Français
  Register!
Forgot your Password?

Work

Data Collection

During the 1990's, volunteers working with The Church of Jesus Christ of Latter-Days Saints (LSD) based in Utah invested approximately 200,000 hours transcribing the manuscript population schedules of the 1881 Cencus of Population into a database. This work was undertaken as part of ongoing project to construct complete genealogical records for all LSD members.

Data Cleaning history / First Stage: University of Ottawa

From September, 2000, to September, 2001, The Institute of Canadian Studies, University of Ottawa, hosted Phase 1 of the 1881 Canadian Census Project. This phase involved a team of history graduate students and volunteers drawn from genealogical societies within the Ottawa area to conduct a series of cleaning and checking tasks to enhance the data. The team verified and corrected the geographic information contained in the file, eliminated duplicate cases, identified missing cases, and standardized the age, birthplace and origin fields for use in the LDS search engines. For this task, they referred to a cleaning manual. During this process, the students learned valuable research skills in the use of MS ACCESS and MS EXCEL in creating and refining historical microdatabases.

Volunteers from genealogical societies within the Ottawa area continued with checks on inconsistencies within the age, marital status and household position fields, and also inserted single lines of missing data into the file. They completed these checks for Prince Edward Island, Nova Scotia, New Brunswick, Quebec, Ontario and the West. Previously missing data that had been transcribed by LDS members in Utah from February to June, 2001, has also been integrated into the 1881 database.

By September, 2001, the 1881 Canadian Census Project had completed the first phase of cleaning and checking for all provinces. The complete 1881 Canadian census file contains 4,278,387 records. In September, 2001, this data was sent to our Latter-Day Saints partners in Utah who then released the 1881 Canadian Census database as a genealogical index through The FamilySearch Internet Genealogy Service.

The LDS have released the genealogical database containing the 1880 Census of the United States on CD-ROM; this data is also available via FamilySearch Internet Genealogy Service. This database was cleaned and checked by our partners, the Minnesota Population Center.

Phase 1 was directed by Lisa Dillon, who at the time held the position of Research Co-ordinator, and Chad Gaffield, Director of the Institute of Canadian Studies. For more information regarding Phase 1 of the 1881 Canadian Census Project, please visit this link.

For further information about Phase 1 of the 1881 Canadian Census Project, please see:

Lisa Dillon. "International Partners, Local Volunteers and Lots of Data: The 1881 Canadian Census Project." History and Computing (forthcoming).

Lisa Dillon and Steven Ruggles. "Creating Historical Snapshots of North America in 1880/1: Collaboration between Historians and the Church of Jesus Christ of Latter-Day Saints on the 1880/1 census databases of the United States and Canada." Genealogical Journal. Vol. 29, No. 3: 107-113.

Data cleanin history / Second stage: Université de Montréal

The headquarters of the 1881 Canadian Census Project is now located in the Département de Démographie at the Université de Montréal where project director, Lisa Dillon, has taken the position of assistant professor. Since the 1960s, the Département de Démographie has been host to a world-renowned research and teaching program in historical demography. With the inclusion of the Registre de la Population du Québec ancien (RPQA), a database comprising of more than 700,000 baptismal, marriage and death certificates registered before 1800, the Programme de Recherche en Démographie Historique (PRDH) directs its focus upon the historical censuses of Canada. Current additional components of the PRDH include the entire 1881 Census of Canada, as well as a 20% population sample drawn from the census of 1852. Thanks to the collaboration of Bertrand Desjardins, Aggregate Researcher and co-founder of the RPQA, phase 2 of this project has benefited greatly from the expertise acquired through the PRDH. This second phase was funded by the North Atlantic Population Project and a subsidy from the National Science Foundation. The initiative was orchestrated by Steven Ruggles, Director of the Minnesota Population Center and creator of the Integrated Public-Use Microdata Series (IPUMS) database, with the objective to successfully clean and refine the data for further use in social science research.

Status encoding: occupations

In 2002, research assistants Samuel Rousseau (Dépt. d'Histoire, Université de Montréal), Sarah Hayford (Dépt. de Démographie, Université de Montréal and Dept. of Demography, University of Pennsylvania) and Cheryl Desroches (Dept. of History, University of Ottawa) were hired to implement the codification of occupations recorded in the 1881 Canadian Census. To begin, they standardized occupations by eliminating erroneous characters (such as slashes, commas, periods or asterisks) as well as typos created during the data entry process. Any additional variations found within the data (such as abbreviations or added information) were left untouched. With this simple standardization pass, the numbers of unique occupational titles were reduced from 42,000 to approximately 24,000. An additional 4,000 occupational strings from the 1871 and 1901 Canadian census databases have also been incorporated into the file in order to enhance compatibility across census years.

Research assistant Cheryl Desroches (University of Ottawa) conducted the standardization and codification of the variables origin, religion and birthplace. Samuel Rousseau has edited the age variables and, under the guidance of Bertrand Desjardins, has also standardized French surnames which were especially poorly transcribed. Louis Duchesne, a research volunteer and employee at l'Institut de la statistique du Québec, conducted the standardization of first names as part of his own research on francophone names. These corrections and standardizations of names were necessary in order to construct household interrelationship variables, as well as to facilitate in future record linkage efforts.

In 2003, Alexandre Bujold (Dépt. d'histoire, Université de Montréal) was hired to work on checking tasks (such as the household number, name cleaning, and further checks in inconsistencies not previously undertaken). This work was finished by France Normand (Dépt. d'histoire, Université de Montréal) in the fall of 2004. The remaining 1881 tasks (imputing and inferring constructed variables) will be done in Minnesota, while the documentation will be written by Lisa Dillon and Evan Roberts with the help of the research assistants.

Last updated: 9/1/2010

Copyright © Département de démographie (UdeM)
All rights reserved