Skip to navigation Skip to content Skip to footer

Digital Promise

Menu

Programs

Current

AIMS EduData

Part of Learning Sciences Research

Advancing Innovative Math Solutions (AIMS) EduData seeks to build sustainable and scalable research infrastructure to advance math learning.

Audiences

Researchers

Learning Environments

K-12

Now accepting applications!

Overview

The Advancing Innovative Math Solutions (AIMS) EduData Initiative, funded by the Gates Foundation, seeks to build sustainable and scalable research infrastructure to advance math learning. AIMS EduData especially encourages applications from graduate students and early career professionals, educators with research roles (such as those working in district research offices), and researchers located at minority-serving institutions or Title I school districts. Successful applicants will conduct original research using data sets provided by leading digital learning platforms (DLPs) and participate in a growing community.

Participating Digital Learning Platforms

Learn more about the datasets.

AIMS EduData Phase II RFP - Apply by December 14, 2026

For education researchers, digital learning platforms and educators

  • Benefits for successful applicants: 
    • Access to deidentified datasets and research tools
    • Opportunities to study student learning, engagement, motivation, persistence, and instructional practice
    • Involvement in a dynamic and supportive community of digital learning platforms and researchers
    • Awards ranging from $50K – $60K

ACCESS THE FULL REQUEST FOR PROPOSALS

Submission Checklist

Before you submit your application, be sure to complete this checklist:

  • A project narrative not exceeding 2000 words, covering: abstract/summary; significance; research design and analytic plan; career and partnership growth; and resources
    • For Second Approach: Integration of Carnegie Learning’s UpGrade tool: include sample system data (can be synthetic – the goal is to determine that your system is instrumented to capture student outcome metrics) 
  • References
  • Resume or CVs for the principal investigator (PI)/project leader, and any other key staff
  • If the PI is a graduate student or postdoc, include a letter of commitment from a faculty advisor at your institution. The advisor should commit to helping navigate IRB approval, data sharing agreements, and the execution of grant agreements and receipt of funds. The letter must specify a concrete time commitment. 
  • A letter of support from your current education partner confirming their intention to participate in this research during the 2026-2027 school year.
  • Budget in the template provided, along with any additional justification (can include additional narrative outside the 2000 word limit if necessary). Download a copy of the template and save your version with the file name convention PI LAST NAME_BUDGET.XLS

Office Hours

  • Carnegie Learning: 
    • Register | Thursday, December 10 at 11 am EST
  • EDSI Benchmark:
    • Register | EDSI Multimodal Math Classroom Dataset – Introduction & Opportunities Webinar | Wednesday, November 18 at 1 pm EST
    • Register | Friday, November 20 at 4 pm EST
    • Register | Wednesday, December 9 at 10 am EST

Frequently Asked Questions

Can I submit or be named on multiple applications?

We will not accept multiple applications from the same Principal Investigator.

Can multiple teams from my organization/department submit applications?

Yes, we will accept multiple applications from organizations and
departments, as long as they have different Principal Investigators.

Can I apply if I am an independent researcher not affiliated with a university, non-profit research institution, or a school district?

Yes. Research must be conducted within the U.S., but there are no
restrictions to affiliation. Keep in mind that you must be able to
demonstrate your ability to comply with data security requirements and
obtain IRB approvals, if applicable.

Can my application include personnel outside the U.S. (e.g., as a consultant or mentor)?

All researchers should be located in the U.S. to avoid complicated international payments.

Is international travel allowable (e.g., to international conferences)?

Yes, as long as the event is justified.

Is there an overhead rate cap?

Yes, the Gates Foundation limits overhead expenses to 15%. See https://docs.gatesfoundation.org/documents/indirect_cost_policy.pdf

How many grants will be awarded this round?

We expect to award up to 7 grants in this round.

When will there be another round of funding opportunities?

We anticipate opening another opportunity to join the AIMS EduData initiative in January 2027.

Can funds be used to compensate participants (e.g., teachers, students) for their time?

Yes, although we expect that most research projects at the $50,000 or below level will primarily use secondary datasets

Can projects be extended beyond the initial funding period?

Planning grants and small research projects may lead to larger funding opportunities. We may allow no cost extensions on a case-by-case basis under unusual circumstances, but proposed projects should be achievable within the time frame permitted and should include timelines.

Are Letters of Support required from collaborators?

Yes, collaborators (such as mentors, research assistants, and partnering districts/schools) should provide letters of support, including the level of effort they are prepared to commit.

What expenses are allowable for planning grants?

Planning grants should be used to cover expenses directly related to proposal development. These expenses may include personnel time and travel to meet with potential partners. We do not expect that conference attendance would be included in these expenses except under unusual circumstances (if you feel that attending a conference is important to your proposal development, please explain why).

