Posters

Section Chairs:
Ken Hardy

University of North Carolina

Chapel Hill, NC
Michael Riddle

Quintiles, Inc.

Research Triangle Park, NC

The Case of the Missing Data

Maribeth Johnson and Mark Litaker, , Medical College of Georgia, Augusta, GA

The MIXED procedure of the SAS System enables examination of correlational structures and variability changes between repeated measurements on experimental units across time. While PROC MIXED has the capacity to handle unbalanced data when the data are missing at random, a question arises as to when the degree of sparseness jeopardizes inference. Simulation is a tool that can be used to try to answer these types of questions. This paper shows how to simula te sets of data where assumptions have been made as to the variance-covariance (V-C) structure of the repeated measurements. Then observations are systematically deleted in a pattern determined from the data in question. The results of the MIXED analyse s on the balanced simulated data sets and those from the analysis of unbalanced sets can be compared to see if and when the underlying assumed V-C structure could be determined. An application of this technique to a longitudinal study of ambulatory blood pressures in children with a family history of hypertension will be shown.

Maribeth Johnson is a research statistician in the Office of Biostatistics at the Medical College of Georgia. She worked previously in the University of Georgia Department of Animal and Dairy Science for 11 years as a statistical programmer. She has M.S. degrees in Animal Breeding and Genetics from VA Tech and in Statistics from UGA. She has been using SAS for 16 years.

Mark Litaker is a research statistician at the Medical College of Georgia, and has used SAS for 9 years. He has a PhD in biostatistics from the University of South Carolina. His research interests include statistical methods in epide miologic studies and the evaluation of diagnostic tests.

Section Chairs:		Ken Hardy University of North Carolina Chapel Hill, NC	Michael Riddle Quintiles, Inc. Research Triangle Park, NC

	The Case of the Missing Data
	Maribeth Johnson and Mark Litaker, , Medical College of Georgia, Augusta, GA
	The MIXED procedure of the SAS System enables examination of correlational structures and variability changes between repeated measurements on experimental units across time. While PROC MIXED has the capacity to handle unbalanced data when the data are missing at random, a question arises as to when the degree of sparseness jeopardizes inference. Simulation is a tool that can be used to try to answer these types of questions. This paper shows how to simula te sets of data where assumptions have been made as to the variance-covariance (V-C) structure of the repeated measurements. Then observations are systematically deleted in a pattern determined from the data in question. The results of the MIXED analyse s on the balanced simulated data sets and those from the analysis of unbalanced sets can be compared to see if and when the underlying assumed V-C structure could be determined. An application of this technique to a longitudinal study of ambulatory blood pressures in children with a family history of hypertension will be shown.
	Maribeth Johnson is a research statistician in the Office of Biostatistics at the Medical College of Georgia. She worked previously in the University of Georgia Department of Animal and Dairy Science for 11 years as a statistical programmer. She has M.S. degrees in Animal Breeding and Genetics from VA Tech and in Statistics from UGA. She has been using SAS for 16 years. Mark Litaker is a research statistician at the Medical College of Georgia, and has used SAS for 9 years. He has a PhD in biostatistics from the University of South Carolina. His research interests include statistical methods in epide miologic studies and the evaluation of diagnostic tests.