Statistics by subject – Statistical methods

Other available resources to support your research.

Help for sorting results
Browse our central repository of key standard concepts, definitions, data sources and methods.
Loading
Loading in progress, please wait...
All (18)

All (18) (18 of 18 results)

  • Articles and reports: 12-001-X199100214501
    Description:

    Although farm surveys carried out by the USDA are used to estimate crop production at the state and national levels, small area estimates at the county level are more useful for local economic decision making. County estimates are also in demand by companies selling fertilizers, pesticides, crop insurance, and farm equipment. Individual states often conduct their own surveys to provide data for county estimates of farm production. Typically, these state surveys are not carried out using probability sampling methods. An additional complication is that states impose the constraint that the sum of county estimates of crop production for all counties in a state be equal to the USDA estimate for that state. Thus, standard small area estimation procedures are not directly applicable to this problem. In this paper, we consider using regression models for obtaining county estimates of wheat production in Kansas. We describe a simulation study comparing the resulting estimates to those obtained using two standard small area estimators: the synthetic and direct estimators. We also compare several strategies for scaling the initial estimates so that they agree with the USDA estimate of the state production total.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214508
    Description:

    The usual dual system estimator for population size can be severely biased, if there is population heterogeneity in the capture probabilities. In this note we investigate the bias of the corresponding variance estimator under heterogeneity. We show that the usual estimator is conservative, i.e., it gives too large values, if the two registration systems are negatively correlated, uncorrelated, or when the correlation is positive, but small. In the case of high positive correlation the usual estimator may yield too low values. Two alternative estimators are proposed. One is conservative under arbitrary heterogeneity. The other is conservative under Gaussian heterogeneity. The methods are applied to occupational disease data from Finland.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214505
    Description:

    The X-11-ARIMA seasonal adjustment method and the Census X-11 variant use a standard ANOVA-F-test to assess the presence of stable seasonality. This F-test is applied to a series consisting of estimated seasonals plus irregulars (residuals) which may be (and often are) autocorrelated, thus violating the basic assumption of the F-test. This limitation has long been known by producers of seasonally adjusted data and the nominal value of the F statistic has been rarely used as a criterion for seasonal adjustment. Instead, producers of seasonally adjusted data have used rules of thumb, such as, F equal to or greater than 7. This paper introduces an exact test which takes into account autocorrelated residuals following an SMA process of the (0, q) (0, Q)_s type. Comparisons of this modified F-test and the standard ANOVA test of X-11-ARIMA are made for a large number of Canadian socio-economic series.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214502
    Description:

    A sample design for the initial selection, sample rotation and updating for sub-annual business surveys is proposed. The sample design is a stratified clustered design, with the stratification being carried out on the basis of industry, geography and size. Sample rotation of the sample units is carried out under time-in and time-out constraints. Updating is with respect to the selection of births (new businesses), removal of deaths (defunct businesses) and implementation of changes in the classification variables used for stratification, i.e. industry, geography and size. A number of alternate estimators, including the simple expansion estimator and Mickey’s (1959) unbiased ratio-type estimator have been evaluated for this design in an empirical study under various survey conditions. The problem of variance estimation has also been considered using the Taylor linearization method and the jackknife technique.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214504
    Description:

    Simple or marginal quota surveys are analyzed using two methods: (1) behaviour modelling (superpopulation model) and prediction estimation, and (2) sample modelling (simple restricted random sampling) and estimation derived from the sample distribution. In both cases the limitations of the theory used to establish the variance formulas and estimates when measuring totals are described. An extension of the quota method (non-proportional quotas) is also briefly described and analyzed. In some cases, this may provide a very significant improvement in survey precision. The advantages of the quota method are compared with those of random sampling. The latter remains indispensable in the case of large scale surveys within the framework of Official Statistics.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214503
    Description:

    Surveys are often conducted of flows of persons, such as: visitors to museums, libraries and parks; voters; shoppers; hospital outpatients; tourists; international travellers; and car occupants. The sample designs for such surveys usually involve sampling in time and space. Methods for sampling flows of human populations are reviewed and illustrated.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214506
    Description:

