The barrier came with the licence. The duty did not.

Three ways to read this

• Fifteen minutes. Read the path through. It stands on its own.
• An hour. Read the path, then open the stop that matches where you are, with its resources beside it.
• A term. Use the seven stops as a reading order through the landscape. Taken in sequence, their resources touch every group on the map.

A note on confidentiality

The case: A mandatory module, a licensed platform, four months before term

Every learner entering the faculty’s laboratories must complete a laboratory safety module before their first session. The module runs on a platform the faculty licenses from a vendor; the licence has two years to run. It consists of a forty-minute video without captions, a set of interactive figures in which hazard symbols and equipment are dragged onto a labelled diagram, and a twenty-minute timed quiz that must be passed at eighty per cent. In the second week of term, each learner also completes a live practical in which laboratory technicians watch them handle glassware and a spill kit, and require them to explain each step aloud before performing it.

In May, the access office writes to the associate dean responsible for the faculty’s teaching laboratories, copying the learning design unit. Without naming anyone, it says that among the learners entering in September are one who uses sign language and will work through interpreters, one who uses a screen reader, and one who communicates through a speech-generating device. It asks whether the module and the practical will be accessible to them, and what the faculty needs in order to make them so.

The vendor’s accessibility conformance report is on file. Against the relevant criteria it says “partially supports” for captions, for keyboard operation of the interactive figures, for adjustable timing, and for the name, role, and value of the controls. The licence was signed by procurement two years ago; the report was attached to the contract and not read.

The request lands with the associate dean, who answers for the module, the practical, and the budget for both. What follows is the work.

Stop 1: The frame

Where the request is now

May. The office’s email and the conformance report, read together before anything is decided.

The question

Is this a request about three learners, or about a module? And what is mine to decide?

What to read

• Why Universal Design Never Eliminates Individual Accommodation — the anchor of the whole landscape. The three learners will need individual arrangements in September whatever the faculty does in May; and the module will still be inaccessible to the next cohort if that is all that is done.
• From Accommodation to Proactive Inclusive Design — the field’s defining reframe: from retrofitting access for individuals after the fact to designing so that fewer people have to ask.
• Models of Disability as Design Decisions — the model implicit in how the request is read decides what gets built. Read as three people to be arranged for, the request produces three workarounds. Read as a module with barriers in it, it produces a different module.
• Designing for the Assistive Technology People Actually Use — why conformant content can still fail in practice, and why a conformance report is the beginning of testing rather than the end of it.
• Disability Categories and the Barriers People Meet — the reference layer, written barrier-first: what a learner who signs, a learner who uses a screen reader, and a learner who uses augmentative communication meet in a video, a figure, a quiz, and a practical.

A good decision looks like

Before replying, the associate dean has the request rewritten, with the learning design unit, as barriers in the module rather than needs of three people: speech in a video with no text equivalent; an interaction that can only be completed with a mouse; a time limit on a quiz that measures recognition, not speed; controls whose names a screen reader cannot announce; and a practical whose safety check is delivered through speech. This is the frame the International Classification of Functioning supplies, in which functioning is the interaction of a health condition with activities, participation, and the environment rather than a property of the person (source 13), and it is the frame disability studies has long insisted on: a barrier is a feature of the environment and the institution, not a deficit in the learner (sources 15, 16). The three learners are the reason the barriers are visible this May; they are not the scope of the work.

The leader’s part is named at the same time. The associate dean answers for whether the module and the practical can be completed by everyone required to complete them, for the time and budget that makes them so, and for what the faculty asks of the vendor and of procurement. The learning design unit holds the evidence about what the platform does and the work of rebuilding it. The associate dean does not decide individual accommodation, which belongs to the office, and does not settle what the safety check requires at the bench without the technicians who run it. Keeping those apart is what makes the fifth stop possible.

