[R] hosmer-lemeshow

Frank E Harrell Jr fharrell at virginia.edu
Fri Nov 22 18:44:46 CET 2002

On Fri, 22 Nov 2002 17:37:55 +0100
stecalza at tiscali.it wrote:

> Hi eveybody.
> I've seen posted before the same question, but no answer. Is there anybody
> who has implemented the Hosmer-Lemeshow test in R?
> TIA,
> Stefano

I think I've posted this before, but the H-L test has disadvantages: low power, sensitive to arbitrary choices of intervals, other issues related to testing the model on the training data.  The Hosmer-Le Cessie test is a one d.f. omnibus test of lack of fit for a binary logistic model that will generally have more power than H-L, and it does not require one to construct intervals of predicted probabilities.  The new test is done by the residuals.lrm function in the Design library (http://hesweb1.med.virginia.edu/biostat/s/Design.html).  Like all omnibus tests, it will have low power against directed alternatives (e.g., nonlinearity, nonadditivity).
Frank E Harrell Jr              Prof. of Biostatistics & Statistics
Div. of Biostatistics & Epidem. Dept. of Health Evaluation Sciences
U. Virginia School of Medicine  http://hesweb1.med.virginia.edu/biostat
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list