An alternative is SD=1/m(sqrt(v[1])+sqrt(v[2])+...+sqrt(v[m]))
The two expressions are not mathematically equivalent, but not sure which is correct (or better). I think you want to average on the SD scale, not the variance scale. Bill H, MS, Wash U, St Louis -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Ly-Mee Yu Sent: Tuesday, July 12, 2005 11:28 AM To: [email protected] Subject: [Impute] Standard deviation of imputed data sets Dear Imputers, I'd be much grateful if someone can confirm me whether the sd of all the imputed data sets is simply: SD=sqrt(1/m(v[1]+v[2]+v[3]+...v[m])) where m=no. of imputation, v[*]=variance of * imputed dataset which is same as you calculated the mean? Many thanks in advance. Lymee _______________________________________ Ly-Mee Yu Medical Statistician Centre for Statistics in Medicine Wolfson College Annexe University of Oxford Linton Road OXFORD, OX2 6UD UK Tel: +44 (0)1865 284405 Fax:+44 (0)1865 284423 Emails: [email protected] [email protected] http://www.csm-oxford.org.uk _______________________________________ _______________________________________________ Impute mailing list [email protected] http://lists.utsouthwestern.edu/mailman/listinfo/impute <br/>The materials in this message are private and may contain Protected Healthcare Information. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.
