Poster #34 - Teacher-Child Closeness and Kindergarten Children’s Internalizing Behavior on Academic Outcomes: A Multilevel Model

• In Event: Poster Session 02

Abstract

Educators have long recognized the importance of internalizing behavior in influencing children’s academic outcomes. Children’s internalizing behavior (e.g., anxiety, social withdrawal, depression) has been associated with academic underachievement (Flook et al., 2005; Kovacs & Devlin, 1998), deficient problem-solving skills, and repercussions that can extend into adulthood (Nock & Kazdin, 2002; Roza et al., 2003). High quality teacher-child relationships can reduce the risks associated with internalizing behavior problems on children’s development, particularly for kindergarten students, who are transitioning to formal schooling and are at a critical time in their cognitive development (Morrison et al., 2019; O’Connor et al., 2011).

While research has investigated the influence of teacher-student relationship quality on academic outcomes (e.g., Flook et al., 2005; Schmitt et al., 2018), less research exists regarding the between-school differences among teacher-child relations and internalizing behavior on academic achievement. However, investigating between-school variance is important given the many contexts (e.g., peer dynamics, school culture) that exist among schools, which could influence the associations teacher-child relationships and internalizing behaviors problems have on a child’s academic achievement.

To address this gap, this study uses multilevel modeling on a large, nationally-representative dataset to examine potential school level associations between teacher-child closeness and kindergarten students’ academic outcomes, as well as the association between academic outcomes and internalizing behavior. Study uses the Early Childhood Longitudinal Study, Kindergarten Class of 2010-11 (ECLS-K:2011), which followed a nationally-representative sample of students from kindergarten through fifth grade and collected data from children, parents, and teachers (response rates 74%-89%). The analytic sample (9,261 children nested within 607 schools) only included students in the spring kindergarten wave without missing values in our variables of interest, and students in schools where 10 to 30 students were sampled in the ECLS-K:2011 data (see Table 1).

Table 2 presents the coefficients from our multilevel models that explain kindergarten spring academic outcomes. Compared with the null model, the full model (Model 2), provided a better fit to the data as it reduced the spring kindergarten’s math and reading between-school variance by 24% and 10%, respectively. Additionally, the residual variance fell by 10% for math and 5.6% for reading when comparing the full and null models. In accordance with expectations, our findings show that internalizing behavior and teacher-child closeness have a negative (Model 2 in Tables 2, 𝛾10) and positive (Model 2 in Tables 2, 𝛾20) association with kindergarten academic outcomes, respectively. However, counter to expectations, the association of teacher-child closeness with academic outcomes varies by school while internalizing behavior does not.

Next steps include more robust analyses such as using the restricted-use ECLS-K:2011 data, applying multiple imputation to address missingness, and employing multilevel structural equation modeling to examine directionality. Such analyses may help inform policymakers in improving teachers’ relationships with their students (e.g., professional development) across all schools to ensure a more equitable schooling experience.

Authors

• Jeffrey Yo, University of California - Los Angeles
  Presenting Author

• Eric Burnheimer, University of California - Los Angeles