The Effects of a Child’s Gender and First Language on the Latent Profiles Emerging from the inCLASS

• In Event: Poster Session 03
  In Poster Session: PS 03 Topic - Education, Schooling

Abstract

As preschool turns into a typical setting in early childhood education (ECE), it seems essential to better understand the nature of children’s preschool experiences. Indeed, the social interactions and learning tasks to which children are exposed on a daily basis in their educational context are essential to their socio-emotional development and their academic success. The Individualized Classroom Assessment Scoring System (inCLASS) is an American observational tool that assesses the social and behavioral interactions of a three to five-year-old child in his educational context (Downer, Booren et al., 2010). However, certain characteristics of the child, such as the child's gender and first language, might affect the level of engagement and social interactions reported by the inCLASS. The present study aims to determine the latent profiles of the four domains of the inCLASS as well as adding specific characteristics of the individual child, such as the child's gender and first language, which may also influence inCLASS scores. The ultimate goal of this study is to get a more accurate picture of the cultural sensitivity of the inCLASS. The sample included 404 children observed with the inCLASS divided in two groups, 201 children (103 girls) from Quebec (Canadian) sample (M= 38.06 months, SD = 2.32) and 203 children (100 girls) from the French sample (M = 40.52 months, SD = 2.43). A Latent Profile Analysis (LPA) was used to create profiles of children’s engagement with teachers, peers, tasks and negative engagement. One, two-, three-, four- and five-profile solutions were examined. Results indicated a four-profile solution for both French (BIC = 2238,2, ABIC = 2133,7, VLMR and Adjusted LRT p value = .319; entropy = .755) and Quebec (BIC = 2563, ABIC = 2458,6, VLMR and Adjusted LRT p value = .186; entropy = .805). When the predictors of gender and first language were added to the four-profile solution discussed above, it was found that children whose first language is French were more likely to be regrouped in the Middle-Negatively-Engaged profile than in the Typically-Engaged, Middle-Positively-Engaged and Low-Typically-Engaged profiles in the French sample. No significant differences were found regarding language in the Quebec sample. This analysis also revealed that boys were more likely than girls to be regrouped in the Middle-Negatively-Engaged profile than in the Middle-Positively-Engaged profile in the French sample. At last, boys were more likely than girls to be regrouped in the Negatively-Engaged profile than in the Positively-Engaged and the Middle-Positively-Engaged profiles in the Quebec sample. The LPA revealed that more boys were among the profiles with the highest level of negative engagement. Those results suggest a gender bias confirmation in the conceptual framework of the inCLASS, where boys could often be disadvantaged compared to girls, particularly regarding aggressive behaviors (Vitello et al., 2012). Furthermore, having a language barrier in the French sample also had an impact on the children being more likely to be clustered in the lower level profiles. The focus on language and academic competencies in the French educational systems could explain this tendency (OECD, 2004; Garnier, 2013).

Authors

• Precilia Hanan, L'Université du Québec à Montréal (UQAM)
  Presenting Author

• Nathalie Bigras, L'Université du Québec à Montréal (UQAM)
  Non-Presenting Author

• Andreanne Gagne, Universite du Quebec a Montreal
  Non-Presenting Author