Bookmark and Share

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.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: One-sided test on difference between correlation coefficients


From   Austin Nichols <austinnichols@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: One-sided test on difference between correlation coefficients
Date   Sat, 30 Jul 2011 11:26:10 -0400

Michael Boehm <michael.boehm1@gmail.com>:
Note that GDP, not the unemployment rate, is the usual definition of
"cyclical" (the unemployment rate might peak a year or more after a
trough in GDP), not that I advocate using GDP. But why not use
regression instead of correlation?

ren acadnewjobs_res a
ren nonacadnewjobs_res n
glm a unem, link(log)
est sto r1
glm n unem, link(log)
est sto r2
suest r1 r2
test [r1_a]unem==[r2_n]unem

Now you can also put other predictors on the right hand side.  You
might also use state-specific unemployment rates and job offer counts,
and cluster by state to allow for serial correlation.

On Sat, Jul 30, 2011 at 11:09 AM, Michael Boehm
<michael.boehm1@gmail.com> wrote:
> Dear all,
>
> My coauthor and I are looking at the correlation between the
> unemployment rate and the number of academic and non-academic job
> offers in the American Economic Association's "Job Offers for
> Economists" publication. We are computing correlations of the kind:
>
> pwcorr acadnewjobs_res nonacadnewjobs_res unem, sig obs
>
>                     | acadn~es nonaca..     unem
> -------------+---------------------------
> acadnewjo~es |   1.0000
>                     |
>                     |       274
>                     |
> nonacadne~es |   0.4460   1.0000
>                     |   0.0000
>                     |        273       273
>                     |
>            unem |  -0.1222  -0.2580   1.0000
>                     |   0.0432   0.0000
>                     |        274       273       409
>
> Where unem is the unemployment rate and acadnewjobs_res and
> nonacadnewjobs_res the seasonally adjusted number of academic and
> non-academic new jobs, respectively, in a given month. We would like
> to test whether the former are less cyclical than the latter, i.e.
> whether the pwcorr(acadnewjobs_res,unem) <
> pwcorr(nonacadnewjobs_res,unem). Does any of you have an idea how to
> nicely do this in Stata?
>
> Thanks,
> Michael

*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index