Page 220 - Second language development of newly arrived migrant kindergarteners - Frederike Groothoff

220 Chapter 9 9.5.2 Instruments and procedure Since we did not find a clear-cut effect of school type in our study on our language measures: being in a segregated facility did not emerge as the better option for our language measures, it might be better for all pupils to start in an inclusive facility. However, being in a segregated facility at the start might have positive effects on other outcomes than language measures. For example, it could be that to start in a segregated facility where everybody is learning the new language helps pupils feel safer in the beginning and supports their socio-emotional development. Therefore, we recommend also including the evaluation of the well-being of the newly arrived migrant pupils in future research and the investigation of the long-term effects this has on their development. With our data we have shown how the narratives of 42 newly arrived migrant pupils in their second language developed. However, if we had been able to assess the participants in their L1 as well, this would have provided us with the possibility to see whether or not these L1 stories would mirror the L2 stories. Furthermore, whether the L1 stories would vary based on cultural context of the L1. Although the creators of the MAIN claim that the metric is culturally appropriate, storytelling conventions are in fact influenced by culture. “Substantial cultural differences influence the ways children structure their narratives. Such differences must be carefully distinguished from individual deficits” (McCabe & Rosenthal Rollins, 1994). Unfortunately, this went beyond the scope of this dissertation, however, researchers should be aware of cultural difference and therefore it is valuable to compare stories (from the same children) in different languages in more detail. In previous research with the MAIN, authors, to our knowledge, never mentioned children’s use of communication strategies like acting out or gesturing. In our first assessments we noticed the participants doing this and we started videotaping the sessions. However, these video recordings could not be included in the present study but can be used in future research. The CLASS is a fairly new instrument and recently Buell, Han, and Vukelich (2017) questioned the CLASS’ accountability as an evaluation tool. They suggest that CLASS results should be interpreted based on classroom composition (it seems for example that a higher percentage of boys influences CLASS results), seasonal fluctuations, and the context of the data collection. Since in the present study CLASS has been used only once with each teacher, and the assessments could not take place within one season, differences in CLASS scores between the teachers should be interpreted with caution. In the present study the CLASS and a snapshot method were used once to investigate the first year of education of migrant pupils newly arrived in the Netherlands. Due to practical limitations including limited time, we could not include multiple observations per teacher or per student, especially not when all of the pupils were dispersed over 35 Mainstream schools in the final part of the study. Taking into account that learning