What type of “graduate student” is eligible to apply?

Graduate students who are currently enrolled in doctoral programs are eligible to apply.

Do you have to apply through your institution’s office of research?

No, though you’ll need to demonstrate IRB approval or a letter of determination and meet data security requirements.

To what extent does the curriculum vitae of the principal investigator influence your evaluation or decision-making process?

We would like to see evidence that you have the skills needed to do the proposed work, and a track record of related work that shows that skill. As noted, we encourage early career scholars to apply.

To what extent must research questions align with the DLP's priority topics, and will the DLP provide support during the initial exploration of datasets (i.e., determining feasibility for analyses) if funded?

Your research questions do not necessarily need to align with the priority topics, if you have a very compelling idea. DLPs will be looking for ideas that could advance how their platform serves students. Attending a platform’s office hours is a great way to ask about alignment or request data from the DLP catalog. Also for DLPs that provide a free, open-access system (e.g., Khan Academy), we recommend exploring it to get familiar with the platform.

For the small grant research projects ranging from $10,000 - $50,000, will a researcher need to have an already existing partnership with a district or can the researcher plan to submit an application on their own?

For the Teachley Math and RightOn Education, a proposal needs to demonstrate a partnership exists or will by the time the funding will be awarded. In other RFPs (e.g., Fall 2026) you can submit an application for a small research grant without a district partner provided you the proposed project can be completed without district data.

What are the key expectations for planning grants regarding deliverables and milestones?

The expectation for planning grants is to develop a complete research proposal with a district partner letter of agreement in place. Awardees will have three months to gain experience working with a DLP’s data set and develop a relationship with a school district. Winning a planning grant doesn’t guarantee access to the same funding source for a larger grant, but it helps build a foundation for pursuing bigger grants from various sources.

Can conferences that fall outside the grant timeline be included in the budget?

Probably, although you would need confirmation from the DLP, if a grant is awarded.

Awarded Projects - Spring 2026

Formative Data, Observation, and Classroom Understanding Study (FOCUS): A Research-Practice Partnership

Abstract: The Partnership for Los Angeles Schools (Partnership), in collaboration with Student Achievement Partners (SAP) and EdLight, proposes a 13-month research initiative to examine how a structured, continuous improvement cycle grounded in formative assessment informs educator instructional decision-making and supports asset-based mathematics teaching. Building on a district-wide professional learning initiative launched in 2024–25, this project will engage teachers across three Partnership schools in a structured improvement cycle. The cycle pairs the EdLight platform, which analyzes handwritten student work using AI, with the California Math Framework Learning Walk tool* to collect classroom-level instructional data. The study addresses a priority question for the Partnership: How do teachers interpret and use formative assessment data, made visible through student work and technology, to inform instructional decision-making in ways that center student mathematical thinking — particularly for historically marginalized learners?

Principal Investigators:

Dr. Alfonso Romero is the Senior Director of Secondary Math at the Partnership for Los Angeles Schools.

Seeing Student Thinking: Using AI-Assisted Formative Assessment to Reduce Mathematical Misconceptions with a Focus on Algebraic Thinking in Grades 6-8 in the School District of Philadelphia

Abstract: Seeing Student Thinking investigates how EdLight’s AI-assisted formative assessment platform supports grades 6–8 mathematics teachers in interpreting student thinking, addressing misconceptions, and improving instruction in the School District of Philadelphia. Led by ImpactSTATS Inc., a national research and evaluation nonprofit, the mixed-methods study will examine the implementation of the platform and its influence on teacher decision-making, instructional practices, student engagement, and mathematics persistence.

Participating educators will engage in Professional Learning Communities (PLCs) with fellow math teachers to collaboratively analyze student work, identify instructional strategies, and reflect on patterns in student thinking. The study will explore how AI can be responsibly integrated into mathematics classrooms to better support educators, improve student learning, expand access to high-quality instruction, and improve outcomes for historically underserved students.

Principal Investigators:

Melodie Baker's headshot

 

 

 

 

Melodie Baker, Ph.D. is the Founder and Executive Director of ImpactSTATS Inc., a nonprofit research and policy organization that uses data, research, and evidence-based strategies to advance equitable education systems.

Kristian Monet's headshot

 

 

 

 

Kristian Monet is a Project Coordinator with ImpactSTATS Inc. and a junior Computer Science major at Howard University with a focus on data science.

Claire E. Cameron's headshot

 

 

 

 

 

Dr. Claire E. Cameron teaches mixed methods, doctoral student writing, and child development online.

Chris Rates' headshot

 

 

 

 

Chris Rates has a PhD in Science Education from the University of Virginia, where he gained expertise in complex systems.

