Teacher-Child Relationships, Gender, and Academic Achievement: A Between- and Within-Child Analysis

• In Event: Promising Scholarship in Education: Dissertation Fellows and Their Research

Abstract

A robust body of research has documented effects of teacher-child relationship quality on academic achievement. Several methodological and theoretical concerns, however, may limit application of this knowledge to educational intervention. First, researchers have often relied on cross-sectional analyses to examine associations between early teacher-child closeness and conflict and later achievement. Additional research on the effects of differences in teacher-child closeness and conflict on achievement over time is needed. Second, no studies have examined how gender moderates links between teacher-child closeness and conflict and achievement. Yet, researchers have theorized a “female advantage” in elementary school, noting that girls tend to form more emotionally supportive relationships with teachers, relative to boys (DiPrete & Jennings, 2012). Given these gaps, the current study will answer three research questions: 1) Do teacher-child closeness and conflict relate to greater overall math and reading achievement in elementary school? 2) Are changes in teacher-child closeness and conflict associated with changes in math and reading achievement in elementary school? and 3) Do between and within child effects of teacher-child closeness and conflict on math and reading achievement in elementary school vary by gender?
Data. I used data from the National Institute for Child and Human Development Study of Early Child Care and Youth Development (NICHD SECCYD). See Table 1 for demographics.
Measures. I used the fifteen item Student Teacher Relationship Scale (STRS; Pianta, 1992) to assess teacher-child closeness and conflict in first, third, and fifth grade. I assessed reading and math achievement with two subscales (Letter–Word ID (reading achievement; Applied Problems (math achievement)) from the Woodcock–Johnson (WJ–R; Woodcock & Johnson, 1989). Covariates were child age, child race, maternal education, marital status, maternal employment, maternal IQ, household income, poverty status, number of children in the household, the HOME (Caldwell & Bradley, 1984), teacher experience, and teacher education.
Analytic Approach & Results. I used hierarchical linear modeling to examine the main research questions. I fitted all models with STATA XTMIXED and used grade as a metric of time. All models adjusted for the outcome variable when children were 54 months old. Because data were missing at random, I conducted a listwise deletion procedure to handle missing data. See Table 2 for descriptive statistics.

Author

• Meghan P. McCormick, New York University