Random Rounding of Data

Most of the data taken from the population census were subjected to random rounding.

The "random rounding" method was applied to all tabulation figures, including totals, in order to prevent the possibility of associating statistical data with any identifiable individual. This technique does not add significant error to the census data. Under this method, all counts over "10" are randomly rounded to a multiple of "5"; consequently, all figures end in either a "5" or a "0". Counts under "10" are randomly rounded to a base of "10". All the figures will therefore take on a value of "0" or "10" after being subjected to the rounding. This method may cause the final data to suffer distorsion, which is even greater when small numbers are rounded. However, the margin of error due to rounding tends to cancel out when the data is aggregated for larger areas. With a view to protecting data confidentiality, some of the data may be deleted (e.g., data on very small municipalities) whenever random rounding does not provide sufficient protection.

This rounding has the following consequences:

• Data may vary from one table to the next. For example, although it is a question of the same data, the total number of families by structure may differ slightly from the total number of families by size since that data has been the subject of two independent random roundings.
• The sum of the parts does not necessarily add up to the total, since each figure is rounded independently. Similarly, the sum of percentages, which is calculated using rounded figures, does not necessarily add up to 100%.
  Example:
  Lone-parent families 615
  Female parent 515
  Male parent 105
• Data on geographic areas that are not available from Statistics Canada and have been obtained by adding up the data on the smaller areas composing them are not as accurate as they would have been if taken directly from Statistic Canada files.

Sampling and Weighting

The 2001 census data were collected either from 100% of the households or on a sample basis (i.e. from a random sample of one in five households), with the data weighted up to provide estimates for the entire population. Most of the data in the tables is of the second type. Consequently, some figures will differ from one table to the next, depending on whether the data were collected on a 100% basis or on a sample basis. Note that, on Indian reserves and in remote areas, all data were collected on a 100% basis.

Selection of Municipalities for Tables

In selecting the municipalities for the tables on the administrative regions:

• The 25 most populated municipalities in the administrative region at the time of the 2001 Census were chosen.
• When an RCM was not represented, the most populated municipality in that RCM was added to the list of the 25 main municipalities in the administrative region.
• Due to the low number of municipalities in the Nord-du-Québec region, all of them were selected.

Sources of Information

The information was taken from the population census conducted on May 15, 2001. Most of the data used to prepare the tables were from Statistic Canada's Topic-based tabulations and Profiles, and the definitions and methodological notes were elaborated using the 2001 Census Dictionary.

Undercoverage

The figures in the tables do not take into account the net undercoverage that took place in the 2001 Census.

One of the goals of the census is to enumerate the entire population on Census Day. Inevitably, however, a number of people will not be counted and others may be counted twice. Net undercoverage is obtained by calculating the difference between the number of people who were not counted and those who were counted twice.

In order to determine the number of people who were not counted or who were counted more than once, Statistics Canada conducts post-censal studies of a representative sample of the population. The results of these studies provide information that is used to rectify the census figures.

Undercoverage of the Aboriginal Population

Undercoverage in the 2001 census was considerably higher among Aboriginal people than among other segments of the population due to the fact that enumeration was not permitted in some Indian reserves and settlements. These geographic areas are called incompletely enumerated Indian reserves and settlements. There were four of them in Québec in the 2001 census: Akwesasne, Lac-Rapide (Rapid Lake ), Kahnawake and Kanesatake.

Data are not available for incompletely enumerated reserves and settlements, and these reserves and settlements are not included in tabulations. While the impact of the missing data tends to be small for province of Québec-level statistics, it can be significant for some smaller areas.

Legal Status

In a number of the tables, when two municipalities in an administrative region have the same name, the symbol of their legal status follows the municipality name, i.e., Senneterre (P) and Senneterre (V), the parish municipality and the town of Senneterre.

For a list with the legal status and geographic code of the most populated municipalities in a given administrative region, click on that administrative region and then on "Geographic Code and Legal Status".

__________________________
Sources: Statistique Canada, 2001 Census Dictionary.
Institut de la statistique du Québec, Direction de l'édition et des communications.

Update: January 29, 2003