Results
All (10)
All (10) ((10 results))
- Articles and reports: 12-001-X198100214319Description:
The problems associated with making analytical inferences from data based on complex sample designs are reviewed. A basic issue is the definition of the parameter of interest and whether it is a superpopulation model parameter or a finite population parameter. General methods based on a generalized Wald Statistics and its modification or on modifications of classical test statistics are discussed. More detail is given on specific methods-on linear models and regression and on categorical data analysis.
Release date: 1981-12-15 - 2. The nonresponse problem ArchivedArticles and reports: 12-001-X198100214320Description:
This paper presents an outline of the nonresponse research which is carried out at the Netherlands Central Bureau of Statistics. The phenomenon of nonresponse is put into a general frame-work. The extent of nonresponse is indicated with figures from a number of CBS-surveys. The use of auxiliary variables is discussed as a means for obtaining information about nonrespondents. These variables can be used either to characterize nonrespondents or as stratification variables in adjustment procedures.
Adjustment for nonresponse bias by means of subgroup weighting is considered in more detail. Finally, the last section lists a number of other methods which also aim at reduction of the bias.
Release date: 1981-12-15 - Articles and reports: 12-001-X198100214321Description:
The problem of specifying and estimating the variance of estimated parameters based on complex sample designs from finite populations is considered. The results of this paper are particularly useful when the parameter estimators cannot be defined explicitly as a function of other statistics from the sample. It is shown how these results can be applied to linear regression, logistic regression and log linear contingency table models.
Release date: 1981-12-15 - Articles and reports: 12-001-X198100214325Description:
The author briefly reviews the factors determining the production of health statistics in Canada, with particular attention to the different sources of data and to the long-standing co-operation among the many agencies involved in the gathering of health-related information.
Release date: 1981-12-15 - 5. Models for estimation of sampling errors ArchivedArticles and reports: 12-001-X198100214326Description:
This paper presents results of an empirical study on fitting log-linear models to data on estimates of characteristics and their coefficients of variation (CV) from the Canadian Labour Force Survey. The characteristics were classified into groups on the basis of design effects and models were fitted to data on estimates of characteristic totals and their CVs over twelve month period. The models can be used in situations where estimates of CV are needed for new characteristics, and for providing more precise estimates of reliability of estimates based on past data. The problem of evaluation of fit of the models is considered.
Release date: 1981-12-15 - 6. For-hire trucking survey: Survey Design ArchivedArticles and reports: 12-001-X198100154932Description:
The methodology of the For-hire Trucking Survey is discussed in this paper. This survey provides good examples of administrative and operational constraints faced by survey statisticians and field data collection teams.
Release date: 1981-06-15 - 7. Survey maintenance - Philosophy and practice ArchivedArticles and reports: 12-001-X198100154933Description:
An aspect of surveys not always given adequate consideration is maintenance.
The scope and importance of survey maintenance are discussed and a case is made for a more scientific methodological approach. Practical applications to various stages of surveys are illustrated by examples from the Canadian Labour Force Survey.
Release date: 1981-06-15 - 8. Imputation in surveys: Coping with reality ArchivedArticles and reports: 12-001-X198100154934Description:
In surveys a response may be incomplete or some items may be inconsistent or, as in the case of two-phase sampling, items may be unavailable. In these cases it may be expedient to impute values for the missing items. While imputation is not a particularly good solution to any specific estimation problem, it does permit the production of arbitrary estimates in a consistent way.
The survey statistician may have to cope with a mixture of numerical and categorical items, subject to a variety of constraints. He should evaluate his technique, especially with respect to bias. He should make sure that imputed items are clearly identified and summary reports produced.
A variety of imputation techniques in current use is described and discussed, with particular reference to the practical problems involved.
Release date: 1981-06-15 - 9. Construction of working probabilities and joint selection probabilities for Fellegi’s PPS sampling scheme ArchivedArticles and reports: 12-001-X198100154935Description:
A FORTRAN Subroutine to obtain the “working probabilities” for Fellegi’s (1963) method of unequal probability sampling is given. The solution is obtained by an iterative procedure where the starting values for the (k+l)th draw “working probabilities” are the solutions for the kth draw “working probabilities” and the iterative procedure is terminated when a prespecified accuracy is achieved. The limitation is that the Subroutine can only be used to obtain up to and including the 5th draw “working probabilities”. It was observed that the convergence occurs very fast in double precision. Therefore all real variables have been declared as double precision. The joint selection probabilities \Pi_{ij}’s i.e. the probability that both the ith and jth units are in the sample are obtained by summing the probabilities of selecting those samples that contain both the ith and jth units. The joint selection probabilities are required for the variance estimation of the Horvitz-Thompson estimator of population total of the characteristic of interest.
Release date: 1981-06-15 - Articles and reports: 12-001-X198100154836Description:
Survey organizations undertake periodic redesigns of continuous surveys. Reasons for such redesigns related to changes in information needs to be satisfied by the survey and changes in public awareness and attitudes towards surveys are discussed in the context of the redesign of the Canadian Labour Force Survey following the 1981 Census. In particular, the importance of close dialogue between users of the survey data and design statisticians at the early stages of the redesign process in order to establish survey objectives is stressed.
Release date: 1981-06-15
Stats in brief (0)
Stats in brief (0) (0 results)
No content available at this time.
