Statalist The Stata Listserver


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

RE: st: Standard error of a sum


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: Standard error of a sum
Date   Wed, 14 Jun 2006 08:22:12 +0100

I think Clive meant -summarize, detail-. 

To repeat what Clive said: please work out how not
to send a winmail.dat attachment to the list
with your postings. This triggers
MIME encoding and means that many
list members will see mailjunk
and not what you intend. This 
is all explained in the Statalist FAQ. 

People typically don't respond to postings
they can't read, and some people are less
likely to respond to postings that breach
suggested format, so this in your own
interests. 

Nick
n.j.cox@durham.ac.uk 

Meloria Meschi

Well summarize gives me mean, standard deviation, the usual stuff.  ci gives the confidence interval around the mean, with the se calculated in the usual way, ie sdev/sqrt(n).

If the sum is say, 1000 and the mean is 50 - I can't surely use the standard error of the mean to construct a confidence interval for the sum.  And also the stdev should refer to the mean - population mean, rather than sample mean.  So how do I get the SE of the sum?  

My version of stata does not allow the display option by the way.   

Clive Nicholas

Meloria Meschi wrote:

> I have a silly problem that's driving me nuts.  I have estimated two
> production function, with AR(1) correction.  I got 7 point forecasts each
> and their standard errors (call each forecast series x1 and x2).  I have
> computed the confidence intervals for each of these forecasts.
>
> Then I have added the two forecasts into one series y=x1 + x2 and produced
> s.e. and confidence intervals for this series as well.  What I need now is
> a confidence interval for the sum of y (ie total production from the two
> sites).  How do I compute the standard error for this total? And does it
> actually make sense to do so?

What does

. summarize y, display

tell you? I'm assuming these functions are continuous, valid measurements
which have no bounds. If so, calculating standard errors should pose no
difficulties. But only _you_ can answer that question!

You're politely reminded not to send attachments to the list.


*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index