Dear Statalisters
I am wondering about the feasibility of an analysis I want to conduct within
Stata. I am far from expert in using Stata so would really just like advice
at this stage on whether what I require is actually viable before investing
the time and effort to actually do it. What I would like to do is to
bootstrap the following:
For a binary variable x with no missing values, draw a random 10% of cases
and recode values to missing.
Use the multiple imputation hotdeck to recover the missing values.
My understanding is that, if the dataset on which the bootstrap is done
represents the true population value for x (and covariates for the hotdeck),
then the bias estimate from the bootstrap would represent the bias of the
hotdeck procedure in recovering these population parameters. Any input
gratefully received,
Patrick
Dr Patrick Sturgis
Department of Sociology
University of Surrey
Guildford
GU2 7XH
tel: 01483 686 974
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
