Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: posting on behalf of Jill Huppert

From   "Sayer, Bryan" <>
To   "''" <>
Subject   st: RE: posting on behalf of Jill Huppert
Date   Thu, 3 Oct 2002 10:45:14 -0400

Send me the COMPLETE SUDAAN commands and I'll tell you if it is exactly
replicable in Stata or not.

Bryan Sayer
Statistician, SSS Inc.

-----Original Message-----
From: baum []
Sent: Thursday, October 03, 2002 10:37 AM
To: StataList
Subject: st: posting on behalf of Jill Huppert

Hi, Statlisters.
I hope you can help.
I have been exploring a publically avail dataset (NHAMCS), which includes 
all weighting variables, and their documentation says I should be able to 
get accurate estimates and SE using STATA.
However, my statistical colleague uses SUDAAN. I have been able to match 
her point estimates, but not SE.
We discovered that she uses many more weighting variables in SUDAAN than 
STATA requires.

So my command looks like:

pweight: patwt
PSU: psum

xi: svylogit outcome var1 var 2 i.var1*var2 patwt, str(stratm1) 
psu(psum)sub(sample) or

Wheras in SUDAAN, Jane uses all of the following:
 stratm psum subfile prostrat su poppsum popsu popvism popprovm
( and sets levels)

Am I missing the finite pop ct variable? (fpc)
Anyone else use STATA for NHAMCS?

Can Stata handle interaction terms in svy data?

Many thanks.
Jill Huppert

Christopher F Baum, Boston College Economics, Chestnut Hill MA 02467 USA             
*   For searches and help try:
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index