Date: Tue, 15 Mar 2005 10:26:32 0600
ReplyTo: Debarchana Ghosh <drg999@ksu.edu>
Sender: "SPSSX(r) Discussion" <SPSSXL@LISTSERV.UGA.EDU>
From: Debarchana Ghosh <drg999@ksu.edu>
Subject: Re: Applicability of HosmerLemeshow?
InReplyTo: <DB33CB380BDB6A43A06FF44519FB4005072EA146@ems4.umassb.net>
ContentType: text/plain; charset=ISO88591
> Hello, My apologies if this is elementary, but is the HosmerLemeshow
> test applicable when I run a binary logistic regression with about
> half
> of the explanatory variables being dichotomous? And if not, how might
> I
> test for model fit in SPSS? Thanks in advance.
>
Yes the HosmerLemeshow test is also applicable for binary logistic
regression.
Hosmer and Lemeshow's Goodness of Fit Test tests the null hypothesis
that the data are generated by the model fitted by the researcher. The
test divides subjects into deciles based on predicted probabilities,
and then computes a chisquare from observed and expected frequencies.
Then a probability (p) value is computed from the chisquare
distribution with 8 degrees of freedom to test the fit of the logistic
model. If the Hosmer and Lemeshow GoodnessofFit test statistic is .05
or less, we accept the null hypothesis that there is no difference
between the observed and modelpredicted values of the dependent. (This
means the model predicts values significantly different from what they
ought to be, which is the observed values). If the HL goodnessoffit
test statistic is greater than .05, as we want for wellfitting models,
we fail to reject the null hypothesis that there is no difference,
implying that the model's estimates fit the data at an acceptable
level.
Hope this helps,
Debarchana
Department of Geography
Kansas State University
