TABLE 3.
Model results | ||||
| ||||
Estimate | Standard error (SE) | Estimate/ SE | p-value | |
Intercept of loneliness → slope of loneliness | –0.14 | 0.08 | –1.85 | 0.07 |
Intercept of depression → | ||||
Slope of depression | –0.07 | 0.12 | –0.60 | 0.55 |
Intercept of loneliness | 0.66 | 0.02 | 29.43 | <0.001 |
Slope of loneliness | –0.31 | 0.10 | –3.02 | 0.003 |
Intercept of social anxiety → | ||||
Slope of social anxiety | 1.68 | 0.36 | 4.62 | <0.001 |
Intercept of loneliness | 0.60 | 0.02 | 28.42 | <0.001 |
Slope of loneliness | –0.54 | 0.15 | –3.70 | <0.001 |
Intercept for depression | 0.61 | 0.02 | 25.58 | <0.001 |
Slope of depression | –0.09 | 0.08 | –1.16 | 0.25 |
Slope of depression → | ||||
Intercept of loneliness | –0.12 | 0.07 | –1.58 | 0.12 |
Slope of loneliness | 1.50 | 0.62 | 2.41 | 0.02 |
Slope of social anxiety → | ||||
Intercept of loneliness | 0.92 | 0.18 | 5.03 | <0.001 |
Slope of loneliness | –0.58 | 0.23 | –2.48 | 0.01 |
Intercept of depression | 0.85 | 0.17 | 5.07 | <0.001 |
Slope of depression | 0.43 | 0.21 | 2.08 | 0.04 |
Intercept of social restrictions → | ||||
Slope of social restrictions | 3.06 | 0.21 | 14.94 | <0.001 |
Intercept of loneliness | –0.16 | 0.11 | –1.41 | 0.16 |
Slope of loneliness | –0.02 | 0.02 | –1.13 | 0.26 |
Intercept of depression | –0.16 | 0.07 | –2.32 | 0.02 |
Slope of depression | 0.00 | 0.01 | 0.32 | 0.75 |
Intercept of social anxiety | –0.05 | 0.03 | –1.57 | 0.12 |
Slope of social anxiety | –0.12 | 0.01 | –2.08 | 0.04 |
Slope of social restrictions → | ||||
Intercept of loneliness | –0.50 | 0.37 | –1.35 | 0.18 |
Slope of loneliness | –0.03 | 0.05 | –0.58 | 0.57 |
Intercept of depression | –0.21 | 0.21 | –1.00 | 0.32 |
Slope of depression | –0.04 | 0.04 | –1.16 | 0.25 |
Intercept of social anxiety | –0.05 | 0.03 | –1.58 | 0.12 |
Slope of social anxiety | –0.07 | 0.02 | –2.84 | 0.01 |
LGCM includes data from the subsample whose country level data on social restristrictions during the first six months of the COVID-19 pandemic could be retrieved (N = 1562). Linear growth models were estimated, with continuous outcomes; models were estimated using the robust maximum likelihood (MLR) estimator, to account for missing data (20).