WP5 Integrative Modelling

Understanding how human development is shaped by many different factors acting together, including biological, social and environmental influences, requires analytical methods that can integrate insights from different disciplines.

In this context, Work Package 5 focuses on developing a new method for analysing multidisciplinary data, called Stacked Domain Learning (SDL).

How SDL Works

SDL integrates different disciplinary perspectives on a shared outcome by translating theories from each discipline into statistical models and combining them within an overarching meta-model. This approach produces weights that indicate the relative contribution of each disciplinary perspective to predicting the outcome.

Integrate the diverse data collected across the consortium

The goal of developing SDL within GUTS is to provide a general framework for integrating the diverse data collected across the consortium. By doing so, it enables a more comprehensive assessment of the influences that are relevant to the challenges young people face today and to their engagement in an increasingly complex society.

Recent Progress and Future Directions

Recent progress includes the completion of an introductory SDL paper with an empirical application to data from the ABCD Study, a project comparable in scope to GUTS. Building on this foundation, future work will focus on extending SDL by incorporating interactions and integrating a wider range of modelling techniques.

WP5:

• Mark De Rooij 
• Hilde Huizenga
• Nienke van Atteveldt
• Gert Stulp
• Marjolein Fokkema
• Zino Brystowski