Table 4.

Mediating effects of perceived social support, loneliness, and social media violence on the associations between problematic social media use and depressive symptoms among college students in Shaanxi, China.

The Paths	Mediating Effect a	95% CI	p-Value b	Percentage of Mediating Effects in the Total Effects (%)
1. Problematic social media use → Perceived social support → Depressive symptoms	0.013	0.011, 0.016	0.007	4.545
2. Problematic social media use → Perceived social support → Loneliness → Depressive symptoms	0.01	0.008, 0.012	0.01	3.497
3. Problematic social media use → Perceived social support → Social media violence → Loneliness → Depressive symptoms	0.001	0.001, 0.001	0.008	0.350
4. Problematic social media use → Perceived social support → Social media violence → Depressive symptoms	0.004	0.003, 0.005	0.009	1.399
5. Problematic social media use → Loneliness → Depressive symptoms	0.06	0.055, 0.065	0.013	20.979
6. Problematic social media use → Social media violence → Loneliness → Depressive symptoms	0.012	0.011, 0.014	0.007	4.196
7. Problematic social media use → Social media violence → Depressive symptoms	0.044	0.039, 0.049	0.007	15.385
Indirect effects	0.143	0.133, 0.156	0.003	50.000
Direct effects	0.143	0.134, 0.158	0.011	50.000
Total effects	0.286	0.270, 0.301	0.007	100.000

^a: Structural equation model was used to analyze the mediating effect of perceived social support, social media violence, and loneliness on the associations between problematic social media use and depressive symptoms. Age and sex were included as covariates. ^b: We used Bonferroni Tests to adjust for multiple comparison effects, where p = 0.007 = 0.05/7 indicates the standard of statistical significance for each mediating effect. These results were not significant after considering multiple comparison effects.