Ten Knowledge Foundations
Universal Design for Learning is a synthesizing framework, so genuine expertise in designing for variability draws on the feeder disciplines below. The strongest design work holds these in view simultaneously — producing solutions that a single-discipline approach cannot. They are ordered roughly by how load-bearing each is.
Learning Sciences & Neuroscience of Learning
UDL is built on the premise that learning engages distinct neural networks — affective, recognition, and strategic — which map onto its three principles. Cognitive load theory, working memory, and how the brain processes novel information ground every decision about flexible instruction.
Adult Learning Theory
Knowles's andragogy and the broader literature on self-directed learning explain how adults differ from younger learners: they bring extensive experience, expect relevance and autonomy, and are often internally motivated. This is the specialization that distinguishes the work from K–12 practice.
Transformative & Experiential Learning
Mezirow's transformative learning, Kolb's experiential cycle, and Schön's reflective practice describe how adults make meaning, revise assumptions, and learn through experience and reflection — central to designing meaningful engagement and expression options for mature learners.
Motivation & Self-Determination Science
Deci and Ryan's self-determination theory, Bandura's self-efficacy, and expectancy-value models underpin the engagement principle. Because adult participation is frequently voluntary, autonomy, competence, and relevance are not enrichment — they are conditions for participation at all.
Disability Studies & Accessibility
UDL grew out of the disability rights movement. The social model of disability, accessibility standards such as WCAG, assistive technology, and accommodation law ground a commitment to removing barriers in the environment rather than locating the deficit in the learner.
Universal Design (the Parent Philosophy)
Ronald Mace's original Universal Design in architecture and product design — and its seven principles — supply the conceptual DNA: design proactively for the full range of human variation rather than retrofitting after exclusion has already occurred.
Instructional & Learning Design
Backward design (Understanding by Design), ADDIE, and the principles of curricular alignment translate UDL ideals into actual course architecture — ensuring goals, methods, materials, and assessments cohere rather than fracturing under added flexibility.
Assessment & Educational Measurement
Validity, authentic assessment, and especially construct-irrelevant variance let a practitioner distinguish when an assessment measures the intended competency versus an incidental barrier — reading speed, tech fluency, or test anxiety. This is the measurement spine of inclusive design.
Educational Technology & Multimedia Learning
Mayer's cognitive theory of multimedia learning, fluency with learning platforms and authoring tools, and emerging AI tooling make multiple means of representation and expression practical rather than prohibitively labor-intensive at scale.
Culturally Responsive & Equity-Centered Pedagogy
Work on inclusive pedagogy, intersectionality, and learner identity addresses the social and cultural barriers learners encounter — not just the cognitive ones. The 2024 UDL Guidelines make this dimension explicit, treating identity as part of variability rather than separate from it.
A Synthesizing Framework, Not a Single Theory
No one of these disciplines is sufficient on its own. The neuroscience explains why variability is the norm; disability studies supplies the ethical and historical commitment; measurement theory determines whether a flexible assessment is still valid; instructional design keeps the whole thing coherent as a course. Expertise is the capacity to draw on the right body of knowledge for the decision at hand, and to recognize when a barrier belongs to one discipline rather than another.
The linchpin skill that ties them together is goal–means separation: stating a learning goal cleanly so that the goal is fixed but the means of reaching and demonstrating it remain flexible. "Write a persuasive essay" smuggles a method into the goal; "construct and defend a persuasive argument" does not. Most design failures trace back to a goal with a method baked into it — and most of the ten foundations exist, in practice, to inform how that separation is made and acted upon.
Why "At Scale" Changes the Problem
Designing for one learner is accommodation. Designing so that one experience serves many learners — across disability, language, prior experience, motivation, and circumstance — is design. The shift from the first to the second is what "at scale" names. It moves inclusion from a reactive, per-person, retrofit cost to a proactive design property, where access is built in once rather than negotiated repeatedly. This is the thesis from which the remaining sections of this resource follow.
How Designing for Variability Works in Practice
The process is a backward-design cycle adapted for learner variability: it begins with the goal, separates that goal from any particular method, and treats each design as a hypothesis to be tested against whether barriers were actually reduced. The stages below represent a comprehensive approach; in practice, depth is calibrated to context. Understanding the full process clarifies what is gained or lost when stages are abbreviated.
Before any design begins, the learning goal is stated cleanly enough that the competency is fixed while the path to it remains open. This is the most consequential step in the entire process: a goal with a method baked into it ("write an essay," "give a live presentation") quietly excludes capable learners for whom that method is itself the barrier.