Articles and reports (10)
Articles and reports (10) ((10 results))
- Articles and reports: 12-001-X198100214319Description:
The problems associated with making analytical inferences from data based on complex sample designs are reviewed. A basic issue is the definition of the parameter of interest and whether it is a superpopulation model parameter or a finite population parameter. General methods based on a generalized Wald Statistics and its modification or on modifications of classical test statistics are discussed. More detail is given on specific methods-on linear models and regression and on categorical data analysis.
Release date: 1981-12-15 - 2. The nonresponse problem ArchivedArticles and reports: 12-001-X198100214320Description:
This paper presents an outline of the nonresponse research which is carried out at the Netherlands Central Bureau of Statistics. The phenomenon of nonresponse is put into a general frame-work. The extent of nonresponse is indicated with figures from a number of CBS-surveys. The use of auxiliary variables is discussed as a means for obtaining information about nonrespondents. These variables can be used either to characterize nonrespondents or as stratification variables in adjustment procedures.
Adjustment for nonresponse bias by means of subgroup weighting is considered in more detail. Finally, the last section lists a number of other methods which also aim at reduction of the bias.
Release date: 1981-12-15 - Articles and reports: 12-001-X198100214321Description:
The problem of specifying and estimating the variance of estimated parameters based on complex sample designs from finite populations is considered. The results of this paper are particularly useful when the parameter estimators cannot be defined explicitly as a function of other statistics from the sample. It is shown how these results can be applied to linear regression, logistic regression and log linear contingency table models.
Release date: 1981-12-15 - Articles and reports: 12-001-X198100214325Description:
The author briefly reviews the factors determining the production of health statistics in Canada, with particular attention to the different sources of data and to the long-standing co-operation among the many agencies involved in the gathering of health-related information.
Release date: 1981-12-15 - 5. Models for estimation of sampling errors ArchivedArticles and reports: 12-001-X198100214326Description:
This paper presents results of an empirical study on fitting log-linear models to data on estimates of characteristics and their coefficients of variation (CV) from the Canadian Labour Force Survey. The characteristics were classified into groups on the basis of design effects and models were fitted to data on estimates of characteristic totals and their CVs over twelve month period. The models can be used in situations where estimates of CV are needed for new characteristics, and for providing more precise estimates of reliability of estimates based on past data. The problem of evaluation of fit of the models is considered.
Release date: 1981-12-15 - 6. For-hire trucking survey: Survey Design ArchivedArticles and reports: 12-001-X198100154932Description:
The methodology of the For-hire Trucking Survey is discussed in this paper. This survey provides good examples of administrative and operational constraints faced by survey statisticians and field data collection teams.
Release date: 1981-06-15 - 7. Survey maintenance - Philosophy and practice ArchivedArticles and reports: 12-001-X198100154933Description:
An aspect of surveys not always given adequate consideration is maintenance.
The scope and importance of survey maintenance are discussed and a case is made for a more scientific methodological approach. Practical applications to various stages of surveys are illustrated by examples from the Canadian Labour Force Survey.
Release date: 1981-06-15 - 8. Imputation in surveys: Coping with reality ArchivedArticles and reports: 12-001-X198100154934Description:
In surveys a response may be incomplete or some items may be inconsistent or, as in the case of two-phase sampling, items may be unavailable. In these cases it may be expedient to impute values for the missing items. While imputation is not a particularly good solution to any specific estimation problem, it does permit the production of arbitrary estimates in a consistent way.
The survey statistician may have to cope with a mixture of numerical and categorical items, subject to a variety of constraints. He should evaluate his technique, especially with respect to bias. He should make sure that imputed items are clearly identified and summary reports produced.
A variety of imputation techniques in current use is described and discussed, with particular reference to the practical problems involved.
Release date: 1981-06-15 - 9. Construction of working probabilities and joint selection probabilities for Fellegi’s PPS sampling scheme ArchivedArticles and reports: 12-001-X198100154935Description:
A FORTRAN Subroutine to obtain the “working probabilities” for Fellegi’s (1963) method of unequal probability sampling is given. The solution is obtained by an iterative procedure where the starting values for the (k+l)th draw “working probabilities” are the solutions for the kth draw “working probabilities” and the iterative procedure is terminated when a prespecified accuracy is achieved. The limitation is that the Subroutine can only be used to obtain up to and including the 5th draw “working probabilities”. It was observed that the convergence occurs very fast in double precision. Therefore all real variables have been declared as double precision. The joint selection probabilities \Pi_{ij}’s i.e. the probability that both the ith and jth units are in the sample are obtained by summing the probabilities of selecting those samples that contain both the ith and jth units. The joint selection probabilities are required for the variance estimation of the Horvitz-Thompson estimator of population total of the characteristic of interest.
Release date: 1981-06-15 - Articles and reports: 12-001-X198100154836Description:
Survey organizations undertake periodic redesigns of continuous surveys. Reasons for such redesigns related to changes in information needs to be satisfied by the survey and changes in public awareness and attitudes towards surveys are discussed in the context of the redesign of the Canadian Labour Force Survey following the 1981 Census. In particular, the importance of close dialogue between users of the survey data and design statisticians at the early stages of the redesign process in order to establish survey objectives is stressed.
Release date: 1981-06-15
Journals and periodicals (0)
Journals and periodicals (0) (0 results)
No content available at this time.
- Date modified: