Maartje Boer

CHAPTER 2 30 Eijnden et al., 2016), the reliability at different levels of problematic SMU has not been investigated. Third, it remains unclear whether the factor structure of the SMD-scale is equal across subpopulations, which is required to reliably compare observed levels of problematic SMU across subpopulations (F. F. Chen, 2007). Because studies suggest that girls, lower-educated adolescents, specific age groups, and immigrant adolescents are more sensitive to developing problematic SMU (Bányai et al., 2017; Ho et al., 2017; Mérelle et al., 2017), it is pivotal to examine whether the scale is measurement invariant across these groups in order to be able to interpret these differences. Fourth, research shows that it is often possible to distinguish subgroups whose members show similar characteristics with regard to a particular behavior (Bányai et al., 2017; Király, Sleczka, et al., 2017; Lemmens et al., 2015; Peeters et al., 2019). It has not been investigated whether the SMD-scale can be used to study subgroups of users, and if so, by which set of criteria these subgroups could be characterized. The identification of such subgroups may provide more insight into the phenomenon of problematic SMU and allow researchers to use the scale to compare subgroups of users on, for example, their wellbeing. Fifth, previously conducted criterion validity analyses on the SMD-scale were limited to assessments of mental health problems (Savci et al., 2018; Van den Eijnden et al., 2016). In order to verify whether the test score interpretations of the scale are valid, associations with other constructs related to adolescents’ daily life functioning should be considered as well, including school functioning and sleep problems. Current Study Given the increasing body of literature showing that problematic SMU is negatively associated with mental health and functioning in important life domains, it is essential that research on problematic SMU uses a psychometrically sound instrument. The present study is the first that uses a large-scale, nationally representative sample of adolescents to validate the nine-item SMD-scale. Data came from 6,626 Dutch secondary school adolescents aged 12-16 years who participated in the Health Behavior in School-aged Children study (HBSC). The present study aimed to investigate the (1) structural validity, (2) reliability, (3) measurement invariance, (4) item score patterns, and (5) criterion validity of the SMD-scale scores. After