Where practice disagreesWhether to arrange for the three learners and leave the module alone until the licence ends. Many faculties would, on the reasonable ground that a two-year licence is a fact and a rebuild is a budget. This path does both and treats neither as optional. Individual arrangements for September cannot wait for a rebuild and are owed regardless; the Supreme Court of Canada has held that equal access to a public service can require positive measures, and did so in a case about sign language interpretation (source 1). And a module that only works for some of the people required to complete it is a barrier to everyone it fails, next year and the year after. The licence sets what can be fixed by when, not whether.

Stop 2: What is owed

Where the request is now

May, the same week. Before the reply to the office is written.

The question

What does the institution owe learners who have not yet arrived, whose duty is a licensed platform, and what does the conformance report actually establish?

What to read

• The Duty to Accommodate in Learning Design — the procedural duty and the substantive one, the unified test for a standard, and accommodation to the point of undue hardship, in the Canadian frame.
• The Case Law Behind Accommodation in Education — six decisions in plain language. Three carry this stop: Eldridge, in which a government’s failure to provide sign language interpretation in a public service was held to deny equal benefit of the law (source 1); Moore, in which the service owed is described as meaningful access to education (source 2); and Meiorin, which places on the institution the burden of showing that further accommodation would be undue hardship (source 3).
• Canadian Accessibility and Accommodation Law for Learning Designers — two separate instruments in plain language: the Human Rights Code, which creates the duty to accommodate, and the Accessible British Columbia Act, which requires a plan. A platform is subject to the first whatever the second says.
• Accessible Procurement and Vendor Accountability — the decisions made before anything is built. An inaccessible platform, once licensed, sets the ceiling on everything designed inside it.
• Selecting Materials, Textbooks and Platforms — reading a conformance claim against what a person meets: what “supports”, “partially supports”, and “does not support” mean, and what a report does not say.

A good decision looks like

The reply to the office opens a process rather than promising an outcome. It says what the faculty will establish and by when, what it needs from the office, which is functional requirements and the lead times for interpreting and testing rather than anything about the learners, and that the laboratory technicians will be part of it. The duty it answers is the institution’s duty to accommodate under the provincial Human Rights Code (source 6), which the institution cannot license away: a contract can allocate work to a vendor, and the duty to the learner remains with the institution that requires the module. The duty is a positive one (source 1), and it calls for individual assessment rather than a blanket rule (source 4), which is why the reply does not say “the platform is accessible” and does not say it is not.

The conformance report is read for what it is. A completed Voluntary Product Accessibility Template is a vendor’s own statement, criterion by criterion, using the terms supports, partially supports, does not support, and not applicable (source 22). “Partially supports” is a claim that there are exceptions, and the report is expected to say what they are. At the associate dean’s request, the learning design unit writes to the vendor for the exceptions in writing, for the version tested, and for the assistive technologies tested with. Conformance to the Web Content Accessibility Guidelines is defined in terms of what is accessibility supported, which means working with the assistive technologies people actually use (source 23), and the report is evidence about the vendor’s testing, not about the module as the three learners will meet it.

Where practice disagreesWhether a vendor’s conformance report is the institution’s evidence. Many institutions treat an attached report as due diligence done. This path treats it as a self-report that narrows the testing, not a substitute for it: the learning design unit tests each “partially supports” criterion with a screen reader and with the keyboard before anything is accepted or rejected, and records the result beside the vendor’s claim. The procurement standard Canada has adopted sets out the requirements a report should be measured against, clause by clause, for web content, documents, and software (source 24), and that is what the test is written to.

Stop 3: Before anyone asks

Where the request is now

June. At the associate dean’s request, the learning design unit takes the module apart.

The question

Which barriers in this module are there by design, which parts of it are essential to safety, and in what order is the work done?

