Bell Curve - Volume of Random Numbers and Number or Bins?

[rlm]

In following the instructions for creating a bell curve in Excel
posted at http://support.microsoft.com/kb/213930, how can a person
determine the size and the number of bins that should be used? Also,
what is the formula for determining the number of random numbers that
should be generated? I am weeding through "Statistics for Dummies" :^)
but am struggling a bit. Finally is the bin size equal to the STDEV?

 

[Mike Middleton]

rlm -

how can a person determine the size and the number of bins that should be
used? <

A general guideline is five to fifteen bins, but analysts often use more
bins if the data set is large. I usually pick a "nice" round number as a
starting point for the bin intervals, and I use a "nice" round number for
the bin width. For example, if the values in a large data set range from 34
to 127, I might use 30,40,50,...,120,130 as the bins.

what is the formula for determining the number of random numbers that
should be generated? <

Classical statisticians have a formula for sample size (number of random
numbers) that depends on the confidence level you specify. But, in general,
a large sample is better than a small one. Also, it's instructive to try
different sample sizes, e.g., to compare 100 with 1,000 with 10,000, so that
you can observe the randomness and how its affected by sample size.

is the bin size equal to the STDEV? <

Not usually.

- Mike
http://www.MikeMiddleton.com