Multi-state models for observed and latent cognitive function in the older population
| Dates: | 20 March 2019 |
|---|
| Times: | 14:00 - 15:00 |
|---|
| What is it: | Seminar |
|---|
| Organiser: | Department of Mathematics |
|---|
| Who is it for: | University staff, External researchers, Adults, Alumni, Current University students |
|---|
| Speaker: | Dr Ardo van Den Hout |
|---|
|
|
Join us for this research seminar, part of the Probability and Statistics seminar series.
Due to the ageing population there is a growing interest in the statistical modelling of cognitive function in old age. When analysing longitudinal data on ageing, lost to follow-up because of death cannot be ignored. One option is to model survival and change of cognitive function jointly by specifying submodels for the two processes and linking these models by individual-specific random effects.
Another option – and the topic of this seminar - is to use a continuous-time multi-state survival model where a series of living states is defined by the level of cognitive function and an additional dead state is included. This multi-state approach is quite general and can be used in many other applications in biostatistics, social statistics, and demography.
The seminar will start with introducing the continuous-time multi-state survival model by discussing model specification, maximum likelihood estimation, and some applications. The second part will present an extension of current methods: a hidden Markov model for modelling bivariate cognitive function, where each state-dependent distribution is a bivariate extension of the binomial distribution.
The methods will be illustrated by using longitudinal data from a UK survey of the older population.
Speaker
Dr Ardo van Den Hout
Organisation: University College London
Biography: See Dr vam Den Hout's research profile
Travel and Contact Information
Find event
Frank Adams 2
Alan Turing Building
Manchester
