26 Jun
2009
26 Jun
'09
1:48 p.m.
Hello Tanja,
Currently, Amelia uses a standard non-parametric bootstrap, without
blocking.In our experience, the standard bootstrap appears to work,
yet we plan to add more complex bootstrapping options in the future.
cheers,
matt.
On Thu, Jun 25, 2009 at 8:05 PM, Tanja
Srebotnjak<tanjas(a)u.washington.edu> wrote:
Hello,
I am trying to find out what specific bootstrapping algorithm is used to
generate the m incomplete data sets, which are then used to get m sets of
point estimates for mu and Sigma. Specifically, are you using a block,
paired, wild or other type of bootstrapping to sample m data sets of size n
from the data D={D_obs, D_mis}?
I checked the documentation and the accompanying paper but could not find
the answer.
Many thanks for any response you can give.
Best,
Tanja
Tanja Srebotnjak, PhD, MSc, Dipl. Stat.
Postgraduate Fellow
Institute for Health Metrics and Evaluation
University of Washington
2301 5th Ave, Suite 600
Seattle, WA 98121
Email: tanjas(a)u.washington.edu
Tel: +1-206-897-2866
www.healthmetricsandevaluation.org
-
Amelia mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive: http://lists.gking.harvard.edu/?info=amelia
More info about Amelia: http://gking.harvard.edu/amelia