|
SPEAKER: |
|
|
TITLE: |
"On the Use of Auxillary Variables for Incomplete Data Problems" |
|
DAY: |
Wednesday, November 19, 1997 |
|
TIME: |
3:30 p.m. [Coffee in BSB-202 at 3:00 p.m.] |
|
PLACE: |
BSB-108 |
Incomplete data is a problem investigators face in a wide variety of research settings. On occasion, studies are designed so that the primary response is only observed in a fraction of the sample. These designs may be selected because it is expensive, unduly invasive, or unethical to observe the primary response. In other situations, a stochastic mechanism may lead to censored values for the primary response (e.g. as in event history analysis). There has recently been considerable interest in the use of auxillary variables to improve efficiency in inferences regarding the distribution of the primary response in the presence of incomplete data. In this talk I will consider a variety of scenarios in medical research in which these problems arise and will review the various methods that have been proposed. The results of a preliminary simulation study will also be discussed, which was designed to provide practical insight into the utility of auxillary variables in settings with censored data.
This research is in progress and is joint with Prof. J.F. Lawless.
Dr. Richard Cook obtained a B.Sc. (Statistics) from McMaster University in 1988 and a M.Math and Ph.D. in Statistics from the University of Waterloo in 1989 and 1993, respectively. His research interests include event history analysis, the design and analysis of clinical trials, and collaborative medical research. Dr. Cook is a Scholar of the Medical Research Council of Canada and is President of the Biostatistics Section of the Statistical Society of Canada.
The following articles have been provided by Dr. Cook to be used as background for his talk. They are on reserve at Thode Library (STATS 770: Statistics Seminar).
[1] Fleming T.R., Prentice R.L, Pepe M.S. and Glidden D. (1994). "Surrogate and Auxillary Endpoints in Clinical Trials, With Potential Applications in Cancer and AIDS Research," STATISTICS IN MEDICINE 13, pp. 955-968.
[2] Finkelstein D.M. and Schoenfeld D.A. (1994). "Analysing Survival in the Presence of an Auxillary Variable," STATISTICS IN MEDICINE 13, pp. 1747-1754.