Re: st: saving the confidence interval as a scalar

[Julian Reif <jreif@uchicago.edu>]

Michael, you can also use -regsave- to save confidence intervals to a dataset. You can install -regsave- with the command -ssc install regsave, replace-

*****************************************
sysuse auto, clear
svyset rep78 [pw = turn]
svy, subpop(if foreign==0): total trunk
regsave, ci
list
******************************************

Julian

On Jul 29, 2011, at 1:33 AM, statalist-digest wrote:

> Date: Thu, 28 Jul 2011 17:56:02 -0400
> From: Steven Samuels <sjsamuels@gmail.com>
> Subject: Re: st: saving the confidence interval as a scalar
> 
> Michael, try the following:
> 
> *****************************************
> sysuse auto, clear
> svyset rep78 [pw = turn]
> svy, subpop(if foreign==0): total trunk
> matrix b = e(b)
> scalar tot = b[1,1]
> matrix v=e(V)
> scalar v =v[1,1]
> scalar bound =  ///
> sqrt(v)*invttail(e(df_r),0.5*(1-c(level)/100))
> scalar  ll = tot-bound
> scalar  ul = tot + bound
> di  c(level) "% Confidence Interval" _newline ///
> "Estimate: " %8.0fc  tot ///
> " ["%7.0fc  ll "," %8.0fc  ul "]"
> ******************************************
> 
> On Jul 28, 2011, at 12:12 PM, Monuteaux, Michael wrote:
> 
> Hello everyone,
> 
> I am estimating various parameters using a survey dataset.  I am running the following command:
> 	svy: total myvar, subpop(adults)
> I am able to save the point estimate as a scalar using the saved matrix, e(b).  But, I would like to also save the upper and lower confidence interval bounds as scalars, as well.  Is it possible to access these values as scalars or a matrix?


*
*   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/