    This paper compares the magnitude and nature of attrition in two separate RDD panel surveys conducted in the City of Chicago (i.e. the surveys were independent studies and were not conducted as part of a planned experiment), each with a between-wave lag of approximately one year. For each survey, sampling at Wave 1 was performed via one-stage (i.e. simple) random-digit dialing. In Study 1, respondents’ names were not elicited; thus, when telephone calls were made at Wave 2 of Study 1 interviewers could not ask for respondents by name. Instead, interviewers asked for respondents by using a gender-age identifier. In Study 2, respondent name identifiers were gathered during Wave 1 and were used in Wave 2 re-contact attempts. The magnitude of the attrition in Study 1 (i.e. the proportion of Wave 1 respondents not re-interviewed at Wave 2) was 47%, whereas in Study 2 it was 43%: a marginal difference in attrition rates. In both surveys, age, race, education and income were significantly related to attrition. Discussion is presented on the trade-off between minimizing attrition vs. minimizing respondent reactivity as potential sources of total survey error. Suggestions for decreasing the size of attrition in RDD panel surveys are discussed.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214500
    Description:

    The Canadian General Social Survey is an annual survey that aims to provide data on the demographic and social characteristics of Canadians. This paper provides an overview of the program, based on the experience of the first five surveys. The objectives of the program, the methodology used, the themes and issues addressed, the program outputs and the plans for the future are all discussed.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214507
    Description:

    For estimating the proportion and total of an item for the present occasion, independent estimates at the current and previous occasions are combined through three different procedures. In the first one, trend over the occasions is utilized. For the second one, the One-Way Random Effects Model is employed. The third procedure uses the Empirical Bayes approach. All the three procedures are seen to perform better than the sample estimates obtained from the data of the current occasion alone. Advantages of these methods and their limitations are discussed. All the procedures are illustrated with the data from the National Health Discharge Survey.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100114525
    Description:

    Electronic data collection utilizing touchtone recognition is in place for a monthly establishment survey at the Bureau of Labor Statistics. The Touchtone Data Entry (TDE) system features digitized phrases requesting respondents to answer questions using the numeric keypad of a touchtone telephone. TDE has substantial implications for lowering survey costs; many labor intensive activities are eliminated. However, little is known about measurement errors associated with this mode of data collection. This study assesses TDE mode error using three sources of data, which allow for analyses of errors associated with selected aspects of the human-machine interface. In addition, instrument design issues associated with mode error are addressed. We conclude by extending the implications of our findings to other surveys.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114521
    Description:

    Marginal and approximate conditional likelihoods are given for the correlation parameters in a normal linear regression model with correlated errors. This general likelihood approach is applied to obtain marginal and approximate conditional likelihoods for the correlation parameters in sampling on successive occasions under both simple random sampling on each occasion and more complex surveys.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114522
    Description:

    Findings from the research and testing of telephone and computer assisted survey methods for household surveys are presented, followed by discussion of how these findings will influence the redesign of household surveys at Statistics Canada during the 1990’s. Significant emphasis is given in the presentation to the Canadian Labour Force Survey.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114519
    Description:

    This paper develops a framework for estimating a system of linear equations with survey data. Pure design-based sample survey theory makes little sense in this context, but some of the techniques developed under this theory can be incorporated into robust model-based estimation strategies. Variance estimators with the form of the single equation “linearization” estimator are nearly unbiased under many complex error structures. Moreover, the inclusion of sampling weights in regression estimation can protect against the possibility of missing regressors. In some situations, however, the existence of missing regressors can make the estimation of a system of equations ambiguous.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114526
    Description:

    A chronic problem in the preparation of time critical estimates is the significant limitations inherent in data collection by mail. To address this issue, the U.S. Bureau of Labor Statistics has conducted an extensive 7 year research effort into the use of computer assisted telephone interviewing (CATI) and the computer assisted self interviewing (CASI) methods of touchtone and voice recognition self-response. This paper will summarize some of the significant results of this research covering both performance and cost data. The paper will conclude with a discussion of a large scale implementation program of these techniques for a monthly sample of 350,000 establishments.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114523
    Description:

