Anna Langener

Science and Engineering
University of Groningen

On 5 September 2024 Anna Langener defended her thesis: From Data Streams to Mental Health Predictions: Improving the use of Passive Measures from Digital Devices at the University of Groningen (RUG).

Summary

Poor mental health is a global concern, with the World Health Organization reporting that one in eight people suffer from a mental disorder. Identification and treatment are hampered by limited access to care and inadequate health insurance coverage. Digital technologies, such as smartphones, offer promising tools for improving mental health through continuous monitoring and timely intervention. These devices can collect rich data on various factors, such as social context and behavior, through active (e.g., questionnaires) and passive (e.g., GPS tracking) methods. Researchers often aim to use this passively collected data to predict mental health outcomes. Despite its potential, passive data collection is still evolving, and current predictive accuracy remains low to moderate. The overall goal of this thesis is therefore to optimize the use of passive measures from digital devices for predicting mental health outcomes.The first part of this thesis focuses on improving the accuracy of predicting mental health outcomes. Results show that combining passive and active data collection methods outperforms passive measures alone, but predictive performance remains low to moderate. Advanced machine learning models also show only moderate success in predicting variability in depressive symptoms. The second part of this thesis focuses on improving the transparency and reproducibility of studies using passive measures. It highlights key challenges researchers face and provides guidance for researchers working with passive measures, for example by proposing a preregistration template. Preregistration involves publicly outlining the study plan before the research begins, which can increase transparency and prevent bias.

Project

Using multimodal data to understand the social environment and its effect on wellbeing
Humans are fundamentally social beings. Social interactions are essential for individuals to develop the skills needed to function in society. The lack of social interactions and social relations are a leading cause of poor mental wellbeing and depression (Dodge et al., 2012; Huang et al., 2019). Thus, it is not surprising that the social context is studied across various disciplines. However, the ability to capture the social environment has been challenging.

First, the social environment is a multidimensional construct with various components (Phongsavan et al., 2006; Yu et al., 2015). Thus, different disciplines are using different operationalizations and conceptualizations of the social environment (Gariépy et al., 2016; Giordano et al., 2011). Even more so, each discipline is developing methods and measures that capture different parts of the social environment relevant for wellbeing (Phongsavan et al., 2006). Research that attempts to capture these different aspects of the social environment simultaneously is lacking.

Next, the social environment is a dynamic concept, yet most methods used involve static measures. Recent innovations have led to methodologies that are able to capture the dynamics (e.g. day-to-day social interactions) of the social environment. Psychologists, for instance, often use experience sampling methods (ESM) to study change within a person, the environment, and social interactions (e.g. Brown et al., 2011; Siewert et al., 2011). Often dynamical psychological network models (such as VAR-models) are used to analyze those collected ESM data (Bringmann et al., 2013). Next, various disciplines are using digital phenotyping (based on, e.g., smartphones) to passively capture the social behavior and interactions of a person (e.g. Eskes et al., 2016; Jaques et al., 2015). Among other methods, machine learning is often used to analyze the high amount of collected data (Burns et al., 2011; Jacobson & Chung, 2020). Further, sociologists are starting to use egocentric networks in a dynamical way to integrate the change of an individual in a dynamic larger social context (e.g. Atherley et al. 2020). Here, researchers often analyze the change of network characteristics between different time points (e.g. Atherley et al. 2020, Feld et al., 2007).

Up to now, these methods have not been integrated. However, combining the strength of those methods might create a more accurate and feasible measurement of social dynamics and its relation to wellbeing. Therefore, in my PhD project, I aim to answer the following questions;

How are ESM, digital phenotyping, and ego networks used to capture the dynamics of the social environment?
What are the similarities and differences of those methods? What are the advantages and disadvantages?
How are those measurements most beneficially combined to capture the social environment?
Which ways are suitable to statistically analyze the multimodal data?

I propose that the integration of these methods will provide a way forward to assess the human social environment and its dynamics in a quantitative manner.

Supervisors
prof. dr. M.J.H. Kas, dr. L.F. Bringmann, dr. G. Stulp

Financed by
PhD Scholarship Program

Period
1 September 2020 – 1 September 2024