Scanning for What Works: Validating Optimal Conditions for AI-Assisted Formative Assessment

Abstract: Every school evaluating EdLight has been asked to judge it through whatever camera happened to be in the room, usually a Chromebook tilted at the wrong angle under bad light. That’s like testing a precision audio tool on speakers built for something else and concluding the tool is weak. KIPP NorCal proposes a two-year study to fix the question itself: under what hardware configurations and implementation conditions does EdLight actually maximize its impact on formative assessment and student math outcomes, especially for students with IEPs, English Learners, and students reading below grade level?

Using KIPP NorCal’s three-tier validation model (Regional AI Task Force to micro-pilot to Catalysts), Year 1 runs a controlled micro-pilot at one school, with trained teachers testing rapid scanners, document cameras, and Chromebook cameras side by side while we measure AI diagnostic accuracy, teacher adoption, and student growth across conditions and subgroups. Year 2 scales the winning configuration through 17 Catalyst teacher-leaders. The deliverable is an implementation playbook (the EdLight Launch Kit) that any similarly structured network can deploy, plus openly licensed findings, making this the first study to treat hardware as an independent variable in AI-assisted formative assessment.

Principal Investigators:

Jennie Dougherty's headshot

 

 

 

 

Jennie Dougherty is the Director of Strategic Initiatives at KIPP NorCal.

Poorvi Goradia Iyengar's headshot

 

 

 

 

Poorvi Goradia Iyengar is the Associate Director of Academics at KIPP Public Schools Northern California and supports 5-8 Math and 5-12 Science implementation across the region.

Misconception Trajectories as Early Indicators of Mathematics Risk: Testing the Predictive Value of Feature-Engineered Cognitive Error Patterns in EdLight

Abstract: Early warning systems in digital learning platforms typically rely on behavioral engagement indicators such as assignment completion, time on task, and click patterns. While these measures detect disengagement, they overlook students who remain behaviorally compliant while accumulating conceptual misconceptions that predict later mathematics failure. This limitation has instructional and equity implications, particularly for students whose persistence masks fragile understanding. This study tests the predictive value of feature-engineered cognitive error patterns derived from EdLight’s longitudinal, AI-annotated corpus of handwritten middle school mathematics work. Using secondary data from grades 6-8, we will engineer temporal features from misconception tags and rubric scores, including error-type slopes, rates of conceptual versus procedural error accumulation, profile volatility, and transitions across misconception categories. These features capture the dynamic structure of student understanding rather than static misconception rates or aggregate performance. Through nested predictive modeling and early-warning timing analyses, we will evaluate whether engineered trajectory features improve prediction of end-of-year mathematics outcomes beyond static cognitive indicators and behavioral engagement metrics. Subgroup calibration and fairness analyses will assess performance across English learners, students with IEPs, and racial or ethnic groups. Findings will inform practitioner-facing guidance that integrates trajectory-based signals into formative assessment routines within EdLight.

Principal Investigators

Nicholas A. Vest's headshot

 

 

 

 

Nicholas A. Vest, Ph.D., is a Postdoctoral Researcher in the School of Teaching and Learning at the University of Florida.

Avery Closser's headshot

 

 

 

 

Avery Closser, Ph.D., is an Assistant Professor of Educational Technology at the University of Florida and a member of the Institute for Advanced Learning Technologies.

Xintian Gao's headshot

 

 

 

 

Xintian Gao is a second-year Ph.D. student in Educational Technology at the University of Florida.

One Practice, Two Outcomes: Using Student Work to Build Teacher Content Knowledge and Precision in Instructional Response

Abstract: Teachers are increasingly expected to use student work to drive instructional decision-making, yet many lack the pedagogical content knowledge necessary to accurately diagnose misconceptions and design targeted instructional responses. Consequently, analysis of student work often results in broad re-teaching rather than precise interventions that address students’ underlying misunderstandings. This mixed-methods practitioner research study investigates whether AI-supported analysis of student work can strengthen both teachers’ diagnostic expertise and their pedagogical content knowledge in middle school mathematics. Conducted across grades 6–8 at East Harlem Scholars Academies, the study examines how integrating EdLight’s AI-powered platform into professional learning communities, coaching, and instructional planning influences teachers’ ability to identify misconceptions, develop standards-aligned instructional responses, and improve student learning outcomes. Rather than positioning artificial intelligence as a replacement for teacher expertise, this research explores AI as a catalyst for deeper teacher learning by anchoring professional development in authentic examples of student thinking. Findings from this study aim to inform emerging models of AI-enhanced professional learning that simultaneously build teacher capacity and improve instructional precision in mathematics classrooms.

Principal Investigators

Martin Palamore's headshot

 

 

 

 

Martin Palamore is the Chief Schools Officer at East Harlem Scholars Academies, where he leads academic strategy and school leadership development across a network of Pre-K–12 schools.

AI-Enabled Learning: Building Teacher and Instructional Leadership Capacity Through Student Work

Abstract: Leap Educational Consulting will partner with secondary schools to study how EdLight supports instructional leaders and teachers in identifying student misconceptions, strengthening feedback cycles, and making data-informed instructional decisions in mathematics.

Principal Investigators

Jerry Silva is one of the founders of Leap and, prior to founding Leap, was a co-Senior Director of Achievement First’s Navigator Middle School Math team.

Jen Caruso is one of the founders and leaders of Leap. Prior to starting Leap, she served as Co-Senior Director of the Middle School Math Team for Navigator.

Shifting Teacher Attention from Scoring to Instruction

Abstract: New Visions proposes to integrate EdLight tools into the collection and analysis of data from interim assessments aligned with high-stakes end-of-year testing. We will then research how well this integration supports teachers shifting their attention from data capture to data analysis that better supports student success on those high-stake assessments.

Among the challenges of HQIM implementation, teachers often face a gap between effective classroom formative assessment and high-stakes end-of-year assessments. To bridge this gap, New Visions has provided annual “Mock Regents” where practice exams were supported with data analysis protocols and instructional guidance. We recently added “Interim Assessments,” starting with Algebra I, which provided shorter, more frequent assessments aligned both with specific curricula and end-of-year assessment expectations. We will look deeply at participating schools’ workflows to measure how well the addition of EdLight tools to interim assessment analysis demands less time for scoring and affords more time for instructional planning, To support effective use of the afforded time, we will test several existing data protocols to determine how they can best support team and teacher analysis of the data as part of their planning for year-end student success.

Principal Investigator

Russell West Jr.'s headshot

 

 

 

 

Russell West Jr. supports the Curriculum & Instruction team at New Visions. Additionally, Russell directs the Astor Center for Public School Libraries and provides legal counsel to the organization’s management team.

Strengthening Data-Driven Instruction Through Integrated Professional Learning and Collaborative Data Cycles

Abstract: Foxborough Regional Charter School(FRCS) is partnering with EdLight to strengthen instructional leadership and improve the quality and consistency of classroom feedback across our K–12 district. Building upon our district’s Instructional Vision, the project focuses on developing a shared understanding of high-quality instruction through aligned classroom observations, evidence-based feedback, and coaching cycles. School and district leaders will use the EdLight platform to calibrate observations, identify instructional trends, and leverage real time data to inform professional learning and continuous improvement.

Through this work, the district aims to build leadership capacity, create greater coherence across all three schools, and ensure that instructional feedback consistently supports teacher growth and improved student outcomes. The project will establish sustainable systems for monitoring implementation, measuring progress, and using instructional data to drive decision making while advancing equitable learning opportunities for every student.

Principal Investigator
Kathleen Foley serves as the Director of Teaching and Learning at Foxborough Regional Charter School (FRCS).

Timing is Everything: A NCARE/Washoe County/Curriculum Associates Collaboration on Digital Platform Usage, Engagement and Mathematical Outcomes

Abstract: Digital learning platforms (DLPs) are widely used to provide personalized math instruction, yet little is known about how students should structure their engagement. This study examines whether within-lesson engagement patterns—such as sustained attention, pacing, idle time, and lesson completion—predict middle school students’ math mastery and growth on Curriculum Associates’ i-Ready Personalized Instruction. Drawing on research on cognitive load, sustained attention, and skill acquisition, we examine whether fragmented or rushed engagement reduces learning, even when total time-on-task is the same.

Using data from all 17 Washoe County, Nevada, middle schools during the 2025–26 school year, we will analyze lesson logs, clickstream data, and assessment outcomes alongside teacher interviews from five schools. Findings will inform digital learning platform design, usage recommendations, and instructional practices to help maximize student learning and promote equitable math outcomes.

Principal Investigators

Amy Auletto's headshot

 

 

 

 

Amy Auletto is an education policy researcher and the Assistant Director of the Northwestern Collaborative for Applied Research in Education (NCARE).

Paul Goren's headshot

 

 

 

 

 

Paul Goren serves as Executive Director of the Northwestern Center for Applied Research in Education (NCARE) at the Northwestern School of Education and Social Policy.

Modeling Educator Report-Use Strategies in i-Ready: Patterns, Context, and Associations with Student Growth

Abstract: Guided by the framework of data-based decision making, this study analyzes the 2024-2025 i-Ready educator report to identify distinct report-use strategies and examine how they evolve across fall, winter, and spring diagnostic cycles. School characteristics will be incorporated to investigate how contextual factors shape strategy prevalence and transitions. Finally, the predictive effect of report-use strategies on student growth will be examined at the school level. By moving beyond simple usage counts, this study conceptualizes report-use through multidimensional engagement indicators. Findings will provide actionable insights for learning platform designs, informing professional learning supports, and strengthening the alignment between diagnostic reporting and instructional decision-making.

