Keyword search

Results

All (13)

All (13) (0 to 10 of 13 results)

1. Rotation group biases in the old and new Labour Force Survey Archived
Articles and reports: 12-001-X197600200001
Description: This paper presents results on rotation group biases in the Canadian Labour Force Survey (LFS). The biases are studied in detail by decomposition into components responsible for the biases. Also, a comparison between the old and the new LFS is done on the basis of 1975 parallel run and differences are analyzed. Some conclusions are drawn and recommendations for other studies presented.
Release date: 1976-12-13
2. Stratification index: Methodology and analysis Archived
Articles and reports: 12-001-X197600200002
Description: To obtain estimates of means or totals for a universe, a sample of units is often drawn to represent the universe and these units are then surveyed. One of the most important procedures used in the selection of the units is that of stratification, whereby the universe is split up into strata and independent samples of units are drawn from each stratum. A stratification index is developed to indicate the approximate fractional reduction in the sampling variance from that which would result if no stratification were undertaken. Also the methodology is extended to examine the effect of stratification on the sampling variance at different levels of stratification through the concept of a summary index. The stratification index is also extended to the case of ratio estimates using independent source data to re-weight the sample data. The index has been applied to the Canadian Labour Force Survey (LFS), a typical multi-stage stratified sample where ratio estimation, using projected age-sex population estimates is applied and empirical data are presented and analyzed.
Release date: 1976-12-13
3. 1974 Survey of Housing Units Archived
Articles and reports: 12-001-X197600200003
Description: The 1974 Survey of Housing Units was carried out by Statistics Canada on behalf of the Central Mortgage and Housing Corporation during the autumn of 1974. Statistics Canada's responsibilities on this project included the design and implementation of all phases of the survey up to and including the production of "clean" micro data tapes. The sponsoring department was in turn responsible for the specification of objectives and data requirements and for the analysis of the resulting data.
This report, which is a modification of the summary report produced by the project team at the conclusion of the project, provides a general description of the survey and the work done by Statistics Canada on the survey.
Release date: 1976-12-13
4. The estimation of total variance in the 1976 Census Archived
Articles and reports: 12-001-X197600200004
Description: Published reports for the 1976 Census will include estimates of Total Variance as indicators of the reliability of the figures in these reports. In order to obtain these estimates of Total Variance, an Interpenetrating Design Experiment was incorporated into the collection methods for a sample of enumeration areas. In this paper we derive the formula for Total Variance in terms of variances due to sampling, correlated response and simple response. We then show how the Total Variance, and its components, can be estimated from the design and we give the estimators that will be used for the 1976 Census. The estimates of sampling and correlated response variance are unbiased but the simple response variance estimate is not.
Release date: 1976-12-13
5. The methodology of the 1971 Reverse Record Check Archived
Articles and reports: 12-001-X197600200005
Description: The 1971 Reverse Record Check is one of the most important studies that were carried out as part of the 1971 Census Evaluation Programme. Its main purpose was to investigate the incidence of under-enumeration in the 1971 Census. To do this, a frame containing all persons who should be enumerated in the 1971 Census was built up from the 1966 Census returns, plus birth and immigrant registrations. A random sample was selected from the frame and each selected person was traced to his current Census address. Current Census returns were then checked to see whether or not the selected person was enumerated. Sample figures were weighted up to the population level to obtain estimates of undercoverage. This paper gives a general description of the methodology of this study, and indicates some of the resulting improvements incorporated for 1976.
Release date: 1976-12-13
6. An approximation to the inverse moments of the hypergeometric distribution Archived
Articles and reports: 12-001-X197600200006
Description: The negative moments of the positive hypergeometric distribution are often approximated by the inverse of the positive moments of this distribution. In this paper, a suitable approximation to the positive hypergeometric distribution is used to obtain the negative moments.
Release date: 1976-12-13
7. Double frame Ontario Pilot Hog Surveys Archived
Articles and reports: 12-001-X197600200007
Description: Three Ontario pilot hog surveys were conducted in 1975 to test a sampling method based on the simultaneous use of two list frames. This paper describes the different aspects of the experience. Particular emphasis is given to the double frame methodology such as discussed by Hartley [11]. Optimal allocation of the sample between frames is considered, with revision for each following survey based on all the accumulated results.
Release date: 1976-12-13
8. Methodology and analysis of the Pilot Auto Exit Survey, 1974 Archived
Articles and reports: 12-001-X197600100001
Description: The 1974 Pilot Auto Exit Survey tested three variants of a handout, mailback questionnaire for U.S. visitors leaving Canada, in a search for a reliable low-cost method of data collection. The sample design was based on a personal interview survey done in Ontario in 1973/74. This design, results of the Pilot, comparison with the Ontario results and some conclusions are presented in this paper.
Release date: 1976-06-14
9. Typical survey data: Estimation and imputation Archived
Articles and reports: 12-001-X197600100002
Description: A special class of missing data problems is discussed, namely that of typical survey data whereby zeros dominate the multivariate response space. Here, techniques which impute means (whether conditional or unconditional) distort rather than improve the quality of the data. A probabilistic model is described which provides reasonable estimates, but also upholds the integrity of the data base. Results are given from a comparative study of the proposed methodology with other estimation/imputation models.
Release date: 1976-06-14
10. Raking ratio estimators Archived
Articles and reports: 12-001-X197600100003
Description: This paper presents large sample results for the bias and variance of raking-ratio estimators for up to four iterations. Estimators of the bias and variance are also presented. An expression for the asymptotic covariance matrix of the maximum likelihood estimators of the cell proportions in a two-way table with known marginals is also given.
Release date: 1976-06-14

