Page 194 - Second language development of newly arrived migrant kindergarteners - Frederike Groothoff

194 Chapter 8 effect of Age1 and Age2, as well as variance within and between pupils components which depends on Age needs to be included, and with this model we continued the analysis. Table 8.17: Fit of Different Polynomials (-2LL) for Changes in MLR score (162 cases) as well as the Comparison of Consecutive Models. Model MLRI: β0ijcons a 736.53 Comparison Models MLRI vs MLRII MLRII vs MLRIII MLRIII vs MLRIV MLRIV vs MLRV MLRV vs MLRVI ΔΧ2 Δdf 19.73 1 54.56 2 13.00 1 6.43 1 6.51 3 p <.001 <.001 <.001b .01 .09ns -2LL MLRII: MLRI + β1Age1ij MLRIII: MLRII + e1ijAge1ij MLRIV: MLRIII + u10jAge1ij MLRV: MLRIV + β2Age2ij MLRVI: MLRV + e2ijAge2ij MLRVII: MLRVI + u2ijAge2ijc a In addition to the intercept, variance individuals are estimated. 716.81 662.24 649.25 642.81 636.30 components for differences within and between b Only the covariance-coefficient between the intercept- and the age-residuals was estimated. C No convergence within 250 iterations. Based on this General Development Model we constructed Figure 8.13, in which both the average development as well as the differences within and between individuals are represented (see Table 6.9 in Appendix 6 for the parameter estimates). The average MLR at an age of 73 months was estimated as 2.26. Each month a child grew older, his MLR score increased by 0.07, but the quadratic function of Age decreased the MLR score with 0.003 (since this is a small effect it is not visible in Figure 8.13).