Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Lowess smoothing using survey command
From
Austin Nichols <[email protected]>
To
[email protected]
Subject
Re: st: Lowess smoothing using survey command
Date
Fri, 13 Jul 2012 15:29:51 -0400
Au, Maria C. <[email protected]>:
-lpoly- allows aweights to get correct point estimates, and rakedw0 is
your individual weight; I assume you do not need SE bars.
Also see -help mfp-
-pspline- (on SSC) is also ripe for modification to allow weights.
On Fri, Jul 13, 2012 at 3:20 PM, Au, Maria C. <[email protected]> wrote:
> Hi all,
>
> I'm using the svy command to analyze a 2-stage stratified complex data and am trying to do a lowess smoothing plot for my continuous against the outcome to see if I need to make splines for my model. However, lowess is not supported by the svy command. Someone has suggested using the "expand" option to get the inverse of the individual weights to create a weighted lowess plot. However, because individual weights are not providing in this public dataset (I'm using jackknife variance estimation method with the following command :svyset [pw=rakedw0], jkrw(rakedw1-rakedw80, multiplier(1)) vce(jack) mse) to run my svy analysis, I cannot use that method either. There are too many observations in my dataset to simply do a scatterplot without smoothing, does anyone know of other ways to conduct exploratory data analysis on continuous covariates against outcome to look for splines using the svy command?
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/