Analysis

All (2) ((2 results))

• 1. The impact of different rotation patterns on the sampling variance of seasonally adjusted and trend estimates Archived

  Articles and reports: 12-001-X20000025536

  Description:

  Many economic and social time series are based on sample surveys which have complex sample designs. The sample design affects the properties of the time series. In particular, the overlap of the sample from period to period affects the variability of the time series of survey estimates and the seasonally adjusted and trend estimates produced from them.

  Release date: 2001-02-28

  ---

• 2. Making unit-level inferences from aggregated data Archived

  Articles and reports: 12-001-X199600114390

  Description:

  Data are often available only as a set of group or area means. However, it is well known that statistical analysis based on such data will often produce results very different from those obtained from analysing the corresponding individual or household data. If the results of area level analyses are thought to apply to the individual level then we risk committing the ecological fallacy. Aggregation or ecological effects arise in part because geographic areas are not comprised of random groupings of people or households but exhibit strong socio-economic differences between areas. The population structure must be incorporated into the statistical model underpinning the analysis if aggregation effects are to be understood. A simple general model is proposed to achieve this and the consequences of the model and its implications for the estimation of population means and covariance matrices are obtained. Furthermore, methods are suggested which can provide unbiased estimates of individual level parameters from aggregated data and so avoid the ecological fallacy. These methods rely on identifying the “grouping variables” that characterise the process that led to the population structure, or at least characterise the area differences. An estimate of the unit level covariance matrix of the grouping variables is required from some source. Data from the 1991 Census of the United Kingdom have been analysed to identify the important grouping variables and evaluate the effectiveness of the proposed adjustment methods for the estimation of covariance matrices and correlation coefficients. These results lead to a suggested strategy for the analysis of aggregated data.

  Release date: 1996-06-14

Stats in brief (0) (0 results)

No content available at this time.

Articles and reports (2) ((2 results))

• 1. The impact of different rotation patterns on the sampling variance of seasonally adjusted and trend estimates Archived

  Articles and reports: 12-001-X20000025536

  Description:

  Many economic and social time series are based on sample surveys which have complex sample designs. The sample design affects the properties of the time series. In particular, the overlap of the sample from period to period affects the variability of the time series of survey estimates and the seasonally adjusted and trend estimates produced from them.

  Release date: 2001-02-28

  ---

• 2. Making unit-level inferences from aggregated data Archived

  Articles and reports: 12-001-X199600114390

  Description:

  Data are often available only as a set of group or area means. However, it is well known that statistical analysis based on such data will often produce results very different from those obtained from analysing the corresponding individual or household data. If the results of area level analyses are thought to apply to the individual level then we risk committing the ecological fallacy. Aggregation or ecological effects arise in part because geographic areas are not comprised of random groupings of people or households but exhibit strong socio-economic differences between areas. The population structure must be incorporated into the statistical model underpinning the analysis if aggregation effects are to be understood. A simple general model is proposed to achieve this and the consequences of the model and its implications for the estimation of population means and covariance matrices are obtained. Furthermore, methods are suggested which can provide unbiased estimates of individual level parameters from aggregated data and so avoid the ecological fallacy. These methods rely on identifying the “grouping variables” that characterise the process that led to the population structure, or at least characterise the area differences. An estimate of the unit level covariance matrix of the grouping variables is required from some source. Data from the 1991 Census of the United Kingdom have been analysed to identify the important grouping variables and evaluate the effectiveness of the proposed adjustment methods for the estimation of covariance matrices and correlation coefficients. These results lead to a suggested strategy for the analysis of aggregated data.

  Release date: 1996-06-14

Journals and periodicals (0) (0 results)

No content available at this time.