This week’s SWE Colloquium (14 November) features a presentation by Manuel Völkle, Humboldt University of Berlin

Timing is everything! An introduction to continuous time dynamic modeling

The primary goal of this presentation is to introduce participants to continuous time dynamic modeling. The secondary goal is to go beyond “yet another” statistical modeling approach and to discuss how better considering the role of time may advance our quest for understanding psychological processes. Continuous time dynamic models are models for the analysis of change that make optimal use of the time structure to infer the development and dynamic relationships among constructs of interest. After distinguishing between static and dynamic models for the analysis of change and a short discussion of their respective advantages and disadvantages, I will introduce the basics of continuous time dynamic modeling in a stepwise fashion. I will highlight the possibility to work with intensive longitudinal data, including the analysis of N = 1 time-series (e.g., dynamic factor models), as well as panel data (T small, N large). Apart from a general introduction, special emphasis will be put on the interpretation and practical implementation of these models. I will end with an overview of selected recent developments, current limitations, and future research directions.