Page 127 - Second language development of newly arrived migrant kindergarteners - Frederike Groothoff

Development of narrative ability 127 It was possible to build a General Development Model for five out of six narrative ability measures. Table 5.18 summarizes the significant effects of the variables on the five narrative aspects. Of the 20 possible significant effects, 12 appeared to be significant. Table 5.18: Overview of Significant Child Characteristics on the Narrative Ability Models. Microstructure Macrostructure NDW GIS MLR SS IST Age +++++ Age2 ++ Exposure to Dutch at School +ac +a +ac +/- Educational Facility -b NDW = Number of Different Words; GIS = Guiraud Index Score; MLR = Measure of Lexical Richness; SS = Story Structure; IST = Internal State Terms. +: positive effect; -: negative effect; +/-: positive fixed effect and negative interaction effect. a Only the fixed main effect of the variable was significant. b Only an interaction effect with Age was found. c When Exposure is added to the model, Age seemed to be no longer significant. Age is significant for five out of six narrative measurements, meaning that there is a development with age for these narrative ability measures, as was predicted based on for example Gagarina (2016) and Bohnacker (2016). For NDW, GIS, and IST this development is linear, while the growth rate for MLR and SS decreases over time. The increase in MLR and SS score is larger for younger pupils than for older individuals. For none of the measures a ceiling effect was observed. Beside Age, Exposure to Dutch at School seemed to be an important factor as well. It influenced all measures except for the number of ISTs. For NDW and MLR the effect of Exposure to Dutch at School even seemed to be of more importance than the effect of Age. The longer a pupil attended a Dutch school, the higher his or her score on the different measures. For SS there was also an interaction between Age and Exposure to Dutch, thus the influence of Exposure of Dutch at School was larger for younger pupils than for older pupils. Altman et al. (2016) found in their study that length of exposure had no influence on SS. However, it is difficult to directly compare their outcomes with the participants in the current study. That is, in the Altman et al. (2016) study, the mean months of exposure was 45.26 (sd. 20), while the participants in our study had a mean of only 29.2 months of exposure (sd. 1.9) by the final session. Nevertheless, in the present study it seems that there was an interaction between Age and Exposure to Dutch at School for SS. Therefore, it might have been that if we had followed our participants for a longer period we would have reached the same conclusion as Altman et al. (2016).