What to read

• Accessibility Triage for Existing Course Inventories — a defensible order of work when everything cannot be fixed at once: what stops a person completing first, what degrades their experience second, what is cosmetic last.
• Accessibility Remediation of Mandatory Training at Organisational Scale — the compliance training everyone must take, made usable by everyone. A safety module is the clearest case: no one can opt out of it, so no one can be left out of it.
• Captions, Transcripts and Audio Description — three artefacts, three sets of users. The video needs all three, and they are production steps with owners, not features of a player.
• Time, Pacing and Deadlines as Design Variables — a time limit is either part of what an assessment measures or a convention it inherited. A twenty-minute quiz on hazard recognition is almost certainly the second.
• Essential Requirements and the Limits of Flexibility — what the module genuinely requires, which is that a learner can recognise a hazard and knows what to do, separated from how the module happens to test it.
• Accessible Documents as Everyday Practice — structure at the point of authoring: the quiz questions, the transcript, and the practical’s procedure sheet are documents before they are anything else.

A good decision looks like

An inventory goes into the file, item by item, with the criterion each barrier fails, who can fix it, and what blocks a learner outright. The video has no captions, which fails a Level A criterion, and no audio description or text alternative for what is shown but not said (source 23). The figures cannot be operated from the keyboard, fail the criterion for name, role, and value, and offer no alternative to dragging (source 23). The quiz’s time limit cannot be turned off, extended, or adjusted, which a Level A criterion requires unless the limit is essential (source 23). Each is sorted: the captions, transcript, and description are the unit’s to produce and can be done now; the time limit is a platform setting the vendor confirms can be changed; the figures cannot be fixed by anything the unit authors and are a vendor matter or a rebuild.

The essential requirements are written at the same time, before any dispute. The associate dean brings together the technicians and the instructors who teach in the laboratories to confirm what the module must establish: that a learner can recognise the hazard classes, identify the correct response to each, and locate and use the spill kit. The eighty per cent threshold is a standard and stays. The twenty minutes is not a standard; nothing in laboratory safety is measured by how quickly a symbol is recognised on a screen, and a time limit that is not part of what is measured is construct-irrelevant variance (source 8). The associate dean removes the limit for every learner rather than extending it for some. The decision sits within the faculty’s own authority, and no one has to ask for it.

The findings have a place in the design framework they come from. Captions, transcripts, and a text alternative to the figures are matters of representation in the Universal Design for Learning Guidelines; a keyboard-operable activity and an untimed quiz are matters of action and expression (source 17). Human variability is the rule in any learning environment rather than the exception (source 18), which is why a module fixed because of three learners will be used by many more without anyone knowing. It will not serve all: a learner who signs at the practical and a learner who uses a speech-generating device at the same bench are owed arrangements that no module design supplies. Design reduces the individual duty; it does not discharge it (source 19). The practical is held for the fifth stop.

Stop 4: When someone asks

Where the request is now

August. Through the office, and only by their own choice, two of the three learners agree to try the rebuilt parts before term.

The question

How is the module tested with the people it has to work for, without making them do the institution’s work or give up more than they choose?

What to read

• Partnering with Disabled People in Learning Design — co-design without extraction: partnership that is compensated, early enough to shape decisions, and never the institution’s only quality check.
• Assistive Technology and Adaptive Strategies — what a screen reader, a speech-generating device, and a signed language each do for a person, and what each needs from content to work.
• Designing for the Assistive Technology People Actually Use — read again, now as the test plan: which screen reader and browser, on which device, against which pages.
• Interpreting, Live Captioning, Note-Taking and Alternate Formats — access services as service design: booking lead times, what the interpreter needs in advance, what the technician contributes.
• Disability Etiquette and Communication in Learning Environments — how the testing session is run, who speaks to whom, and how a conversation through an interpreter or a device is conducted without the person being talked past.

A good decision looks like

Expert testing came first. Before any learner was asked, two of the unit’s designers worked through the captioned video, the quiz, and the rebuilt figures with a screen reader and with the keyboard alone, against the inventory from the third stop, and fixed what they found. Learners test what the unit believes already works. The office makes the invitation, the learners decide, and the decision changes nothing about what they are owed in September.

