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: correct confidence intervals of -mean- ?
From
Tirthankar Chakravarty <[email protected]>
To
[email protected]
Subject
Re: st: correct confidence intervals of -mean- ?
Date
Sat, 6 Mar 2010 19:05:38 +0530
Hi Dirk,
I assume the behaviour you are looking for is given by:
mean price if rep78==1
etc.
The reason for the CIs being the way they are in the output of -mean-
is because the degrees of freedom takes into account the calculation
of a covariance matrix with observations in other subpopulations
(albeit with covariance restricted to zero). The manual makes this
clear for the estimation of means, but not for the covariance
estimates. Just a guess though.
T
2010/3/6 Dirk Enzmann <[email protected]>:
> Very carefully I want to ask: Are the confidence intervals given by -mean-
> really correct?
>
> Below I compare the results of -mean- with the results of a different
> procedure:
>
> * --------------------------------------
> sysuse auto, clear
> mean price, over(rep78)
>
> local df = e(df_r)
> display "degrees of freedom = n-1 = `df'"
> * Note that here df is the number of the total sample - 1!
>
> * alternative route:
> collapse (mean) mprice=price (sd) sdprice=price (count) nprice=price if
> (rep78 < .), by(rep78)
> * calculate the confidence intervals the -mean- way:
> gen ci95la = mprice - invttail(`df',.025)*sdprice/sqrt(nprice)
> gen ci95ua = mprice + invttail(`df',.025)*sdprice/sqrt(nprice)
> * calculate the confidence intervals a different way:
> gen ci95lb = mprice - invttail(nprice-1,.025)*sdprice/sqrt(nprice)
> gen ci95ub = mprice + invttail(nprice-1,.025)*sdprice/sqrt(nprice)
> * compare both sets of confidence intervals (..a vs ..b):
> list
> * --------------------------------------
>
> My question: Which procedure is correct?
>
> *************************************************
> Dr. Dirk Enzmann
> Institute of Criminal Sciences
> Dept. of Criminology
> Schlueterstr. 28
> D-20146 Hamburg
> Germany
>
> phone: +49-(0)40-42838.7498 (office)
> +49-(0)40-42838.4591 (Mrs Billon)
> fax: +49-(0)40-42838.2344
> email: [email protected]
> www:
> http://www2.jura.uni-hamburg.de/instkrim/kriminologie/Mitarbeiter/Enzmann/Enzmann.html
> *************************************************
> *
> * 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/
>
--
To every ω-consistent recursive class κ of formulae there correspond
recursive class signs r, such that neither v Gen r nor Neg(v Gen r)
belongs to Flg(κ) (where v is the free variable of r).
*
* 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/