This two-day workshop, presented by University of Michigan Professors Roderick Little and Trivellore Raghunathan, and CSCAR's staff Brenda Gillespie, Kathy Welch, and Brady West, will provide an overview of approaches to statistical analysis in the presence of missing data and statistical computing tools for dealing with missing data in statistical analysis.

Statistical Analysis With Missing Data will address the following issues:

• Types of nonresponse and missing data mechanisms
• Simple approaches: Complete case and available case
• Weighting methods for unit nonresponse
• Imputation methods for item nonresponse
• Model-Based methods
• Longitudinal analysis and attrition
• Implementing methods in statistical software
• Applications to survey data

There will also be hands-on computer lab sessions to implement missing data methods on real data sets.

INSTRUCTORS:

                     Roderick Little is Professor in the Department of Biostatistics and the Department of Statistics at the University of Michigan. He is past Coordinating and Applications Editor of the Journal of the American Statistical Association. He is coauthor of Statistical Analysis with Missing Data (R.J.A. Little and D.B. Rubin, 1987), which has become the standard statistical text on the subject.

Trivellore Raghunathan is Professor in the Department of Biostatistics and Research Professor in the Survey Research Center, Institute for Social Research.

Brenda Gillespie is the associate director of CSCAR and Assistant Professor in the Department of Biostatistics at the University of Michigan.  She has extensive experience as a statistical consultant, and specializes in the various methods for analysis of censored data.

Kathy Welch is a Statistical Consultant and primary SAS Consultant for CSCAR.  She consults at CSCAR on linear regression, Poisson regression, and analysis of cluster or longitudinal data, as well as other topics.  Kathy also teaches a course in statistical computing for the School of Public Health, and workshops on SAS for CSCAR and various academic departments and public agencies.

Brady West is the primary SPSS consultant at CSCAR, and consults on topics in Stata as well as a variety of other statistical software packages.

AUDIENCE:

Anyone interested in modern methods for handling nonresponse and missing data in statistical analysis. This workshop will emphasize concepts and applications over technical details, but will presume a familiarity with basic statistical concepts, including probability distributions, estimation, maximum likelihood and regression.

PROVISIONS:

The participant will receive outlines and a bibliography. Professor Little's book, Statistical Analysis with Missing Data is recommended but not required. Refreshments will be provided. Lunch not provided.

DATE: Tuesday and Wednesday, May 17 and 18, 2005

TIME: 8:30 a.m. - 5:00 p.m.

             Morning Lecture: 8:30 a.m. - 12:00 noon

             Afternoon Lab: 1:00 p.m. - 5:00 p.m.

LOCATION: Morning Lecture: Steven M. Ross School of Business, Room K1310

                 Afternoon Lab: Steven M. Ross School of Business, Room E0402

FEE: Registration until May 2, 2005:
          $250 for University of Michigan affiliated faculty, staff and students
        $600 for others
        Registration after May 2, 2005:
        $300 for University of Michigan affiliated faculty, staff and students
        $720 for others

Please make checks payable to CSCAR-University of Michigan, or give the University of Michigan Project/Grant or Short Code to be billed. Send checks to: CSCAR, 3550 Rackham Bldg., University of Michigan, 915 E. Washington St., Ann Arbor MI 48109-1070.

Registration: Call CSCAR (734) 764-7828, ext. 0.  Enrollment is limited.