BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Virtual BSU Seminar: 'Parameterizing Causal Marginal Models' - Pro
 f Robin Evans\, University of Oxford
DTSTART:20200616T130000Z
DTEND:20200616T140000Z
UID:TALK142747@talks.cam.ac.uk
CONTACT:Alison Quenault
DESCRIPTION:Many statistical problems in causal inference involve a probab
 ility distribution other than the one from which data are actually observe
 d\; as an additional complication\, the quantity of interest is often a ma
 rginal quantity of this other probability distribution.  This creates many
  practical complications for statistical inference\, even where the proble
 m is non-parametrically identified. \n\nNaive attempts to specify a model 
 parametrically can lead to unwanted consequences such as incompatible para
 metric assumptions or the so-called 'g-null paradox'. As a consequence it 
 is difficult to perform likelihood-based inference\, or even to simulate f
 rom the model in a general way.  We argue that this occurs because insuffi
 cient thought is given to which parts of the model are truly of interest\,
  what nuisance parameters remain to be specified given the parts of intere
 st\, and what other aspects of the distribution are redundant and can be d
 erived from the rest.  We provide a recipe for constructing a smooth\, non
 -redundant parameterization using (identifiable) causal quantities of inte
 rest. \n\nWe adapt some existing marginal parameterizations to causal mode
 ls for both continuous and discrete data\, allowing us to parameterize a w
 ide range of causal models including marginal structural models (MSMs)\, C
 ox MSMs and structural nested models. This makes it easy to simulate from 
 and fit models\; to introduce possibly high-dimensional individual-level c
 ovariates\; and to include additional assumptions such as stationarity or 
 symmetry. Our approach is likelihood-based\, and therefore also amenable t
 o a fully Bayesian approach.  We also provide an R package that implements
  these routines.\n
LOCATION:Virtual Seminar 
END:VEVENT
END:VCALENDAR