[Phys-L] Re: Average earlier or average later?

[Brian Whatcott <betwys1@SBCGLOBAL.NET>]

At 08:53 AM 9/9/2005, you wrote:
> 1) Suppose the true relation is strongly nonlinear, such as y=A*x^10.
> 2) Suppose the the distribution of experimentally measured x is
> gaussian (due to random errors of measurements).
> 3) The distribution of the corresponding y values (calculated from
> individual x) will not be Gaussian; it will be skewed.
>
> For that reason averaging at the level of what is experimentally
> measured seems to be preferable.
> Ludwik Kowalski


There is the famous cautionary example of curve fitting, by translating
  raw data to logs, then best-fitting a straight line.

That weights the errors of small values unequally with the errors of
  large data values.

Better to fit with a test curve, and minimize the errors from the curve
directly.    This is much the same point that Ludwik makes.



Brian Whatcott    Altus OK    Eureka!
_______________________________________________
Phys-L mailing list
Phys-L@electron.physics.buffalo.edu
https://www.physics.buffalo.edu/mailman/listinfo/phys-l