✍️ Scout Briefing: 9/24/2026 Thursday
Today’s scan is deliberately broader than yesterday’s early-literacy concentration. Two state stories provide a useful contrast: Massachusetts has just reported continued weakness in English language arts while Tennessee is drawing attention for sustained reading recovery, reminding us that “Science of Reading state” is not itself an outcome. Research adds two important challenges to common practice: third-grade struggling readers made stronger accuracy gains with scaffolded complex text than with simplified controlled text, and a multiyear knowledge-building intervention produced effects that transferred into reading, mathematics, and later grades. Outside instruction, the literacy workforce and market are shifting too: teacher preparation has improved sharply but remains uneven, AI tutoring is running into a human problem—students often do not use it—and a major publisher’s move into K–12 dyslexia screening shows assessment becoming more tightly integrated with curriculum portfolios. The human-centered item today sits outside the classroom: community organizations are increasingly building trained tutoring capacity around individual children, which makes visible the labor required to turn “high-impact tutoring” from a policy phrase into repeated instruction.
1. Massachusetts’ new results are a reminder that literacy legislation and literacy outcomes run on different clocks
Finding. Massachusetts released 2026 Massachusetts Comprehensive Assessment System (MCAS) results this week. English language arts performance in grades 3–8 slipped about two percentage points from 2025, while grade 10 English language arts fell more sharply. The state has also enacted a new evidence-based literacy law, but its major instructional requirements do not take effect until the 2027–28 school year.
Why it matters. It is tempting to read state test results as immediate verdicts on current literacy policy. Massachusetts is a useful warning against that. Assessment results describe students who learned under several years of prior conditions; legislation, curriculum change, teacher preparation, and professional learning take time to reach classrooms and longer to reach statewide outcome data.
Important qualification. MCAS English language arts is broader than a direct test of reading skill. Massachusetts also remains a comparatively high-performing state nationally, so declining state results should not be interpreted as evidence that its students are uniformly weak readers.
Best source. Massachusetts Department of Elementary and Secondary Education, 2026 MCAS Achievement Results; Massachusetts early-literacy law.
https://profiles.doe.mass.edu/statereport/mcas.aspx
2. Tennessee offers a contrasting state story: reading recovery appears tied to a system of materials, training, tutoring, and implementation support—not one reform
Finding. Tennessee ranked fourth among 35 states in reading recovery from 2022 to 2025 in the 2026 Education Scorecard, and its 2025–26 state assessment results showed third-grade English language arts achievement at its highest level since at least 2017. State and independent accounts point to a package that includes high-quality instructional materials, teacher professional learning, preparation requirements, summer supports, tutoring, and regional implementation networks.
Why it matters. The interesting lesson is not simply that Tennessee adopted Science of Reading policy. Its experience is increasingly being used as evidence that literacy improvement is a systems problem: curriculum, adult learning, intervention, assessment, and implementation capacity have to work together over time.
Important qualification. The available evidence does not isolate which Tennessee policy caused the gains. State comparisons are observational, and changes in student populations, assessment systems, pandemic recovery, and other reforms can contribute to trends.
Best source. Tennessee Department of Education, Education Scorecard recognition and 2025–26 TCAP results; Harvard Center for Education Policy Research, Education Scorecard.
https://cepr.harvard.edu/education-recovery-scorecard
3. A third-grade intervention study challenges the instinct to give struggling readers easier text
Finding. A 2026 Reading Research Quarterly study compared 110 third graders reading below the 40th percentile. One group received Read Like Us, which paired explicit multisyllabic-word instruction with informational texts written at roughly grades 4–5 complexity; the active comparison group used Corrective Reading with more controlled text. After 60 sessions, the complex-text group made larger gains in reading accuracy and was 2.79 times more likely to reach 98 percent accuracy. Oral reading fluency outcomes were equivalent.
Why it matters. Struggling readers are often protected from difficult text on the assumption that easier material is necessary until foundational skills catch up. This study suggests a different possibility: with strong scaffolding, challenging content and explicit word work can coexist, giving students access to richer vocabulary and knowledge while they continue developing accuracy.
