A paper from the '70s or so compared a large number of complex sample data
sets and found no predictable efficiency of error-estimation from bootstrap
vs. linearization. I don't remember the exact cite, but one author was Karol
Krotki.
arnold
Arnold H. Levinson, PhD
Associate Scientist
Center for Research Methodology and Biometrics
AMC Cancer Research Center
303.777.8801
fax 303.239.3394
levinsona@amc.org
Assistant Professor
Department of Preventive Medicine and Biometrics
University of Colorado School of Medicine
Arnold.Levinson@uchsc.edu
>
> Date: Thu, 31 Oct 2002 09:22:48 -0500
> From: "Sayer, Bryan" <BSayer@s-3.com>
> Subject: st: RE: complex standard errors
>
> I have seen something recently, but I can't remember exactly where. But I
> suggest you try this question on the Survey Research Methods list serve.
> Which is at SRMSNET@LISTSERV.UMD.EDU. I think there is an access point
from
> the ASA website (www.amstat.org)
>
> Bryan Sayer
> Statistician, SSS Inc.
>
> - -----Original Message-----
> From: Patrick Sturgis
> To: Statalist (E-mail)
> Sent: 10/30/02 5:59 AM
> Subject: st: complex standard errors
>
> Can anybody provide me with a cite that compares the efficiency of
> bootstrap
> and Taylor linearization for standard errors under complex survey
> designs
> (clustered, stratified with unequal weights)? In anticipation,
>
> Patrick
>
> Dr Patrick Sturgis
> Department of Sociology
> University of Surrey
> Guildford
> Surrey GU2 7XH
> United Kingdom
> tel: +44 (0)1483 686 973
> fax: +44(0)1483 689551
*
* 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/