The work here is distinguishing the construct — the actual capability being developed — from the conventional way it has historically been taught or demonstrated. Done well, it makes every later decision about options easier, because the options follow naturally once the goal is clean.
- Learning goals stated as fixed competencies, free of embedded methods
- Explicit separation of construct-relevant skill from incidental format
- Shared agreement with stakeholders on what is, and is not, being measured
This stage characterizes the actual range of learners — across prior experience, language, sensory and motor access, attention, motivation, and circumstance — rather than designing for an imagined typical participant. It draws on human-centered research methods: interviews, observation, and review of existing performance data.
The analytical core is distinguishing construct-relevant difficulty — the desirable struggle that is the learning — from construct-irrelevant barriers that block some learners for reasons unrelated to the goal. Removing the second without flattening the first is a matter of expert judgment, not a checklist.
- A profile of the learner range, not a single persona
- An inventory of predictable barriers, sorted by which are construct-relevant
- Evidence about where current designs already fail at the edges
In backward design, assessment is settled before activities. Here the question is how learners will demonstrate the fixed competency through varied means, while the assessment continues to measure the intended construct and nothing incidental. Construct-irrelevant variance — when scores are influenced by factors irrelevant to the construct — is the central threat to validity this stage exists to remove.
Flexible, authentic assessment widens the means of demonstration without widening what is being measured. An assessment that quietly rewards reading speed or tech fluency produces invalid signals about who can actually do the work.
- Assessment options aligned to the goal, with the construct held constant
- A documented rationale for why each option measures the same competency
- Rubrics that evaluate the construct rather than the format chosen
With the goal and assessment fixed, the design provides multiple means of engagement (the "why" of learning), representation (the "what"), and action and expression (the "how"). Options are placed deliberately where the barrier analysis showed they were needed — not scattered for their own sake.
A key judgment here is knowing how many options is enough. Unlimited choice imposes its own cognitive load, so the work is offering meaningful alternatives at the points of real friction while keeping the path navigable. The three principles provide the structure; cognitive load theory provides the restraint.
- A learning architecture with options mapped to identified barriers
- Engagement, representation, and expression alternatives that share one goal
- A storyboard or prototype for stakeholder and learner review
Development produces the actual materials: multimodal content, accessible documents and media, and platform builds. Accessibility is built in, not retrofitted — alt text, caption quality, keyboard navigability, color contrast, and flexible pacing are part of construction rather than a later remediation pass.
Standards such as WCAG function as the floor, not the ceiling: conformance is necessary but does not by itself guarantee that anyone is actually included. Sound multimedia design principles govern how representations are built across modalities so that flexibility aids rather than overwhelms.
- Standards-conformant, multimodal learning materials
- Accessible media — captions, alt text, contrast-compliant visuals, navigable structure
- Production approaches that make added options sustainable rather than one-off
The design meets real learners — ideally first in a pilot — and feedback is gathered from across the range, with particular attention to those at the edges the original analysis identified. Involving learners who experience the barriers directly consistently improves relevance and reduces post-launch revision.
This stage treats co-design as a method rather than a courtesy: the people whose access is most often overlooked are the ones best positioned to reveal where a design still falls short.
- Pilot findings on where options were used and where barriers persisted
- Learner feedback weighted toward the edges of the range
- A prioritized list of refinements before full release
UDL is a cyclical process, not a one-time build. Evaluation asks the question the whole effort exists to answer: did the options that were built actually reduce barriers, and for whom? Evidence comes from completion and performance patterns across the learner range, the uptake of provided options, and direct learner report — not from the assertion that a design is inclusive.
Findings feed back into the goal and barrier analysis, and the cycle repeats. Ending on evidence rather than aspiration is what distinguishes rigorous practice from good intentions.
- Evidence on barrier reduction disaggregated across the learner range
- Analysis of which options were used, and which went unused
- A prioritized iteration plan that returns to Stage 1
What Shapes a Learner-Variability Project
No two projects are identical. Decisions about how the goal is framed, when access is built, how options are bounded, and how validity is protected determine whether a design genuinely reaches the full range — or only appears to. These dimensions also shape how the work is structured collaboratively with the organization.
Goal–Means Separation
Have we separated the competency from the method?The most consequential decision in the work. A goal that embeds a method ("write," "present live") quietly excludes learners for whom that method is the barrier. Stating the goal as a fixed competency with flexible means is what makes every downstream option coherent rather than arbitrary.
