Some Practical Issues Related to the Integration of Data from Sample Surveys

The users of official statistics data expect multivariate estimates at a low level of aggregation. However, due to financial restrictions it is impossible to carry out studies on a sample large enough to meet the demand for low-level aggregation of results. At the same time, respondents’ burden prev...

Full description

Saved in:
Bibliographic Details
Published inStatistika (Prague, Czech Republic) Vol. 95; no. 1; pp. 60 - 75
Main Author Wojciech Roszka
Format Journal Article
LanguageEnglish
Published Czech Statistical Office 01.01.2015
Subjects
Data fusion
data integration
multiple imputation
quality assessment
sample survey
statistical matching
Online AccessGet full text

Cover

More Information
Summary:The users of official statistics data expect multivariate estimates at a low level of aggregation. However, due to financial restrictions it is impossible to carry out studies on a sample large enough to meet the demand for low-level aggregation of results. At the same time, respondents’ burden prevents creation of long questionnaires covering many aspects of socio-economic life. Statistical methods for data integration seem to provide a solution to such problems. These methods involve fusion of distinct data sources to be available in one set, which enables joint observation of variables from both files and estimations based on the sample size being the sum of sizes of integrated sources.
ISSN:1804-8765
1804-8765