    Application of recent developments in computer technology allow national statistical offices to produce high quality statistics in an efficient way. At the Netherlands Central Bureau of Statistics (CBS) an increasing use is made of microcomputers in all steps of the statistical production process. This paper discusses the role of software and hardware in data collection, data editing, tabulation, and analysis. To avoid the negative effects of uncontrolled de-centralized data processing, the importance of integration is stressed. This makes the statistical production process easier to manage, and moreover it increases its efficiency. The Blaise System, developed by the CBS, is discussed as a data processing tool that encourages integration. Using a description of the survey questionnaire, this system is able to automatically generate various computer programs for data collection (CAPI or CATI), or data entry and data editing (CADI). The system can also create interfaces to other packages. Particularly, the link between Blaise and the internally developed packages Bascula (for weighting) and Abacus (for tabulation) is described. In this way the Blaise System controls and co-ordinates, and therefore integrates, a large part of the survey process.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114520
    Description:

    The current Survey of Employment, Payroll and Hours, conducted by the Labour Division of Statistics Canada is a major monthly survey collecting data from a large sample of business establishments. This paper describes the methodology of the survey. The description of the stratification, sample size determination and allocation procedures is brief, whereas the description of the rotation procedure is more detailed because of its complexity. Some of the possible simplifications of the design are also highlighted.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114518
    Description:

    This article is a critical review of recent developments in the evaluation of the reliability of provisional national accounts estimates. First, we will sketch a theoretical outline of the process used to produce successive estimates of an aggregate, and will reflect upon its implications regarding the design of the analyses of errors in provisional data. Second, particular attention will be focused upon the choice of elementary measurements of errors and suitable integrative accuracy indices, and the impact of revisions on constant price aggregates and implicit deflationary factors. Finally, the results of some empirical analysis on discordances between provisional and revised estimates will be summarily discussed on the basis of a comparison of national accounts data in Canada, Italy, and the United States.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114524
    Description:

    The Mitofsky-Waksberg procedure is an efficient method for selecting a self-weighting, random digit dialing (RDD) sample of households. The Mitofsky-Waksberg procedure is sequential, requiring a constant number of households be selected from each cluster. In this article, a modified Mitofsky-Waksberg procedure which is not self-weighting or sequential is described. The bias and variance for estimates derived from the modified procedure are investigated. Suggestions on circumstances which might favor the modified procedure over the standard Mitofsky-Waksberg procedure are provided.

    Release date: 1991-06-14

Data (0)

Data (0) (0 results)

Your search for "" found no results in this section of the site.

You may try:

Analysis (18)

Analysis (18) (18 of 18 results)

  • Articles and reports: 12-001-X199100214501
    Description:

    Although farm surveys carried out by the USDA are used to estimate crop production at the state and national levels, small area estimates at the county level are more useful for local economic decision making. County estimates are also in demand by companies selling fertilizers, pesticides, crop insurance, and farm equipment. Individual states often conduct their own surveys to provide data for county estimates of farm production. Typically, these state surveys are not carried out using probability sampling methods. An additional complication is that states impose the constraint that the sum of county estimates of crop production for all counties in a state be equal to the USDA estimate for that state. Thus, standard small area estimation procedures are not directly applicable to this problem. In this paper, we consider using regression models for obtaining county estimates of wheat production in Kansas. We describe a simulation study comparing the resulting estimates to those obtained using two standard small area estimators: the synthetic and direct estimators. We also compare several strategies for scaling the initial estimates so that they agree with the USDA estimate of the state production total.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214508
    Description:

    The usual dual system estimator for population size can be severely biased, if there is population heterogeneity in the capture probabilities. In this note we investigate the bias of the corresponding variance estimator under heterogeneity. We show that the usual estimator is conservative, i.e., it gives too large values, if the two registration systems are negatively correlated, uncorrelated, or when the correlation is positive, but small. In the case of high positive correlation the usual estimator may yield too low values. Two alternative estimators are proposed. One is conservative under arbitrary heterogeneity. The other is conservative under Gaussian heterogeneity. The methods are applied to occupational disease data from Finland.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214505
    Description:

    The X-11-ARIMA seasonal adjustment method and the Census X-11 variant use a standard ANOVA-F-test to assess the presence of stable seasonality. This F-test is applied to a series consisting of estimated seasonals plus irregulars (residuals) which may be (and often are) autocorrelated, thus violating the basic assumption of the F-test. This limitation has long been known by producers of seasonally adjusted data and the nominal value of the F statistic has been rarely used as a criterion for seasonal adjustment. Instead, producers of seasonally adjusted data have used rules of thumb, such as, F equal to or greater than 7. This paper introduces an exact test which takes into account autocorrelated residuals following an SMA process of the (0, q) (0, Q)_s type. Comparisons of this modified F-test and the standard ANOVA test of X-11-ARIMA are made for a large number of Canadian socio-economic series.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214502
    Description:

