|
Unit of Analysis
| Individuals |
|
Universe
| The census is divided into four universes (sets): 1. Population (i.e. persons): Provides information on the characteristics of Canada's population; 2. Families: pertains to the characteristics of family units. Both census and economic families are included; 3. Households: Covers the characteristics of a person or a group who occupy a private dwelling; 4. Dwellings: Describes the characteristics of dwelling units in Canada. The census enumerates the entire Canadian population, which consists of Canadian citizens (by birth and by naturalization), landed immigrants, and non-permanent residents together with family members who live with them. Non-permanent residents are persons living in Canada who have a Minister's permit, a student or employment authorization, or who are claiming refugee status. The census also counts Canadian citizens and landed immigrants who are temporarily outside the country on Census Day. This includes federal and provincial government employees working outside Canada, Canadian embassy staff posted to other countries, members of the Canadian Armed Forces stationed abroad, and all Canadian crew members of merchant vessels. Because persons outside the country are enumerated, the Census of Canada is considered a modified de jure census. Some variables are collected for the entire universe; others are collected for a 20% sample only. |
|
Time Method
| Cross-sectional |
|
Data Collector
| Statistics Canada |
|
Frequency
| Quinquennial |
|
Sampling Procedure
| In the 1996 Census, 80% of households received a short questionnaire (2A) containing seven questions, while 20% were given the 55-question long form (2B). The long Census form (2B) is collected on a sample basis. The approach used in sampling areas is a systematic 1 in 5 sample selection of households. Initially, the weight for each sampled household is simply 5 - because a 1 in 5 sample was selected. However, although we know the sample contains almost one-fifth of all dwellings, we cannot be certain that all sub-groups (such as males aged 20 to 24) will be properly represented. To ensure that the most important sub-groups are consistently estimated, a weighting procedure known as "two step generalized least squares estimation procedure" is used. This method constrains the sample weights to known population counts for the most important sub-groups for both the collection unit (known as the Enumeration Area) and for a larger geographic unit (known as the Weighting Area). The estimation method will adjust the initial household weight as little as possible. |
|
Collection Mode
| This stage of the census process ensured that each of the 10.9 million dwellings in Canada was enumerated. To ensure the best possible coverage, the country is divided into small geographic areas called enumeration areas (EAs). Each census representative is responsible for at least one EA. The optimal number of households in an EA ranges from 125 in rural areas to 440 in urban areas. In the 1996 Census, there were 49,361 enumeration areas in Canada, and 38,000 people engaged in collecting the data. a) Self-enumeration: In 1996, census representatives dropped off census questionnaires at approximately 98% of households. These households were asked to complete the questionnaire themselves on May 14, and mail it in. Short Questionnaire (Form 2A): - given to four fifths of all households in mail-back enumeration area; - contains seven questions: name, relationship to Person 1, date of birth, sex, legal marital status, common-law status and first language leaned in childhood. Long-Questionnaire (Form 2B): - given to every fifth household in mail-back enumeration area; - contains 55 questions, including the seven questions on Form 2A as well as questions on labour force activity, income, education, activity limitations, citizenship, housing, ethnic origin, and so on. b) Interview: Roughly 2% of households were enumerated by interview; in other words, a census representative visited the household and completed the questionnaire during an interview. This method was used in remote and northern areas, and on most Indian reserves. It was also used in the town centres of large urban areas, where residents are more difficult to enumerate. In addition, some remote northern areas were enumerated in February and March of 1996. This advance census was carried out in areas where communities disperse in the spring and migrate to their hunting and fishing grounds. |
|
Characteristics of Data Collection Situation
| To support collection activities, an extensive communications program was established to make Canadians aware of the census and the importance of taking part in it. In addition, to respond to people's questions and concerns about this vast undertaking, the Census Help Line (CHL), a free, nation-wide, multilingual service, was available to all respondents. |
|
Weighting
| Weights are calculated at the household level. Weighting was done by a method known as calibration or regression estimation. Calibration estimation starts with initial weights of approximately 5 and then adjusts them by the smallest possible amount need to agreement between the sample estimates (e.g. number of males, number of people aged 15 to 19) and the actual population counts. |
|
Response Rate
| The basic unit used in the calculation of the response rate is the household. (This includes private and collective households, 2Cs, random additions for misclassified vacants and Form 4 households. It excludes unoccupied dwellings and dwellings occupied only by foreign or temporary residents.) In the 1996 Census, there were 10,803,628 in-scope households on the database, of which 88,641 were total non-response (Form 4) households. There were 10,714,987 households with at least some data. The overall response rate was therefore 10,714,987 / 10,803,628 = 99.18%. |
|
Estimates of Sampling Error
| The Generalised Least Squares Estimation Procedure (GLSEP) is used to produce the census estimates. The GLSEP reduces the standard errors of the sample estimates and ensures that the estimates agree with known population counts. |
