The Data Analysis Support Core (DASC) aims to help Stress in Action’s researchers make the most of their research. This requires an alignment between one’s research question, data, and analytical technique. Hence, the DASC is concerned with:

• Raising awareness about how to best align these components of empirical research and how to make methodological decisions,
• Identifying data analytical challenges and developing new techniques to tackle these, by innovating the individual techniques and by combining dynamic modelling and machine learning techniques.

In 2024, the DASC held monthly online meetings to discuss progress, challenges, and opportunities for cross-fertilization between projects. We also reviewed updates on the 14 consultancy requests from the consortium, covering topics such as the application of statistical and AI techniques and evaluations of methodological approaches.

Within the DASC, four PhD candidates and one postdoctoral researcher worked on a variety of advanced research topics, submitting five papers for publication. Their projects included ensemble learning for dynamic prediction, using Deep Latent Variable Models for longitudinal data analysis, addressing timeseries analysis challenges, expanding joint models for intensive longitudinal data, and exploring dynamic prediction in digital phenotyping. Additional research included the development of a mask initialization algorithm for deep learning model explanations, as well as the establishment of power analyses for g-methods and causal effects in panel data.

As each senior DASC member is also part of one of the three RTs, short lines of communication are guaranteed. Several of the research projects the DASC members are involved in include collaborations with researchers from specific RTs, allowing them to study their data or collaborating to answer their research questions.

In addition, during the SiA educational programme in November two sessions were held to explain both longitudinal and AI techniques, along with the associated methodological challenges, in order to better equip the junior researchers to apply such techniques.

Looking ahead, we plan to actively seek out use cases within the RTs to apply our methodological innovations.

DASC members: Jeroen Mulder, Sten Willemsen, Yong Zhang, Bülent Ündes, Fridtjof Petersen, Laura Bringmann, Dimitris Rizopoulos, Nina van Gerwen.