Adaptive Teaching and Inclusive Design: Evidence, Components, and What Works

Adaptive teaching and inclusive design are the evidence-informed teaching methods most often treated as self-evidently good and most often unevidenced in their umbrella form. The underlying idea — adjust teaching responsively, reduce barriers by design — is ethically compelling and pedagogically defensible. The direct causal evidence that the full branded frameworks produce the effects their advocates claim is thinner than the evidence for their individual components. This review sets out what the research actually says, where the components are better-evidenced than the umbrella, and how to build an adaptive, inclusive practice from mechanisms the evidence supports.

What adaptive teaching and inclusive design are

Adaptive teaching means being responsive to evidence about how pupils are learning and adjusting teaching accordingly. In current English practice, "adaptive teaching" has increasingly replaced "differentiation" — which acquired unfortunate associations with permanently tracking students into different tasks, worksheets, or ability groups.

Inclusive design means reducing barriers from the outset so that one shared learning journey is accessible to the widest range of learners. The most widely cited framework is Universal Design for Learning (UDL), from the Center for Applied Special Technology (CAST). UNESCO frames inclusion as the identification and removal of barriers so all learners can participate and achieve together.

The two ideas overlap. Inclusive design anticipates barriers before the lesson; adaptive teaching responds to them in the lesson. Both hold shared ambitious goals and vary the supports rather than the targets.

What the research actually shows

The evidence here is more fragmented than for the other methods in this library.

A meta-analysis of UDL-aligned interventions reported a moderate positive effect of g ≈ 0.43 from pre-kindergarten to adulthood. A 2023 systematic review and meta-analysis also concluded that UDL can improve outcomes.

A 2024 critical analysis argued, however, that the research evidence specifically supporting CAST's full guideline architecture — its three principles and their sub-guidelines — remains weak and often really supports narrower techniques rather than the full framework. That is an important distinction: the components often work; the branded superstructure does not always get independently validated.

For differentiation / adaptation more broadly, the primary-school review literature found small-to-moderate benefits when differentiation involved computer-supported tools or sat within wider reform. Ability grouping alone — the old-school version of "differentiation" — had a small negative effect for lower-attaining pupils, a finding that has been consistent across decades.

Secondary evidence is thinner. One review notes a very small overall effect for adaptive teaching and substantial knowledge gaps. A 2025 systematic review of adaptive learning, instruction, and teaching found 69 studies, mostly in elementary mathematics and mostly system-delivered (through adaptive software) — highlighting how underdeveloped the classroom-level adaptive-teaching research base still is.

The most defensible synthesis: the components of adaptive teaching and inclusive design (explicit instruction, scaffolding, formative assessment, flexible grouping, structured support) are better evidenced than the umbrella labels. Build the practice from the components. Treat the umbrella as a design framework, not as a single intervention with its own effect size.

The EEF's five-a-day

The Education Endowment Foundation's widely cited "five-a-day" framework for SEND-inclusive teaching consolidates the components. A high-quality, inclusive lesson typically includes:

• Explicit instruction — clearly explained content, modelled, practised (see our Explicit Instruction Evidence review)
• Cognitive and metacognitive strategies — teaching learners how to think about the task (see Metacognition and Self-Regulated Learning Evidence)
• Scaffolding — temporary support structures that are gradually faded
• Flexible grouping — purposeful, short-term groups formed for specific learning needs
• Using technology — where it genuinely supports access, retrieval, or application

Each of these five has its own independent evidence base. Together, they are the most defensible starting point for adaptive, inclusive practice.

UNESCO's inclusion framing complements this by focusing on identifying and removing barriers — linguistic, sensory, cognitive, cultural, motivational — before they produce exclusion. Good inclusive design anticipates, not retrofits.

Core principles

The principles that bind these ideas together are more stable than any of the branded frameworks.

• Maintain ambitious common goals. Same ambition for all; vary the supports, not the destination.
• Anticipate barriers before the lesson. Which students will struggle with the language, with the representation, with the working memory load, with the motivation.
• Provide flexible routes to the same learning. Not different tasks; the same task made accessible by scaffolds, representations, or supports.
• Use formative evidence to adapt support, grouping, and scaffolds in real time. The adaptation is responsive, not fixed.
• Judge success partly by who can access, participate, and achieve. If the lesson "worked" but a third of the class couldn't engage, it didn't work.

Where differentiation went wrong

It's worth naming explicitly why the word "differentiation" fell out of favour. In practice, differentiation often degraded into:

• parallel worksheets of different difficulty, which turned into permanent tracks
• "bronze / silver / gold" tasks where students self-selected downward
• ability groups that crystallised into permanent streaming and produced the negative effect for lower-attainers in the primary evidence
• unsustainable planning demands — three versions of every resource

Adaptive teaching as a reframing aims to fix this: same task, varied supports; flexible grouping not fixed streaming; adaptation based on in-lesson evidence rather than on fixed ability judgements.

Classroom examples across phases

Primary. Year 2 story retelling. All pupils work toward the same learning goal — retell the story in sequence with key details. Supports vary: picture prompts for those who need a visual anchor, oral rehearsal in pairs before writing for those who need to talk first, optional sentence frames for those who need the structure. The goal is the same; the access route varies.

Secondary. Year 9 mathematics. All pupils solve the same algebra problem set. Scaffold density varies across the room — some use manipulatives, some use pictorial representations, some work directly with symbols. Flexible grouping: the teacher forms a short-term small group mid-lesson for three students whose mini-whiteboard responses reveal a shared misconception, reteaches, and releases them back. No fixed tables, no labels.

Tertiary. First-year law seminar on a case analysis. Captioned video summary of the case before the seminar. Reading guide with vocabulary definitions. Varied participation routes: contribute in writing via a shared doc, speak in pairs before speaking to the whole seminar, or speak directly. The case analysis is the same; the access routes vary.

Where adaptive and inclusive teaching fails

The failure modes are avoidable but common.

• Confusing adaptation with permanent tracking. If your "adaptive" groups haven't changed for a term, they aren't adaptive. They are streaming.
• Offering too many choices without clarity of goal. If students don't know what the lesson is aiming at, giving them three ways to get there doesn't help. Keep the goal sharp; vary the support.
• Bespoke tasks for every pupil. Unsustainable and usually not necessary. Same task, varied scaffolds, is the default.
• Retrofitting access instead of designing it in. Captions added after complaints, reading guides produced the night before. Inclusive design is anticipatory.
• Treating the umbrella framework as the intervention. UDL, adaptive teaching, and differentiation are design frameworks. The effect comes from the underlying mechanisms — explicit instruction, scaffolding, formative assessment, flexible grouping, technology — not from adopting the label.

Best fit and resource needs

Best fit: all age phases, especially SEND-inclusive and linguistically diverse classrooms. The framework approach is broadly applicable; the components are evidence-backed across phases.

Teacher requirements are moderate to high. The practice demands teacher judgement, specialist collaboration with SENCOs and learning support staff, and sometimes technology or accessibility tools. It is not materials-cheap but it is judgement-heavy.

Evaluate with disaggregated progress data, accessibility checks, subgroup scrutiny, and pupil voice. Progress data aggregated across the whole class will hide inclusion failures. Subgroup analysis — SEND, EAL, pupils eligible for free school meals — is what reveals whether the inclusive design actually worked.

How TAyumira supports adaptive teaching and inclusive design

TAyumira supports adaptive teaching as one of its ten research-backed teaching methods, and every lesson it generates includes inclusive design prompts. When you use it, you get:

• A shared ambitious objective with varied support routes (scaffolds, representations, optional sentence frames) rather than different tasks
• An anticipated-barriers section: linguistic, sensory, cognitive, and motivational barriers identified before the lesson
• The EEF five-a-day prompts built into the plan: explicit instruction, cognitive and metacognitive strategies, scaffolding, flexible grouping, technology
• Automatic caption and accessibility checks on presenter outputs
• Disaggregated analytics: subgroup performance visible in the post-lesson view

Start for free — the Free tier covers the full workflow.

FAQ

What is the effect size of Universal Design for Learning?

A meta-analysis of UDL-aligned interventions reports a moderate positive effect of g ≈ 0.43 from pre-kindergarten to adulthood. A 2024 critical analysis has argued that the evidence specifically supporting CAST's full guideline architecture remains weaker than the umbrella-level claims suggest, and that much of the effect sits in component techniques rather than in the branded framework.

What is the difference between differentiation and adaptive teaching?

Differentiation, as practised in many schools, often came to mean fixed ability groups and parallel worksheets — which the primary-school evidence links to small negative effects for lower-attaining pupils. Adaptive teaching keeps shared ambitious goals, varies the supports not the destination, and uses flexible short-term grouping based on in-lesson evidence rather than fixed ability judgements.

Why did "differentiation" fall out of favour?

Because in practice it often drifted into permanent tracking, parallel-task worksheets, and unsustainable planning demands, with a small negative effect on lower-attaining pupils in primary education research. Adaptive teaching is an explicit reframing designed to avoid those failure modes.

What is the EEF "five-a-day"?

The EEF's framework for inclusive teaching: explicit instruction, cognitive and metacognitive strategies, scaffolding, flexible grouping, and technology. Each has its own evidence base. Together they are a defensible starting point for SEND-inclusive teaching.

Is Universal Design for Learning the same as adaptive teaching?

They overlap substantially but come from different traditions. UDL is an American framework from CAST, built around three principles (multiple means of engagement, representation, and action/expression). Adaptive teaching is the preferred UK practice term in recent years. Both share the goal of ambitious common learning with varied support; both are better-evidenced at the component level than at the umbrella level.

Related evidence reviews

• Explicit Instruction Evidence
• Formative Assessment Evidence
• Metacognition and Self-Regulated Learning Evidence
• Cooperative Learning Evidence

Sources

• EEF. Moving from differentiation to adaptive teaching.
• EEF. Supporting high-quality teaching for pupils with SEND.
• CAST. UDL guidelines and evidence pages.
• UNESCO. Inclusion in education resources.
• MDPI Behavioural Sciences review — UDL evidence.
• Sciencedirect — adaptive teaching evidence base review.

Build adaptive and inclusive practice one component at a time

Pick one of the EEF five-a-day components and deepen your practice on it for a term. Scaffolding is often the single highest-impact starting point: shared ambitious goals, varied supports, gradual fading. If you want lesson plans that come with the five-a-day prompts built in, create a free TAyumira account.