value) is based on a
nondirectional alternative, which means that it's two-sided. It's actually
one-tailed, but two-sided.

David Nichols
Principal Support Statistician and
Manager of Statistical Support
SPSS Inc.

> -----Original Message-----
> From: lee boonsiew [mailto:leebsiew@HOTMAIL.COM]
> Sent: Sunday, December 12, 1999 10:15 AM
> To: SPSSX-L@LISTSERV.UGA.EDU
> Subject: [SPSSX-L] likelihood test
>
>
> Dear members,
>
> if i have -2LL of 138.58592 without independents and -2LL of
> 105.712 when
> include independents. the following are my results:
>
>                       chi square    df     significant
> model                  32.874       6        .000
> block                  32.874       6        .000
> step                   32.874       6        .000
>
> Why do i have model, block and step with same chi-square? what is the
> difference among them? Do i have to check back on the chi
> square table to
> see whether it is significant or not? Is the model tested on
> 2-tailed test
> or one tail??
>
> How do I present my results using the above values??
>
> LEE
> University MAlaysia SArawak
>
> ______________________________________________________
> Get Your Private, Free Email at http://www.hotmail.com
>