Table 3.
Linear Mixed‐Effects Models Predicting Loneliness
| Models (DV: Loneliness) | AIC |
Log likelihood (p‐value) |
|
|---|---|---|---|
| Demographics | Loneliness ˜ session + gender + age + (1|ID) | 1714 | −850.995 |
| Demographics and controls | Loneliness ˜ session + age + screen time + family social support + friend social support + (1|ID) | 1382 |
−681.99 (p < .001) |
| Online social experiences | Loneliness ˜ session + online experiences + gender + age + screen time + family social support + friend social support + (1|ID) | 1257.8 |
−617.88 ( p < .001***) |
| Full interaction | Loneliness ˜ session + positive online × negative online + age + gender + screen time + family social support + friend social support + (1|ID) | 1259.0 |
−617.52 (p = .40) |
| Interactions between online social experiences and session | Loneliness ˜ session × positive online experiences + negative online experiences + gender + age + screen time + family social support + friend social support + (1|ID) | 1259.5 |
−614.76 (p = .14) |
| Four‐way interaction with gender | Loneliness ˜ session × positive online experiences × negative online experiences × gender + age + screen time + family social support + friend social support + (1|ID) | 1271.5 |
−613.78 (p = .96) |
| Four‐way interaction with age | Loneliness ˜ session × positive online experiences × negative online experiences × age + gender + screen time + family social support + friend social support + (1|ID) | 1266.8 |
−611.37 (p < .001***) |
Preferred model (in bold) was selected based on AIC values by sequentially comparing nested models. Models with interaction effects also included a main effect for these factors.
AIC = Akaike information criterion.
*p < .05, **p < .01, ***p < .001.