Bookmark and Share

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: RE: RE: discrepancy in the number of groups


From   "EUN BUM CHO" <[email protected]>
To   <[email protected]>
Subject   st: RE: RE: discrepancy in the number of groups
Date   Wed, 2 Mar 2011 21:55:17 -1000

When I ran the two-level random intercept model (firms at level 2 and observations at level 1), firms (i.e., number of groups) were correctly counted as 369.
But when I ran the three-level random intercept model (industries at level 3, firms at level 2, and observations at level 1), firms (number of level groups) were counted as 370.
Both share the same sample and I really cannot sort it out why this discrepancy arises. 
Any idea? 


-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of EUN BUM CHO
Sent: Tuesday, March 01, 2011 10:08 PM
To: [email protected]
Subject: st: RE: discrepancy in the number of groups

I counted the number of groups in the xtmixed model and found that 369 is the right number.
I wonder why xtmixed reports 370 when 369 is the right number of groups.
Any ideas?

++++++++++++++++++++++++++
Eunbum Cho
Department of Management
Shidler College of Business
University of Hawaii at Manoa
Tel: 1-808-265-9618
[email protected]
Ua Mau ke Ea o ka ʻĀina i ka Pono
++++++++++++++++++++++++++





-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of EUN BUM CHO
Sent: Monday, February 28, 2011 9:30 PM
To: [email protected]
Subject: st: discrepancy in the number of groups

Dear Stata experts!

My data is firm level data, with firms nested within industries.
I ran the xtmixed and xtreg commands using Stata 10.1.
When I looked at the outputs, I found a strange discrepancy in the reported number of groups (firms) between two models.

xtmixed reports that the number of groups is 370, while xtreg reports that the number of groups is 369.
They are analyzed with the same sample, thus both should report the same group number.
I cannot find out why xtmixed has one more group than the xtreg does, when they share the same sample.
Can anyone think of the possible reason for this?

I attached the syntaxes and partial outputs for your reference.

KSIC represents unique industry codes.
id uniquely identifies firms.

xtmixed F.EXPI y1982-y2001 AGE SIZE IFDI ROA CAPI AVSLACK RESLACK POSLACK DSUB NCOMP DMSIZE
DMGR LEADER DMCOM DMPER CL TECH MKT TOP5 TOP30 GROUP || KSIC:, || id:

Mixed-effects REML regression                   Number of obs      =      2410

--------------------------------------------------------------------------------
                          |   No. of           Observations per Group
 Group Variable |   Groups    Minimum    Average    Maximum
--------------------------------------------------------------------------------
                 KSIC |      162          1       14.9         70
                       id |      370          1        6.5         17
--------------------------------------------------------------------------------


xtreg F.EXPI y1982-y2001 AGE SIZE IFDI ROA CAPI AVSLACK RESLACK POSLACK DSUB 
NCOMP DMSIZE DMGR LEADER DMCOM DMPER CL TECH MKT TOP5 TOP30 GROUP, re

Random-effects GLS regression           Number of obs      =      2410
Group variable: id                               Number of groups   =       369

R-sq:  within  = 0.1141                         Obs per group: min =         1
       between = 0.1326                                        avg =       6.5
       overall = 0.1232                                        max =        17

++++++++++++++++++++++++++
Eunbum Cho
Department of Management
Shidler College of Business
University of Hawaii at Manoa
Tel: 1-808-265-9618
[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/


*
*   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/


*
*   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 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index