Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: oglm program now available at ssc


From   Richard Williams <Richard.A.Williams.5@ND.edu>
To   statalist@hsphsun2.harvard.edu, statalist@hsphsun2.harvard.edu
Subject   Re: st: oglm program now available at ssc
Date   Fri, 20 Jan 2006 10:57:20 -0500

At 05:33 AM 1/20/2006, Phil Schumm wrote:
On Jan 19, 2006, at 11:17 AM, Richard Williams wrote:
-oglm- was inspired by SPSS's PLUM command and its results have
been verified by checking it against PLUM.

I think it's worth noting here that PLUM was originally a program
written by Peter McCullagh, whose 1980 paper (Regression Models for
Ordinal Data, JRSS, B 42, 109-142) laid out much of the foundation
for ordinal regression models.  I have no idea how SPSS came by this,
or whether their implementation bears any resemblance to the original.
I wonder about this too. :) There was a previous thread about the differences between Stata & SPSS in the naming of links. In particular, what SPSS calls nloglog and cloglog are called cloglog and loglog in Stata's glm and cloglog programs (and now oglm). Joe Hilbe explained the differences between Stata & SPSS in this post:

http://www.stata.com/statalist/archive/2006-01/msg00049.html

Now, as to why the differences exist in the first place, I don't know. I'll have to look at the paper Phil cites. However, in Generalized Linear Models, 2nd edition, also by McCullagh (and Nelder), I swear the formula they give for the ordinal cloglog model on p. 153 is the same as the formula oglm uses. My guess is that, for whatever reason, SPSS decided to deviate from what was in these original articles. It may not be that SPSS is wrong, but it is different, and people need to be aware of those differences. I figured that since olgm was a Stata program, I should follow Stata's naming conventions.

In any event, Stata is in good company in its naming conventions. As Joe Hilbe wrote to me in a private message,

"I double checked the code in cloglog (which I wrote in the first place) and in glm, and compared sample models using Stata, LogXact, LIMDEP, and Systat. The cloglog models produced identical results."


-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX: (574)288-4373
HOME: (574)289-5227
EMAIL: Richard.A.Williams.5@ND.Edu
WWW (personal): http://www.nd.edu/~rwilliam
WWW (department): http://www.nd.edu/~soc
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/




© Copyright 1996–2021 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index