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

RE: st: missing z-values when using dprobit

From   Maarten buis <>
Subject   RE: st: missing z-values when using dprobit
Date   Mon, 26 Nov 2007 22:54:09 +0000 (GMT)

--- "Ohinata, Asako" <> asked:
> I am currently running dprobit with many dummy variables as well as
> some continuous variables (i.e. age etc). When I run the
> regression, some of the z-values are missing although standard
> errors are estimated.

--- Maarten buis answered 
> The standard error is in the example is very very small
> (0.0000000000000000000000000000000000000000000000000000000
> 00000000000000286), i.e. de facto zero. So I am not surprized the
> z-score is missing. Is kid3plus the reference category?

--- "Ohinata, Asako" <> replied:
> Yes, it is a reference category. I think the main problem is that the
> model contains too many dummy variables as when I drop some of them,
> all the standard errors are estimated. I am wondering, however,
> if there are any solutions other than dropping variables.

You can remove the constant with the -noconst- option (notice that now
the interpretation of your coefficients change), otherwise you always
have to leave out the reference category (you can leave it in but than
Stata will throw it out). By leaving out the reference category you can
determine which category is the reference category. In principle it
doesn't matter which category is the reference, but in practice
choosing your reference smartly can greatly improve the
interpretability of your coefficients, especially when you are
including interaction effects.

-- Maarten

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

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

Yahoo! Answers - Got a question? Someone out there knows the answer. Try it
*   For searches and help try:

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