Weighting and estimation
Results
All (8)
All (8) ((8 results))
- Articles and reports: 12-001-X198600214445Description:
A problem of estimating monthly movements in rents based on data collected every four months is explored. Five alternative composite estimators of the rent index are presented and justified, both from an intuitive as well as theoretical point of view. An empirical study testing and comparing the proposed methods is described and summarized. Recommendations are put forth.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214446Description:
This paper discusses the influence of the sampling design on the estimation of a linear regression model. Particularly, sampling designs will be discussed which are dependent on the values of the endogenous variable in the population: endogenous (or “informative”) designs. A consistent estimator of the regression coefficients is given. Its variance is the sum of a sampling design component and a disturbance term component. Also, model-free regression is briefly discussed. The model-free regression estimator is the same as the model estimator in the case of an endogenous design.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214449Description:
Nearly all surveys and censuses are subject to two types of nonresponse: unit (total) and item (partial). Several methods of compensating for nonresponse have been developed in an attempt to reduce the bias associated with nonresponse. This paper summarizes the nonresponse adjustment procedures used at the U.S. Census Bureau, focusing on unit nonresponse. Some discussion of current and future research in this area is also included.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214451Description:
The Canadian Census of Construction (COC) uses a complex plan for sampling small businesses (those having a gross income of less than $750,000). Stratified samples are drawn from overlapping frames. Two subsamples are selected independently from one of the samples, and more detailed information is collected on the businesses in the subsamples. There are two possible methods of estimating totals for the variables collected in the subsamples. The first approach is to determine weights based on sampling rates. A number of different weights must be used. The second approach is to impute values to the businesses included in the sample but not in the subsamples. This approach creates a complete “rectangular” sample file, and a single weight may then be used to produce estimates for the population. This “large-scale imputation” technique is presently applied for the Census of Construction. The purpose of the study is to compare the figures obtained using various estimation techniques with the estimates produced by means of large-scale imputation.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214462Description:
In the presence of unit nonresponse, two types of variables can sometimes be observed for units in the “intended” sample s, namely, (a) variables used to estimate the response mechanism (the response probabilities), (b) variables (here called co-variates) that explain the variable of interest, in the usual regression theory sense. This paper, based on Särndal and Swensson (1985 a, b), discusses nonresponse adjusted estimators with and without explicit involvement of co-variates. We conclude that the presence of strong co-variates in an estimator induces several favourable properties. Among other things, estimators making use of co-variates are considerably more resistant to nonresponse bias. We discuss the calculation of standard error and valid confidence intervals for estimators involving co-variates. The structure of the standard error is examined and discussed.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214463Description:
The procedure of subsampling the nonrespondents suggested by Hansen and Hurwitz (1946) is considered. Post-stratification prior to the subsampling is examined. For the mean of a characteristic of interest, ratio estimators suitable for different practical situations are proposed and their merits are examined. Suitable ratio estimators are also suggested for the situations in which the Hard-Core are present.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600114441Description:
The analysis of survey data becomes difficult in the presence of incomplete responses. By the use of the maximum likelihood method, estimators for the parameters of interest and test statistics can be generated. In this paper the maximum likelihood estimators are given for the case where the data is considered missing at random. A method for imputing the missing values is considered along with the problem of estimating the change points in the mean. Possible extensions of the results to structured covariances and to non-randomly incomplete data are also proposed.
Release date: 1986-06-16 - Articles and reports: 12-001-X198600114443Description:
When the technique of adjustment using weighting classes is applied to compensate for the effect of non-response, several questions arise that call for precise and quantified answers: How does the choice of the variables used for definition of the classes affect total root-mean-square error, in particular non-response bias and sampling variance? What rule and what procedure should be followed in choosing the adjustment variables? On the basis of what criterion can the optimal sizes for the weighting classes be established? Finally, when this procedure is applied to compensate for non-response with respect to specific elements of a questionnaire, how can strongly correlated ancillary variables be used effectively when they themselves are affected by non-response? This article is addressed to those professionals working at a practical level who are seeking guidelines.
Release date: 1986-06-16
Data (0)
Data (0) (0 results)
No content available at this time.
