Table 5.
Parameter estimates for HLM analysis in Online Addiction Behavior, Problems Caused by Computer Use, and Time Spent Online as function of time (treatment effect)
| Online Addiction Behavior | Problems Caused by Computer Use | Time Spent Online | |||||
|---|---|---|---|---|---|---|---|
| Parameter | Model 0 | Model 1 | Model 0 | Model 1 | Model 0 | Model 1 | |
| Fixed effects | |||||||
| Initial status | Intercept (γ00) | 10.81 | 18.75 | 31.91 | 37.64 | 1.72 | 2.19 |
| (0.50) | (1.11) | (1.02) | (1.75) | (0.08) | (0.21) | ||
| p < .001 | p < .001 | p < .001 | p < .001 | p < .001 | p < .001 | ||
| Rate of change | Slope (γ10) | −6.32 | −4.85 | −0.31 | |||
| (0.84) | (1.22) | (0.13) | |||||
| p < .001 | p < .001 | p = .026 | |||||
| Variance components | |||||||
| Level 1 | Within-person ( ) |
25.98 | 12.09 | 31.77 | 20.41 | 0.26 | 0.22 |
| (6.17) | (3.41) | (8.66) | (5.70) | (0.75) | (0.06) | ||
| p < .001 | p < .001 | p < .001 | p < .001 | p < .001 | p < .001 | ||
| Level 2 | Initial status ( ) |
0.67 | 6.54 | 55.41 | 63.10 | 0.04 | 0.07 |
| (5.12) | (3.79) | (14.37) | (13.46) | (0.06) | (0.06) | ||
| p < .001 | p < .001 | ||||||
| Model fit parameters | |||||||
 |
.53 | .36 | .15 | ||||
| −2 log-likelihood | 661.01 | 618.83 | 746.87 | 734.18 | 82.35 | 77.30 | |
| AIC | 667.01 | 626.83 | 752.87 | 742.18 | 88.35 | 85.30 | |
Note. Parameter estimates for Time Spent Online are log-transformed using the ln(x)-function and need to be retransformed using the ex-function for interpretation. Standard errors are displayed in parentheses. The rate of change displays the amount of change per observation. 
: estimates the proportion of explained within-person variation (level 1); HLM: hierarchical linear models; AIC: Akaike’s information criterion.