Journals of Gerontology Series A: Biological Sciences and Medical Sciences, Vol 50, Issue 3 B177-B185, Copyright © 1995 by The Gerontological Society of America
Survival estimates and sample size: what can we conclude?
R Shouman and M Witten
Department of Applications Research and Development, University of Texas System, USA.
Attempts to understand aging processes often involve life-span measurements
from which a survival curve is constructed and model parameters estimated.
The parameter estimates are then compared, and conclusions concerning the
underlying biological processes are subsequently deduced, based upon the
magnitude of the parameter differences. In this article we discuss the role
of sample size and sample fluctuation on the parameter estimates and the
profound effect that these factors may play in our arrival at meaningful
biological conclusions. We then extend this discussion to examine one
methodology that can help select sample sizes for specific parametric
survival models.
