Sampling and Weighting Technical Report, Census of Population, 2021

Release date: October 4, 2023

Table of contents

Introduction

1. Census data collection

1.1 Census delivery methods

1.2 Census wave approach

1.3 Occupancy verification and follow-up activities for the 2021 Census

1.4 Census of Population questionnaires

1.5 Collection response rate

2. Sampling

2.1 Long-form sample universe

2.2 Long-form sampling design

3. Census data processing

3.1 Introduction

3.2 Receipt and registration

3.3 Scanning and keying from images

3.4 Coverage edits, completion edits and failed edit follow-up

3.5 Coding

3.6 Classification and non‑response adjustments for unoccupied and non‑response dwellings

3.7 Use of administrative data

3.8 Edit and imputation

3.9 Non‑response

3.10 Weighting

3.11 Final response rates

4. Estimation from the census long-form sample

4.1 Considerations in the choice of an estimation procedure

4.2 Weighting areas

4.3 Design weights

4.4 Coverage and total non‑response adjustment

4.5 Final calibration

4.6 Details on the selection of constraints

5. Evaluation of the weighting procedures

5.1 Distribution of the weights

5.2 Discrepancies between census counts and long-form estimates, Canada

6. Variance estimation

6.1 Elements to consider in choosing a variance estimation method

6.2 Variance estimator

6.3 Replicate weight adjustment

7. Statistical inference

7.1 Confidence intervals and their interpretation

7.2 Constructing confidence intervals

7.3 Student’s confidence interval

7.4 Modified Wilson confidence interval for proportions

7.5 Modified Wilson confidence interval for counts

8. Conclusion

Appendix A – Glossary

Appendix B – The history of sampling in the Canadian census

Appendix C – Constraints used in or excluded from the weighting process

Bibliography

Date modified: