|Title:||Coaching to Improve Common Core Aligned Mathematics Instruction in Tennessee|
|Principal Investigator:||Russell, Jennifer||Awardee:||University of Pittsburgh|
|Program:||Continuous Improvement Research in Education [Program Details]|
|Award Period:||3 years (7/1/2014–6/30/2017)||Award Amount:||$2,496,261|
|Type:||Other Goal||Award Number:||R305H140112|
Name of Partners: The Learning Research and Development Center at the University of Pittsburgh (LRDC); the Institute for Learning at the University of Pittsburgh (IFL); Tennessee Department of Education (TDOE)
Purpose: The Common Core State Standards in Mathematics (CCSS-M) set high standards for procedural fluency, conceptual understanding, and reasoning, and call upon teachers to instruct in ways that support students as they work to meet these ambitious standards. As districts and states work toward implementing the CCSS-M, they are revisiting typical professional development for mathematics teachers.
This partnership seeks to improve the in-service training of Tennessee Grades 3-8 mathematics teachers as a route to improving the math skills and test performance of all Tennessee students. TDOE has conducted summer trainings for teachers over the past three years using a “Common Core Coaching Model” (CCM) developed in partnership with the University of Pittsburgh. The partnership now hopes to build out an offshoot of this initiative to create an ongoing state training model that can take place throughout the year. By deepening the training of mathematics coaches, the partnership hopes to achieve a network of highly-trained coaches that can be expanded over time to reach across the entire state. The partnership, with active involvement from the Tennessee Department of Education (TDOE), intends to build systemic capacity within TDOE to support teachers’ enactment of instruction aligned to the CCSS-M standards, and to carry out future improvement work.
Project Activities: The partnership will coordinate the activities of leaders from the TDOE, researchers at the University of Pittsburgh’s Learning Research and Development Center (LRDC), as well as trainers from its Institute for Learning (IFL). During the first two years of the project, the partners will implement the Core Coaching Model—an approach to training coaches developed and implemented by IFL—with 30 “Core Coaches.” After initial implementation, the partners will employ a continuous improvement process to collect data and revise the model. During year three of the project (the pilot study), the partners will implement the revised CCM with 30 new coaches, and compare their students’ performance to students in classes coached by similar non-CCM teachers.
Products: The products of this project will be a fully revised Core Coaching Model and reports for policymakers and practitioners. Peer-reviewed publications will also be produced.
Setting: The project will take place in urban, suburban, and rural districts and schools across Tennessee. Core Coaches will work with teachers across the state in the schools where they teach.
Sample: During the improvement process, the project will train 30 Core Coaches. These 30 Core Coaches will extensively document all coaching activities with two teachers. During the pilot study, the project will train 30 additional Core Coaches and conduct primary data collection with approximately 280 Grades 3-8 mathematics teachers—180 treatment (on average, 6 per coach) and 100 comparison teachers. The 100 comparison teachers will be recruited for the study to demonstrate balance with the treated sample across school contexts. Researchers will also use propensity score stratification methods to create matches between our 180 treated teachers and 1,080 comparison teachers in order to investigate secondary data already collected by the state (a 1:6 ratio representing approximately 10% of the teachers at these grade levels in TN). The project will collect demographic and performance data on approximately 25,000 students taught by these teachers.
Approach: The partnership will build upon the Common Core Coaching Model, which was developed for a series of summer trainings that directly targets improving the capacity of TN mathematics teachers to implement the Common Core in Grades 3 through 8. The CCM incorporates research-based findings regarding the critical role of content (mathematics) and how coaches and teachers learn to improve their practice. It includes three key components: (1) a framework of Essential Coaching Practices (based on IFL’s Content-Focused Coaching program); (2) tools to specify and scaffold coach uptake of the Essential Practices (e.g., a discussion protocol to guide post-observation conferences); and (3) tools to specify and scaffold teacher uptake of CCSS-M aligned mathematics instruction (e.g., lesson guides for aligned task sets). The partnership will test the feasibility and effectiveness of the CCM model in ongoing coaching throughout the school year.
Research Design and Method: This project includes an improvement process as well as a pilot study to investigate the CCM’s promise of efficacy. During the first two years of the project, the team will implement six plan-do-study-act cycles, each lasting three months. During this improvement phase of the project, researchers will assess whether and how the CCM leads to measurable and meaningful changes in how coaches carry out their work. The researchers will measure how usable the coaches find the CCM, as well as the feasibility of the model in different settings across Tennessee. Throughout the project, the researchers will also assess whether the CCM is associated with measurable changes in teacher practice. The pilot study will employ a matched sample quasi-experimental design to assess whether the CCM is associated with improvements in: 1) teacher perceptions of their access to quality coaching; 2a) teacher instruction as measured through cognitive demand of assignment tasks and maintenance of cognitive demand during video observations (n=280 teachers); 2b) teacher instruction as measured through items on an annual survey (n=1,160 teachers); and 3) student learning on the state accountability test.
Control Condition: Comparison teachers recruited for primary data collection will be sampled from schools with access to coaches in order to compare the effects of CCM against “business-as-usual” coaching for our matched comparison sample. Comparison teachers for our secondary data analysis may have received coaching from a non-CCM coach or no coaching at all, and students will have received instruction from a teacher who has received non-CCM coaching or no coaching at all.
Key Measures: To measure usability and feasibility of the CCM for coaches and teachers, the partnership will collect coaching logs, transcripts from coaching sessions, instructional artifacts, teacher surveys, and videotapes of instruction. LRDC and IFL partners will code these data sources to build analytical measures indicating the quality of coaching and instructional events. To measure student performance, the partnership will collect student classroom work and administrative data on student performance on the state-mandated math assessment.
Data Analytic Strategy: During the improvement process phase, the partnership will analyze incoming data to make refinements to the CCM. Ongoing analysis of measures of coach and teacher practice will explore trends within and across coaches over time, including, for example, descriptive statistics of survey items, run charts to examine variation in coaching processes over time (e.g., time between classroom observations), and qualitative analysis of transcripts of coaching sessions and classroom discourse. In the pilot study, researchers will utilize in-depth instructional data from approximately 280 teachers (180 treated teachers and 100 matched comparison teachers) in one set of analyses and will utilize secondary data from existing data collection efforts in a second set of analyses with approximately 1,160 teachers (180 treated teachers and 1,080 matched comparison teachers). The research team will examine a series of regression/multi-level hierarchical linear models to examine the effects of the CCM on proximal outcomes (i.e., coaching intensity, coaching quality, and instructional practice) and on distal outcomes (i.e., student learning on state tests).