Page 116 - Second language development of newly arrived migrant kindergarteners - Frederike Groothoff
P. 116

116 Chapter 5    5 4,5 4 3,5 3 2,5 2 1,5 1 0,5 0           Age in Months Mean growth MLR Growth MLR 80% within - Within 80% - 80W%itwhitnhi8n0+% + 80B%etbweetwenee8n0-% - 80B%etbweetwenee8n0+% + Figure 5.8: Graphical Representation of the General Development Model Measure of Lexical Richness with 80% Reliability. Figure 5.8 shows that the effect of Age on MLR scores is significant. This is a quadratic relation. Furthermore, the differences within individuals (represented by the striped lines in Figure 5.8) also depend on Age; we can give a more precise estimate of the Measure of Lexical Richness for younger pupils as compared to older pupils. The differences between individuals (represented by the dotted lines in Figure 5.8) are a function of Age as well; differences between younger individuals are smaller than differences between of older pupils. A likelihood ratio test showed that the main effect of Exposure to Dutch at School contributed significantly to the fit of the model to the observed data (ΔΧ2 (MLR7) = 9.25; df = 1; p = 0.002; see Table 5.14): the longer the Exposure to Dutch at School, the higher the score on the MLR. No interaction between Age and Exposure to Dutch at School was found (MLR8). Table 5.14: Fit of Different Models (-2LL) for Changes in Measure of Lexical Richness (168 cases) with Addition of Exposure to Dutch at School (EDS) as Explanatory Variable. Comparison -2LL Models     ΔΧ2 Δdf p      Model MLR4: β0ijcons + β1ijAge1ij +     656.22         β2Age2ij MLR7: MLR4 + β3EDSij MLR8: MLR7 + β4Age*EDSij 646.97 MLR4 vs MLR7 646.81 MLR7 vs MLR8 9.25 1 .002 0.16 1 .69ns  51 55 59 63 67 71 75 79 83 87 91 95 99 103 Measure of Lexical Richness MLR Score Score 


































































































   114   115   116   117   118