Maartje Boer

VALIDATION OF THE SMD-SCALE 29 2 lying about time spent on SMU). These nine criteria for problematic SMUwere adopted from the DSM-5 definition of internet gaming disorder (American Psychiatric Association, 2013; Lemmens et al., 2015). By adding three additional criteria to the six core criteria of addiction, the nine-item SMD-scale provides a more comprehensive operationalization of problematic SMU. The SMD-scale was developed based on a confirmatory factor analysis (CFA) on data from a 27-itemquestionnaire assessed among 10- to 17-year-old Dutch adolescents, which included three items for each of the nine criteria (Lemmens et al., 2015; Van den Eijnden et al., 2016). The nine-item SMD-scale consists of the items that showed the highest factor loading per criterion. The nine items can be regarded as nine subdimensions, yet together, they intend to reflect one overarching dimension (Van den Eijnden et al., 2016). Indeed, CFA on the nine-item scale demonstrated solid structural validity for a unidimensional (i.e., one-factor) model, with acceptable internal consistency of the test scores. Also, higher scores were associated with higher reports of compulsive internet use, self-declared social media addiction, and mental health problems, indicating good convergent and criterion validity of the test score interpretations (Van den Eijnden et al., 2016). An adapted version of the SMD-scale with polytomous instead of dichotomous response scales was validated among a sample of 553 Turkish adolescents aged 14-18 (Savci et al., 2018). In this study, Exploratory Factor Analysis (EFA) also identified one dimension, and internal consistency of the test scores was acceptable. Also, the convergent and criterion validity of the test score interpretations was adequate (Savci et al., 2018). Although these studies indicated that the SMD- scale has appropriate psychometric properties, important validation steps remain unaddressed. First, the structural validity of the SMD-scale score interpretations has not been explored in a nationally representative sample. Although the scale aims to measure one overarching dimension problematic SMU (Van den Eijnden et al., 2016), exploring possible multidimensionality is crucial to enhance our understanding of problematic SMU. Furthermore, the use of the sum-score of the nine items to assess adolescents’ level of problematic SMU is only justified when the scale measures one underlying dimension to which all nine items substantially contribute. Second, although the test scores of the SMD-scale were found to have acceptable internal consistency (Savci et al., 2018; Van den