A Provisional Agenda for Research on Improving the Quality of Mathematics Teaching on a Large Scale Paul Cobb and the MIST Team Vanderbilt University University of Washington Michigan State University McGill University Purpose • Outline a set of issues that need to be addressed if research is to provide adequate guidance for large-scale instructional improvement efforts in mathematics – Across a large urban district Overview • Preamble: what counts as high-quality instruction • Background: ongoing work as a setting for appreciating the limitations of current research • Proposal for issues that need to be addressed What Counts as Instructional Quality? • Has to be justified in terms of students’ learning of mathematics that is worth knowing – Conceptual understanding as well as procedural fluency – Justifying solutions, evaluating the reasonableness of solutions, generalizing from solutions, making connections among multiple representations of mathematical ideas Research on Students’ Mathematical Learning • Rigorous mathematical tasks • Individual or small group work • Whole class discussion – Teacher presses students to: • Explain and justify their reasoning • Make connections between different solutions Goals for Teachers Learning • High-leverage instructional practices – Planning and conducting productive whole class discussions – Setting up rigorous mathematical tasks • Specific types of knowledge implicated in the enactment of these practices – Mathematical knowledge for teaching – Vision of high-quality mathematics instruction – View of students’ mathematical capabilities Challenge for Districts • How to organize, support, and press for teacher learning across the entire system – What guidance can research provide? Background: MIST Project • Four-year collaboration with four large urban districts – 360,000 students – 2007-2011 • Continued collaboration with two districts – 180,000 students – 2011-2015 • Investigate (and support) the districts’ instructional improvement efforts in middlegrades mathematics Background: MIST Project • High proportion of students from traditionally underserved groups – Limited financial resources – High teacher turn over – High proportion of novice teachers • Atypical in one respect: – Aiming at ambitious goals for student learning and thus for teachers’ instructional practices District Participants • 30 middle-grades mathematics teachers in 610 schools in each district • Mathematics coaches • School leaders – Principals, assistant principals • District leaders – Across central office units that had a stake in mathematics teaching and learning Collaboration with Districts October • Interview district leaders to document current strategies for improving middle-school mathematics JanuaryMarch • Audio-recorded interviews with the 200 participants to document how the districts’ strategies are actually playing out in schools and classrooms Collaboration with Districts FebruaryMay • Analyze transcripts of the 200 interviews • Identify and explain gaps between each district’s intended and implemented improvement strategies • Develop a detailed report for leaders in each district • Shared findings and made actionable recommendations • Meet with district leaders to discuss our findings and recommendations May Collaboration with Districts • District leaders attempt to act on our recommendations to a significant extent • Become co-designers of district improvement strategies – Participants in as well as observers of the districts’ instructional improvement efforts Collaboration with Districts • Formulating recommendations: Have to address concrete organizational design problems • Occasion to appreciate – The types of problems that district leaders have to address – Extent to which current research can provide guidance – hence this talk Research Goal • Develop an empirically grounded theory of action for instructional improvement at scale – Can inform other districts’ instructional improvement efforts Ongoing Analyses • Initial conjectures about supports and accountability relations – Drew on then available literature • Conjectures informed initial recommendations to districts • District leaders acted on recommendations – opportunity to test and revise conjectures Retrospective Analyses • On-line surveys for teachers, coaches, and school leaders • Video-recordings of two consecutive lessons in the 120 participating teachers’ classrooms – Coded using the Instructional Quality Assessment (IQA) • Assessments of teachers’ and coaches’ Mathematical Knowledge for Teaching (MKT) • Video-recordings of select district teacher professional development • Audio-recordings of teacher collaborative planning time • Student achievement data Theory of Action for Instructional Improvement at Scale • A coherent instructional system: – Instructional materials + professional development + assessments to inform instruction + additional supports for struggling students • Mathematics coaches’ practices in providing jobembedded support for teachers’ learning Theory of Action for Instructional Improvement at Scale • School leaders’ practices as instructional leaders in mathematics • District leaders’ practices in supporting the development of school-level capacity for instructional improvement Research Team Paul Cobb Kara Jackson Ilana Horn Tom Smith Erin Henrick Ken Frank Research Team Jessica Rigby Jonee Wilson Brooks Rosenquist Britnie Kane Brette Garner Emily Kern Mahtab Nazemi Mollie Appelgate Adrian Larbi-Cherif Charlotte Munoz Jason Brasel Seth Hunter Megan Webster I-Chien Chen Explicit goals for students’ mathematics learning Vision of high-quality instruction Teacher