Learning : Research Talks
Learning : Research Talks
We regularly give research talks at universities, government institutions, and professional society meetings. Below are some of the talks that we have given that relate to the foundations, methods, or application of statistics to clinical trials.
On the use of prior information for flexibly monitoring group sequential survival trials.
This talk proposes a Bayesian framework for estimating the proportion of maximal statistical information at the time of an interim analysis when testing is based upon a weighted survival statistic. The methods are used in conjunction with the constrained boundaries approach of Burington and Emerson (2003) to efficiently implement a group sequential stopping rule.
Presenter & Location: Gillen DL, Design and Analysis of Experiments Conference (Columbia MO; October, 2009).
A constrained boundaries approach for monitoring time to event outcomes.
This talk explores the use of the constrained boundaries approach (Burington and Emerson, 2003) for monitoring clinical trials when the test statistic exhibits a non-linear information growth pattern.
Presenter & Location: Gillen DL, Eastern North American Region (ENAR) of the International Biometric Society (Arlington VA; March, 2008).
This talk proposes an alternative to the Cox estimate of the hazard ratio that is independent of the underlying censoring patterns exhibited in an RCT.
Presenter & Location: Gillen DL, Department of Epidemiology, University of California Irvine (Irvine CA; March, 2009).