For comparison,
finds confidence intervals for μ and σ
from averages and from (individual, classical) items.
[This is based on the real case of samples of bags ('items') and
their total weights (leading to 'averages').]
The items are simulated from a Gaussian
(μ, σ) distribution.
Draws plots of ('pdf's): Student's 't'
fE (experimental),
and fT (theoretical) for
nbig and nsmall
; and
'chi-square' fE (experimental),
and fT (theoretical) for
nbig and nsmall. | 
