Note. Results of divergent validity analyses for social media mindsets constructed using latent variable modeling. To test whether agency and valence mindsets were distinct from problematic social media use (Bergen Social Media Addiction Scale, Andreassen et al., 2016), general self-efficacy (New General Self-Efficacy Scale, Chen et al., 2001), and trait optimism (Revised Orientation to Life Test, Scheier et al., 1994), we compared structural equation models that treated mindsets as either a distinct construct from each measure (the two factor model) or as a reflection of a preexisting construct (the one factor model). Five model fit indices were used to compare models: chi-square test (goodness of fit of factor model), CFI (Comparative Fit Index, higher scores indicate better fit), TLI (Tucker–Lewis Index, higher scores indicate better fit), RMSEA (root mean squared error of approximation, lower scores indicate better fit), and SRMR (standardized root mean square residual, lower scores indicate better fit). Two-factor models treating social media mindsets as distinct constructs were better fitting on all indices. All models were constructed using the lavaan package in R.