How do we know whether early learning programs are working for the families they’re meant to serve? It’s a harder question than it sounds. In many places, the data needed to answer it is spread across agencies, stored in legacy systems, or never linked to what happens once children reach kindergarten and beyond.
Two years ago, the Gates Foundation partnered with the Harvard Strategic Data Project (SDP) to support 13 fellows working inside state and district agencies across the country.
Their charge was to dig into two big questions:
- Do preschool supply and enrollment processes meet family needs?
- Are children reaching developmental and learning milestones from preschool through 3rd grade?
The fellows’ projects ranged from examining whether a state assessment predicts fifth-grade outcomes to redesigning a pre-K report card with families and educators. We asked the fellows what they found, what surprised them, and why this work matters. Their answers, lightly edited for length, are below.
Fellows’ findings are already shaping decisions: expanding parent education access, rebuilding data systems so programs’ data can be linked, redesigning a pre-K report card for families, and supporting a citywide child care commitment. Others hope their findings will lead to wider use of an assessment in kindergarten or small changes to enrollment. Fellows also noted limits, including a 25% cut to one state’s preschool grant.
Robert Carr, California Department of Education
“For my SDP Capstone Project, I investigated the predictive validity of the DRDP (2015) Preschool version (CDE 2015). The analyses found that CSPP students with higher DRDP (2015) scores were more likely to score higher on assessments of English Language Arts/Literacy and Mathematics in third through fifth grade. Given the findings from my SDP Capstone Project, I hope that California educational policymakers and practitioners will consider an expanded use of the DRDP in settings like Transitional Kindergarten and Kindergarten. I hope that the findings of this study are helpful in emphasizing that efforts to improve early literacy skills can support learning in other domains, like math, while efforts to improve early math skills can support learning in other domains such as literacy.”
Melinda Wengrin, Washington State Education Research & Data Center
“In sharing my findings, reviewers noted that enrolling children in the Learning Assistance Program (LAP) during kindergarten is often not prioritized, though my analysis shows that kindergarten LAP participation is associated with stronger third-grade assessment outcomes. Stakeholders also expressed interest in the correlation I found between accessing summer child care before kindergarten and sustained progress toward kindergarten assessment benchmarks. One noted that current childcare access supports often deprioritize children who are about to enter kindergarten because of short enrollment windows, suggesting that these findings may warrant reconsidering how barriers are managed for this group … While my paper is not yet published, my hope is that agencies can use these insights to make small, practical adjustments to program design or enrollment processes.”
Heather McCabe, Washington State Education Research & Data Center
“We are making an effort to create and sustain a sense of community amongst early learning researchers in our state. It is really exciting that with collaboration, tools and resources we can get our early learning data in a position to support a research agenda that answers important policy questions.”
Molly Tuck, South Carolina First Steps
“My capstone on how kindergarten readiness outcomes are strengthened by program stacking (specifically 4K and parent education programs) has impacted both our programming plans as well as how we collect and manage data. From a programming perspective, we are being more intentional about providing greater access to specific parent education programs across the state. As far as data operations, my capstone illustrated the power of better understanding the combination of early childhood programming ahead of kindergarten. Therefore we have transformed our data collection systems to allow different programs’ data to talk to each other more easily.”
Yuzhu Xia, Boston Public Schools
“At Boston Public Schools, our findings prompted us to rethink the purpose of the Pre-K report card and how it can best serve children, educators, and families across our mixed-delivery system. Rather than continuing to use the report card as an aggregated measure of student skill development across settings, we are redesigning it as a more accessible, family-centered communication tool grounded in shared learning goals … Our next step is to prepare for districtwide implementation in SY 2027–28.”
“Our findings prompted us to rethink the purpose of the Pre-K report card and how it can best serve children, educators, and families across our mixed-delivery system.” – Yuzhu Xia, Boston Public Schools
Fellows were surprised that early measures predicted outcomes years later, and that interest in the work was higher than expected. They also shared how much engaging educators, families, and other users upfront helped shape their work and found that users want simple visualization tools rather than complex dashboards. Fellows noted that their work reinforced that early childhood data systems still lag far behind K–12, which wasn’t necessarily a surprise, but was held up by their experience.
Robert Carr, California Department of Education
“I was surprised to find that the DRDP Preschool assessment is predictive of student achievement outcomes all the way through fifth grade. That’s six years later! When preschool teachers complete the DRDP, what they are telling us about children and child development really matters for later success throughout elementary school.”
Melinda Wengrin, Washington State Education Research & Data Center
“It surprised me that I found a significant correlation between the Learning Assistance Program in kindergarten and third grade test scores. A lot of research shows that the positive correlation of early interventions fades by third grade.”
