Re: Testing for product of Gaussians

Firstly thanks to you both - I hope you don't mind my using your
fornames in the below.

Stephen - Thanks for the mathworld pointer - that's very helpful -
there's always a name for these sort of things but due to lack of
statistical knowledge I never know what it is nor how to find it. I did
think of using Chi squared given I have expected and observed counts
for each bin, but rejected it out of hand based on my limited knowledge
of stats because I thought that unless I only tested the very centre of
the plane the low expected counts for bins corresponding to large
radial distances would skew the final statistic. Maybe I am
misunderstanding?

Herman - Thanks, I'll look into the Kolmogorov-Smirnov test, which
sounds ideal for multiple purposes as you can tell it what you
distribution you are expecting, and sounds sufficient for my needs.
However, as a point of interest, does its flexibility make it a less
powerful test - once again based on my limited knowledge I would have
thought that it would be possible to produce a more specific test when
you know the distribution you are testing for in advance rather than
"plugging it in" to a generic framework as your "distribution free"
comment implies? And when the statisticians of old were devising out
particular tests for particular distributions I would have thought a
test for normality would be pretty much the first one they would
produce?

.

Relevant Pages

  • Re: Explaining significance to laypersons: why the tails?
    ... >> the same time sufficiently convincing rationale for why statisticians ... >A good example is looking at the binomial distribution with large N and ... which does not mean much to the layperson; ...
    (sci.stat.math)
  • Re: Explaining significance to laypersons: why the tails?
    ... > the same time sufficiently convincing rationale for why statisticians ... A good example is looking at the binomial distribution with large N and ... Perhaps you could use the flipping of a fair coin 2000 times. ... so if we just the probability of that result and compared it ...
    (sci.stat.math)
  • Re: Dynamical Systems and Expansion-Contraction
    ... >>From Osher Doctorow ... academic statisticians) point out in one of their early 1970s volumes ... bivariate normal/Gaussian distribution with correlation coefficient not ...
    (sci.physics)
  • Re: wow... Lidle did not...
    ... Vinnie S. wrote: ... not expecting you do have the ambition to do it. ... discussed the distribution of runs per inning based on a pitcher's ERA. ... If the probability ...
    (alt.sports.baseball.ny-yankees)
  • Re: "Extended" probability problem
    ... large bottles have an expected wastage of 1.0%. ... The exact answer for a normal distribution would involve integrating ... calculation you would be expecting. ...
    (sci.math)