Database Quality Management

Given a set of integer pairs, DQM calculates two measures:
1. the probability that the rst elements in the pairs are similar to a given distribution,1 and
2. the probability the ratio of the rst and second elements of the pairs are similar, given that a success in the first element is greater than a success
in the second.
Similarity is de fined statistically: two sets of numbers are similar if they are derived from the same distribution. If it is more likely the distributions are
dierent, DQM also calculates whether it is more likely the dis-similaries are due to general differences, or if there exists a single outlying number or pair. Further, if an outlier is shown to be responsible for the dis-similarity, DQM determines whether the outlying number or pair's ratio is an excess or de ficit.