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]
st: significance of correlation coeff accounting for clustering
From
Raymond Lim <[email protected]>
To
[email protected]
Subject
st: significance of correlation coeff accounting for clustering
Date
Thu, 31 May 2012 16:42:11 -0400
Hello Statalisters,
I'm calculating the correlation coefficient of an individual-level
variable and a group-level variable (for example, student test scores
and school budget). Is it possible to clustering the variance based on
group ID? I believe the correlation coefficient itself won't change
with the adjustment, but the significance level will.
One hack method I did was regress the individual-level variables on
the group-level variables. The square root of R^2 gives the
correlation coefficient and the value above (Prob > F) gives the
significance. These values correspond to those given by pwcorr, sig.
Thus, I was planning to repeat this regression and cluster on the
group ID. Is there a name of this type of correlation?
Thanks,
-Raymond
--
Raymond Lim
Columbia University
B.A. Economics 2009, M.A. Statistics 2012
[email protected]
*
* 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/