Principal Investigator

Dr. Yuxi Qiu's headshot

Dr. Yuxi Qiu is an Assistant Professor of Research Methodology at Florida International University.

Integrating Struggle Quality into Knowledge Tracing Models of Mathematics Learning.

Abstract: This project leverages de-identified student log data from middle school mathematics classrooms using Khan Academy to integrate theory-informed indicators of struggle quality into knowledge tracing models. Using time-stamped records of performance, hint usage, retries, and time-on-task, we will develop extended models that distinguish productive from unproductive struggle states and examine how these states predict persistence, mastery acquisition, and disengagement. We will also investigate whether these relationships vary across content modalities (e.g., practice exercises and quizzes). By embedding theoretically grounded measures of struggle into mastery modeling, this study advances understanding of persistence in digital mathematics environments while generating actionable insights to improve detection and support of student struggle at scale.

Principal Investigator

Yang Shi's headshot

 

 

 

 

Yang Shi is an Assistant Professor at Utah State University.

Interpretable Teacher Engagement Profiles and Causal Effects in Math Matrix

Abstract: This project examines how teacher engagement in online professional learning can be measured and how distinct engagement patterns relate to teacher learning outcomes in the Math Matrix micro-credential platform. Using existing platform, gradebook, assessment, and survey data, the project will construct interpretable engagement indicators that capture dimensions such as content coverage, activity intensity, pacing, persistence, and course completion. These indicators will be used to identify teacher engagement profiles and examine their impacts on outcomes including post-assessment performance, learning gains, mastery, completion, and survey-reported transfer to practice. By combining theory-guided feature engineering, exploratory data review, latent profile analyses, and causal-inference approaches for multi-valued engagement profiles, the project aims to move beyond descriptive platform metrics and generate decision-ready evidence about which forms of engagement appear most productive for supporting teacher learning.

Principal Investigators

Wei Li's headshot

 

 

 

 

Wei Li is an Associate Professor in the Research and Evaluation Methodology Program at the University of Florida.

Katherine Strickland's headshot

 

 

 

 

Katherine Strickland, Ph.D., is a Postdoctoral Researcher in the Research and Evaluation Methodology program at the University of Florida.

Early Detection of Teacher Disengagement in Competency-Based Professional Development: A Machine Learning Analysis of the Math Matrix Micro-Credential

Abstract: Teacher disengagement in digital professional development programs is poorly understood and rarely detected until it is too late to address. This study uses the Math Matrix Micro-Credential, an asynchronous competency-based PD platform developed by the Lastinger Center for Learning, to examine this problem directly. Drawing on behavioral trace data, competency-based assessment records, and teacher background information from the Math Matrix platform, we pursue three aims. We first map engagement trajectory patterns and their relationship to mastery outcomes, then identify early behavioral indicators that predict disengagement or mastery failure within the first two to three weeks of participation, and finally examine whether predicted risk profiles differ systematically across teacher characteristics, including certification pathway, years of experience, grade band, and role. We use trajectory modeling to identify engagement patterns over time and machine learning models to detect early warning signals. Findings will provide actionable evidence for designing timely interventions within the Math Matrix platform and contribute broader insight into how digital professional development systems can identify and support teachers at risk of disengagement.

Principal Investigators

Anna Kye's headshot

Anna Kye, PhD, is a Postdoctoral Scholar jointly appointed in the Postsecondary Education Research Institute (PERI) and the Division of Teaching Excellence and Innovation (DTEI) at the University of California, Irvine.

Connecting Teacher Cognition and Engagement in Math Matrix: A Joint Modeling Study of Asynchronous Mathematics Professional Learning

Abstract: This study proposes a cognitive – engagement joint modeling framework for Math Matrix Micro-Credential data that integrates evidence of teachers’ conceptual understanding and misconceptions with behavioral engagement patterns in a unified analytical pipeline. Rather than treating platform use and teacher learning as separate outcomes, the project examines how assessment performance, assignment attempts, available rubric or response data, cumulative content visits, time spent on course pages, survey responses, and teacher background characteristics jointly relate to course mastery and micro-credential completion. Analyses combine engagement profile clustering, item- and assignment-level performance analysis, graph-based representation of teacher-concept-misconception relationships, and mixed-effects predictive modeling. Where open-ended data are available, human-guided coding and validated language-model-assisted analysis will identify recurring themes in teacher conceptions. The novelty of the project lies in connecting teacher cognition and engagement behavior within an interpretable, data-feasible modeling framework. Findings will provide the Lastinger Center with actionable evidence for improving course design, identifying differentiated teacher supports, and advancing research on asynchronous professional learning at scale.

