Submitted on behalf of Susanne Raessler. The list owner,
T. Robert Harris Associate Professor, Biostatistics The University of Texas School of Public Health at Houston Dallas Regional Campus 5323 Harry Hines Blvd., v8.112 Dallas TX 75390-9128 [email protected] (214) 648-1776, fax (214) 648-1081 >>> R?ssler Susanne <[email protected]> 6/23/2004 10:42:55 AM >>> A very simple but hopefully illustrative simulation I did recently is published under R?ssler S. (2004), The Impact of Multiple Imputation for DACSEIS, DACSEIS Research Paper No 5 download available at http://www.dacseis.de/ -> research Hope that helps a bit, Susanne -------------------------------------------------------------------------- PD Dr. Susanne R?ssler Institute of Employment Research (IAB) Regensburger Str. 104 D-90478 N?rnberg, Germany Tel: +49 911-179-3084 Fax: +49 911-179-3297 email: [email protected] > -----Urspr?ngliche Nachricht----- > Von: David Judkins [mailto:[email protected]] > Gesendet: Mittwoch, 23. Juni 2004 16:01 > An: [email protected] > Betreff: RE: [Impute] Multiple imputation references > > > There was an interesting exercise of different teams using > different techniques on NHANES data that was reported in a > special session at the 1993 JSM. Pp 292-311 of the 1993 SRMS > Proceedings. > > > David Judkins > Senior Statistician > Westat > 1650 Research Boulevard > Rockville, MD 20854 > (301) 315-5970 > [email protected] > > -----Original Message----- > From: [email protected] > [mailto:[email protected]] On Behalf Of > Ofer Harel > Sent: Tuesday, June 15, 2004 10:32 AM > To: [email protected] > Subject: [Impute] Multiple imputation refrences > > Good day, > I am looking for some references, preferably using medical > examples or simulations, in which there is a use of both MI > and ad-hoc techniques (case deletion, single imputation etc) > in which there is proof that using the different methods > gives different results. In other words I am looking to cite > papers that showed MI is superior. Any suggestions? > > Thanks in advance, > Ofer > > ************************************ > Ofer Harel, Ph.D > Postdoctoral Fellow > Department of Biostatistics > School of Public Health > University of Washington > > Biostatistics Unit > HSR&D Center of Excellence > VA Puget Sound Health Care System > 1660 South Columbian Way, 1/424 > Seattle, WA 98108 > phone: 206-277-1027 > Fax: 206-764-2935 > e-mail: [email protected] > ************************************* > > > > > > > _______________________________________________ > Impute mailing list > [email protected] > http://lists.utsouthwestern.edu/mailman/listin> fo/impute > > > _______________________________________________ > > Impute mailing list > [email protected] > http://lists.utsouthwestern.edu/mailman/listin> fo/impute > -------------- next part -------------- A non-text attachment was scrubbed... Name: Header Type: application/octet-stream Size: 1294 bytes Desc: not available Url : http://lists.utsouthwestern.edu/pipermail/impute/attachments/20040623/386224f2/Header.obj From susanne.raessler <@t> wiso.uni-erlangen.de Mon Jun 28 01:26:52 2004 From: susanne.raessler <@t> wiso.uni-erlangen.de (Susanne Raessler) Date: Sun Jun 26 08:25:02 2005 Subject: [Impute] Symposium on Multisource Databases, July 22 Message-ID: <4.2.0.58.20040628082354.01e49...@amelia> Dear All, Below please find the announcement of a symposium which might be of interest for you. For further information and registration please use http://www.statistik.wiso.uni-erlangen.de/ - Aktuelles - Symposium All best wishes, Susanne Symposium on Multisource Databases July 22nd, 2004 in Nuremberg, Germany Reports from Academic and Practice Part I Linked Employer-Employee Databases Chair: Claus Schnabel (University of Erlangen-Nuremberg) 09:00 John M. Abowd (Cornell University, Census Bureau, USA) Integration of individual and employer data using a job link 09:45 Till von Wachter / Stefan Bender (Columbia University, USA / IAB) In the right place at the wrong time: The role of firms and luck in young workers careers 10:15 Martyn J. Andrews / Thorsten Schank / Richard Upward (University of Manchester, UK / University of Erlangen-Nuremberg / University of Nottingham, UK) High wage workers and low wage firms: Negative assortative matching or statistical artefact? Part II Statistical Matching in Practice Chair: Susanne R?ssler (Institute of Employment Research) 11:15 Julia Lane (Urban Institute, Census Bureau, USA) New statistical products using data from multiple sources 11:45 Gerhard Paa? (Fraunhofer Institute for Autonomous Intelligent Systems) Linking web