Co-Principal Investigators: Papageorge, Nicholas; Lindsay, Constance

Purpose: The goal of this project is to further the understanding of how teachers improve over time—and specifically how white teachers can increase their racial competency and become more effective teachers to students of color—by bridging 3 previously distinct literatures in education research: the returns to teaching experience, teacher peer effects, and the impact of same-race teachers. Specifically, researchers will answer two broad research questions: First, does the racial makeup of a teacher's same-grade colleagues affect their teaching effectiveness or persistence in the classroom? For example, do black colleagues increase white teachers' racial competence and effectiveness teaching black students? Do black teachers benefit from having black peers in terms of teaching effectiveness or remaining in the school or profession? And if so, through what mechanisms do these peer effects operate? Second, do the returns to teaching experience depend on the demographics of the classrooms in which that experience was accrued? For example, does white teachers' effectiveness teaching black students increase with repeated exposure to black students? If so, why does this happen?

Project Activities: The research team will use a mixed-methods research design to answer the proposed research questions. The quantitative phase of the analysis will apply quasi-experimental panel-data techniques to administrative data from the state of North Carolina. For question 1, the team will estimate the effect of Black and Latino peers on white teachers' effects on Black and Latino students' achievement, school attendance, and school suspensions; Black and Latino teachers' overall effectiveness in the classroom, and Black and Latino teachers' absences and retention in the classroom. The team will then investigate possible channels by testing for differential effects by teachers' experience and whether teacher peer effects are persistent over time. For question 2, researchers will test whether the returns to teaching experience vary by the demographic composition of the classrooms in which that experience was accrued, and how similar those classrooms were to the current classroom. Student outcomes include test scores, absences, and suspensions.

Products: The main products will include two academic manuscripts, one on each research question, of the type typically published in high-impact, peer-reviewed journals. A nontechnical summary and discussion of the project's main results will be published as well.

Structured Abstract

Setting: This project will take place in all public elementary schools in North Carolina.

Sample: Participants include all third-through fifth-graders in North Carolina Public Schools.

Factors: For question 1, the factor under study is the racial make-up of teachers' same-grade peers, as measured by either an indicator for "at least one" Black / Latino peer or the share of peers who are Black/Latino. For question 2, the factor under study is a nuanced definition of teaching experience that distinguishes between total years of teaching experience and years of experience teaching in classrooms that are similar, in terms of racial composition, socioeconomic status, and other student characteristics, to the current classroom. Researchers will define similarity in a few different ways to ensure that the results are robust to how this is done (e.g., presence/absence versus count versus proportion of classroom students who are black).

Research Design and Methods: Researchers will use a mixed methods design in which quantitative analyses aim to identify and quantify the association between peers/classroom experience and teachers' capacity to improve student outcomes and qualitative analyses aim to identify the mechanisms through which those associations might operate. Specifically, the quantitative analyses use and expand on existing value-added models of student development in which the research team will reconceptualize how teachers' peer-quality and experience are coded and enter the model; the qualitative analysis is based on teacher focus groups from 12 diverse schools. Researchers will use open-ended interviews to identify how, when, and under what circumstances teachers learn from their peers and learn from the classroom experience.

Control Condition: Given the nature of the design, there is no control condition in this study.

Key Measures: Key learner outcomes include performance on end-of-grade standardized math and reading tests, the count of annual absences, the count of suspensions, and indicators for "chronically absent" and "ever suspended." Educator outcomes, which might represent channels through which learners are affected, include the count of annual teacher absences and indicators for whether the teacher changed grades, changed schools, or left teaching in the following year.

Data Analytic Strategy: The research team will conduct quantitative analyses use linear and poisson multi-way fixed effects regression models and cluster-robust standard errors. The qualitative analysis will have the audio recordings transcribed and will use grounded theory and open coding to identify common themes and ideas that arise organically in the data.