Principal Investigators

Dr. Bo Pei is an Assistant Professor in the Instructional Technology Program, Department of Educational and Psychological Studies, College of Education, at the University of South Florida.

Uncovering Teacher Engagement and Learning in the Math Matrix Micro-Credential

Abstract: Our project, “Uncovering Teacher Engagement and Learning in the Math Matrix Micro-Credential”, investigates how mathematics teachers engage with a competency-based digital professional learning platform and how their learning develops over time. The goal is to generate actionable insights that improve the platform, including better scaffolding of difficult concepts and more personalized support for different groups of teachers. Ultimately, this work will support the design of more effective, adaptive professional learning systems for mathematics educators.

Using data from the Math Matrix Micro-Credential, we examine how teachers progress through targeted skills and identify factors that influence engagement and learning. Our analysis focuses on three areas: patterns of teacher engagement, common mathematical misconceptions, and growth in mastery of key competencies. We also explore how these patterns vary by teacher characteristics such as experience level, grade level taught, and engagement level.

Principal Investigators

Dr. Hailey Kuang's headshot

 

 

 

 

Dr. Hailey Kuang is an Assistant Professor of Measurement and Statistics at Florida State University.

Dr. Secil Caskurlu's headshot

 

 

 

 

 

Dr. Secil Caskurlu is an Assistant Professor of Instructional Systems and Learning Technologies at Florida State University.

An Investigation of Teacher Knowledge Growth and Transfer Across Teacher Characteristics in the Math Matrix Micro-credential

Abstract: This study will investigate whether teacher characteristics can predict teachers’ learning gains and systematic patterns of misconception persistence in the Math Matrix Micro-credential. Furthermore, AI-powered recommendation models will be trained and validated to identify content areas with a high risk of persistent misconceptions. The study will use teacher characteristics to build theoretically grounded teacher profiles and examine them with assessment and end-of-course datasets to determine which teacher profiles are most vulnerable to misconception persistence and whether risk patterns vary systematically across content areas and courses. Findings will advance Math Matrix’s personalization strategy by predicting which teacher profiles show the greatest knowledge growth and which topics pose a greater risk of persistent misconceptions. This study will also contribute to the broader conversation about how asynchronous Digital Learning Platforms (DLPs) can be designed to serve heterogeneous teacher populations.

Principal Investigators

Shristi Shrestha's headshot

Shristi Shrestha is a Ph.D. student in Human Computer Interaction at Iowa State University with research interests in AI-augmented learning environments.

Jiyeong Yi's headshot

Jiyeong Yi is an Associate Professor at Iowa State University, where she develops instructional frameworks to support Emergent Bilinguals in rigorous mathematics.

Digital Scaffolding in Action: How Video Engagement Trajectories Shape Mathematics Learning in Math Nation

Abstract: This study examines how variations in video engagement, including timing, duration, sequencing, tutor selection, and content choices, are associated with mathematics learning. Grounded in a scaffolding-oriented perspective, instructional video is conceptualized as a form of digital support through which students access and interact with mathematical content. Leveraging large-scale activity log data, this study will identify distinct patterns of video engagement and examine how these patterns relate to assessment performance. In addition, the study will explore which aspects of engagement are most strongly associated with learning and whether early engagement behaviors are indicative of later performance. Findings will provide actionable insights into how students use instructional video within a digital learning platform and which patterns of engagement are most supportive of mathematics learning. Results will inform the ongoing development of Math Nation and contribute to broader efforts to improve the design and effectiveness of digital learning environments in mathematics education.

Principal Investigators

Dr. Qingli Lei's headshot

Dr. Qingli Lei is an Assistant Professor in the Department of Special Education at the University of Illinois Chicago.

The Association Between Patterns of Student Video Engagement and Mathematics Performance in Math Nation

Abstract: Math Nation’s (MN) documented achievement gains in Florida, including a 2.9-point increase in FAST math scores and a 2.6-percentage-point increase in Level 5 proficiency in 2022-23, raise a natural next question: which patterns of student engagement with the platform’s features, especially videos, drive those gains? This project uses event-level video log data from ~140,000 middle school students in grades 6–8 to investigate how students engage with instructional videos over time and how these engagement patterns relate to their adaptive engagement and mathematics performance. The project will use latent class analysis to identify distinct engagement profiles from students’ video-use behaviors and sequential pattern mining to characterize common engagement patterns over time, including language selection in video viewing.

Principal Investigators

Sarah Clerjuste's headshot

Sarah Clerjuste is a Postdoctoral Research Associate in the RISE Program at the University of Wisconsin–Madison’s Wisconsin Center for Education Research.

