Dear colleagues,
Some weeks ago I queried the list about the meanings we might assign to the
statement that the probability of an event is 50%. I received lots of
responses; most of the useful ones were copied to the list. I appreciate the
effort that folks put into their answers, even the ones that scolded me for
appearing to be writing just to stop the EBH list from debating the results
of the trial of intercessory prayer.
I am hard pressed to report a consensus among those who responded, giving
the divergent nature of the comments. A (very) few items were agreed upon by
most, if not all, respondents:
a. We can usefully distinguish between 3 things:
1. The truth of the matter [i.e., patient really does or does not have
disorder or event really will or will not occur] which is unknown to us at
the time of probability estimation (or ever!);
2. The probability of that event, which can be understood as the frequency
with which we'd be correct if we concluded the event will occur from what we
know at the time [and this changes with our state of knowledge]; and,
3. Our certainty/uncertainty about this prediction, which might be variably
represented with confidence intervals around a probability estimate, or
measures of entropy, or in other ways.
b. For estimating the probability of a given disorder as the explanation for
a particular patient's illness, a probability of 50% actually represents a
relatively high value.
Again, I appreciate everyone's efforts to get me out of the dark.
Cheers!
WSR
W. Scott Richardson, M.D.
Audie L. Murphy Memorial Veterans Hospital *******************
7400 Merton Minter Blvd. Even in
San Antonio, TX 78284 the owl's swoop
T: (210) 617-5314 no sound
F: (210) 567-4423 M. Lerman
Email: [log in to unmask] *******************
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|