Table 5.
Linear regression analysis showing associations between daily time spent on social media, negative social media-related experiences, general self-efficacy, and depressive symptoms by gender.
| Independent variables |
Boys (n = 11,574) |
Girls (n = 12,900) |
|---|---|---|
| Social media use | β | β |
| Daily time spent on social media | 0.02 | 0.05*** |
|
Explained variance (Adjusted R2) |
1.3% *** | 2.6% *** |
| Negative social media- | ||
| related experiences | ||
| Experienced online bullying | 0.22*** | 0.19*** |
| Arguments or conflict in the family | 0.02 | 0.09*** |
| Not enough sleep | 0.15*** | 0.11*** |
| Pressure toward having followers and likes | 0.13*** | 0.13*** |
|
Explained variance (Adjusted R2) |
14.4% *** | 15.0% *** |
| Psychological resources | ||
| General self-efficacy | −0.22*** | −0.32*** |
|
Explained variance (Adjusted R2) |
18.7% *** | 24.5% *** |
| Sociodemographic | ||
| variables | ||
| Higher age | 0.12*** | 0.12*** |
| Centrality | −0.04*** | −0.02** |
| Parents with higher education | −0.03** | −0.03*** |
|
Explained variance (Adjusted R2) |
20.5% *** | 26.1% *** |
Dependent variable is depressive symptom level. Higher values on “centrality” indicates more rural. Table content is standardized β values. The models show direct associations when all variables are included.
***
p < 0.001,
**
p < 0.01. Bold is used to separate explained variance from beta values.