View the most recent version.

2006 Census Technical Report: Coverage

1. Estimates of population coverage error

2. Census universes

3. Population coverage error

4. Census data collection

5. Census data processing

6. Dwelling Classification Survey

7. Reverse Record Check

8. Census Overcoverage Study

9. Estimation

10. Evaluation of Coverage Studies

11. Historical estimates of population coverage error

12. Special topics

Appendix A. Reverse Record Check Survey questionnaires

Appendix B. Acronyms

Bibliography

 

Introduction

The 2006 Census required the participation of the entire population of Canada, some 32.5 million people distributed over a territory of nine million square kilometres. Although there are high quality standards governing the collection and processing of the data, it is not possible to eliminate all errors. In order to help users assess the usefulness of census data for their purposes, the 2006 Census technical reports detail the conceptual framework and definitions used in conducting the census, as well as the data collection and processing procedures employed. The principal sources of error, including where possible the size of these errors, are described, as are any unusual circumstances which might limit the usefulness or interpretation of census data. With this information, users can determine the risks involved in basing conclusions or decisions on census data.

This technical report addresses 2006 Census population coverage errors. There are two types of coverage error. Population undercoverage refers to the error of excluding someone who should have been enumerated. Population overcoverage refers to the error of either enumerating someone more than once or including someone who should not have been enumerated. The latter error is considered negligible. Undercoverage is more common than overcoverage. The net impact of undercoverage and overcoverage on the size of a population of interest is population net undercoverage. Net undercoverage is calculated as the number of persons excluded who should have been enumerated (undercoverage) less the number of excess enumerations of persons enumerated more than once (overcoverage). Coverage errors are one of the most important types of error since they affect not only the accuracy of the counts of the various census universes, but also the accuracy of all of the census data describing the characteristics of these universes.

Users of census data should be aware that the presence of coverage error in the 2006 Census means that census products may present the results of a less than complete enumeration and/or include duplicate enumerations. Undercoverage, for example, is highest for young adult males. Users are directed to Section 1 to obtain estimates of 2006 Census population coverage error for a variety of demographic and geographic levels and groupings.

Section 2 gives the 2006 Census conceptual framework and definitions for the population universe, the dwelling universe, and the usual place of residence. This is what the census is trying to measure. Section 3 discusses coverage error, sources of coverage error, census practices that minimize coverage error, the conceptual framework for measuring coverage error, and an introduction to the census coverage studies. Section 4 and Section 5 describe how the 2006 Census was done from the frames to data collection to editing, coding, imputation, and weighting.

Census coverage error is measured by three studies. The 2006 Dwelling Classification Survey (DCS) addressed coverage error resulting from dwelling occupancy classification error. Census data were adjusted for this type of coverage error. The 2006 Reverse Record Check (RRC) measured population undercoverage. The 2006 Census Overcoverage Study (COS) measured population overcoverage. Census data are not adjusted for the population coverage error measured by the RRC and the COS. Rather, estimates of net undercoverage are used in the production of Statistics Canada's demographic estimates of population.

The 2006 Census coverage studies differ from the 2001 Census coverage studies:

1. The 2006 Census was the first time that the names of persons listed on all of the census forms were available in electronic format. This change greatly increased the efficiency of coverage studies since matching could include the name and not be restricted to demographic characteristics.
2. A new coverage study, the COS, was designed to exploit the use of an individual's name for identifying overcoverage. The COS was able to evaluate overcoverage resulting from persons being enumerated more than once with a high degree of accuracy.
3. The measurement of population overcoverage was dropped from the RRC. Consequently, much less field collection was required since only those persons that could not be easily found on the census database were sent to the field.
4. There is a change in terminology. What used to be called 'gross undercoverage' is now 'undercoverage.' The more complete label is '2006 Census population undercoverage.'
5. The Automated Match Study (AMS) was carried out for the 2001 Census and has been in place as a coverage study since the 1991 Census. The AMS was repeated for the 2006 Census but the results were primarily used for evaluating the COS.

Section 6 describes the methodology and results of the 2006 Dwelling Classification Survey (DCS). This survey, carried out after census non-response follow-up, provides information used in the census to account for persons living in non-response dwellings and in occupied dwellings misclassified as unoccupied. This is done by imputing persons onto the census database via the whole household imputation (WHI) procedure. The number of persons added in WHI is a key input for the estimates of population coverage error.

Estimates of coverage error are produced only for the census population universe. Section 7 describes the methodology and results of the 2006 RRC. Section 8 describes the methodology and results of the 2006 COS. Section 9 shows how the results of the RRC and the COS are combined with census data to produce estimates of population coverage error and their associated estimated standard errors. Given the extensive use made of estimates of net undercoverage, it is important to undertake critical and detailed evaluations. Section 10 presents the results of evaluations done for the RRC and the COS as well as an evaluation of the error of closure. The error of closure is the difference between demographic estimates and census counts adjusted for net undercoverage.

Statistics Canada has conducted census population coverage studies since the Reverse Record Check methodology was first applied to the 1961 Census¹. The historical perspective from the 1971 Census to the 2006 Census is given in Section 11.

Section 12 presents additional topics. We examine the degree to which all persons who should have been enumerated were not and population coverage error for the Aboriginal identity population.

Appendix A contains the 2006 RRC Survey questionnaires and Appendix B lists all of the acronyms used in this report.

This report has been prepared by Mark Armstrong, Karen Bruce, Colleen Clark, Peter Dick, Heather Farr, Josée Morel, Karen Switzer, Alain Théberge and Christian Thibault, members of the Social Survey Methods Division, and Denis Morissette from the Demography Division. Normand Laniel and David Dolson from the Social Survey Methods Division contributed valuable comments on earlier drafts that improved the content and readability of the final report. The support of members of the Census Operations Division, the Demography Division, and the Social Survey Methods Division is noted with appreciation.

You can find additional information on census concepts, variables and geography in the 2006 Census Dictionary. You can find additional information on the complete census process in the 2006 Census reference materials.

Note:

1. The first RRC was conducted in 1961 but there was no frame of persons missed in the previous census. The 1966 RRC used the results of the 1961 RRC to construct the frame of persons missed in the 1961 Census.

Next page