Q/Hampel web service for PT providers
Computation of mean and standard deviation with the robust Q/Hampel method
This web application enables users to estimate the mean and standard deviation from a data set. The statistical parameters are computed by means of the Q/Hampel method, a robust algorithm useful for data sets with outliers.
Q/Hampel method
The Q/Hampel method is a robust statistical method for the computation of mean and standard deviation. In a first step, the standard deviation is estimated by means of the Q method, which is based on the consideration of pairwise absolute differences. This standard deviation is then used for the Hampel estimation of the mean, which is based on the principle of winsorizing.
The Q/Hampel method is more robust than the algorithm A+S, and has already been adopted in the new version of ISO 13528.
It is implemented in PROLab Plus, a software package for the evaluation of interlaboratory studies.