The Statistics menu option computes the probabilities that the values of the selected field quantity are within the interval:
If the field quantity's values are distributed normally, the resulting probabilities follow the 68–95–99.7 rule. This measure can be used as a simple test for outliers if the probability distribution is assumed normal and as a normality test if the probability distribution is potentially not normal.
Given a cumulative distribution function F(Y)
, the
probability p
for the k-sigma-interval is defined as:
where Y
denotes the quantity, j
denotes the mesh position, and kl
and
ku
denote the lower and upper
scaling factor of the standard deviation σ.
The CDF is approximated from the given sample values as described in Calculating Quantile Values.
Note:
The computation of the CDF requires at least 20 input samples.
If you do not have a lower bound, use a lower factor of -8.
If you do not have an upper bound, use a lower factor of +8.