BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY: Default priors and model parametrization - Nancy Reid\, Universit
 y of Toronto
DTSTART:20090507T150000Z
DTEND:20090507T160000Z
UID:TALK18115@talks.cam.ac.uk
CONTACT:8047
DESCRIPTION: This talk presents the development of classes of priors that\
 nensure calibration of the resulting posterior inferences.  These\npriors 
 are built using asymptotic properties of likelihood inference\nand locatio
 n model approximations to general models.  The role of\nparameterization o
 f the model in obtaining calibrated posterior\ninference for sub-parameter
 s is described\, and the proposed priors are\nrelated to Jeffreys' prior a
 nd the Welch-Peers approach. Connections\nare made to so-called strong mat
 ching priors\, which are data dependent\npriors derived by equating poster
 ior marginal probabilities to\nconditional $p$-values\, and the importance
  of targetting the prior on\nthe parameter of interest.\n\n\n\n
LOCATION:MR5\, CMS\, Wilberforce Road\, Cambridge\, CB3 0WB
END:VEVENT
END:VCALENDAR