professional development Instructional materials Component 1: Coherent Instructional System (Bryk et al., 2010; Newmann et al., 2001) Additional supports for struggling students Assessments to inform ongoing instructional improvement Explicit goals for students’ learning Vision of high-quality instruction Instructional materials Component 1: Coherent Instructional System Vision of high-quality instruction: • Small set of high-leverage practices that are potentially learnable in the context of highquality professional development Instructional materials: • Grounded in student learning trajectories that aim at significant mathematical ideas • Pull-out Professional Development (PD): • Specific PD designs – promising findings • Grounded in classroom practice – pedagogies of investigation and enactment • Most work in pre-service – have extrapolate to in-service • Teacher Collaborative Time (TCT) • Most researcher-led – potentially productive types of activities • Naturally occurring – characteristics of productive teacher groups Teacher professional development Component 1: Coherent Instructional System Component 1: Coherent Instructional System Assessments to inform instruction • Formative assessment systems – Aligned with ambitious goals for students’ learning – Grounded in trajectories of students’ learning • Goal: support struggling students to participate effectively in mainstream instruction Component 1: Coherent Instructional System Additional supports for struggling students Coherent Instructional System • Collaborating districts: fragments of a coherent instructional system – Strengths: explicit goals for students’ learning, vision of high-quality instruction, instructional materials – Challenge: teacher professional development – district capacity – Challenge: TCT – expertise + leadership of meetings – district capacity – Weakness: additional supports for struggling students – not aligned with mainstream classroom instruction Needed Research: Developing District Capacity • Researchers typically assume full responsibility for “building” particular elements • The problem of scale involves supporting districts’ development of the capacity to create, coordinate, and sustain the elements of such a system Developing District Capacity: Sacrificial Offering • Example: co-designing and co-leading PD for coaches with district mathematics specialists – Support the development coaches’ capacity to design and lead high-quality teacher PD – Gradual hand over of responsibility to district mathematics specialists • Overall goal: Investigate how to support districts’ development of capacity to develop and sustain a cadre of mathematics coaches Needed Research: Interrelations Between Elements of the System • Current research typically focuses on the individual elements of a coherent instructional system • Also need to investigate interrelations between various the elements, and between elements and other components of ToA – Which are preconditions for the development of other elements/components? Developing District Capacity: Sacrificial Offering • Example: co-designing and co-leading PD for school leaders – School leaders press for instructional improvement – Coaches support teachers in meeting those expectations • Investigate development of aligned support and press for teachers’ improvement of their instructional practices Mathematics Coaching • Finding: Teachers’ improvement of their instructional practices depends crucially on their access to colleagues who have already developed accomplished practices • Three of our four districts: Small proportion of accomplished teachers • Critical role of coaches as more accomplished colleagues Mathematics Coaching • Design and lead pull-out PD • Work with groups of teachers during TCT – Current research + our findings indicate importance of leadership/expertise in TCT • Support teachers one-on-one in their classrooms – Build on pull-out PD and work in TCT Current Research on Coaching • Provides little guidance on: – Types of activities in which coaches might engage teachers – Coaches practices as they enact these activities – Supporting the development of a cadre of accomplished mathematics coaches • One of the collaborating districts: four years of sustained professional development for coaches • Only slightly ahead of the teachers they were expected to support Needed Research: Delineating Goals for Coaches’ Learning • Develop testable conjectures about potentially productive coaching practices • Working with groups of teachers: • Draw on studies of researcher-led and naturally occurring PLCs – Kazemi and Franke: Potentially productive activities – Horn: Press teachers of key issues Needed Research: Delineating Goals for Coaches’ Learning • Working with teachers in their classrooms: – Draw on research on teacher learning and professional development • Points to importance of modeling and especially coteaching • Observation/feedback at specific points in teachers’ development – e.g., Tuning their enactment specific practices Needed Research: Delineating Goals for Coaches’ Learning • Important to explicate the forms of knowledgeability implicated in the enactment of proposed practices – Researchers do not typically report what they needed to know Needed Research: Specifying Forms of Knowledgeability • Mathematical knowledge for teaching (MKT) • Additional candidate: an envisioned trajectory for teachers’ learning – Classroom management – Student engagement – Teacher questioning • Informs decisions about which aspects of practice to work on with teachers Needed Research: Designs for Supporting Coaches’ Learning • Both settings: – Draw on