Maartje Boer

SUMMARY AND DISCUSSION 307 9 may be undesirable to consider diagnostic recognition for each candidate addictive behavior. Therefore, should a diagnostic recognition of addiction- like SMU be warranted, including a more general diagnostic classification, such as ‘Internet-related addiction’, may be more tenable. Extending Theoretical Knowledge on Problematic SMU So far, little is known about individual differences in the effect of SMU problems. Multilevel modelling techniques applied to longitudinal data allow to explore this so-called heterogeneity . In our longitudinal chapters, we did not investigate heterogeneity for the within-person association between SMU problems and wellbeing (Chapters 5, 6, 8). Instead, we assumed that the association was homogeneous (i.e., the same across adolescents), while in psychology, person-specific heterogeneity in associations is often more realistic (Keijsers & Roekel, 2018). Accordingly, our findings (Chapter 8), as well as earlier experience sampling studies (Beyens, Pouwels, Valkenburg, et al., 2020; Beyens, Pouwels, Van Driel, et al., 2020; Valkenburg, Beyens, et al., 2021), demonstrated that the within-person association between adolescents’ SMU intensity andwellbeing differed substantially across adolescents. For example, one study showed that, while on average momentary passive SMU was not associated with momentary changes in wellbeing, for 46% of adolescents passive SMU increased wellbeing, for 10% passive SMU decreased wellbeing, and for 44% there was no association (Beyens, Pouwels, Valkenburg, et al., 2020). Future longitudinal studies adopting multilevel analytical techniques to study individual variation in within-person associations between SMU problems and wellbeing provide more knowledge on the robustness of the association across adolescents. Furthermore, with intensive longitudinal data (e.g., 100 measurements per adolescent), more advanced multilevel techniques, such as dynamic structural equation modelling (Asparouhov et al., 2017), allow researchers to study heterogeneity in the effect of SMU problems on wellbeing and vice versa in one model (Beyens, Pouwels, Van Driel, et al., 2020; Valkenburg, Beyens, et al., 2021; Valkenburg, Pouwels, et al., 2021). Not only does the association between SMU problems and wellbeing possibly differ across adolescents, it may also differ across time frames. Chapter 6 yielded rather small effect sizes for the effects of SMU problems