|
The protection of the confidentiality of personal
Census information is paramount, and a number of protection
measures will be used to prevent the inadvertent disclosure of
information about identifiable individuals. In particular:
 |
a sample of Census records will be swapped between geographical
locations in the output database, adding a small amount of
uncertainty to Census outputs.
|
|
minimum thresholds of numbers of households and individuals will
apply before output is produced. These will be 40
households and 100 individuals for Census Area Statistics
and 400 households and 1,000 people for Standard Tables.
Desirable Census output that falls below these threshold
criteria will be examined on a case-by-case basis.
|
|
the smallest area for which commissioned output will be released is
a Census Output Area which contains approximately 125
households. All other areas will be amalgamations of
these areas.
|
|
all outputs will be subject to small cell adjustment, that is, for
those cells in output tables where the cell count is
small, the count will be adjusted. |
o
only small counts are adjusted
o
information on what constitutes a small cell count cannot be
provided as this may compromise confidentiality protection
o
the magnitude of the adjustment will generally have little impact
on the conclusions that can validly be drawn from the data
o
totals and subtotals in tables are calculated as the sum of the
adjusted data so that all tables are internally additive; within
tables, totals and subtotals are the sum of the adjusted
constituent counts
o
tables are independently adjusted; this means that counts of the
same population in two different tables may not necessarily be
the same
o
tables for higher geographical levels are independently adjusted,
and, therefore, will not necessarily be the sum of the lower
geographical component units
o
caution should be taken in interpreting small cell counts as they
are impacted to a greater proportion than larger cells
o
some small cells will be randomly altered to zero. Caution should
be exercised in deducing that there are no people or households
in an area having particular characteristics
o
when calculating proportions, percentages or ratios from
cross-classified or small area tables, the random error
introduced can be ignored except when very small cells are
involved, in which case the impact on percentages and ratios can
be significant
Click
HERE for more detail on
small cell adjustment (option 2). Although this paper is written
from an England and Wales perspective, the small cell adjustment
methodology is the same as that used in Northern Ireland.
All output tables have been assessed for disclosure risk and data
utility. A general principle has been applied to the design of
all the 2001 Census Output of making the average cell count in a
table greater than or equal to one.
A condition of use included in all end user licenses is that the
Census material shall not be used to attempt to derive
information relating to an identified person or household nor
shall a claim be made that such information has been obtained or
derived.
|
