In order to promote more rigorous and consistent evaluation of cell adjustment methods, this page provides a link to four benchmark sets of pre-adjustment tabular counts. All are based upon 1991 England and Wales Census outputs which, although subject to pre-release data perturbation, retain their original distributions of counts of all sizes. Each provides representative population and geographical coverage of England and Wales. These benchmark datasets are freely available for use subject to inclusion of the following acknowlegement of copyright and source in any outputs that arise:

"Benchmark data derived from 1991 England & Wales Census Small Area Statistics and freely available from http://pcwww.liv.ac.uk/~william/SDC/. Census output is Crown copyright and is reproduced with the permission of the Controller of HMSO."

Description of benchmark counts (table layouts)

Benchmark n=1
1000 enumeration districts (average population c. 300)

Benchmark n=16
1000 zones, each constructed through the aggregation of 16 geographically neighbouring enumeration districts.

Benchmark n=46
1000 zones, each constructed through the aggregation of 16 geographically neighbouring enumeration districts.

Baseline Benchmark
5350 enumeration districts, representing a 5% population-weighted sample from the 1991 England and Wales Census.