Dr. Ana Stephens' headshot

Dr. Ana Stephens is a researcher at the Wisconsin Center for Education Research at the University of Wisconsin–Madison.

Martha Wagner Alibali's headshot

Martha Wagner Alibali is the Susan Goldin-Meadow Professor of Psychology and an affiliate of the Wisconsin Center for Education Research at the University of Wisconsin–Madison.

Awarded Projects - Spring 2025

An AI-Driven Approach to Predicting Short-Term Teacher Learning Outcomes from Curriculum Pathways and Engagement in Math Matrix

Abstract: This study will investigate how variations in curriculum navigation and behavioral engagement relate to short-term learning gains, and to build machine learning models that predict immediate learning outcomes. We detail a two-phase approach: first, identifying and quantifying different learning pathway profiles among participants; second, training and evaluating predictive models to forecast post-assessment performance. This project will yield insights into which engagement behaviors and content navigation strategies correlate with better short-term learning in the Math Matrix program, and demonstrate how AI techniques can enhance educational outcomes by early identification of at-risk learners.

Principal Investigators:

Shuman Wang

Shuman Wang is a Ph.D. student in curriculum and teacher education at Stanford University.

Ari Jiayu An

Ari Jiayu An is a master’s student in education data science at Stanford University.

Early Signals, Timely Support: Using Machine Learning to Identify and Support Striving Learners in Personalized Learning Platforms

Abstract: This study employs Machine Learning (ML) techniques to predict math achievement trajectories for striving learners by analyzing engagement and persistence patterns within the i-Ready platform. Despite the potential of personalized instructional platforms to accelerate academic growth, significant variability exists in outcomes among users, some students make remarkable gains while others stagnate. By developing predictive models based on early-stage engagement indicators and persistence behaviors, this research distinguishes students likely to accelerate from those at risk of stagnation. The findings might provide educators with actionable insights for timely, targeted interventions and support Curriculum Associates in enhancing i-Ready’s responsiveness through improved early-warning tools. This work addresses both practical and methodological gaps in educational research, demonstrating how predictive analytics can transform support for striving learners when intervention matters most.

Principal Investigators:

Eter Mjavanadze

Eter Mjavanadze is a doctoral student at George Mason University.

Angela Miller, Assistant Professor, CEHD. Photo by Creative Services/George Mason University

Angela Miller is an associate professor in research methods and educational psychology at George Mason University.

Understanding Teacher Motivation in Digital Learning Platforms: A Situated Expectancy-Value Theory Analysis of the Math Matrix DLP

Abstract: This study investigates teacher engagement in the Math Matrix Digital Learning Platform (DLP) to understand factors that drive or hinder motivation. Situated expectancy-value theory (SEVT) is a motivational theory suggesting that an individual’s perceived competence and values predict decision-making and outcomes. Thus, we will use SEVT to examine teachers’ takeaways and challenges of using the DLP and how engagement and teacher characteristics (i.e., grade level, gender, race) predict competence beliefs, perceived usefulness, instructional strategy implementation, and ability to motivate their students. Findings will inform instructional design of the Math Matrix DLP to ensure effective teaching practices and student learning. This study will contribute to the broader conversation about the effectiveness of DLPS as research infrastructure. Leveraging platform-generated data will provide insights into the relation between teacher engagement with the platform and teacher motivation and whether patterns vary based on important teacher characteristics (e.g., grade level taught, experience level, etc.). These findings will not only inform the refinement of the Math Matrix DLP but also aligns with the goals of the DLP to offer guidance for the broader design and implementation of DLPs in mathematics education. The findings from this research will offer valuable insights into how DLPs can be optimized to improve mathematics instruction through enhancing teacher motivation, ultimately supporting student learning outcomes.

Principal Investigators:

Patrick Beymer

Dr. Patrick Beymer is an assistant professor of psychology at the University of Cincinnati.

Jessica Gladstone

Dr. Jessica Gladstone is an assistant professor of educational psychology at the University of Illinois, Urbana-Champaign.

Balancing Learning and Uncertainty: Algorithmic Approaches to Personalized Education through Knowledge Tracing and Reinforcement Learning

Abstract: The project aims to develop methodology and algorithms that optimize learning paths for improved efficiency. It will integrate robust measurement and knowledge tracing models to estimate students’ abilities along with their associated measurement errors. A Deep Q-Network-based recommendation system will guide learning by balancing skill improvement with reductions in measurement uncertainty, resulting in highly personalized learning paths. In addition, the proposed methodology will help us understand transfer of learning to new topics and estimate rates of acquisition and forgetting over time. This work directly supports PI’s (Yikai EK Lu’s) trajectory as an early-career researcher transitioning into a tenure-track assistant professor role.