Data (0)

Data (0) (0 results)

No content available at this time.

Analysis (13)

Analysis (13) (10 to 20 of 13 results)

11. Methodology of the Labour Force Survey re-interview program Archived
Articles and reports: 12-001-X197600100004
Description: With the recent review of the Labour Force Survey, several periphexal projects have been redesigned. This is the case with the LFS re-interview program which will for the coming years be oriented toward the measurement of response errors. This paper describes the new design of the program and discusses how data will be analysed to achieve the objectives.
Release date: 1976-06-14
12. The multivariate Behrens-Fisher problem with application to the Labour Force Survey Archived
Articles and reports: 12-001-X197600100005
Description: This paper presents the Behrens-Fisher problem and gives an overview of the major solutions brought forward to this date. The aim of the paper is to use the most appropriate approach to the problem for testing sets of six month Labour Force Survey data against those of a pilot study. This is done since in many cases (such as Methods Test Panel studies) studies are conducted for six consecutive months and comparisons are required on the basis of those sets of six month data. Empirical results are also given by testing Methods Test Panel Phase III data against corresponding Labour Force Survey data.
Release date: 1976-06-14
13. Analysis of design effects and variance components in multi-stage sample surveys Archived
Articles and reports: 12-001-X197600100006
Description: Multi-stage statistical surveys as a means of obtaining socioeconomic characteristics for the population have been in use for many years. Each survey requires an extensive and precise sample design which is governed by the cost structure for obtaining the data and the variance of the characteristic data between units at various stages of sampling. The authors analyzed variance components derived from one month's data of the Canadian Labour Force Survey and examined the variance that would have resulted under different allocation strategies in Table 6 and for different average sizes of units in Table 7. The percentage components of variance, the design effects by stage of sampling and population variances between units of the various stages, as well as measures of homogeneity for households within stages, are derived and shown in Tables 2 to 5.
The analysis was carried out for the Canadian Labour Force Survey, but the methodology of component of variance estimation (Gray [4]) and the methods used to analyze the results of a particular survey are readily applied to any multi-stage statistical sample survey, where Horvitz-Thompsen estimators and ratio estimation are applied.
Release date: 1976-06-14

Reference (0)

Reference (0) (0 results)

No content available at this time.

Date modified:: 2026-04-26