Speaker: Sacha Epskamp
Title: The dynamics of psychology: Applying network estimation on psychological data
Date: Thursday October 1, 12-1 pm
Venue: AS4/02-08 (Psychology Department Meeting Room)
Abstract:
Recent years have seen an emergence of the dynamical systems approach to psychology, in which constructs, e.g., depression, are not seen as common cause to their indicators but rather as emergent behavior in a dynamical system: e.g., insomnia -> fatigue -> concentration problems -> excessive worrying -> insomnia. Estimating such relationships in psychological data results in network models that portray the full system of pairwise relations in the data: the full picture. State-of-the-art estimation techniques of network structures on psychological data have been implemented in easy to use software. In this talk, I will show three empirical examples. The first example will show an undirected network structure of personality inventory items estimated from cross-sectional data. The second example will show how limited longitudinal data of a single subject can be used to estimate temporal and contemporaneous network structures usable in clinical practice. Finally, I will describe the results of multi-level estimation of temporal effects between psychopathology and personality items.
About the Speaker:
After obtaining my master’s degree in 2012 majoring in psychological methods and minoring in computational science, I received a personal research talent grant for a 4 year PhD project. In this project I work on integrating network modeling techniques in the field of psychometrics, with applications in clinical, social and developmental psychology. In addition to theoretical work I have developed several software package to facilitate the use of our methods to empirical researchers. Besides my PhD research, I teach at the University of Amsterdam on programming and network analysis, have done statistical consultancy work and work one day per week for the university-originated company Oefenweb in which novel psychometric modeling techniques are used to adaptively measure and educate elementary school children on mathematical and language skills.
