![]() | |
![]() |
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
![]() |
Data quality2001 Census of Agriculture concepts, methodology and data qualityUsing the following information will ensure a clear understanding of the basic concepts that define the data provided in this product, and of the underlying census methodology and key aspects of the data quality. It will give you a better understanding of how the data can be effectively used and analysed according to their strengths and limitations. The information may be particularly important when making comparisons with data from other surveys or sources of information, and in drawing conclusions regarding change over time. Data sources and
methodology
Data sources and methodologyThe Census of Agriculture collects and disseminates a wide range of data on the agriculture industry such as number and type of farms, farm operator characteristics, business operating arrangements, land management practices, crop areas, numbers of livestock and poultry, farm capital, operating expenses and receipts, and farm machinery and equipment. These data provide a comprehensive picture of the agriculture industry across Canada every five years at the national, provincial and territorial levels as well as at lower levels of geography. General methodologyTarget population The Census of Agriculture also collects and disseminates data pertaining to a related sub-population farm operators. In 2001, "farm operators" was defined as those persons responsible for the day-to-day management decisions made in the operation of a census farm or agricultural operation. Up to three farm operators could be reported per farm. Prior to the 1991 Census of Agriculture, the farm operator referred to only one person responsible for the day-to-day decisions made in running an agricultural operation. Collection Data processing
Reference periodThe Census of Agriculture has been conducted concurrently with the Census of Population every five years since 1951. The 2001 Census of Agriculture was conducted on May 15, 2001.
RevisionsData from the Census of Agriculture are not subject to revision.
AdjustmentsData from the Census of Agriculture are not subject to seasonal adjustments or benchmarking to other data sources.
Concepts and variables measuredFor a full description of census concepts, derived variables and geographic levels, please click on Census terms and Geographic definitions.
Data accuracyAn integral part of each Census of Agriculture is the implementation of new or enhanced methods, procedures and technologies that improve not only the collection, but also the processing, validation and dissemination of the data. New methods, procedures and technologies adopted for the 2001 Census of Agriculture included the Farm Coverage Follow-up Survey, the Coverage Evaluation Survey and the use of intelligent character recognition (ICR) technology. In addition, to help ensure that data from the 2001 Census of Agriculture would be of consistently high quality, improved quality assurance and control procedures were incorporated into each of the collection and data processing stages. Primarily as a result of adopting these methods,
procedures and technologies, the 2001 Census of Agriculture data are of
very good quality, with the major commodities generally being of the highest
quality. A response rate of 98% and an estimated undercoverage rate of
farms of 5.6% also contributed to the overall success of the 2001 Census
of Agriculture. Note that, over half of the estimated undercoverage was
of farms with sales below $10,000 in 2000. As a result, the undercoverage
rate for major commodities is generally below With projects as large and complex as the Censuses of Agriculture and Population, the estimates produced from them are inevitably subject to a certain degree of error. Knowing the types of errors that can occur and how they affect specific variables can help users assess the data's usefulness for their particular applications as well as the risks involved in basing conclusions or decisions on them. Errors can arise at virtually every stage of the census process, from preparing materials, through collecting data, to processing. Moreover, errors may be more predominant in certain areas of the country or vary according to the characteristic being measured. Some errors occur at random, and when individual responses are aggregated for a sufficiently large group they tend to cancel each other out. For errors of this nature, the larger the group, the more accurate the corresponding estimate. For this reason, data users are advised to be cautious when using estimates based on a small number of responses. Some errors, however, might occur more systematically and result in "biased" estimates. Because the bias from such errors is persistent no matter how large the group for which responses are aggregated, and because bias is particularly difficult to measure, systematic errors are a more serious problem for most data users than random errors. The most common types of errors are described below. Coverage errors Non-response errors Response errors Processing errors
Comparability of data and related sourcesThe data validation process identified some instances in which data either were not directly comparable to those from previous censuses or were of reduced quality, primarily because of coverage or response errors. After thoroughly investigating each case, notes were developed to identify the variables affected and explain the situation associated with each. Following each Census of Agriculture, other agricultural surveys use Census of Agriculture data as a basis, or benchmark, for the production of regularly published estimates of the agriculture industry.
Other quality indicators and assessments
Coverage Evaluation SurveyThe purpose of the Coverage Evaluation Survey (CES) is to estimate the coverage of the 2001 Census of Agriculture that was conducted on May 15, 2001. Coverage is a problem that affects the quality of estimates of all censuses. For the Census of Agriculture, coverage errors occur when farms are missed, incorrectly included or double counted. The CES measures the level of coverage and is one way to assess the quality of the Census of Agriculture estimates. The CES selects a random sample of smaller
farm operations from Statistic Canada's Farm Register for which no Census
of Agriculture questionnaire was received. The survey uses a short questionnaire
to collect key information about the operating status and the size of
the farm. Please note that there are no estimates of undercoverage
for Yukon Territory, the Northwest Territories and Nunavut. Table 1. Farm undercoverage: breakdown by province
Table 2.Farm undercoverage: breakdown by total gross farm receipts
|