Page 141 - Second language development of newly arrived migrant kindergarteners - Frederike Groothoff

Pedagogical practices: focus on teacher behavior 141 The first conclusion of this chapter is about the overall pattern of teacher behavior in the kindergartens in the present study. Overall, the teachers in this study showed a comparable pattern in their scores over the ten dimensions which were found in other studies (cf. Mashburn, Justice, Downer, & Pianta, 2009; Pakarinen et al., 2010; and the Dutch studies: Leseman & Veen, 2016; Henrichs & Leseman, 2016; Veen et al., 2017). The highest score on the CLASS was found on Negative Climate, meaning that there was very little negativity in the classrooms in this study. Buell, Han, and Vukelich state that “in general, classrooms score highest in the Emotional Support domain, and lowest in the Instructional Support domain” (Buell et al., 2017, p. 1636). We found this similar pattern in our data, highest scores on Emotional Support and lowest score for instructional support, with no significant difference in the pattern between the two school types. Even though the general pattern of the observed quality mirrors that of previous studies, the teachers in our study seem to have overall lower CLASS scores than in international (Mashburn et al., 2009; Pakarinen et al., 2010; Osborn, 2012) and national studies (Henrichs & Leseman, 2016; Veen et al., 2017; Slot et al., 2017). The difference between our study and previous studies was the fact that we looked at classrooms with newly arrived migrant pupils, and thus beginner learners of a second language. A possible explanation for the lower process quality in our study could be that teachers have different, maybe lower, expectations for newly arrived migrant kindergarteners compared to monolingual pupils and therefore that they might set lower goals, ask simpler questions, and want to have more control on the process as a teacher instead of giving pupils the opportunity to show initiatives. To illustrate this, a necessary precondition for the use of rich language by the teachers, and the occurrence of, for example, back and forth conversations (indicators for Language Modeling), is the expectation that the teachers will be understood by the pupils in class and can elicit output from them. It might be so that when teachers do not expect to be understood by the pupils or when they expect that pupils do not have enough Dutch vocabulary to respond, they simplify their speech, using shorter sentences and more frequent words or demand other replies and actions from the pupils. To score higher on process quality teachers could instead have used language that is rich, with a lot of scaffolding around it to make the language also understood by pupils with lower proficiency in Dutch. As can be seen in the boxplot of the observational data in Appendix 4, there was a lot of variation in teachers’ scores, also within each school type and even between teachers within one school. It would we recommendable that there is constant professionalization based on observations in the classrooms, time for teachers to discuss the observations, and time for teachers to visit colleagues who can show good examples. A tool such as the CLASS can function as a guide line with all the indicators of good practices for each dimension. For example, a school or a teacher can choose one dimension to improve during