I am an enthusiastic and determined Associate Professor in Statistics at the Department of Methodology and Statistics (M&S) at Utrecht University. I am driven by my belief in the urgent need for novel statistical methods for intensive longitudinal data. Facilitated by ubiquitous mobile devices, intensive longitudinal data refers to measurements collected with a high frequency over a prolonged period of time on a group of individuals. Recognition of the scientific value of intensive longitudinal data is evidenced by its use in publications, which increased from ~250 to >2.500 between 2000 and 2023 (PubMed). These data provide unique opportunities to study within-person behavioural phenomena that unfold over time. I develop methods to unravel these personalised processes. 

From PhD to current position 
After completing a Research Master in Methodology and Statistics (cum laude; Utrecht University), I pursued an interdisciplinary PhD combining statistics and neuroscience (VU University Amsterdam). Since my PhD, I have been passionate about developing novel statistical methods to fit complex datasets. Fitting methods are essential for the advancement of science, as they ensure that data are used optimally and that reliable research conclusions are drawn. During my PhD, I extended the everyday statistical toolbox of neuroscientists. After positions as junior researcher (Max Planck Institute, Berlin) and data scientist (TNO), I returned to Utrecht University as assistant professor and since 2022 associate professor. Here I focused my interest on a specific type of complex data: intensive longitudinal data. 
 

Research programme
Currently, I have two lines of research: 

Personalized latent dynamics 
Since 2018, I have been pioneering and promoting the multilevel hidden Markov model (HMM; a longitudinal latent mixture modelling approach) as a novel method for summarising complex latent dynamics in social and behavioural intensive longitudinal data. The novelty lies in applying a method commonly used in other research fields and extending it to infer personalised dynamics, facilitated by the multilevel framework. Within this line of research I supervise(d) 3 PhD students.

Real-time prediction 
Since 2021, I have started a second line of research in collaboration with the Department of Cardiology, University Medical Center (UMC) Utrecht. The project uses intensive longitudinal data from heart failure patients and is partly funded by a public-private partnership via two Health-Holland grants (2 PhD students). Here we develop and apply novel machine learning algorithms for real-time prediction of heart failure deterioration.