%0 Generic
%A Stroe-Kunold, Esther
%D 2010
%F heidok:10670
%K multivariate time series analysis , cointegration methodology, dynamic systems
%R 10.11588/heidok.00010670
%T Analyzing Dynamic Process Systems : Cointegration Methodology as a Tool of Psychological Research
%U https://archiv.ub.uni-heidelberg.de/volltextserver/10670/
%X The present paper-based thesis puts forward cointegration methodology as a multivariate tool of psychological research. Aiming at familiarizing psychologists with the toolbox of cointegration techniques, the studies conducted here provide strategies to analyze complex dynamic process systems. Within the framework of these systems, integrated processes displaying an unpredictable course due to stochastic trends interact over time. If these non-stationary series are co-integrated, their interaction is driven by common stochastic trends with the systems returning to stable equilibrium states in the long run. Vector error-correction (VEC) modeling, a frequently used representation of cointegrated systems, allows insights into these short- and long-term dynamics at a glance. The objectives of this thesis are (a) to adapt this econometric approach to psychological circumstances based on conceptual considerations; (b) to provide a systematic investigation of the mathematical models behind integrated and cointegrated processes as well as their VEC representation, thus clarifying how their parameters are to be interpreted from a psychological perspective; and (c) to address issues of research practice such as spurious relations or long memory characteristics. By means of simulated as well as empirical data from different domains of psychology, this work is designed as a step-by-step guideline inducing psychological applications.