Important qualification. This was a matched-sample quasi-experimental study, not a randomized controlled trial, and it compared two specific interventions. It does not mean every struggling reader should simply be handed grade-above text without substantial instructional support.
CRSL connection. For grades 3–5 intervention, this is worth watching because it argues against making remediation intellectually smaller. Foundational support and meaningful, knowledge-rich reading may be complementary rather than sequential.
Best source. Downs, Hiebert, Conradi Smith & Martz, Reading Research Quarterly, 2026.
https://ila.onlinelibrary.wiley.com/doi/full/10.1002/rrq.70117
4. Long-term evidence strengthens the case for knowledge building as part of literacy instruction
Finding. A longitudinal randomized controlled trial of the Model of Reading Engagement (MORE) followed 2,870 students across 30 elementary schools. Students received a sustained, spiraled content-literacy intervention from grades 1 through 3 that connected science topics, vocabulary, schema, and wide reading. Students in the full intervention outperformed peers on science vocabulary, science reading comprehension, domain-general reading comprehension, and mathematics; effects in reading and mathematics were still evident 14 months later in grade 4.
Why it matters. This addresses one of literacy research’s harder questions: transfer. A child can improve on material closely aligned to an intervention without becoming broadly better at understanding unfamiliar text. The MORE findings suggest that sustained knowledge building can produce effects beyond the exact content practiced and beyond the year in which it was taught.
Important qualification. MORE is a specific multiyear intervention with coherent content and substantial implementation support. The findings do not establish that adding disconnected background-knowledge activities to any reading program will create the same effects.
CRSL connection. This provides stronger evidence for treating vocabulary, knowledge, and comprehension as cumulative work across grades rather than enrichment added after children learn to decode.
Best source. Center for Education Policy Research at Harvard University, Time to Transfer: Long-Term Effects of a Sustained and Spiraled Content Literacy Intervention in the Elementary Grades, 2026.
5. Teacher preparation in reading has improved quickly—but the national pipeline is still split almost in half
Finding. The National Council on Teacher Quality reviewed more than 700 elementary teacher-preparation programs in 2026. Fifty-three percent earned its top grade for covering the five core components of scientifically based reading instruction and avoiding practices the organization considers unsupported, up from 26 percent in 2023. At the same time, 47 percent still fell short, and one in five programs continued to teach methods NCTQ identifies as discredited.
Why it matters. Schools are trying to implement new literacy expectations while receiving novice teachers with very different levels of preparation. That means districts may need to treat professional learning not merely as curriculum orientation but, for some teachers, as completion of knowledge they did not receive before entering the classroom.
Important qualification. NCTQ evaluates programs through its own evidence framework and course-material review; its grades are not direct measures of how well graduates teach or how their students perform. The sharp improvement nevertheless documents substantial change in what preparation programs say they teach.
CRSL connection. This helps explain why teacher-facing materials need to work across a wide range of prior knowledge. A manual that assumes every teacher already understands phonemic awareness, phonics, fluency, vocabulary, and comprehension may be operating in a workforce that does not yet support that assumption.
Best source. National Council on Teacher Quality, Teacher Prep Review: Decoding Progress in Reading Preparation, June 2026.
https://www.nctq.org/research-insights/decoding-progress-in-reading-preparation/
6. AI tutoring is running into an implementation problem that has little to do with the sophistication of the AI
Finding. Researchers at Stanford’s National Student Support Accelerator report that evidence for artificial-intelligence tutoring at U.S. scale remains limited and that one of the clearest problems is low student uptake. In reporting on tools including Amira and Khanmigo, executive director Susanna Loeb said the largest finding so far is that students often do not use the tools enough. Providers point to prior studies showing benefits when intelligent tutoring systems are implemented well.
Why it matters. A tool can be technically capable and still have little educational effect if students do not engage with it consistently. This shifts part of the AI-in-education question from “Can the software teach?” to “Under what human conditions will students actually use it enough to matter?”
Contrary evidence. The broader research literature on intelligent tutoring systems includes positive effects, and vendors argue that purpose-built systems differ substantially from generic generative AI. The unresolved issue is less whether digital tutoring can ever work than whether current AI systems can produce reliable participation and outcomes at ordinary-school scale.
