Maartje Boer

VALIDATION OF THE SMD-SCALE 53 2 suggesting that the association between age and problematic SMU was non-linear. This non-linear association may explain why previous research on problematic SMU in adolescents found only a small effect size of age or no age differences at all (Bányai et al., 2017; Ho et al., 2017; Mérelle et al., 2017). There may be several reasons why girls, lower educated, 15-year-olds, and non-Western adolescents reported higher levels of problematic SMU. Girls may find it more important to maintain and expand social relationships and to express or validate their thoughts and feelings than boys (Kuss & Griffiths, 2011, 2017). This may make girls more vulnerable to developing problematic SMU, as social media facilitates fulfilling these needs (Kuss & Griffiths, 2011, 2017). In addition, Dutch adolescents with a low educational level or with a non-Western backgroundare relatively likely to come fromlowsocioeconomic status families (Central Bureau for Statistics, 2017, 2018a). Adolescents with low socioeconomic status backgrounds are more sensitive to engaging in risky behavior in general than adolescents with high socioeconomic status backgrounds, possibly related to lower support from family, cognitive challenges, or limited self- control (Inchley et al., 2016; Stevens et al., 2018). Similarly, adolescents with a low educational level or with a non-Western background may be more sensitive to developing problematic SMU. Further, the finding that the level of problematic SMU was highest among 15-year-olds implies that there may be an increased risk of problematic SMU during this stage of adolescence. The popularity of social media during adolescence may reach its peak at this age, which may make social media harder to resist. However, empirical research is required to examine the mechanisms underlying the differences found in the present study. In addition, the observed proportions of positive scores on the problematic SMU criteria were rather low ( < 30%). Consequently, the scale’s sum-scores showed a skewed distribution, indicating that many adolescents did not endorse any criteria, and aminority endorsedmany criteria. This finding suggests that higher levels of problematic SMU are relatively uncommon, which is in line with previously reported prevalence rates of problematic SMU and other problematic internet-related behaviors, including internet gaming disorder and internet addiction (Andreassen, 2015; Kuss et al., 2014; Lemmens et al., 2015). While intense SMU, indicated by very frequent use of social media, is common among contemporary adolescents (Anderson & Jiang, 2018), scholars emphasize that a rather small proportion of social media users may