Daniel Wilde <[email protected]>:
Something is wrong with your mail server or client software. Both of
these messages appear to have been sent within one second of each
other, as you can see at:
http://www.stata.com/statalist/archive/2008-01/
You can do all this within Stata, though it is not clear to me that it
is a good idea. Try this (untested) approach for starters:
g a=.
g b=.
g p=.
levelsof country, loc(c)
foreach i of local c {
qui reg data301 time if country==`i'
replace p=2*ttail(e(df_r),abs(_b[time]/_se[time])) if country==`i'
su time if time==1990 & e(sample), meanonly
if r(N)==0 {
replace time=1990 if country==`i'
predict A if country==`i'
}
else g A=data301 if country==`i'
replace time=2015 if country==`i'
predict B if country==`i'
replace a=A if country==`i'
replace b=B if country==`i'
drop A B
}
bys country: drop if _n>1
keep country p a b
g c="no significant trend" if p>0.1
replace c="significant upward trend" if b>a & p< 0.1
replace c="significant downward trend" if b<a & p< 0.1 & b > a/3
replace c="achieved 2/3 reduction" if p<.1 & b < a/3
list, noo clean
On Jan 18, 2008 9:10 AM, Daniel Wilde <[email protected]> wrote:
> All,
>
> I just thought I would repost the below message because I didn't get any
> responses yesterday. Is there anything about it that's unclear? Or is it
> just a boring question?
>
> Kind Regards,
>
> Daniel Wilde
>
> ------------ Forwarded Message ------------
> Date: 16 January 2008 19:37 +0000
> From: Daniel Wilde <[email protected]>
> To: [email protected]
> Subject: Has the time trend resulted in a 2/3 reduction between year ?? and
> year ??, can Stata tell you yes or no.
>
*
* 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/