Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

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

RE: st: arega and post regression tests

From   Nick Cox <>
To   "''" <>
Subject   RE: st: arega and post regression tests
Date   Wed, 2 Nov 2011 12:45:42 +0000

No; you don't need to do that as you can use factor variable notation. 


Ozgur Ozdemir

Hi Kit,
my industry codes are one on single column, do you mean that I need to create one dummy variable for each industry code first and then I can use normal regression ?

> From:

> I have been requested to use areg in my regression due to large dummy variable (15 industry codes in a sample of 1100 firms) set...
> The requestor is being silly. What's wrong with having 15 additional coefficients in a sample of 1100? The areg command is estimating those same coefficients; it is just not displaying them.
> Using a table-producing routine such as built-in -estimates table-, you can produce output not showing the dummy coefficients with drop() or keep(). Same goes for user-written routines like Ben Jann's -estout- on SSC. Use -regress- and show the requestor the edited output.

*   For searches and help try:

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