Friday, April 7, 2017

Another Look at the Spring Equinox Statistics


  I redid the statistics for the deviations from fit formula and got something a little more normal using fewer intervals


Microsoft's histograms are a little difficult to do. For this one I set the underflow and overflow values at -16 and 16 respectively and used 6.4 for the interval width. One can compare the observed values with the expected values and their standard deviations.


One would expect most of the observed values to be within 3 standard deviations of the expected values. Under the circumstances one cannot rule out a normal distribution for the deviations.

Supplemental (Apr 8): I used numerical integration to evaluate the probability integrals for the expected values and their deviations. A more accurate calculation shows that for the 16< interval the expected value is 1.6.

No comments: