Description:

The Anderson-Darling K-sample test can be used to test whether
several independent random samples of various sizes come from the
same but unspecified continuous distribution.It is a rank test and
consistent against all alternatives.A low to moderate number of tied
observations can be tolerated.The combination of such tests can be
used to test whether M groups of samples (with K allowed to vary from
group to group) come from respective common distributions, which may
vary from group to group.This is useful in testing for treatment
effects in randomized (incomplete) block designs or in examining
whether several laboratories perform equally well when asked to
measure a sufficient number of test speciments from different batches
or materials.