Statalist


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

Re: st: ARMA(1,1) with multiple panels


From   Austin Nichols <austinnichols@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: ARMA(1,1) with multiple panels
Date   Thu, 31 Dec 2009 13:49:50 -0500

Andrew <KLINKA@email.chop.edu>:

This seems to be a to be a question born of confusion--can you tell us
what you are regressing on what, how you think the error process
works, and what you hope to estimate?  If you are worried about serial
correlation, you should simply cluster by hospital, assuming you have
many of them.  You could even do this for every pair of consecutive
years, for 7 separate estimates.

On Thu, Dec 31, 2009 at 10:32 AM, Klink, Andrew <KLINKA@email.chop.edu> wrote:
> The panels are not balanced ­ some hospitals have only two years of data
> (each with 12 months) and some hospitals have all 8 years of data.
>
> -Andy
>
> Date: Wed, 23 Dec 2009 08:38:40 -0600
> From: Charles Koss <hqtiger@gmail.com>
> Subject: Re: st: ARMA(1,1) with multiple panels
>
> are the panels balanced?
>
> - --
> Charles Koss
> http://charlesonnet.blogspot.com/
>
> On Mon, Dec 21, 2009 at 10:43 AM, Klink, Andrew <KLINKA@email.chop.edu>
> wrote:
>> I have a question regarding ARMA(1,1) with multiple panels (in my case they
>> are hospitals that each have data for years 2001-2008). I am looking for a
>> way to estimate ARMA(1,1) over all hospitals at once, instead of estimating
>> ARMA(1,1) for each hospital for each year. In other words, I would like to
>> estimate ARMA(1,1) by year for all hospitals (i.e., one estimate for each
>> year). I keep getting an error for multiple panels. I'm using Stata 10.1 on
>> a Mac.
>>
>> Thank you,
>> -Andy

*
*   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   |   What's new   |   Site index