Page 80 - Second language development of newly arrived migrant kindergarteners - Frederike Groothoff

80 Chapter 4 Table 4.3: Fit of Different Polynomials (-2LL) for Changes in PPVT Score (168 cases) as well as the Comparison of Consecutive Model Models. Comparison -2LL Models ΔΧ2 Δdf p PPVT0: β0ijcons a PPVT1: PPVT0 + PPVT2: PPVT1 + PPVT3: PPVT2 + PPVT4: PPVT3 + PPVT5: PPVT4 + PPVT6: PPVT5 + a In addition to individuals are estimated. b In order to circumvent numerical issues in parameter estimation, the explanatory variable age was centered around the grand mean. c Only the covariance-coefficient was estimated. d No convergence within 250 iterations. Based on this General Development Model both the average development and the differences within and between individuals are represented in Figure 4.2 (see Table 1 in Appendix 1 for the parameter estimates). The average receptive vocabulary at an age of 72 months was estimated as 64.39. With each month a child grew older, his receptive vocabulary increased by 1.60, but the quadratic function of Age decreased the score by 0.14. β1 Age1ijb e1ij Age1ij u10j Age1ij β2 Age2ij e2ij Age2ij u20j Age2id 1540.51 1414.63 PPVT0 vs PPVT1 1390.09 PPVT1 vs PPVT2 1386.09 PPVT2 vs PPVT3 1377.92 PPVT3 vs PPVT4 1375.17 PPVT4 vs PPVT5 125.87 1 <.001 24.54 2 <.001 4.00 1c .046 8.18 1 .004 2.74 3 .43ns the intercept, variance components for differences within and between