Independent Evaluation Studies
When funders and partners ask for more than testimonials, rigorous third-party evaluation turns promising programs into evidence-backed ones.
Designing for the Students Furthest from Opportunity and Showing It Works
The Challenge
All Means All (AMA) is a 15-month leadership fellowship that coaches teams of school and district leaders to redesign their schools around the students furthest from opportunity, including students with disabilities and multilingual learners. The program's central conviction is that when leadership teams design for the students who face the steepest barriers, every student in the building benefits.
AMA had seen strong results in its self-reported data from participating leaders across three cohorts spanning 35 schools in 10 states. But the organization had reached a point where third-party validation was essential. Funders were asking questions: Does your model produce measurable changes in student outcomes? Can improvements for students with disabilities come at the expense of other students? And can you back this up with something beyond participant testimonials?
Without independent evidence, AMA could not confidently answer these questions. And without answers, scaling into new districts and securing the funding to sustain that growth was increasingly difficult.
The Solution
The Research & Insights Lab designed and conducted an independent evaluation that combined rigorous quantitative analysis with rich qualitative case studies to build a comprehensive picture of AMA's impact.
On the quantitative side, TRAIL collected three years of publicly available student outcome data for AMA partner schools and built matched comparison groups using propensity score matching to identify non-AMA schools with similar demographics, size, prior performance, and geographic context. Using a difference-in-differences approach, the study tracked changes in ELA proficiency, math proficiency, suspension rates, chronic absenteeism, and attendance across multiple student subgroups, including students with disabilities, Black students, Latino students, English learners, and economically disadvantaged students.
On the qualitative side, TRAIL conducted in-depth case studies of four AMA partner systems, drawing on recorded interviews with school and district leaders to document how AMA's framework translates into practice and how leaders describe the shifts in adult beliefs and behaviors that drive student outcomes.
The Results
The evaluation produced the first independent, matched-comparison evidence that AMA's model shows promising signs of impact.
Math proficiency was the strongest, most consistent signal. AMA schools' growth outpaced matched schools', and the effect strengthened after the program ended, suggesting that the changes AMA catalyzes in adult practice are sustained.
The qualitative case studies documented how AMA's Habits of Mind framework and the "First Will, Then Skill" theory of change translate into observable shifts in how adults think about students, how schools organize instruction, and how leadership teams make decisions.
Why This Matters
The question administrators and funders ask most often about inclusive education is "Aren't professional development opportunities focused on inclusive practices only beneficial to a small group of students?" This evaluation was built to answer that question. The evidence says no. Schools that centered students furthest from opportunity posted gains for all students. For foundations investing in inclusive education and for school systems considering AMA's model, this study provides the kind of independent, matched-comparison evidence that moves the conversation from anecdote to data.
Independent Evidence for an Inclusive Education Partnership Model Across 32 Schools in Washington State
The Challenge
A young nonprofit organization had grown from a program within a charter school association into an independent organization delivering support to schools to bolster their practices related to inclusive education. Their model provides schools with services such as coaching, workshops, technical assistance, and compliance support tailored to each school's needs, with a particular focus on improving outcomes for students with disabilities and multilingual learners.
This organization had developed a framework of inclusive practice indicators and had seen positive signals of their impact. But the organization had reached a point where rigorous, independent evaluation was essential. Prospective district and school partners were facing tight budgets and needed evidence that the partnership model produced measurable changes in student outcomes. Funders wanted more than anecdotal accounts. And the organization's leadership needed to understand which of their service delivery approaches were producing the strongest results to refine the model as they grew.
The Solution
TRAIL designed a quasi-experimental evaluation using a difference-in-differences design with propensity score matching. Using publicly available data, the study compared 32 partner schools against matched comparison schools across four outcome domains: ELA and math proficiency, regular attendance, exclusionary discipline, and restraint and isolation. Outcomes were disaggregated by student subgroups including students with disabilities, English language learners, economically disadvantaged students, and racial/ethnic groups.
The study ran multiple statistical models across two cohorts. TRAIL also conducted a set of interviews with participants to produce case studies of promising practices: descriptive accounts of how the support schools received translates into positive outcomes.
The Results
Attendance emerged as the strongest, most consistent signal across the portfolio, with partner schools outpacing matched comparison schools in attendance gains. Students with disabilities showed the most dramatic individual gains, directly aligned with the organization's core mission of inclusion.
The in-depth qualitative case studies connected key features of the organization's approach to these positive outcomes.
Why This Matters
For organizations whose impact model operates at the school level through partnerships and coaching, building a credible evidence base is structurally challenging: sample sizes may be small, school contexts vary enormously, and the causal chain from partnership activities to student outcomes runs through many intermediary steps. This evaluation demonstrates how studying multiple sources of data can help organizations identify meaningful patterns even within these constraints, and how pairing quantitative outcome analysis with qualitative analysis produces insights that are actionable for program refinement, not just reporting.
Building the Evidence Base for a STEM Curriculum That Reignites Teaching and Learning
The Challenge
The League of Young Inventors had developed a hands-on, NGSS-aligned STEM curriculum that guides elementary students through the engineering design process, encouraging them to test, iterate, and refine their ideas. The curriculum also includes teacher training and digital lesson support to make STEM accessible for educators at all experience levels.
LYI's team could see that their curriculum was working: teachers were enthusiastic and students were engaged. But they lacked the independent, structured evidence that school partners and funders require. Without credible evaluation data, they were relying on anecdote rather than evidence.
The Solution
TRAIL designed and conducted a mixed-methods evaluation study examining how the LYI curriculum was implemented in 8 NYC elementary schools and the outcomes for both students and educators. The study used quantitative data from teacher surveys and qualitative data from focus groups with teachers and school administrators, analyzing changes in student engagement and teacher confidence using a post-and-pre-retrospective design.
The Results
The evaluation documented significant improvements across every outcome area measured.
Student engagement: 97% of teachers reported students were more engaged than in traditional lessons.
STEM confidence: 95% of teachers reported that their students showed greater confidence in collaboration, communication, problem-solving, creativity, scientific reasoning, and iterative learning.
Teacher confidence: 100% of teachers reported feeling confident teaching STEM after participating, a 28 percentage point increase from before the curriculum.
Teachers described the transformation in focus groups: struggling students became engaged and developed creative solutions to design problems. Knowing that failure and iteration were not only acceptable but encouraged gave students a confidence they did not normally have in school.
Why This Matters
For curriculum organizations and edtech companies, credible third-party evaluation data is the difference between a promising product and a fundable one. LYI now has an independent evidence base it can share with school partners, funders, and investors that is grounded in recognized research methods rather than internal metrics. The evaluation also surfaced specific implementation insights that inform how LYI supports teachers in future rollouts.