    A sample design for the initial selection, sample rotation and updating for sub-annual business surveys is proposed. The sample design is a stratified clustered design, with the stratification being carried out on the basis of industry, geography and size. Sample rotation of the sample units is carried out under time-in and time-out constraints. Updating is with respect to the selection of births (new businesses), removal of deaths (defunct businesses) and implementation of changes in the classification variables used for stratification, i.e. industry, geography and size. A number of alternate estimators, including the simple expansion estimator and Mickey’s (1959) unbiased ratio-type estimator have been evaluated for this design in an empirical study under various survey conditions. The problem of variance estimation has also been considered using the Taylor linearization method and the jackknife technique.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214504
    Description:

    Simple or marginal quota surveys are analyzed using two methods: (1) behaviour modelling (superpopulation model) and prediction estimation, and (2) sample modelling (simple restricted random sampling) and estimation derived from the sample distribution. In both cases the limitations of the theory used to establish the variance formulas and estimates when measuring totals are described. An extension of the quota method (non-proportional quotas) is also briefly described and analyzed. In some cases, this may provide a very significant improvement in survey precision. The advantages of the quota method are compared with those of random sampling. The latter remains indispensable in the case of large scale surveys within the framework of Official Statistics.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214503
    Description:

    Surveys are often conducted of flows of persons, such as: visitors to museums, libraries and parks; voters; shoppers; hospital outpatients; tourists; international travellers; and car occupants. The sample designs for such surveys usually involve sampling in time and space. Methods for sampling flows of human populations are reviewed and illustrated.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214506
    Description:

    This paper compares the magnitude and nature of attrition in two separate RDD panel surveys conducted in the City of Chicago (i.e. the surveys were independent studies and were not conducted as part of a planned experiment), each with a between-wave lag of approximately one year. For each survey, sampling at Wave 1 was performed via one-stage (i.e. simple) random-digit dialing. In Study 1, respondents’ names were not elicited; thus, when telephone calls were made at Wave 2 of Study 1 interviewers could not ask for respondents by name. Instead, interviewers asked for respondents by using a gender-age identifier. In Study 2, respondent name identifiers were gathered during Wave 1 and were used in Wave 2 re-contact attempts. The magnitude of the attrition in Study 1 (i.e. the proportion of Wave 1 respondents not re-interviewed at Wave 2) was 47%, whereas in Study 2 it was 43%: a marginal difference in attrition rates. In both surveys, age, race, education and income were significantly related to attrition. Discussion is presented on the trade-off between minimizing attrition vs. minimizing respondent reactivity as potential sources of total survey error. Suggestions for decreasing the size of attrition in RDD panel surveys are discussed.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214500
    Description:

    The Canadian General Social Survey is an annual survey that aims to provide data on the demographic and social characteristics of Canadians. This paper provides an overview of the program, based on the experience of the first five surveys. The objectives of the program, the methodology used, the themes and issues addressed, the program outputs and the plans for the future are all discussed.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100214507
    Description:

    For estimating the proportion and total of an item for the present occasion, independent estimates at the current and previous occasions are combined through three different procedures. In the first one, trend over the occasions is utilized. For the second one, the One-Way Random Effects Model is employed. The third procedure uses the Empirical Bayes approach. All the three procedures are seen to perform better than the sample estimates obtained from the data of the current occasion alone. Advantages of these methods and their limitations are discussed. All the procedures are illustrated with the data from the National Health Discharge Survey.

    Release date: 1991-12-16

  • Articles and reports: 12-001-X199100114525
    Description:

    Electronic data collection utilizing touchtone recognition is in place for a monthly establishment survey at the Bureau of Labor Statistics. The Touchtone Data Entry (TDE) system features digitized phrases requesting respondents to answer questions using the numeric keypad of a touchtone telephone. TDE has substantial implications for lowering survey costs; many labor intensive activities are eliminated. However, little is known about measurement errors associated with this mode of data collection. This study assesses TDE mode error using three sources of data, which allow for analyses of errors associated with selected aspects of the human-machine interface. In addition, instrument design issues associated with mode error are addressed. We conclude by extending the implications of our findings to other surveys.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114521
    Description:

