Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.

# st: RE: Calculate confidence intervals for log-log models

 From "Seed, Paul" To "statalist@hsphsun2.harvard.edu" Subject st: RE: Calculate confidence intervals for log-log models Date Wed, 6 Mar 2013 10:26:14 +0000

```The same result can be got in a simpler way.

Consider a well-known example:
*********** Start of Stata code ***********
clear
sysuse auto
gen ln_price = ln(price)
gen ln_mpg = ln(mpg)

regress ln_mpg ln_price
* effect of a 10% increase in price on mpg.
lincom `=ln(1.1)'*ln_price , eform
* The mpg falls on average by 3.4% to 96.6% of the original value.

* Alternative method
nlcom 1.1^_b[ln_price]
* The same confidence interval; but the test (t in particular)
* no longer agrees with that given by the regression.

***********  End of Stata code  ***********

*******************************************************************

Date: Tue, 5 Mar 2013 20:58:57 +0000
From: Jason Ferris <j.ferris@uq.edu.au>
Subject: st: RE: Calculate confidence intervals for log-log models

Hi  Nick and Rich,

Nick - I was going to run a series of 'changes' in the original variable (as you suggested) by dividing the original variable by a series of constants (representing the percentage value changes) that I want.  I thought I could get this directly by using lincom - but didn't appreciate it was only about 'linear combinations'.

Rich - thank you for suggestion - this gives me exactly what I am looking for.

- ------------------------------------------------------------------------------
lny |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
- -------------+----------------------------------------------------------------
lnx |   2.105722   .0817271    25.77   0.000     1.942802    2.268642
_cons |  -3.032146   .4274653    -7.09   0.000    -3.884282   -2.180009
- ------------------------------------------------------------------------------

. nlcom 1.1^_b[lnx]

_nl_1:  1.1^_b[lnx]

- ------------------------------------------------------------------------------
lny |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
- -------------+----------------------------------------------------------------
_nl_1 |   1.222254   .0095207   128.38   0.000     1.203275    1.241233
- ------------------------------------------------------------------------------

Cheers,

Jason

***************************************************
Paul T Seed, Senior Lecturer in Medical Statistics,
Division of Women's Health, King's College London
Women's Health Academic Centre, King's Health Partners
(+44) (0) 20 7188 3642.

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/
```