CRSL connection. This is useful evidence for the principle that technology has to be evaluated inside the human work surrounding it. Placement, supervision, dosage, and teacher involvement may matter as much as the capability of the software.
Best source. The 74, September 1, 2026, reporting research from Stanford University’s National Student Support Accelerator and provider responses.
https://www.the74million.org/article/ai-tutors-not-yet-a-replacement-for-humans-research-says/
7. Literacy’s supply side is consolidating: McGraw Hill is integrating a Stanford-developed dyslexia screener into its K–12 portfolio
Finding. McGraw Hill announced in August that it had secured exclusive integration of ROAR—Rapid Online Assessment of Reading—along with the technology and intellectual property underlying the Stanford-developed assessment model. ROAR is designed to screen foundational reading skills across the K–12 span, including risk associated with dyslexia. McGraw Hill says it will connect the screener with its broader literacy curriculum and learning resources.
Why it matters. This is not only a product announcement. It illustrates a market pattern worth watching: assessment, curriculum, intervention resources, data, and professional support can increasingly sit inside the same commercial ecosystem. That may simplify implementation for districts, but it can also make assessment choices more closely tied to vendor platforms.
Important qualification. The announcement comes from McGraw Hill and describes intended integration, not evidence that combining ROAR with McGraw Hill curriculum improves student outcomes. ROAR’s research base and the commercial value of an integrated portfolio are separate questions.
Best source. McGraw Hill, August 11, 2026, announcement on ROAR integration.
8. Human-centered story: “high-impact tutoring” ultimately depends on someone showing up for the same child again and again
Finding. Across several communities this fall, literacy organizations are recruiting and training adults to provide repeated one-to-one or small-group reading support. Michigan Education Corps, for example, places trained interventionists with K–3 students during the school day using evidence-based scripted interventions and onsite coaching. In North Carolina, Read and Feed pairs volunteers with K–5 students at schools, churches, and apartment communities and provides tutor office hours for progress checks, literacy games, and individual student concerns.
Why it matters. Policy discussions often compress tutoring into a funding category. At child level, it is a relationship plus a schedule: an adult has to be recruited, trained, coached, matched with students, present week after week, and equipped to respond when a particular child is not progressing. That implementation labor is part of the intervention.
Important qualification. These are practitioner and workforce signals, not comparative effectiveness evidence for either organization. They are useful because they show what tutoring infrastructure looks like when translated from policy into daily practice.
CRSL connection. For intervention materials, the user may not always be a classroom teacher or reading specialist. Programs that can be used faithfully by well-trained tutors or paraprofessionals may reach children through a wider instructional workforce.
Best source. Michigan Education Corps 2026–27 reading tutor recruitment; Read and Feed 2026 volunteer/tutoring calendar.
https://career.grinnell.edu/jobs/michigan-education-corps-reading-tutor-2026-2027-service-year/
Scout is noticing…
Supported observation — literacy policy is entering its accountability phase. Massachusetts and Tennessee illustrate why the next national question is not simply which states have passed reading laws, but whether implementation systems mature enough—and remain in place long enough—to change outcomes.
Supported observation — the evidence base is pushing against false either/or choices. The third-grade complex-text study and the MORE trial both resist a narrow sequence in which foundational skills must be completed before students encounter rich content, vocabulary, knowledge, and comprehension work.
Supported observation — adult capacity remains a limiting variable. Teacher preparation, tutoring infrastructure, coaching, and AI uptake all point to the same constraint from different directions: an instructional resource does not implement itself.
Early signal — assessment may become a more strategic part of literacy-company portfolios. McGraw Hill’s ROAR integration is one example, not yet a trend by itself. Scout should watch whether curriculum companies increasingly acquire or tightly integrate screeners, diagnostics, intervention tools, and progress data.
Early signal — the strongest state literacy stories may increasingly be judged by durability rather than initial reform. Tennessee’s gains are attracting attention because multiple pieces of its literacy system have persisted across several years. Massachusetts and Michigan will provide useful tests of whether newer reforms develop comparable implementation depth.
