Keyword search

This document presents the geographical distribution of federal government expenditures and staff for the fiscal year 2004/2005.

We discuss methods for the analysis of case-control studies in which the controls are drawn using a complex sample survey. The most straightforward method is the standard survey approach based on weighted versions of population estimating equations. We also look at more efficient methods and compare their robustness to model mis-specification in simple cases. Case-control family studies, where the within-cluster structure is of interest in its own right, are also discussed briefly.

Calibration weighting can be used to adjust for unit nonresponse and/or coverage errors under appropriate quasi-randomization models. Alternative calibration adjustments that are asymptotically identical in a purely sampling context can diverge when used in this manner. Introducing instrumental variables into calibration weighting makes it possible for nonresponse (say) to be a function of a set of characteristics other than those in the calibration vector. When the calibration adjustment has a nonlinear form, a variant of the jackknife can remove the need for iteration in variance estimation.

The theory of multiple imputation for missing data requires that imputations be made conditional on the sampling design. However, most standard software packages for performing model-based multiple imputation assume simple random samples, leading many practitioners not to account for complex sample design features, such as stratification and clustering, in their imputations. Theory predicts that analyses of such multiply-imputed data sets can yield biased estimates from the design-based perspective. In this article, we illustrate through simulation that (i) the bias can be severe when the design features are related to the survey variables of interest, and (ii) the bias can be reduced by controlling for the design features in the imputation models. The simulations also illustrate that conditioning on irrelevant design features in the imputation models can yield conservative inferences, provided that the models include other relevant predictors. These results suggest a prescription for imputers: the safest course of action is to include design variables in the specification of imputation models. Using real data, we demonstrate a simple approach for incorporating complex design features that can be used with some of the standard software packages for creating multiple imputations.

In this article, we propose a Bernoulli-type bootstrap method that can easily handle multi-stage stratified designs where sampling fractions are large, provided simple random sampling without replacement is used at each stage. The method provides a set of replicate weights which yield consistent variance estimates for both smooth and non-smooth estimators. The method's strength is in its simplicity. It can easily be extended to any number of stages without much complication. The main idea is to either keep or replace a sampling unit at each stage with preassigned probabilities, to construct the bootstrap sample. A limited simulation study is presented to evaluate performance and, as an illustration, we apply the method to the 1997 Japanese National Survey of Prices.

In this paper, the geometric, optimization-based, and Lavallée and Hidiroglou (LH) approaches to stratification are compared. The geometric stratification method is an approximation, whereas the other two approaches, which employ numerical methods to perform stratification, may be seen as optimal stratification methods. The algorithm of the geometric stratification is very simple compared to the two other approaches, but it does not take into account the construction of a take-all stratum, which is usually constructed when a positively skewed population is stratified. In the optimization-based stratification, one may consider any form of optimization function and its constraints. In a comparative numerical study based on five positively skewed artificial populations, the optimization approach was more efficient in each of the cases studied compared to the geometric stratification. In addition, the geometric and optimization approaches are compared with the LH algorithm. In this comparison, the geometric stratification approach was found to be less efficient than the LH algorithm, whereas efficiency of the optimization approach was similar to the efficiency of the LH algorithm. Nevertheless, strata boundaries evaluated via the geometric stratification may be seen as efficient starting points for the optimization approach.

To select a survey sample, it happens that one does not have a frame containing the desired collection units, but rather another frame of units linked in a certain way to the list of collection units. It can then be considered to select a sample from the available frame in order to produce an estimate for the desired target population by using the links existing between the two. This can be designated by Indirect Sampling.

Estimation for the target population surveyed by Indirect Sampling can constitute a big challenge, in particular if the links between the units of the two are not one-to-one. The problem comes especially from the difficulty to associate a selection probability, or an estimation weight, to the surveyed units of the target population. In order to solve this type of estimation problem, the Generalized Weight Share Method (GWSM) has been developed by Lavallée (1995) and Lavallée (2002). The GWSM provides an estimation weight for every surveyed unit of the target population.

