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

Re: st: Missing values test

From   Maarten buis <[email protected]>
To   [email protected]
Subject   Re: st: Missing values test
Date   Sun, 2 Dec 2007 09:52:20 +0000 (GMT)

--- Constantin Colonescu <[email protected]> wrote:
> I run OLS on a list of 10 regressors and the cumulated missing values
> pile up to about 30 percent of my sample. I know I can use -ice- and
> -impute- to generate the missing values, but I would like to make
> sure that my missing values are MRA (missing at random). I was
> unable to find out how to test for MRA. 

The MAR assumption states that the probability of missingness only
depends on observed information and not on unobserved information, e.g.
that probability does not depend on the missing values. Consequently
there is no test, because that would require access to the unobserved
information to check if that information influences the probability of
missingness, and unobserved information is by definition not observed.

-- 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

Sent from Yahoo! - the World's favourite mail

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index