Crystal Smit

General Introduction 1 17 of this dissertation is the ASSIST intervention (A Stop Smoking In Schools Trial), one of the best known and large-scale social network interventions (Campbell et al., 2008). In the ASSIST intervention, young adolescents nominated classmates based on five sociometric nomination questions (e.g., “Who are good leaders in sports and other group activities at your school?” and “Who do you look up to in Year 8 at your school?”). Those who received the most nominations were trained as influence agents to have informal conversations with their peers to encourage them to not to smoke during 10 intervention weeks (Starkey, Audrey, Holliday, Moore, & Campbell, 2009). The ASSIST social network intervention was effective in reducing the prevalence of smoking up to two years after the start of the intervention (Campbell et al., 2008). Until the start of this PhD project, the social network approach had not been adopted to address healthy drinking behaviors in children. Given the important role of peer influence mechanisms, especially when it comes to healthy drinking behaviors, this dissertation followed the social network approach (Valente, 2012) to develop the so-called Share H 2 O intervention. In this social network intervention, influence agents were trained to informally diffuse messages about water consumption — as an alternative for SSBs — among their peers. To do so, we incorporated a novel, theory-based approach to motivate the influence agents. Integrating Self-Determination Theory in Social Network Interventions The social network intervention literature mainly focuses on the process of selecting the most successful influence agents to diffuse the message and behavior. Specifically, the focus is on investigating the most effective sociometric nomination questions and selection criteria to identify the influence agents (Ott, Light, Clark, & Barnett, 2018; Starkey et al., 2009; Valente, 2012; Valente & Pumpuang, 2008; Woudenberg et al., 2019 ). As yet, there is hardly any attention in the literature on how to exactly to get the influence agents to diffuse the message and behavior. In other words, how to optimally motivate the influence agents to perform and promote the intended health-related behavior among their peers.