This paper first describes Indirect Sampling, which constitutes the foundations of the GWSM. Second, an overview of the GWSM is given where we formulate the GWSM in a theoretical framework using matrix notation. Third, we present some properties of the GWSM such as unbiasedness and transitivity. Fourth, we consider the special case where the links between the two populations are expressed by indicator variables. Fifth, some special typical linkages are studied to assess their impact on the GWSM. Finally, we consider the problem of optimality. We obtain optimal weights in a weak sense (for specific values of the variable of interest), and conditions for which these weights are also optimal in a strong sense and independent of the variable of interest.

A survey of tourist visits originating intra and extra-region in Brittany was needed. For concrete material reasons, "border surveys" could no longer be used. The major problem is the lack of a sampling frame that allows for direct contact with tourists. This problem was addressed by applying the indirect sampling method, the weighting for which is obtained using the generalized weight share method developed recently by Lavallée (1995), Lavallée (2002), Deville (1999) and also presented recently in Lavallée and Caron (2001). This article shows how to adapt the method to the survey. A number of extensions are required. One of the extensions, designed to estimate the total of a population from which a Bernouilli sample has been taken, will be developed.

Félix-Medina and Thompson (2004) proposed a variant of Link-tracing sampling in which it is assumed that a portion of the population, not necessarily the major portion, is covered by a frame of disjoint sites where members of the population can be found with high probabilities. A sample of sites is selected and the people in each of the selected sites are asked to nominate other members of the population. They proposed maximum likelihood estimators of the population sizes which perform acceptably provided that for each site the probability that a member is nominated by that site, called the nomination probability, is not small. In this research we consider Félix-Medina and Thompson's variant and propose three sets of estimators of the population sizes derived under the Bayesian approach. Two of the sets of estimators were obtained using improper prior distributions of the population sizes, and the other using Poisson prior distributions. However, we use the Bayesian approach only to assist us in the construction of estimators, while inferences about the population sizes are made under the frequentist approach. We propose two types of partly design-based variance estimators and confidence intervals. One of them is obtained using a bootstrap and the other using the delta method along with the assumption of asymptotic normality. The results of a simulation study indicate that (i) when the nomination probabilities are not small each of the proposed sets of estimators performs well and very similarly to maximum likelihood estimators; (ii) when the nomination probabilities are small the set of estimators derived using Poisson prior distributions still performs acceptably and does not have the problems of bias that maximum likelihood estimators have, and (iii) the previous results do not depend on the size of the fraction of the population covered by the frame.

Survey sampling to estimate a Consumer Price Index (CPI) is quite complicated, generally requiring a combination of data from at least two surveys: one giving prices, one giving expenditure weights. Fundamentally different approaches to the sampling process - probability sampling and purposive sampling - have each been strongly advocated and are used by different countries in the collection of price data. By constructing a small "world" of purchases and prices from scanner data on cereal and then simulating various sampling and estimation techniques, we compare the results of two design and estimation approaches: the probability approach of the United States and the purposive approach of the United Kingdom. For the same amount of information collected, but given the use of different estimators, the United Kingdom's methods appear to offer better overall accuracy in targeting a population superlative consumer price index.

To provide users with a complete picture of the activities associated with the Couriers and Local Messengers industry in Canada.

This bulletin presents the final set of tables which contain salary information for the year 2004-2005. This information is collected annually under the University and College Academic Staff System and has a reference date of October 1st. Therefore, the data reflect employment in universities as of that date. Each university must authorize Statistics Canada to release their information. However, information for institutions that have less than 100 full-time staff are not included.

The statistics presented in this bulletin are for the year ending on August 31 and for the period from 2002 to 2005. The following text contains references to previous periods when it is useful to set the industry's performance in a historical context.

To provide users with a complete picture of the financial and operational activities associated with Small For-hire Motor Carriers of Freight and Owner Operators in Canada.

Although revenues were higher in 2004 compared to 2003, the Canadian bus and urban transit industries had a difficult year in 2004. Their net income dropped nearly 17% as expenses rose faster than revenues.

