RE: [Impute] Nonnormality

[David Judkins]

A lack of normality is a good reason to explore semi-parametric imputation procedures.  Please see Section 2 of Chapter 22 (by myself, David Marker and Marianne Winglee) in Survey Nonresponse, Eds. R. M. Groves, D. A. Dillman, E. L. Eltinge, and R. J. A. Little.  New York: Wiley.    David Judkins Senior Statistician Westat 1650 Research Boulevard Rockville, MD 20854 (301) 315-5970 [EMAIL PROTECTED]       -----Original Message----- From: Cole, Jason Ph.D. [mailto:[EMAIL PROTECTED] Sent: Monday, March 15, 2004 1:22 PM To: 'Imputation Listserve ([EMAIL PROTECTED])' Subject: [Impute] Nonnormality   Hell all,   I'm curious about the level of effect nonnormal distributions have on imputation's efficacy.  Specifically, I understand that the EM algorithm works on the assumption of normally distributed data.  However, if one imputes before transformation of skewed data, are the estimates markedly distorted?    Thanks for any input,   Jason   ************************************************************** Jason C. Cole, PhD Statistician Department of Psychiatry and Biobehavioral Sciences Cousins Center for Psychoneuroimmunology 300 UCLA Medical Plaza, Room 3148 Los Angeles, CA  90095-7057 Tel:   310 267 4390 FAX: 310 794 9247 E-mail: [EMAIL PROTECTED] http://www.cousinspni.org **************************************************************

_______________________________________________
Impute mailing list
[EMAIL PROTECTED]
http://lists.utsouthwestern.edu/mailman/listinfo/impute