Statalist The Stata Listserver

[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: contrast with probit

From   Maarten buis <>
Subject   Re: st: contrast with probit
Date   Tue, 15 Aug 2006 22:55:31 +0100 (BST)

--- Silke Humbert asked:
> I am using probit with linearly dependent explanatory variables
> (i. e. age categories 0-30, 30-50, >50, coded 0,1). If am doing
> the regression with all these explanatory variables Stata
> naturally drops one if these variables.
> The problem now is that I don't want the coefficients of the
> regression relative to the variable dropped but rather to the total
> average.

--- Dimitriy V. Masterov answered:
> (1) Omit the constant, so that you can include the full set of
> dummies:
> probit outcome d1 d2 d3, nocons
> (2) Use the linear combination command to construct coefficients you
> want:
> probit outcome d2 d3
> lincom _cons+d2

I also thought about that, but the reference to the contrast to the
total average made me think more of effect coding. You can do it
yourself, but Michael Mitchell and Phil Ender wrote -xi3- to do that
for you, see the example below. You'll have to install -xi3- first, by
typing -findit xi3- and follow the instructions from there on.


*-------begin example-------
sysuse auto, clear
reg price mpg foreign
recode rep78 1=3 2=3

/*_Irep78_4 and _Irep78_5 are
  deviations of those categories
  from the grand mean which is 
  the constant)*/
xi3: reg price e.rep78

/* -(_Irep78_4 + _Irep78_5) is the 
  deviation of category 3 from the
  grand mean */
lincom -(_Irep78_4 + _Irep78_5)
*--------end example---------

Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting adress:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

The all-new Yahoo! Mail goes wherever you go - free your email address from your Internet provider.
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index