This publication is a statistical summary of the dairy situation in Canada and the provinces, including farm sales of milk for fluid and manufacturing purposes; cash receipts from farm sales; and production and stocks of creamery butter, cheddar cheese and other dairy products and by-products.

The statistics presented in this Bulletin are for the fiscal year ending August 31 and cover the period from 2002 to 2005. The text includes references to earlier periods when it is useful to put the industry's recent performance in a historical context.

This report provides trends on public school enrolments, educators and expenditures. It uses figures provided by provincial and territorial departments of education on public elementary and secondary schools.

This publication presents a series of tabulations produced from the General Social Survey on time use of Canadians. It includes information on average amounts of time spent on various activities by sex, by age, by selected role groups.

This bulletin contains salary information of full-time teaching staff at Canadian universities for the academic year 2005/2006. Information is provided for institutions that have determined salaries for the period and have responded to the survey by June 2006. This information is collected annually under the University and College Academic Staff Survey and has a reference date of October 1st. Therefore, the data reflect employment in universities as of that date. Each university must authorize Statistics Canada to release their information. However, information for institutions that have less than 100 full-time staff (and who responded to the survey by June 2006) are not included in this bulletin. This information is available by special request to Client Services, Culture, Tourism and the Centre for Education Statistics (see Contact information, below).

This document presents the geographical distribution of federal government expenditures and staff for the fiscal year 2004/2005.

We discuss methods for the analysis of case-control studies in which the controls are drawn using a complex sample survey. The most straightforward method is the standard survey approach based on weighted versions of population estimating equations. We also look at more efficient methods and compare their robustness to model mis-specification in simple cases. Case-control family studies, where the within-cluster structure is of interest in its own right, are also discussed briefly.

Calibration weighting can be used to adjust for unit nonresponse and/or coverage errors under appropriate quasi-randomization models. Alternative calibration adjustments that are asymptotically identical in a purely sampling context can diverge when used in this manner. Introducing instrumental variables into calibration weighting makes it possible for nonresponse (say) to be a function of a set of characteristics other than those in the calibration vector. When the calibration adjustment has a nonlinear form, a variant of the jackknife can remove the need for iteration in variance estimation.

The theory of multiple imputation for missing data requires that imputations be made conditional on the sampling design. However, most standard software packages for performing model-based multiple imputation assume simple random samples, leading many practitioners not to account for complex sample design features, such as stratification and clustering, in their imputations. Theory predicts that analyses of such multiply-imputed data sets can yield biased estimates from the design-based perspective. In this article, we illustrate through simulation that (i) the bias can be severe when the design features are related to the survey variables of interest, and (ii) the bias can be reduced by controlling for the design features in the imputation models. The simulations also illustrate that conditioning on irrelevant design features in the imputation models can yield conservative inferences, provided that the models include other relevant predictors. These results suggest a prescription for imputers: the safest course of action is to include design variables in the specification of imputation models. Using real data, we demonstrate a simple approach for incorporating complex design features that can be used with some of the standard software packages for creating multiple imputations.

In this article, we propose a Bernoulli-type bootstrap method that can easily handle multi-stage stratified designs where sampling fractions are large, provided simple random sampling without replacement is used at each stage. The method provides a set of replicate weights which yield consistent variance estimates for both smooth and non-smooth estimators. The method's strength is in its simplicity. It can easily be extended to any number of stages without much complication. The main idea is to either keep or replace a sampling unit at each stage with preassigned probabilities, to construct the bootstrap sample. A limited simulation study is presented to evaluate performance and, as an illustration, we apply the method to the 1997 Japanese National Survey of Prices.

In this paper, the geometric, optimization-based, and Lavallée and Hidiroglou (LH) approaches to stratification are compared. The geometric stratification method is an approximation, whereas the other two approaches, which employ numerical methods to perform stratification, may be seen as optimal stratification methods. The algorithm of the geometric stratification is very simple compared to the two other approaches, but it does not take into account the construction of a take-all stratum, which is usually constructed when a positively skewed population is stratified. In the optimization-based stratification, one may consider any form of optimization function and its constraints. In a comparative numerical study based on five positively skewed artificial populations, the optimization approach was more efficient in each of the cases studied compared to the geometric stratification. In addition, the geometric and optimization approaches are compared with the LH algorithm. In this comparison, the geometric stratification approach was found to be less efficient than the LH algorithm, whereas efficiency of the optimization approach was similar to the efficiency of the LH algorithm. Nevertheless, strata boundaries evaluated via the geometric stratification may be seen as efficient starting points for the optimization approach.

