Table 2.
Parameter estimates of unconditional models for the total, girls’, and boys’ sample
| Class (class sizes in %) | Intercepta (SE) | Slopeb (SE) |
|---|---|---|
| Total | ||
| 1. Very high-stable (1.1 %) | 3.51*** (0.14) | −0.13 ns (0.10) |
| 2. High-decreasing (8.1 %) | 2.75*** (0.06) | −0.12*** (0.02) |
| 3. Moderate-slightly increasing (28.3 %) | 2.01*** (0.03) | 0.03* (0.01) |
| 4. Low-increasing (62.5 %) | 1.32*** (0.01) | 0.10*** (0.01) |
| Girls | ||
| 1. Very high-stable (1.3 %) | 3.62*** (0.21) | −0.12 ns (0.11) |
| 2. High-slightly decreasing (9.7 %) | 2.82*** (0.08) | −0.09*** (0.03) |
| 3. Moderate-slightly increasing (32.7 %) | 2.06*** (0.03) | 0.05** (0.02) |
| 4. Low-increasing (56.3 %) | 1.33*** (0.01) | 0.14*** (0.01) |
| Boys | ||
| 1. High-strong decreasing (4.8 %) | 2.89*** (0.10) | −0.22*** (0.05) |
| 2. Moderate-stable (25.6 %) | 2.00*** (0.05) | −0.01 ns (0.02) |
| 3. Low-slightly increasing (69.6 %) | 1.32*** (0.01) | 0.07*** (0.01) |
* p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001
aVariances of intercepts were fixed at zero
bVariances of slope factors were constrained to be equal across classes (total: 0.021; girls: 0.024: boys: 0.015) and were all significant (p ≤ 0.001)