Page 173 - Second language development of newly arrived migrant kindergarteners - Frederike Groothoff

The influence of the school learning environment 173 There was an interaction effect of Positive Climate, although the fixed main effect of Positive Climate was not significant. The significant interaction effect between Age and Positive Climate means that the relation between Age and PPVT scores differs per degree of Positive Climate. However, for the interpretation this interaction effect a word of caution is warranted as the standard errors in the last model appeared to be large. Therefore, we are not sure about the specific values for each parameter. Looking at Table 6.1 in Appendix 6 with the estimated parameters we see that when Positive Climate was introduced into the growth model of PPVT this seems to overshadow the effect of Age. Since the fixed main effect of Positive Climate is not significant we can only say that the influence of Positive Climate is larger for older pupils than for younger pupils. We however do refrain from plotting the model in a graph because this would be difficult to interpret. A larger sample is necessary to confirm the effect. The addition of the variable Regard for Student Perspectives as a fixed main effect did not improve the general model of growth in PPVT. It seems that Regard for Student Perspectives is not a significant predictor of differences in PPVT scores or growth. Therefore, we could not show that there are differences in PPVT scores or growth between pupils in a classroom with a teacher that scored high on Regard for Students Perspectives and pupils in classrooms with a teacher that scored low on Regard for Students Perspectives. 8.2.2 Pedagogical practices: focusing on focal pupil experiences The pedagogical practices focusing on focal pupils’ experiences were obtained by observing the focal pupils with a snapshot method. Due to logistical issues, some observations of the focal pupils took place after the first (or first two) PPVT assessments. The PPVT scores which were obtained before the Snapshot observation were taken out of the analyses. In total 127 cases could be taken into account, and therefore a new developmental model has been modelled with these 127 cases. From the comparison between the consecutive models (see Table 8.3) it is apparent that a model with a fixed linear component – allowing for differences in Age – fit the data better than a model with only an intercept (ΔΧ2 (PPVTII) = 85.85; df = 1; p < .001). However, the variance within individuals did not depend on Age (PPVTIII), while the variance between individuals was a function of Age (ΔΧ2 (PPVTIV) = 9.59; df = 2; p = .01). Finally, adding Age2 to the model did not improve the fit significantly (PPVTv). Hence, in the final model (PPVTDEF ) a fixed effect of Age, as well as a variance between pupils component which depends on Age needs to be included, and with this model we continued the analysis.