Co-Principal Investigators and Co-Investigator: Ken Bollen (UNC at Chapel Hill); Sean Kelly (University of Pittsburgh); Esteban Aucejo (Arizona State University); Patrick Coate (American Institute for Economic Research)

Purpose: This project will explore how classroom composition (based on student initial achievement, race, socioeconomic status, and gender) is related to different teaching practices and how teaching practice is linked to classroom composition and student outcomes. Researchers will explore how teachers adapt to different classroom compositions, whether these adaptations predict student outcomes, and which teacher characteristics predict adaptation. Findings from this project will have implications for teacher professional development, tracking, and for addressing inequality in student outcomes.

Project Activities: Using the Measuring Effective Teachers (MET) study dataset, researchers seek to leverage the research design leading to this dataset to explore the effects of teaching and teacher adaptivity independent of the common matching of teachers to students found in naturalistic studies. Researchers will use a first-differenced regression model to study how changes in classroom composition affect teaching practice.

Products: Researchers will produce preliminary evidence of potentially promising practices and peer-reviewed publications

Structured Abstract

Setting: The Measures of Effective Teaching (MET) longitudinal database consists of data from 6 large urban school districts (New York City, NY; Charlotte-Mecklenburg, NC: Denver, CO; Memphis, TN; Dallas, TX; and Hillsborough, FL).

Sample: The MET database includes information drawn from over 1500 Grade 4 to 9 teachers and their 44,500 students from nearly 300 schools over 2 academic years (2009/10 and 2010/11).

Intervention: The malleable factors examined in this study include teaching practices and classroom composition.

Research Design and Methods: In this secondary data analysis study, researchers will use the Measuring Effective Teachers (MET) study dataset. In the MET Study, in the first year teachers were not randomly assigned to classroom and then in the second year teachers were randomly assigned to classroom to address possible violations of assumptions of value-added modeling estimates and other approaches to measuring the effects of teaching on student outcomes. In this current study, researchers seek to leverage the research design leading to the MET Study data to examine the benefits of teaching practice and teacher adaptivity independent of the common matching of teachers to students found in naturalistic studies. The longitudinal nature permits exploration of teacher adaptivity to changes in classroom composition.

Control Condition: Due to the exploratory nature of the research design, there is no control condition.

Key Measures: Measures of teaching practice include observational coding using the Framework for Teaching (FFT), Classroom Assessment Scoring System (CLASS), Mathematical Quality of Instruction (MQI), and Protocol for Language Arts Teaching Observations (PLATO) rubrics and the Tripod student surveys. Student achievement measures include math and English language arts assessments. Student effort, enjoyment of class, persistence, and growth mindset will be measured using student self-report surveys.

Data Analytic Strategy: Researchers will use a first-differenced regression model to study how changes in classroom composition are related to teaching practice.