BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:When can we quantify uncertainty? - David Spiegelhalter\, Statisti
 cal Laboratory\, University of Cambridge
DTSTART:20091014T153000Z
DTEND:20091014T170000Z
UID:TALK20908@talks.cam.ac.uk
CONTACT:Richard Samworth
DESCRIPTION:Statisticians are trained to deal with uncertainties that can 
 be quantified by probability distributions or confidence levels (depending
  on one's philosophy).  But when modelling\, for example\, finance\, epide
 mics\, or\nclimate change\, there may be substantial doubts about how the 
 world works and what might happen.  In 1921 Frank Knight distinguished 'ri
 sk' (quantifiable) from 'uncertainty' (unquantifiable)\n\n[http://en.wikip
 edia.org/wiki/Knightian_uncertainty] \n\nand this distinction continues in
  critiques of the modelling process\, where such deeper uncertainties prov
 ide an argument for the 'precautionary principle' see\, for example\, Stir
 ling (2007)\n\n[http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1852
 772&blobtype=pdf]\n.  \n\nIn climate change\, the IPCC attempted to distin
 guish different 'types' of scientific uncertainty in their 4th assessment 
 report\n\n[http://www.climatechange.cn/qikan/manage/wenzhang/2005-142.pdf]
 \, \n\nwhereas the recent UK Climate Impact Programme assessments are full
 y probabilistic [see\n\nhttp://ukclimateprojections.defra.gov.uk/content/v
 iew/1394/543/ \n\nto see how East Anglia might be in a few years].\n\nI wo
 uld like to discuss how statisticians can deal with the limits of quantifi
 cation in uncertain models of complex phenomena.  I will tentatively sugge
 st an analytic structure and hope people might suggest improvements.  It m
 ay help to have a look at the links provided.\n
LOCATION:MR5\, CMS
END:VEVENT
END:VCALENDAR