very limited literature on coach PD – Extrapolate from work on teacher learning and teacher PD Needed Research: Test and Revise Designs for Coaches’ Learning • Design experiments to: – Test and refine conjectures about supports for coaches’ development of target practices – Assess in terms of improvements in: • Coaches’ practices • Quality of teachers’ classroom instruction • Student learning Coach PD: A Sacrificial Offering MIST & District Math Leaders collaboratively plan for upcoming session Coach PD Session MIST views videorecording of pilot PD in light of goals for coaches’ learning Coaches lead PD with pilot group School Instructional Leadership • Standards-based reform: Principals and assistant principals increasingly expected to act as instructional leaders in specific content areas – Manage instruction rather than manage around instruction Current Research School Instructional Leadership • No consensus on what school leaders need to know and be able to do in order to be effective instructional leaders in mathematics – General content-independent characteristics of high-quality instruction • Observe instruction and provide feedback – MKT, student mathematical learning, high-quality mathematics instruction, teacher learning • Coach mathematics teachers Findings • Interviews – vision of high-quality mathematics instruction (VHQMI) – Form rather than function views – Consistent with teachers’ accounts of the feedback they receive from school leaders • Extensive professional development – Focused on general, content-independent characteristics of high-quality instruction Initial Findings • General characteristics of high-quality inquiryoriented instruction – Too abstract – not able to connect to concrete instructional practices • MKT, student mathematical learning, highquality mathematics instruction, teacher learning – Beyond the capacity of most districts Needed Research: Delineating Goals for School Leaders’ Learning • Develop testable conjectures about potentially productive school leadership practices • Justify in terms of: – Direct support/press for teachers’ learning – Indirect support – developing conditions for teacher learning Current Bets • Identify and capitalize on instructional expertise in the school • Observe mathematics instruction and provide feedback • Participate in teacher collaborative time (TCT) • Support coach to support teachers’ learning School Leadership Routine Attend Teacher Collaborative Time Observe Classroom Instruction Meet with Mathematics Coach School Leader And Coach: Quality of Individual Teachers’ Instruction + How to Support Instructional Improvement + Jointly Plan for Teacher Collaborative Time Needed Research: Delineating Goals for School Leaders’ Learning • Important to explicate the forms of knowledgeability implicated in the enactment of the proposed practices – Vision of high-quality mathematics instruction • Observing instruction and giving feedback • Identifying and leveraging instructional expertise Needed Research: Designs for Supporting School Leaders’ Learning • Extrapolate from work on teacher learning and teacher PD – Vision of high-quality mathematics instruction • Distinguish between high- and low-rigor tasks • Distinguish between strong and weak enactments of specific high-leverage instructional practices Needed Research: Test and Revise Designs for School Leaders’ Learning • Design experiments to: – Test and refine conjectures about supports for school leaders’ development of target practices – Assess in terms of improvements in: • • • • Instructional leaders’ practices Direct and indirect supports for teachers’ learning Quality of classroom instruction Student learning Coaching and School Instructional Leadership • Important to take account of relations between members of different role groups – Coaches’ effectiveness in supporting teachers’ learning depends on relationship with school leaders – Coach is a potential support for school leaders’ learning District Leadership • Also important to take account of coaches’ and school leaders’ relations with district leaders – Supports for their learning – What they are held accountable for • Potential tension between raising student achievement in the short term and improving the quality of classroom instruction in the long term Why Does Current Research Provide Only Limited Guidance? • Math education, teacher education, and the learning sciences: – Student learning, instructional activities and tools – Teaching, teacher professional development, teacher collaborative time • Typically bracket out the school and district contexts in which teachers’ learning occurs Why Does Current Research Provide Only Limited Guidance? • Educational policy and leadership – Typically instructionally agnostic • Content + vision of high quality instruction matter – Research in policy and leadership can be relevant • Have to read through lens of what counts as highquality mathematics instruction Supporting Instructional Improvement at Scale • A problem of both teacher learning and organizational learning – Need to specify goals for organizational as well and teachers’ learning • Differentiate between organizational change and organizational learning Organizational Learning • Conjectured practices, social relations, tools, and routines are provisional goals for organizational learning – Provisional goal for the organizational learning of a district: The creation and ongoing refinement of a coherent instructional system – Provisional goal for the organizational learning of a school: Establishment of the school leadership routine Papers and instruments downloadable at: http://vanderbi.lt/mist