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: iccvar
From
Nick Cox <[email protected]>
To
[email protected]
Subject
Re: st: iccvar
Date
Thu, 13 Dec 2012 11:47:36 +0000
If you want a plain correlation, it is given by -correlate-. That
would have no relation to -xtmixed-. You could calculate it for the
panels lumped together, or for panels separately. What makes sense
would depend on your scientific goals. If that isn't what you want,
you will need to be much more specific.
Nick
On Thu, Dec 13, 2012 at 11:35 AM, megan rossi <[email protected]> wrote:
> Thanks, however I am after the correlation coefficient from panel data, not the ICC.In other words I am looking for a command to use following xtmixed that will give me the the correlation coefficient for my dependent and independent variable. Any ideas?
[email protected]
>> It seems that -quickicc- (from SSC) does *not* calculate the ICC of
>> random slope models. But you can use -xtmrho- (from SSC), instead -
>> although -xtmrho- does not calculate the standard error.
>>
>> Stata example (runable from do-file window):
>>
>> * --------------------------------------------
>> cap which quickicc
>> if _rc ssc install quickicc
>>
>> cap which xtmrho
>> if _rc ssc install xtmrho
>>
>> webuse nlswork
>> xtmixed ln_wage grade age c.age#c.age ttl_exp ///
>> tenure c.tenure#c.tenure || idcode: tenure
>> xtmrho
>> quickicc
>> * --------------------------------------------
>>
>> BTW: Note that in the first "official" example for -xtmixed- (see -h
>> xtmixed-) the variables ln_wage and idcode are abbreviated to ln_w and
>> id which - depending on your varabbrev setting (see -h set_varabbrev-)
>> might give you an error message if you try to run it (this should be
>> corrected!).
*
* 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/