usage information with statistical surveys 12:15 Raimund Wildner (GfK) Data fusion practice in marketing research Part III Data Combination Methodology Chair: Johann Bacher (University of Erlangen-Nuremberg) 13:45 Geert Ridder (University of Southern California, USA) The econometrics of data combination 14:15 Rainer Schnell (University of Konstanz) Record linkage using error prone strings 14:45 Nathaniel Schenker (National Center of Health Statistics, USA) Combining Information from multiple surveys for small-area estimation: A Bayesian approach Part IV Confidentiality Issues and Summary Chair: Jutta Allmendinger (Institute of Employment Research) 15:45 Roderick J.A. Little (University of Michigan, USA) Statistical disclosure control in microdata 16:30 Donald B. Rubin (Harvard University, USA) Concluding integrating comments (Regular fee 100 Euros including breaks and lunch) Or simply print out his email, add your name and fax it to: Lehrstuhl f?r Statistik und ?konometrie - Statistics Symposium - Lange Gasse 20 90403 Nuernberg - Germany Fax ++49 (0)911 5302 277 --------------------------------------------------------------------- PD Dr. Susanne R?ssler Department of Statistics and Econometrics Faculty of Business Administration, Economics and Social Sciences Friedrich-Alexander-University Erlangen-Nuremberg Lange Gasse 20 D-90403 Nuremberg Tel: +49-911-5302-276 Fax:+49-911-5302-277 email: [email protected] -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.utsouthwestern.edu/pipermail/impute/attachments/20040628/0525ae9d/attachment.htm From meng <@t> stat.harvard.edu Thu Jun 24 18:01:30 2004 From: meng <@t> stat.harvard.edu (Xiaoli Meng) Date: Sun Jun 26 08:25:02 2005 Subject: [Impute] Re: Message from "impute" mailing list In-Reply-To: <[email protected]> References: <[email protected]> Message-ID: <[email protected]> Dear Alan (and Vumani), Thanks for letting me know. I am on the road, but here is my quick "answer". The test Don and I developed was based on the usual large-sample argument, namely, the log-likelihood is approximately quadratic. When that assumption fails, negative values can occur. But then that serves as a useful warning that the usual likelihood test based on chi^2 reference distribution should not be trusted. As should be clear from the derivations given in our paper, the accuracy of our approximation depends on parametrization, because the normal approximation depends on it. So one thing could be done is to try different parametrizations -- anything that leads to negative value should not be adopted (but of course positive values themselves do not imply good approximation!). Hope this is useful -- Don may have more to add. Cheers to all, Xiao-Li On Wed, 23 Jun 2004, Alan Zaslavsky wrote: > XL, > > In case you don't follow this list, here is a request for information that > might interest you. You can respond to the list and the sender, if you wish. > > Message: 1 > Date: Wed, 23 Jun 2004 11:22:30 +0000 > From: "Vumani Dlamini" <[email protected]> > Subject: [Impute] negative pooled likelihood > To: [email protected] > Message-ID: <[email protected]> > > I am using multiple imputation for a logistic regression problem I have. The > response and one of my varbale is fully observed and am trying select the > set of model which best describe the data. I am using the likelihood ratio > test statistics proposed by Meng & Rubin (1992), and am getting negative > differences in the pooled likelihood for some of the models. > > If I fit the different models to each of the data sets, the most complex > model has the lowest deviance, but when I use the pooled coefficients this > is not necessarily the case. This leads for some model to a negative value > in the mean of d_{L} resulting in a negative value in D_{L}. Is this common? > > An example of my output is given below. > d'0(1) = 427.0232 > d'1(1) = 518.6282 > > d'0(2) = 425.6645 > d'1(2) = 518.6282 > > d'0(3) = 436.4400 > d'1(3) = 518.6282 > d'0(g) is the deviance of the most complex model for imputation g. d'1(g) is > the deviance of the model incorporating only the fully observed variable in > imputation g. > > Below is the likelihood from the pooled coefficients: > d_L(1) = 521.2215 > d_L(1) = 518.6282 > > d_L(2) = 638.0552 > d_L(2) = 518.6282 > > d_L(3) = 494.4705 > d_L(3) = 518.6282 > Notice that for the simpler model the likelihood is always the same given > that the variables is fully observed, but for the pooled data the most > complex model sometimes has a higher likelihood. > > Thanks for your help. > > Vumani Dlamini > Central Statistical Office > Swaziland >
