Go to Statistics Portal

SHUTTLE ALGORITHM

Statistics Directorate    
Definition:
A method for finding lower and upper cell bounds by iterating through dependencies between cell counts.

Context:
There exist many dependencies between individual counts and aggregations of counts in contingency tables. Where not all individual counts are known, but some aggregated counts are known, the dependencies can be used to make inferences about the missing counts.

The Shuttle algorithm constructs a specific subset of the many possible dependencies and recursively iterates through them in order to find bounds on missing counts. As many dependencies will involve unknown counts, the dependencies need to be expressed in terms of inequalities involving lower and upper bounds, rather than simple equalities. The algorithm ends when a complete iteration fails to tighten the bounds on any cell counts.

Source Publication:
Statistics Netherlands, Statistics Canada, Germany FSO, University of Manchester, 2005, Glossary of Statistical Disclosure Control, incorporated in paper presented at Joint UNECE/Eurostat work session on statistical data confidentiality, Geneva, 9-11 October 2005.

Statistical Theme: Methodological information (metadata)

Created on Tuesday, November 22, 2005

Last updated on Thursday, December 1, 2005