In addition to Paul von Hippel I would say that in future research you
might be using the covariate in some (unforeseen) analysis of the
imputed data. Iit has been shown that in order to obtain
a proper estimate of the relationship between variables suffering from
missing values and
another (complete) relevant variable, this variable should be entered in
the imputation
model; the strength of the relationship will be underestimated if such a
variable is left out
(see, e.g., Little, 1992, Schafer, 1997, pp. 140-143, or Schafer &
Olsen, 1998). Therefore,
to get an unbiased estimate of this relationship in subsequent analysis,
the covariate should be included in the
imputation model.
Little, R. J. A.(1992). Regression with missing X's: A review. Journal
of the American
Statistical Association, 87, 1227{1237.
Schafer, J. L.(1997). Analysis of incomplete multivariate data. London:
Chapman and
Hall.
Schafer, J. L., & Olsen, M. K.(1998). Multiple imputation for
multivariate missing data
problems: A data analyst's perspective. Multivariate Behavioral
Research, 33,
545{571.
Niels Smits
Research Methodology,
Statistics and Data-analysis
Faculty of Psychology and Education
Free University Amsterdam
Van der Boechorststraat 1
1081 BT Amsterdam
The Netherlands
Tel: +31 (0)20 5988713
Secr: +31 (0)20 5988757
Fax: +31 (0)20 5988758
Paul von Hippel wrote:
> My understanding is that you can get bias from having too few
> variables but not from having too many.
>
>
> _______________________________________________
> Impute mailing list
> [email protected]
> http://lists.utsouthwestern.edu/mailman/listinfo/impute
>
