NPR - A statistical nigtmare

2008

It is rather common to use administrative data in the official statistics thanks to their economic and technical advantages. It is indeed more economic to use administrative archives than to gather data through traditional sample surveys or censuses; the outputs are more quickly issued and the information can be issued on very small areas. No biases such as the memory effect, the interviewer effect, the question formulation effect, the reticence to answer, etc. can be found. Only few answers are missing in the archive while the correction process is simpler and the information on sensitive variables (e.g. income) more accurate. On the other hand, the non-primary variables are often of a lower quality. In addition the “administrative” definition and the “statistical” definition of the units generally differ; but in case they do not differ, the administrative reference universe is different from the statistical one. Nonetheless, it is now impossible not to use administrative data. In ...

2015

Civil registration statistics is very important in many developing countries; it is among the best sources of information on population change. A well organized system for collecting such information contributes enough to development of a country. A qualitative research shows that statistics are expensive for developing countries and one of solutions is the development of administrative data collection based. The study will focus on Rwanda an African country where the economy is emerging and the requirement of data is high especially on population change and the economy. Descriptive analysis will show how data from civil registration records will be used to produce indicators of population usefully in health, social protection and education program. In this study a sample survey conducted in Rwanda for the user satisfactory of civil registration data shows the importance of civil registration statistics as a tool for monitoring the systems of good governance and a key component in p...

Elżbieta Gołata New paradigm in statistics and population census quality The paper refers to shift in methods applied in statistics: from traditional censuses through sample survey to administrative data, as a new paradigm in statistical surveys. The research is conducted with respect to population census. Special attention is put to assessing quality of ‘new generation’ population census which use multiple data sources. Register-based censuses in conjunction with sample survey are considered as well as opportunities and statistical challenges of the future. Despite long tradition and well-developed research methodology, censuses do not provide ‘perfect’ results. Census data quality is discussed with respect to criteria resulting from its aim, applied census method, number of data sources used for evaluation and types of errors. First of all, census may be burdened with non-random: coverage and content errors. Due to different methods for conducting censuses, also random errors and ...

2018

According to Istat modernisation programme the Registers Integrated System (RIS) is the core of the data production process. The RIS enables to enlarge the level of the analysis and the quality of the information collected linking at micro-level the economic and social phenomena. In this view, the Base Register on Individuals and Households (BRIH) that identifies the usually-resident population will be enriched by the information collected by the other Base Registers (enterprises, addresses, activities) and by the thematic ones (Labour, Educational level, etc.). BRIH will be the common target for Continuous Census and Demographic and Social Statistics. To improve the accuracy of the population outputs in terms of timeliness, coverage and consistency it has been developed an information system of micro demographic accounting based on a longitudinal statistical register of individuals (Anvis). Anvis continuously monitors population changes by linking the population stock with outflows ...

Statistical Journal of the IAOS, 2020

Statistical Journal of the IAOS

Statistics in Transition New Series, 2016

in: Rudolfovo číslo: Rudolfu Procházkovi k 65. narozeninám kolegové a přátelé (Loskotová, Irena ed.) Brno-Praha, 2021