The Classroom Assessment Scoring system in a 5-Year-Old kindergarten in Quebec: Observation and Interview with the teachers.

Classroom Assessment Scoring System [CLASS] (Pianta et al., 2008) is an observation tool to measure the quality of classroom interactions, a key element of academic success (Sabol et al., 2013). While this tool is widely used in different studies worldwide (Mashburn et al., 2008. Pakarinen et al., 2010), we currently know very little about how teachers define the quality of those interactions. However, it is important to study these representations because they tend to influence the teaching practices (Ahn, 2005; McMullen et al., 2005; McMullen et al., 2006; Stipek et al., 1997; Wishard et al., 2003). Consequently, the objectives of this research is 1) to evaluate the quality of the interactions in a 5-Year-Old kindergarten in Quebec (Canada) with the CLASS tool, in a francophone environment that differs from the US context; 2) to examine how teachers perceive those interactions in the classroom in order to draw a parallel with observational data. Data were collected during the winter of 2014 in 12 classrooms of the Commission scolaire de la Capitale (Quebec) (N = 118 children). As previously observed in US studies (Center for Advanced Study of Teaching and Learning, 2013), descriptive analyses indicate that the quality of interactions in kindergartens is moderate (between 3 and 5), except for "concept development" which yields low results (M = 2.70, SD = 0.69 [see table 1]). The analysis of semi-structured interviews with the TAMS software report on occurrences related to the dimensions of CLASS emerging from the teachers’ representations. Figure 1 shows that "concept development" yielded the lowest level of quality in observed measures, and it is also the dimension that was least mentioned by the teachers during the interviews. More specifically, several elements of the sub-dimensions related to "concept development" were not mentioned, and the sub-dimension "creation" never emerged in the teachers’ representations.In conclusion, these results can be used as a spr