To select a survey sample, it happens that one does not have a frame containing the desired collection units, but rather another frame of units linked in a certain way to the list of collection units. It can then be considered to select a sample from the available frame in order to produce an estimate for the desired target population by using the links existing between the two. This can be designated by Indirect Sampling.

Estimation for the target population surveyed by Indirect Sampling can constitute a big challenge, in particular if the links between the units of the two are not one-to-one. The problem comes especially from the difficulty to associate a selection probability, or an estimation weight, to the surveyed units of the target population. In order to solve this type of estimation problem, the Generalized Weight Share Method (GWSM) has been developed by Lavallée (1995) and Lavallée (2002). The GWSM provides an estimation weight for every surveyed unit of the target population.

This paper first describes Indirect Sampling, which constitutes the foundations of the GWSM. Second, an overview of the GWSM is given where we formulate the GWSM in a theoretical framework using matrix notation. Third, we present some properties of the GWSM such as unbiasedness and transitivity. Fourth, we consider the special case where the links between the two populations are expressed by indicator variables. Fifth, some special typical linkages are studied to assess their impact on the GWSM. Finally, we consider the problem of optimality. We obtain optimal weights in a weak sense (for specific values of the variable of interest), and conditions for which these weights are also optimal in a strong sense and independent of the variable of interest.

A survey of tourist visits originating intra and extra-region in Brittany was needed. For concrete material reasons, "border surveys" could no longer be used. The major problem is the lack of a sampling frame that allows for direct contact with tourists. This problem was addressed by applying the indirect sampling method, the weighting for which is obtained using the generalized weight share method developed recently by Lavallée (1995), Lavallée (2002), Deville (1999) and also presented recently in Lavallée and Caron (2001). This article shows how to adapt the method to the survey. A number of extensions are required. One of the extensions, designed to estimate the total of a population from which a Bernouilli sample has been taken, will be developed.

Félix-Medina and Thompson (2004) proposed a variant of Link-tracing sampling in which it is assumed that a portion of the population, not necessarily the major portion, is covered by a frame of disjoint sites where members of the population can be found with high probabilities. A sample of sites is selected and the people in each of the selected sites are asked to nominate other members of the population. They proposed maximum likelihood estimators of the population sizes which perform acceptably provided that for each site the probability that a member is nominated by that site, called the nomination probability, is not small. In this research we consider Félix-Medina and Thompson's variant and propose three sets of estimators of the population sizes derived under the Bayesian approach. Two of the sets of estimators were obtained using improper prior distributions of the population sizes, and the other using Poisson prior distributions. However, we use the Bayesian approach only to assist us in the construction of estimators, while inferences about the population sizes are made under the frequentist approach. We propose two types of partly design-based variance estimators and confidence intervals. One of them is obtained using a bootstrap and the other using the delta method along with the assumption of asymptotic normality. The results of a simulation study indicate that (i) when the nomination probabilities are not small each of the proposed sets of estimators performs well and very similarly to maximum likelihood estimators; (ii) when the nomination probabilities are small the set of estimators derived using Poisson prior distributions still performs acceptably and does not have the problems of bias that maximum likelihood estimators have, and (iii) the previous results do not depend on the size of the fraction of the population covered by the frame.

Survey sampling to estimate a Consumer Price Index (CPI) is quite complicated, generally requiring a combination of data from at least two surveys: one giving prices, one giving expenditure weights. Fundamentally different approaches to the sampling process - probability sampling and purposive sampling - have each been strongly advocated and are used by different countries in the collection of price data. By constructing a small "world" of purchases and prices from scanner data on cereal and then simulating various sampling and estimation techniques, we compare the results of two design and estimation approaches: the probability approach of the United States and the purposive approach of the United Kingdom. For the same amount of information collected, but given the use of different estimators, the United Kingdom's methods appear to offer better overall accuracy in targeting a population superlative consumer price index.

