next up previous
Next: Partial adjustments of belief Up: Adjusting beliefs in stages Previous: Adjusting beliefs in stages

Performing the initial adjustment

 

If we adjust tex2html_wrap_inline8554 solely by the Doctor's fasting measurement tex2html_wrap_inline8546 , we find that as a source of information tex2html_wrap_inline8546 is almost as effectual as tex2html_wrap_inline8546 and tex2html_wrap_inline8550 combined. For example the resolution for tex2html_wrap_inline8554 is tex2html_wrap_inline9272 , whereas we had earlier tex2html_wrap_inline9274 . Our adjusted expectations are

eqnarray1022

showing relatively small (1.17 standard deviations) changes from her initial expectations. Her adjusted variances are approximately

eqnarray1028

which are only slightly larger than those shown in section 4.2 for the full adjustment. We see also a size ratio of

displaymath9258

implying no conflict between data and prior belief as sources of information in the sense that this value is close to unity, its expectation.



David Wooff
Thu Oct 15 12:20:04 BST 1998