Statalist The Stata Listserver

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

Re: st: Re: SW logisitc

From   "Shi Huang" <[email protected]>
To   [email protected]
Subject   Re: st: Re: SW logisitc
Date   Tue, 12 Sep 2006 13:28:57 -0400

Hi, Rodrigo,

Thank you so much for help. It works.

BTW, Do you know how to force some variables stay in the model when
running backward selection?


On 9/12/06, Rodrigo A. Alfaro <[email protected]> wrote:

It seems to me that you are using -svylogit-... right?? Check in the manual
the formula related with the weight that you are using. I am guessing that
the weights are independent of the number of variables. Therefore, you can
run -sw logit- using weights. Example: (1) Suppose that you are
using -svyset [pw=wt], then create a new variable that describe the weights
(weight is a fair name), (2) Check that -svylogit y x1 x2- gives you the
same as -logit y x1 x2 [pw=weight]-, (3) Run your model as -sw logit

I hope this helps

----- Original Message -----
From: "Shi Huang" <[email protected]>
To: <[email protected]>
Sent: Tuesday, September 12, 2006 11:02 AM
Subject: st: SW logisitc


I am doing backward logistic regression with survey data.
Unfortunately,  SW is not supported by svy. Which means I have to do
this manually. According to, backward LR is not
appropriate if I understand correctly.  So I have to use backward
Wald. Anybody can help me how to get Wald test? Or any command to do

Thanks so much,

*   For searches and help try:

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index