Page 98 - Second language development of newly arrived migrant kindergarteners - Frederike Groothoff

98 Chapter 5 the first assessments of the narratives were later than that of the receptive vocabulary assessment, and hence the amount of Exposure to Dutch at School was different. Table 5.3: Descriptive Characteristics of the Participants During the Narrative Assessments at Session 0 (N = participating pupils). Gender Exposure to Dutch at School Boys Girls (%) (%) DL2-school 53.1 46.9 (in months) Mean sd. 4.50 1.29 5.50 2.32 4.74 1.62 (N = 32) Mainstream school (N = 10) Total (N = 42) 40.0 60.0 50.0 50.0 Microstructure and macrostructure Each narrative was digitally recorded (audio and video) for later transcription and analysis. The samples were transcribed using a sample of the conventions of CLAN/CHAT (MacWhinney, 2000). The analysis of the story will be in twofold: both the microstructure and the macrostructure of the stories were analyzed. Microstructure For the analysis of the microstructure, different lexical measures can be investigated (e.g. Iluz-Cohen & Walters, 2012; Muñoz, Gillam, Peña, & Gulley-Faehnle, 2003; Pearson, 2002; Reuterskiöld, Hansson, & Sahlén, 2011; Reuterskiöld-Wagner, Sahlén, & Nettelbladt, 1999; Uccelli & Páez, 2007). Gagarina et al. (2015) suggested ten features that could serve as the basis for analyzing the microstructure of narratives. These suggested features included length and lexis, syntactic complexity and discourse cohesion, and code- switching. For our purpose, however, we only used one of these suggested (namely, Number of Different Words) features and added two (namely, Guiraud Index Score and Measure of lexical Richness), since we focused more on vocabulary than on sentences. Microstructure was analyzed in three parts, namely two measures for lexical diversity (Number of Different Words and the Guiraud Index Score) and one for lexical richness (Measure of lexical Richness). Lexical diversity Lexical diversity was measured by means of Number of Different Words (NDW) and the Guiraud Index Score (GIS). NDW was used because this was reported in many studies using the MAIN (e.g., Lindgren, 2018; Altman et al., 2016; Kapalková et al., 2016; Tsimpli, Peristeri, & Andreou, 2016).