[Zurück]


Zeitschriftenartikel:

A. Alfons, M. Templ:
"Estimation of Social Exclusion Indicators from Complex Surveys: The R Package laeken";
Journal of Statistical Software, Vol. 54 (2013), Issue 15, Sep 2013; S. 1 - 25.



Kurzfassung englisch:
Units sampled from nite populations typically come with di erent inclusion probabilities.
Together with additional preprocessing steps of the raw data, this yields unequal
sampling weights of the observations. Whenever indicators are estimated from such complex
samples, the corresponding sampling weights have to be taken into account. In
addition, many indicators su er from a strong in
uence of outliers, which are a common
problem in real-world data. The R package laeken is an object-oriented toolkit for the
estimation of indicators from complex survey samples via standard or robust methods.
In particular the most widely used social exclusion and poverty indicators are implemented
in the package. A general calibrated bootstrap method to estimate the variance
of indicators for common survey designs is included as well. Furthermore, the package
contains synthetically generated close-to-reality data for the European Union Statistics
on Income and Living Conditions and the Structure of Earnings Survey, which are used
in the code examples throughout the paper. Even though the paper is focused on showing
the functionality of package laeken, it also provides a brief mathematical description of
the implemented indicator methodology.

Schlagworte:
indicators, robust estimation, sample weights, survey methodology, R


Elektronische Version der Publikation:
http://www.jstatsoft.org/v54/i15


Erstellt aus der Publikationsdatenbank der Technischen Universitšt Wien.