Analysis (8)
Analysis (8) ((8 results))
- Articles and reports: 12-001-X198600214445Description:
A problem of estimating monthly movements in rents based on data collected every four months is explored. Five alternative composite estimators of the rent index are presented and justified, both from an intuitive as well as theoretical point of view. An empirical study testing and comparing the proposed methods is described and summarized. Recommendations are put forth.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214446Description:
This paper discusses the influence of the sampling design on the estimation of a linear regression model. Particularly, sampling designs will be discussed which are dependent on the values of the endogenous variable in the population: endogenous (or “informative”) designs. A consistent estimator of the regression coefficients is given. Its variance is the sum of a sampling design component and a disturbance term component. Also, model-free regression is briefly discussed. The model-free regression estimator is the same as the model estimator in the case of an endogenous design.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214449Description:
Nearly all surveys and censuses are subject to two types of nonresponse: unit (total) and item (partial). Several methods of compensating for nonresponse have been developed in an attempt to reduce the bias associated with nonresponse. This paper summarizes the nonresponse adjustment procedures used at the U.S. Census Bureau, focusing on unit nonresponse. Some discussion of current and future research in this area is also included.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214451Description:
The Canadian Census of Construction (COC) uses a complex plan for sampling small businesses (those having a gross income of less than $750,000). Stratified samples are drawn from overlapping frames. Two subsamples are selected independently from one of the samples, and more detailed information is collected on the businesses in the subsamples. There are two possible methods of estimating totals for the variables collected in the subsamples. The first approach is to determine weights based on sampling rates. A number of different weights must be used. The second approach is to impute values to the businesses included in the sample but not in the subsamples. This approach creates a complete “rectangular” sample file, and a single weight may then be used to produce estimates for the population. This “large-scale imputation” technique is presently applied for the Census of Construction. The purpose of the study is to compare the figures obtained using various estimation techniques with the estimates produced by means of large-scale imputation.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214462Description:
In the presence of unit nonresponse, two types of variables can sometimes be observed for units in the “intended” sample s, namely, (a) variables used to estimate the response mechanism (the response probabilities), (b) variables (here called co-variates) that explain the variable of interest, in the usual regression theory sense. This paper, based on Särndal and Swensson (1985 a, b), discusses nonresponse adjusted estimators with and without explicit involvement of co-variates. We conclude that the presence of strong co-variates in an estimator induces several favourable properties. Among other things, estimators making use of co-variates are considerably more resistant to nonresponse bias. We discuss the calculation of standard error and valid confidence intervals for estimators involving co-variates. The structure of the standard error is examined and discussed.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600214463Description:
The procedure of subsampling the nonrespondents suggested by Hansen and Hurwitz (1946) is considered. Post-stratification prior to the subsampling is examined. For the mean of a characteristic of interest, ratio estimators suitable for different practical situations are proposed and their merits are examined. Suitable ratio estimators are also suggested for the situations in which the Hard-Core are present.
Release date: 1986-12-15 - Articles and reports: 12-001-X198600114441Description:
The analysis of survey data becomes difficult in the presence of incomplete responses. By the use of the maximum likelihood method, estimators for the parameters of interest and test statistics can be generated. In this paper the maximum likelihood estimators are given for the case where the data is considered missing at random. A method for imputing the missing values is considered along with the problem of estimating the change points in the mean. Possible extensions of the results to structured covariances and to non-randomly incomplete data are also proposed.
Release date: 1986-06-16 - Articles and reports: 12-001-X198600114443Description:
When the technique of adjustment using weighting classes is applied to compensate for the effect of non-response, several questions arise that call for precise and quantified answers: How does the choice of the variables used for definition of the classes affect total root-mean-square error, in particular non-response bias and sampling variance? What rule and what procedure should be followed in choosing the adjustment variables? On the basis of what criterion can the optimal sizes for the weighting classes be established? Finally, when this procedure is applied to compensate for non-response with respect to specific elements of a questionnaire, how can strongly correlated ancillary variables be used effectively when they themselves are affected by non-response? This article is addressed to those professionals working at a practical level who are seeking guidelines.
Release date: 1986-06-16
Reference (0)
Reference (0) (0 results)
No content available at this time.
- Date modified: