Print

Print


Dr Ivonne Solis-Trapala will be presenting a methodology seminar on "Graphical Markov models and recent applications", as part of our Biostatistics seminar series at Arthritis Research UK Primary Care Centre, Keele University.

Monday 5th June 2-4pm in Dinwoodie lecture theatre, Primary care sciences, David Weatherall building, Keele University


Abstract: We present an overview of graphical Markov models as a useful statistical modelling strategy for studies aiming to describe complex interrelationships among many variables. The talk will focus the discussion on Sequences of Regressions, a subclass of graphical Markov models, and recent applications. Firstly, in a study involving postmenopausal women, we separate the mechanical effects of body composition, including fat, from physiological non-mechanical effects on bone outcomes, to show how, contrary to general belief, the production of fat hormones may increase the risk of bone fracture. Secondly, in a study involving dialysis patients, we present a modelling strategy to gain understanding on the intricate contributions of patient comorbidities and hospital practice patterns to the association between treatment modality and patient outcomes. We provide general guidance on how to fit the model in R.

You may leave the list at any time by sending the command

SIGNOFF allstat

to [log in to unmask], leaving the subject line blank.