Proactive vs. Retrofit
Are we building access in, or accommodating after?The reactive model — wait for disclosure, then retrofit — is slow, stigmatizing, costly, and legally exposed. Designing access in from the start reduces the number of retrofits ever needed. This reframe, from accommodation to proactive design, is the defining move of the field.
Modality & Multimodality
Which options reduce barriers, and which just add noise?Multiple representations and pathways serve different learners — but each must earn its place against a barrier identified in analysis. Options added for their own sake increase cognitive load without increasing access. Modality follows the analysis; it does not precede it.
Learner Co-Design
Are the people who experience the barriers in the room?Co-design with learners — especially those whose access is most often overlooked — improves relevance and surfaces barriers a design team cannot see from the outside. It is particularly valuable when designing for communities not well represented among the designers or subject-matter experts.
Assessment Validity
Is the assessment measuring the competency or an incidental barrier?Flexible assessment must still measure the intended construct. Construct-irrelevant variance — scores shaped by reading speed, tech fluency, or anxiety — produces invalid signals about who can do the work. Protecting validity while widening the means of demonstration is core expert judgment.
Scale Economics
How do we add options without multiplying production cost?Flexibility has historically meant more production labor, which is why personalization is often cut as unaffordable. UDL as the design logic — paired with current tooling, including AI — is what makes multiple representations and pathways economical rather than prohibitive at scale.
Core Competencies in Designing for Variability
Where the foundations describe what an expert knows, competencies describe what an expert can reliably do — knowledge, skill, and disposition fused into demonstrable practice. They are not equally weighted: the design competencies in the second cluster are the heart of the craft, and everything else exists to serve them.
Framework Fluency
Using the UDL framework rather than reciting it — the three principles, the current guidelines and considerations, and crucially the rationale behind them. An expert can explain to a skeptical subject-matter expert why designing for variability is the realistic default rather than an accommodation afterthought, and can locate precisely where in an experience a given barrier occurs.
Goal–Means Separation
The linchpin skill of the discipline: stating a learning goal so the competency is fixed but the means of reaching and demonstrating it stay flexible. Most design failures trace to a goal with a method baked into it. A designer who has internalized this tends to generate the right options, assessments, and barriers-to-remove almost as downstream consequences.
Barrier Identification
Distinguishing construct-relevant difficulty — the desirable struggle that is the learning — from construct-irrelevant barriers that block some learners for reasons unrelated to the goal. Removing the second without flattening the first is genuine expert judgment, not a checklist exercise.
Proactive Option Design & Alignment
Building flexible pathways and multiple means in from the start — and knowing how many options is enough, since unlimited choice creates its own cognitive load. Backward design keeps goals, assessments, methods, and materials coherent so that added flexibility does not fracture the experience.
Accessible, Multimodal Production
Actually making accessible, multimodal materials: meeting accessibility standards (alt text, captions, accessible documents and navigation), designing representation across modalities using sound multimedia principles, and fluency with the relevant tools — the platform, authoring software, assistive technologies, and increasingly AI tools that make flexibility scalable.
Engagement & Facilitation, Tuned for Adults
Designing for motivation — autonomy, relevance, competence, and value — in a way that respects what adult learners bring: substantial prior experience, the expectation of self-direction, and usually voluntary participation. The engagement principle does the most work with adults and is the one most often underdeveloped in practice.
Valid, Authentic Assessment & Evaluation
Two distinct abilities: designing flexible, authentic assessment that lets learners demonstrate mastery through varied means while still measuring the intended construct; and evaluating one's own designs — gathering evidence on whether the options built actually reduced barriers, then iterating. Every design is treated as a hypothesis to be tested.
Collaboration, Advocacy & Reflective Practice
Expertise rarely operates solo. This includes co-design with subject-matter experts, accessibility staff, and learners themselves; advocacy and change leadership, since adopting UDL is usually an institutional shift that must be coached and championed; and reflective practice anchored in a real equity commitment, so the work does not decay into a compliance checklist.
What Designing for Variability Produces
Rigorous design for learner variability creates change at two horizons. The shorter-term outcomes are tangible and measurable within a program. The longer-term shift changes how an organization builds and sustains access over time.
The design works at the edges, not just the middle
A program built for the range rather than the average reaches learners who were previously served poorly or not at all — without a separate accommodation process for each one.
Assessments produce valid signals
When construct-irrelevant variance is removed, results reflect the actual competency rather than reading speed, tech fluency, or anxiety — improving the quality of downstream decisions about who can do the work.
