> My own belief is that you *can* use single-shot serum values to
> predict disease risk: but you have to use group these with a value
> that has areliable distribution: ex. group serum cholesterol by
> subject's bodyweight, which does not vary appreciably among
> individuals.
The amount of people needed to get an accurate group mean is not often
discussed, but I have calculated it to be the following:
Ng= Ni * Sb^2/(Sw^2*N)
Where,
Ng=number of days required to estimate group mean
Ni=number of days required to estimate individual true intake
Sb=variability between
Sw=variability within
N=number of subjects.
usually validation studies will give you the Sw^2/Sb^2 ratio and also
the number of days required to individual mean. Therefore the above
equation can be used to calculate either: a)amount of subjects needed
for an accurate group mean after a single draw with measurement error
or b)amount of days needed.
I have not come across this in the literature, not sure why.. I guess
people aren't interested in group means, they are more interested in
individuals..
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
