|
by Patty Pitts
Some Canadians voiced concern
over Statistics Canada contracting
the Canadian division of U.S. company
Lockheed Martin to supply
scanners to read some forms in last
month’s national census.
But the Canadian government
used American technology to help
it crunch census data as far back
as 1891, says a UVic historian involved
in a national research project
to make past census data available
for future public use.
“In 1891, the government used
the hollerith, a punch card tabulating
machine, for the first time
instead of tabulating data by hand,”
says Peter Baskerville of the Canadian
century research infrastructure
project. He says concerns over
accuracy and security remain paramount
in the current multi-year,
multi-university initiative.
“Anyone working on this project
must undergo a criminal record
check and swear a confidentiality
oath to Statistics Canada, which
regularly checks our facility to ensure
it’s secure. Any data we send
to Statistics Canada undergoes
military-level encryption to ensure
the information remains secure. We
operate on a closed network.”
Baskerville and his UVic colleagues
are converting data from
microfilm and microfiche of the
1911, 1921, 1931, 1941 and 1951
censuses to a computer-compatible
format. The team is also putting the
data in context—finding a consistent
way to interpret ever-changing
terminology by referencing newspapers,
parliamentary debates, even
cartoons.
“The 1901 census was the last
one to use the term ‘colour’ and the
1911 census was the last one to use ‘crazy or lunatic,’ for example,” says
Baskerville. “If a user of our public
samples wants to understand the
meanings behind such usage we
provide data from debates of the
time in machine-readable form.
“We’re also making it possible
for users to compare, in a systematic
way, information across the
early 20th-century censuses. As
well, we’re providing maps for all
censuses so that the information
can be analysed from a geographic
perspective.”
The UVic team is responsible for
entry of census data north and west
of Ontario. Once completed, the
data are sent to Statistics Canada
for “anonymizing,” to strip away
any information that could lead
to individual identification. Data
entry for 1911 and 1921 is complete
and will be available to the
public in a format it can use online
by 2007.
The $15 million in funding
for the project (UVic’s share is
$2.1 million) is from the Canada
Foundation for Innovation, the first
time the agency has provided major
funding to a non-science project.
“From a historical perspective,
nothing can match the information
provided by these census data,” says
Baskerville. “It informs government
and influences their policies.
For disadvantaged groups, it’s a
mechanism to prove they exist and
deserve attention. A lack of accurate
census data would be a tragedy for
Canadian society.”
|