Fewer retrofits are required
Access designed in from the start reduces the volume of reactive, per-person accommodations — lowering cost, delay, and the stigma attached to disclosure.
Evidence replaces assertion
Disaggregated data on barrier reduction lets a team show that a design is inclusive rather than claim it — and shows precisely where it is not yet.
Inclusion becomes a design property, not a service
Access shifts from something negotiated repeatedly, learner by learner, to something engineered into the experience once — changing the economics and the ethics of the work.
Personalization becomes affordable at scale
UDL as the design logic, paired with current tooling, makes multiple pathways economical rather than prohibitive — so flexibility survives budget scrutiny instead of being cut first.
A culture of iteration takes hold
Treating each design as a tested hypothesis builds an organizational habit of measuring barrier reduction and refining — rather than declaring a build finished and inclusive.
Equity moves beyond compliance
When standards are treated as the floor and design as the real work, the practice addresses social, cultural, and identity barriers alongside cognitive ones — rather than clearing legal review while including no one new.
Where Is Your Organization in This Work?
This reflection is intended to surface useful questions about how your organization currently approaches learner variability. There are no right or wrong answers — choose the response that most honestly reflects your situation.
Academic & Professional Citations
The knowledge claims, frameworks, and evidence in this resource draw on established scholarship and professional practice. Sources are grouped by the foundation they primarily support, and all are independently verifiable.
CAST. Universal Design for Learning Guidelines, Version 3.0. Wakefield, MA: CAST, 2024. Released July 30, 2024; the current version of the three principles and their considerations.
udlguidelines.cast.orgMeyer, A., Rose, D. H., & Gordon, D. Universal Design for Learning: Theory and Practice. Wakefield, MA: CAST Professional Publishing, 2014.
udltheorypractice.cast.orgSweller, J. "Cognitive Load During Problem Solving: Effects on Learning." Cognitive Science, 12(2), 257–285, 1988. Source for the limits of working memory in instructional design.
Mayer, R. E. Multimedia Learning. 2nd ed. Cambridge: Cambridge University Press, 2009. Principles for designing representation across modalities.
Rose, T. The End of Average: How We Succeed in a World That Values Sameness. New York: HarperOne, 2016. Source for the argument that the "average learner" is a statistical artifact.
Knowles, M. S., Holton, E. F., & Swanson, R. A. The Adult Learner: The Definitive Classic in Adult Education and Human Resource Development. 8th ed. London: Routledge, 2015.
Mezirow, J. Transformative Dimensions of Adult Learning. San Francisco: Jossey-Bass, 1991.
Kolb, D. A. Experiential Learning: Experience as the Source of Learning and Development. Englewood Cliffs, NJ: Prentice Hall, 1984.
Schön, D. A. The Reflective Practitioner: How Professionals Think in Action. New York: Basic Books, 1983.
Deci, E. L., & Ryan, R. M. Intrinsic Motivation and Self-Determination in Human Behavior. New York: Plenum Press, 1985.
Bandura, A. Self-Efficacy: The Exercise of Control. New York: W. H. Freeman, 1997.
World Wide Web Consortium (W3C). Web Content Accessibility Guidelines (WCAG) 2.2. W3C Recommendation, October 2023.
w3.org/TR/WCAG22/Sins Invalid. Skin, Tooth, and Bone — The Basis of Movement Is Our People: A Disability Justice Primer. 2nd ed. Sins Invalid, 2019. Source for the principles of disability justice (Patty Berne et al.).
sinsinvalid.orgConnell, B. R., Jones, M., Mace, R., Mueller, J., Mullick, A., Ostroff, E., Sanford, J., Steinfeld, E., Story, M., & Vanderheiden, G. The Principles of Universal Design, Version 2.0. Raleigh, NC: Center for Universal Design, North Carolina State University, 1997.
design.ncsu.edu/research/center-for-universal-designWiggins, G., & McTighe, J. Understanding by Design. 2nd ed. Alexandria, VA: ASCD, 2005. Source for backward design and constructive alignment.
Messick, S. "Validity." In R. L. Linn (Ed.), Educational Measurement, 3rd ed., pp. 13–103. New York: American Council on Education / Macmillan, 1989. Foundational source for construct-irrelevant variance as a threat to validity.
Gay, G. Culturally Responsive Teaching: Theory, Research, and Practice. 3rd ed. New York: Teachers College Press, 2018.
Ladson-Billings, G. "Toward a Theory of Culturally Relevant Pedagogy." American Educational Research Journal, 32(3), 465–491, 1995.