Renáta Németh
What
I’m really inspired by is discovering different application areas of
statistics, understanding their epistemological differences, and adapting
methodological knowledge of a field to other ones. Meanwhile, my aim is to
support sociological knowledge discovery in empirical research.
At
present, my main research interest is automated
text analytics (text mining), since I’m ascertain about its
untapped sociological opportunities. Topics we
are working on: discursive framing of depression in online health communities,
corruption in online editorial media, robustness studies in text analytics.
My
previous researches concerned:
-
causality in
social sciences. The axiom of “correlation does not imply causation” was put in
a new light by some statistical results of the last decades. My main question
was how these results can contribute to the social science research
-
marginal
loglinear models, ie. to models which
restrict certain marginal distributions in the contingency table. This research
involved developing parameterisation of graphical models (eg.
path models) for categorical data. I found this approach to be applicable in
the area of social sciences, since graphically represented causal models for
categorical data are widely used by sociologists
-
survey
methodology, eg. desing-based
variance estimation and survey sampling
-
empirical sociology: social mobility and social
inequalities in health.