Principal Investigators:

Yikai "EK" Lu

Yikai “EK” Lu is a doctoral candidate at the University of Notre Dame.

Ying "Alison" Cheng. Photo by Peter Ringenberg/University of Notre Dame

Dr. Ying (“Alison”) Cheng is professor of psychology at University of Notre Dame.

TRACE-LLMs: Tracing and Assessing Cognition Enhanced by Large Language Models in Math Matrix Micro-Credential

Abstract: Knowledge tracing (KT) predicts learning performance by analyzing past behaviors to enable adaptive instruction and precision education, particularly within digital platforms such as Math Matrix Micro-Credential. Traditional KT models primarily rely on sequential patterns (e.g., temporal sequence of assessment responses), often downplaying valuable semantic insights from textual data, such as problem similarity, embedded concepts, and prerequisite relationships. Meanwhile, prior KT methods have largely prioritized predictive accuracy, offering limited transparency into the underlying decision-making processes and leaving educators with few actionable insights. Leveraging the rich assessment data, open-ended responses, and user behaviors from Math Matrix Micro-Credential, we propose TRACELLMs: an explainable KT framework that fuses sequential interaction data, assessment text, and teachers’ historical learning records to evaluate conceptual understanding and deliver actionable feedback. We will benchmark TRACE-LLMs against existing KT methods to demonstrate its feasibility, robustness, and scalability. Successful validation of TRACE-LLMs will enable its integration into Math Matrix Micro-Credential for precise tracking of teacher competency growth, AI-powered mentoring, and personalized learning pathways, ultimately advancing high-quality, adaptive, and scalable professional learning experiences.

Principal Investigator:

Chenglu Li

Dr. Chenglu Li is an assistant professor of learning sciences at the University of Utah.

Improving Persistence through Adaptive Goal Setting in i-Ready

Abstract: Carnegie Mellon University researchers are developing a new approach to enhance student engagement and persistence in mathematics using the i-Ready adaptive learning platform. Led by PhD student Conrad Borchers in collaboration with Co-PIs Vincent Aleven, Ken Koedinger, and Danielle Thomas, the project explores how regular, personalized goals and feedback can help middle school students stay motivated and on track toward long-term proficiency growth targets. Building on prior success from the PLUS tutoring project, where goal setting increased engagement by 25% and skill mastery by 40%, this initiative seeks to design scalable goal-setting tools integrated into i-Ready’s existing features.

The planning grant will support the creation of persistence analytics and adaptive goal recommendations using i-Ready’s lesson and diagnostic data. In partnership with Curriculum Associates and a school district to be selected, the project aims to empower teachers to guide students in setting and achieving goals, including through i-Ready’s “Data Chats” and growth tracking tools. Ultimately, the project advances research on student motivation and self-regulated learning while aligning with AIMS EduData’s mission to generate actionable insights from digital learning platforms.

Principal Investigator:

Conrad Borchers

Conrad Borchers is a Ph.D. student at the Human-Computer Interaction Institute (HCII) at Carnegie Mellon University’s School of Computer Science

Leveraging AI for Equity: A Research-Practice Partnership Evaluating Khanmigo-Driven Math Interventions in High-Variance Classrooms

Abstract: KIPP Team and Family, in partnership with KIPP New Jersey (Camden), will conduct a research-practice partnership with Khan Academy to evaluate AI-powered math interventions for underperforming students in high-variance classrooms. The partnership will investigate how Khanmigo can be leveraged alongside I-Ready diagnostic data to create targeted interventions that address this extreme variance, particularly for students performing below grade level within the Illustrative Mathematics curriculum, among other research questions.

Principal Investigator:

Kevin Shaw

Kevin Shaw is the director of AI, innovation, and strategy at KIPP Team and Family.

Exploring Teacher Engagement in Professional Learning: Patterns, Predictors, and Outcomes in the Math Matrix

Abstract: This project explores how teachers interact with and benefit from online professional learning (PL), particularly in mathematics education. As more educators enter teaching through various pathways, ongoing PL becomes essential for ensuring high-quality instruction. Our project focuses on understanding teachers’ engagement, how they participate, think about, and value their learning experiences, and its impact on their teaching effectiveness. We’ll use survey and online activity data to identify patterns of engagement and connect them to teachers’ learning outcomes and perceptions. Insights from this project will help create better, more impactful online PL experiences that enhance mathematics teaching and student success.

Principal Investigators:

John Ojeogwu

Dr. John Chukwunonso Ojeogwu is a STEM education postdoctoral researcher at the Desert Research Institute (DRI) and the University of Nevada Las Vegas.

Erin Smith

Erin Smith is an associate professor of mathematics education at University of Nevada, Las Vegas.