 |
 |
Dear All * **Re: Another date for your diary * In The Queen's University Belfast on Wednesday December 9th 2015, at 3pm, Dr. Kevin Burke of the University of Limerick will present the following talk on a new class of models for analyzing survival data. (The room will announced later.) *_ Title_*: Non-proportional hazard effects *_Abstract_*: In the field of survival analysis, the Proportional Hazards (PH) model is the most popular regression model; however, non-PH covariate effects are often observed in practice. One classical explanation for non-PH effects is the presence of an unobservable subject-specific random effect (which multiplies the hazard function) - the so-called "frailty model". In this model covariates may have proportional effects at the individual level but are non-proportional at the population (i.e., marginal) level. An alternative explanation is that covariates truly do exhibit non-PH effects at the individual level (and, hence, at the population level). We will introduce the Multi-Parameter Regression (MPR) model - a fully parametric survival model which handles individual level non-PH covariate effects - and compare this with the frailty approach. Finally, we will combine the MPR and frailty approaches and consider the implications of this more general approach in the context of real data. We hope to see there. Note the 3pm start. Gilbert Hon Sec. -- _____________________________ Prof. Gilbert MacKenzie, Centre of Biostatistics, Dept. of Mathematics & Statistics, University of Limerick, Limerick Ireland CBS ~ http://www.ul.ie/biostatistics BIO-SI ~ http://www.ul.ie/bio-si Gilbert ~ http://www.staff.ul.ie/mackenzieg Email: [log in to unmask] _________________________ You may leave the list at any time by sending the command SIGNOFF allstat to [log in to unmask], leaving the subject line blank.