Virginia Spinks, Massachusetts Department of Early Education and Care
“One surprise was how much work there is still to be done (especially regarding data governance and infrastructure) to bring early childhood data systems into the same maturity level as elementary/secondary data systems. But I also was pleasantly surprised at the receptivity of self-proclaimed “non-data people” to being stewards of good data quality and utilizing data to inform implementation when communities/users can see themselves reflected in the data product.”
Melissa Nelson, DC Office of the State Superintendent of Education
“I was surprised by the level of interest in this work. I expected that a couple of local education agencies (LEAs) would express interest, and then we had responses from about two-thirds of LEAs in Washington, DC with pre-K programs. Whether or not they ultimately had capacity to participate in the research project, stakeholders are interested in better understanding how well their youngest students are learning across grade transitions, as well as how their data fits into the larger data landscape.”
Yuzhu Xia, Boston Public Schools
“What surprised me most was how much the project changed once we looked beyond the data itself and listened closely to the people using the tool. We initially wanted to understand whether report card ratings could provide a meaningful, comparable measure of children’s skill development across different settings. But direct assessments, educator feedback, and structured conversations with families and educators revealed that the report card’s greatest value was actually in helping adults understand children’s learning and communicate about learning goals … It reminded me that when data do not answer the question we hoped they would, the answer is not always to find a more sophisticated analytical method. Sometimes we need to reconsider whether the tool was designed to answer that question in the first place.”
Paige Kaliski, Chicago Public Schools
“What surprised me is that most data users want simple, straightforward data tools—oftentimes just one big visualization and a table—and not exciting, shiny dashboards with lots of features. Data users want straightforward access to information, not overly complex visualization or web design.”
“I was surprised to find that the DRDP Preschool assessment is predictive of student achievement outcomes all the way through fifth grade.” – Robert Carr, California Department of Education
Fellows described fragmented early childhood data systems that rarely connect to each other or to K–12, making it hard to understand what’s working and where more investment is needed to support children on their learning journey. They saw investment as a path to more equitable, better-informed decisions and stressed that data analysts must engage the people who will use the data.
Virginia Spinks, Massachusetts Department of Early Education and Care
“It’s incredibly important to invest in early childhood data systems, because they can tell us how resources are being spent, how quality in state-funded classrooms is improving or not, and the equitability of these trends. In a landscape where early childhood education is so financially inaccessible, yet so critical to our entire society, investing in these systems, hopefully, will help contribute to more stakeholdership in them.”
Heather McCabe, Washington State Education Research & Data Center
“In the education sector, early learning is underdiscussed. This is an area of work that calls for more collaboration because data sources and systems are so fractured.”
Molly Tuck, South Carolina First Steps
“The early education data landscape feels like the great unknown in a lot of ways. Many early childhood education agencies are forced to navigate legacy systems and outdated data collection methods. The systems rarely talk to each other, which makes it difficult to tell the longitudinal data stories that the field deserves. Furthermore, there is very little cross-state alignment or standardization in the assessments valued by the field, so when we talk about data, we are not telling the same story … By investing in the right data systems and personnel capacity to fully utilize the data, the early childhood field as a whole can benefit from a more wide-spread acknowledgement of the effects of high quality early learning experiences on students’ long-term academic trajectories.”
Cristal Cisneros, Denver Preschool Program
“While enrollment figures, provider capacity, and waitlist numbers help quantify the early childhood education landscape, they often fail to capture the lived experiences, decision-making processes, and systemic barriers that shape families’ access to care. Qualitative data provides critical context by illuminating factors such as workforce challenges, transportation constraints, cultural and linguistic preferences, affordability trade-offs, and the unique needs of children and families across Denver’s diverse communities. By elevating the voices of parents, providers, and community stakeholders, qualitative research helps explain not only where gaps exist, but why they exist, enabling more responsive and equitable policy and program solutions.”
Paige Kaliski, Chicago Public Schools
“The most important unit of change in a school system is the school itself, and if we don’t meaningfully engage with the people working with schools to understand how to design data products, we’ll never accomplish anything. Data analysts could produce endless dashboards with endless features that are never utilized for anything if they don’t actually talk to people about what they need.”
“By elevating the voices of parents, providers, and community stakeholders, qualitative research helps explain not only where gaps exist, but why they exist.” – Cristal Cisneros, Denver Preschool Program
To put these lessons into practice, explore two new Harvard SDP early education toolkits, both informed by the fellows’ work. The first toolkit helps agencies assess whether preschool programs match family needs, and the second focuses on linking Pre-K – 2nd grade data to understand children’s outcomes over time.