The Survey of Income and Labour Dynamics (SLID) interview is conducted using computer-assisted interviewing (CAI). CAI is paperless interviewing. This document is therefore a written approximation of the CAI interview, or the questionnaire.

In previous years, SLID conducted a Labour interview each January and a separate Income interview in May. In 2005 (reference year 2004) the two interviews were combined and collected in one interview in January.

A labour and income interview is collected for all respondents 16 years of age and over. Respondents have the option of answering income questions during the interview, or of giving Statistics Canada permission to use their income tax records.

In January 2005, data was collected for reference year 2004 from panels 3 and 4. Panel 3, in its sixth and final year, consisted of approximately 17,000 households and panel 4, in its third year, also consisted of approximately 17,000 households.

This document outlines the structure of the January 2005 Labour and Income interview (for the 2004 reference year) including question wording, possible responses, and flows of questions.

This document examines the functional limitations, physical, emotional and social, related to the musculoskeletal conditions having the largest impact on the health of Canadians. These functional limitations are described and classified using the Classification and Measurement System of Functional Health (CLAMES).

These descriptions and classifications are the first step in a new approach to measuring the health of Canadians that examines what factors are adversely affecting population health and how to address them. This document also provides health professionals, advocacy groups, and individual Canadians with an overview of how living with certain musculoskeletal conditions affects day-to-day functioning.

The National Accounts Advisory Committee reviews and gives advice on the concepts, methods, plans, standards as well as results associated with Statistics Canada's System of National Accounts.

This guide discusses the collection and interpretation of statistical data on Canada's trade in culture goods. This guide has been restructured and simplified to better meet the needs of data users. This version replaces Culture Goods Trade Estimates: Methodology and Technical Notes, Catalogue no. 81-595-MIE2004020.

In previous years, the Survey of Labour and Income Dynamics (SLID) conducted a Labour interview each January and a separate Income interview in May. In 2005 (reference year 2004) the two interviews were combined and collected in one interview in January.

The data are collected using computer-assisted interviewing. Thus there are no paper questionnaires required for data collection. The questions, responses and interview flow for Labour and Income are documented in other SLID research papers. This document presents the information for the 2005 Entry Exit portion of the Labour Income interview (for the 2004 reference year).

The Entry Exit Component consists of five separate modules. The Entry module is the first set of data collected. It is information collected to update the place of residence, housing conditions and expenses, as well as the household composition. For each person identified in Entry, the Demographics module collects (or updates) the person's name, date of birth, sex and marital status. Then the Relationships module identifies (or updates) the relationship between each respondent and every other household member. The Exit module includes questions on who to contact for the next interview and the names, phone numbers and addresses of two contacts to be used only if future tracing of respondents is required. An overview of the Tracing component is also included in this document.

Statistics Canada collects and publishes a large amount of data on all modes of transportation. For example, do you know the level of shipments of commodities last quarter? Where are the key access points to the United States and which commodities are moving through them? How can you determine market share? This guide will familiarize you with the sources for answers to these questions and more and show you how to access them. It will allow you to take advantage of what Statistics Canada has to offer you. The guide is divided into two parts. Part I contains a description of each survey at Statistics Canada that has transportation related information. Each survey is listed with the survey name, a person to contact, phone number and fax number, a brief description of the transportation related information in the survey, the periodicity of the survey and the publication catalogue number, name and price where the information can be found.

This paper introduces and explains modifications made to the Labour Force Survey estimates in January 2006. Some of these modifications include changes to the population estimates, improvements to the public and private sector estimates and historical updates to several small Census Agglomerations (CA).

The objective of this study was to estimate and map livestock manure production by the environmental geography unit known as the Sub-Sub-Drainage Area (SSDA) in 2001 and to calculate the change in manure production over the 1981-2001 period.