Computes the
point estimates (PE) and confidence intervals (CI) of
μ and σ for unequal size
(or equal size) samples, from their sums alone.
The items are assumed Gaussian.
It uses the minimum sample size ansatz, thus
leading to distorted CIs (as in '(b)', below).
Example data can be inserted (copy, paste)
in the textarea, above, such as:
(a) equal size, 4 samples (n ≡ 20),
columns 1, 2;
(b) "industrial", unequal Case50, 50 samples
(144 ≤ n ≤ 1296), columns 3, 2.
Draws a plot of the averages.
*An example file with the data
in Table above is here (for download):
UnequalSize.txt. |