    Marginal and approximate conditional likelihoods are given for the correlation parameters in a normal linear regression model with correlated errors. This general likelihood approach is applied to obtain marginal and approximate conditional likelihoods for the correlation parameters in sampling on successive occasions under both simple random sampling on each occasion and more complex surveys.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114522
    Description:

    Findings from the research and testing of telephone and computer assisted survey methods for household surveys are presented, followed by discussion of how these findings will influence the redesign of household surveys at Statistics Canada during the 1990’s. Significant emphasis is given in the presentation to the Canadian Labour Force Survey.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114519
    Description:

    This paper develops a framework for estimating a system of linear equations with survey data. Pure design-based sample survey theory makes little sense in this context, but some of the techniques developed under this theory can be incorporated into robust model-based estimation strategies. Variance estimators with the form of the single equation “linearization” estimator are nearly unbiased under many complex error structures. Moreover, the inclusion of sampling weights in regression estimation can protect against the possibility of missing regressors. In some situations, however, the existence of missing regressors can make the estimation of a system of equations ambiguous.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114526
    Description:

    A chronic problem in the preparation of time critical estimates is the significant limitations inherent in data collection by mail. To address this issue, the U.S. Bureau of Labor Statistics has conducted an extensive 7 year research effort into the use of computer assisted telephone interviewing (CATI) and the computer assisted self interviewing (CASI) methods of touchtone and voice recognition self-response. This paper will summarize some of the significant results of this research covering both performance and cost data. The paper will conclude with a discussion of a large scale implementation program of these techniques for a monthly sample of 350,000 establishments.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114523
    Description:

    Application of recent developments in computer technology allow national statistical offices to produce high quality statistics in an efficient way. At the Netherlands Central Bureau of Statistics (CBS) an increasing use is made of microcomputers in all steps of the statistical production process. This paper discusses the role of software and hardware in data collection, data editing, tabulation, and analysis. To avoid the negative effects of uncontrolled de-centralized data processing, the importance of integration is stressed. This makes the statistical production process easier to manage, and moreover it increases its efficiency. The Blaise System, developed by the CBS, is discussed as a data processing tool that encourages integration. Using a description of the survey questionnaire, this system is able to automatically generate various computer programs for data collection (CAPI or CATI), or data entry and data editing (CADI). The system can also create interfaces to other packages. Particularly, the link between Blaise and the internally developed packages Bascula (for weighting) and Abacus (for tabulation) is described. In this way the Blaise System controls and co-ordinates, and therefore integrates, a large part of the survey process.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114520
    Description:

    The current Survey of Employment, Payroll and Hours, conducted by the Labour Division of Statistics Canada is a major monthly survey collecting data from a large sample of business establishments. This paper describes the methodology of the survey. The description of the stratification, sample size determination and allocation procedures is brief, whereas the description of the rotation procedure is more detailed because of its complexity. Some of the possible simplifications of the design are also highlighted.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114518
    Description:

    This article is a critical review of recent developments in the evaluation of the reliability of provisional national accounts estimates. First, we will sketch a theoretical outline of the process used to produce successive estimates of an aggregate, and will reflect upon its implications regarding the design of the analyses of errors in provisional data. Second, particular attention will be focused upon the choice of elementary measurements of errors and suitable integrative accuracy indices, and the impact of revisions on constant price aggregates and implicit deflationary factors. Finally, the results of some empirical analysis on discordances between provisional and revised estimates will be summarily discussed on the basis of a comparison of national accounts data in Canada, Italy, and the United States.

    Release date: 1991-06-14

  • Articles and reports: 12-001-X199100114524
    Description:

    The Mitofsky-Waksberg procedure is an efficient method for selecting a self-weighting, random digit dialing (RDD) sample of households. The Mitofsky-Waksberg procedure is sequential, requiring a constant number of households be selected from each cluster. In this article, a modified Mitofsky-Waksberg procedure which is not self-weighting or sequential is described. The bias and variance for estimates derived from the modified procedure are investigated. Suggestions on circumstances which might favor the modified procedure over the standard Mitofsky-Waksberg procedure are provided.

    Release date: 1991-06-14

Reference (0)

Reference (0) (0 results)

Your search for "" found no results in this section of the site.

You may try:

Date modified: