Historical Research

Historical research at GESIS seeks to systematically describe and analyze past societies through the usage of theories, formal methods and quantitative data. The data is generally not obtained through surveys but by census and processing historical sources.

Problems of transregional and long-term comparability of historical statistics

Understanding social development processes in historical perspective means that in the acquisition and archiving of research data, quantitative studies from the social and economic history, the historical demography as well as research results on German historical statistics need to be considered which depict the historical development in the form of a time-series. In light of the vast abundance of topic-specific individual reports there arises a series of methodological problems in transregional, long-term comparisons.

Time series analysis: Trends, cycles, structural breaks, outliers

Time series analysis looks at short-term and long-term changes in a series, regularity in margin of fluctuation, structural breaks and historical peculiarities, the weight of specific parameters in dynamic relation to time series of different indicators as well as predictability of historical developments. Various time series analysis methods are evaluated on the basis of longer historical time series with respect to these epistemological interests. The focus is on structural time series models, filtering, ARIMA and ARFIMA models as well as stochastic regression models.

Criminality, criminal prosecution, penal system

The interrelationship between criminality, criminal prosecution, and the penal system has long been the focus of criminology. Patterns of development and influencing factors for criminality, criminal prosecution and the penal system are analyzed in cooperation with the Kriminologischen Zentralstelle (KrimZ - Central Office of Criminology) in Wiesbaden on the basis of long historical time series (back through 1970) in order to develop prognosis models.