Introduction
Statistics Canada’s Office Business Architecture (BOA) initiative implements measures to reduce operating costs and improve the quality assurance and delivery of new statistical programs. The proposed development of the Integrated Business Statistics Program (IBSP) is one way for Statistics Canada’s business statistics programs to achieve these goals. The PISE aims to transform the existing Unified Enterprise Survey (UES) platform into a generalized AOB -compliant modelto produce business statistics. This model will encompass all stages of a survey, from the sampling frame to data dissemination. The initial objective was to apply the PISE model to all business-related programs except the awards program and the international trade program and canadian phone number format. The PISE proposal is based on six pillars that will give rise to the expected efficiency gains, namely:
the use of tax data for estimating financial data;
improving governance in all sectors involved in the production of statistical data, in particular strengthening change management.
The foundations of the PISE are described in detail in a document from the Business Statistics Division (ESD, 2009). According to this document, the PISE must be sufficiently unified, harmonized and flexible to:
integrating new surveys;
This document summarizes the sampling methodology adopted for the PISE . It contains a detailed description of the main sampling steps and a discussion of the methodology used.
Two-phase sampling
In the case of certain surveys that will be integrated into the PISE , there are situations where the information available in the BR is not up to date or does not meet the needs of the survey. For example, the industry classification in the BR may not always be up-to-date for all surveyed units or additional information, such as the goods or services produced by businesses, may be needed to make the sampling process more efficient.
In order to deal with these situations, the PISEoffers survey programs the option of using a two-phase sampling strategy. Surveys that opt for two-phase sampling are encouraged to join the first main phase, which is called the Survey of Economic Activity, Expenditure and Products (EAEDP). Some surveys for which a two-phase sampling plan is not suitable will have the option of using a single-phase sampling strategy. The objectives of any second-phase surveys must be known to ensure that the first-phase sample (on which second-phase surveys depend) will be large enough to meet the requirements of these surveys. The size of the
Sampling frame
With respect to the information contained in the sampling frame, one of the fundamental assumptions of the PISE maintains that all the necessary information from the BR or from survey data is accessible and up to date at the time of sampling. first phase. No updating of the information contained in the sampling frame is permitted during the sampling process.
Use of additional information
To create the sampling frame of a survey, the use of information from previous survey cycles or that from the first phase of the survey can be useful. This applies, for example, to second-phase surveys that use the EAEDP as the first phase. When creating the sampling frame for these second-phase surveys, the sampling units that should be included will be those selected from the EAEDP sample .
In addition, it may be necessary to have access to sample files from other surveys to manage the coordination of samples between surveys and thus control response burden. It could also be useful to have access to sample files from previous editions of the same survey to manage sample rotation between editions of this survey. In addition, for all second-phase surveys, information from the first-phase survey will be needed to define the sampling frame.
Hierarchy and size of units
In the BR , the business structure includes four levels of statistical entity, namely enterprise, company, establishment and location. For each survey, it is necessary to define the level at which the proportions of the total size that come from the unit (contributions) will be considered at the time of sampling (level of the targeted operating entity), as well as the level of the unit that will be selected at the time of sampling (sampling unit level) and at which the estimate will be made.
Targeted operating entities and sampling units
The targeted operating entity and sampling unit levels are indicated in the metadata and can correspond to any statistical entity level in the BR . The only constraint is that the level of the sampling unit cannot be lower in the hierarchy than the level of the targeted operating entity.
Size measures of operating entities targeted for sampling
For targeted operating entities, the default size metric is based on revenue. Size concepts other than income can also be used and come directly from the BR or be derived based on variables present in the BR . The user could also implement an income-based size function different from the default function. It is also possible to use a combination of different size concepts, such as income and assets, income and capacity,
