Co-Principal Investigators: Christy Murray, Elizabeth Stevens, and Greg Roberts

Purpose: This project will develop, refine, and pilot a professional development model designed to support middle school teacher implementation of reading practices known to improve literacy for middle grades students.  

Project Activities: The research team will carry out an iterative development process in two phases over two years at three middle schools. Throughout the development process, researchers will (1) engage stakeholders and end-users via monthly school leadership team meetings, bi-monthly or post-iteration advisory council meetings about the components and structure of the PD model; (2) identify and document changes to the model, using formative data; (3) document reasons for changes and the extent to which they may be conditioned on district, school, classroom, or student-level factors; and (4) specify a final model, including any variations due to contextual or student factors. In the final year of the project, the team will test the promise of the fully developed intervention via a pilot study. In this study, the team will randomly assign 6 schools to treatment or control.

Products: Researchers will produce a fully developed intervention for middle school teachers (as well as their schools) and peer-reviewed publications.

Structured Abstract

Setting: Participating schools will be located in rural, urban, and suburban settings in Texas and California.

Sample: Participants will include 3 middle schools (with 57 ELA, science, and social studies teachers and their 1,860 students) during the iterative development phase (Years 1 and 2) and 6 middle schools (with 114 ELA, science, and social studies teachers and their 3,720 students) during the pilot study (Year 3).

Intervention: Components of this year-long, campus-wide intervention will include (1) a well-defined set of research-based practices; (2) high-quality resources that align with the research-based practices; (3) a strategic planning process; (4) ongoing and embedded professional learning; (5) follow-up; and (6) capacity building to support sustainability.

Research Design and Methods: Iterative development will occur in two phases over two years at three middle schools. Throughout the development process, researchers will (1) engage stakeholders and end-users via monthly school leadership team meetings, bi-monthly or post-iteration advisory council meetings about the components and structure of the PD model; (2) identify and document changes to the model, using formative data; (3) document reasons for changes and the extent to which they may be conditioned on district, school, classroom, or student-level factors; and (4) specify a final model, including any variations due to contextual or student factors.

The pilot will utilize an underpowered randomized controlled design in which a new set of schools are randomly assigned to treatment or business-as-usual.

To examine and refine consistency of PD across sites in the final pilot study, a fidelity checklist for the initial and follow-up PD sessions will be drafted during Year 1, refined during Year 2, and finalized for use with the pilot study.

Control Condition: In the control condition, teachers will receive business-as-usual professional development.

Key Measures: Researchers will measure student reading comprehension using the Gates-MacGinitie Reading Comprehension Test. They will measure student risk factors associated with school dropout via student attendance, behavior, and coursework grades. The team will measure fidelity of implementation and instructional quality by coding audio-recordings of teacher-delivered instruction.

Data Analytic Strategy: Researchers will analyze data using multilevel regression to examine the promise of the intervention to improve teacher use of evidence-based teaching practices, students' reading comprehension, and students' risk for subsequent dropout). Model specification will also provide a basis for estimating plausible population-level values for subsequent analyses of statistical power (i.e., for purposes of powering a Goal 3 with treatment at level 3 of the model).