Innovating Technologies for Improving Student Mental Health

Common mental health problems (e.g., depression and anxiety) have their first onset during adolescence. Moreover, increasing levels of independence and other stressors warrant early intervention and prevention during the student years. This need is further stressed when considering the impact of mental health problems on educational trajectories. Due to the specific context of the student population, as well as their need for autonomy and privacy, digital mental health interventions (DMHIs) have been proposed as a suitable solution to improve student mental health. Next to DMHIs provided via the internet, other innovations, such as virtual reality (VR), have been explored in this dissertation, together with novel methodological approaches (e.g., Bayesian statistics, machine learning; ML). The aim of this dissertation therefore was to assess how certain innovative and novel approaches in technology and methodology could be used to improve the mental health of students (and potentially their academic performance). Following a brief foreword and a general introduction to the topic in chapter 1, a meta-analysis on six randomised controlled trials (RCTs; N = 2,428) investigated the effects of DMHls on students' academic performance in chapter 2. Results showed small but non-significant effects on academic performance (g = 0.26; p = .05), and small but significant effects for depression and anxiety (respectively g = -0.24 and -0.02; p :s .01), favouring DMHIs. Results were confirmed by a short intermezzo using Bayesian meta-analysis on the academic performance measures. Chapter 3 provides the study protocol for the Dutch lCare Prevent randomised controlled trial (RCT), which set out to investigate the effectiveness of this DMHI for improving depression and anxiety in students. The encountered recruitment challenges and a basic Bayesian analysis of the intervention's effectiveness is provided in chapter 4. Results indicated that the intervention was unlikely to be effective. Building on this experience, chapter 5 describes the development of a prediction model for students’ readiness to change mental health problems. Using the ML algorithm LightGBM, we were able to develop such a model and outlined important variables to consider (e,g, mental health symptoms, social- environmental factors, and international student status) when predicting readiness to change mental health problems in students. In chapter 6, we conducted an experimental investigation of Lunchroom Zondag, a VR application for cognitive restructuring in depression, in a group of students. We showed that - compared to a face-to-face condition - the VR application elicited more physiological arousal. The latter is considered important for skill acquisition and treatment succes. Finally, chapter 7 provides an overall discussion and highlights the need for a balanced consideration between societal- and individual-level factors when aiming to improve the mental health of students. The conclusion is: technological and methodological innovation can play a significant role in improving student mental health.