The sessions are paid, at the associate dean’s direction, at the rate the faculty pays any consultant, are short, and are built on the principles of a trauma-informed approach: safety; trustworthiness and transparency; collaboration and mutuality; empowerment, voice, and choice (source 14). Each learner is told what is being tested, why, who will see the notes, and what happens next. The screen-reader user finds that the rebuilt figure announces each hazard symbol but not the drop zone, which the expert test missed; the learner who signs, through an interpreter the office booked, asks whether the practical’s procedure sheet can be provided in advance, which no one had thought of. The notes record what was found and what will change. They do not record anything about the people, and they are kept by the unit, not the office (source 5).

Where practice disagreesWhether to ask learners to test at all. Some units hold that any such request puts a learner in the position of doing unpaid accessibility work for the institution that failed them, and decline to ask. This path holds the concern and sets conditions rather than a prohibition: expert testing first, compensation, an invitation that is easy to refuse, and a record that shows the institution acted on what it heard. The disability justice principle of leadership by those most affected (source 20) and the longer tradition that research and practice on disability be governed by disabled people (source 21) are both arguments for asking, on those terms, rather than for deciding on their behalf.

Stop 5: The hard call

Where the request is now

Late August. The vendor has added captions support and an adjustable timer, and says the figures will not be keyboard-operable before the licence ends. The technicians say the spoken explanation at the bench is a safety check and cannot be dropped.

The question

Remediate, rebuild, or retire the figures; and what does the practical’s spoken check measure, and what preserves it?

What to read

• Essential Requirements and the Limits of Flexibility — the most demanding judgement in the field: when flexibility preserves what is being measured, and when it alters it.
• Access in Studio, Shop and Lab-Based Learning — protecting the safety requirements that are genuine while examining the ones that are only habit. The check is genuine; speech is the habit.
• Eliminating Construct-Irrelevant Barriers in Assessment — the validity principle underneath all of it: a result should reflect capability, not incidental obstacles.
• Accessible Live and Virtual Facilitation — access in the room: how a technician conducts a check with a learner who signs through an interpreter, or who answers through a device, at the pace the device sets.
• Faculty Development That Changes Accessible Practice — the technician’s brief is faculty development in miniature: what they need to be able to do, in the room, on the day.
• Funding Design for Accessible Learning Programs — writing the budget line so that access is the deliverable. A rebuild funded as “accessibility fixes” is cut first; one funded as the module itself is not.

A good decision looks like

The associate dean decides to rebuild the figures and funds the work. The platform’s drag-and-drop activity is replaced, for every learner, by an activity the unit authors and hosts in the learning management system: the same hazard symbols, the same diagram, operable by keyboard and by pointer, with each control named so that a screen reader announces it, and with a single-pointer alternative to dragging as the current guidelines require (source 23). The platform’s version is retired rather than kept as a parallel route. The reasoning is recorded, with the vendor’s written statement, for the licence renewal.

The associate dean reasons the practical through with the technicians rather than deciding it for them. Validity belongs to the interpretation of a score, not to the test itself (source 9), so the question is what the spoken check is taken to establish. Put in plain words, as Meiorin asks it: what does the technician need to know before the learner picks up the glassware? The technicians answer it themselves: that the learner knows the step, the hazard, and the response. Speech is how they have always learned that, not what they are checking. The knowledge is a bona fide requirement of the standard; the channel is construct-irrelevant variance (sources 3, 8). The check stays, for every learner. The learner who signs gives it through an interpreter booked for the session; the learner who uses a speech-generating device gives it through the device, and the technician waits for the device. Neither is a lowered standard. The brief that goes to the technicians says so in those words, and says how long to wait, how to address the learner and not the interpreter, and where the procedure sheet in advance came from.

Where practice disagreesWhether a separate accessible version of the figures would have done. The guidelines permit a conforming alternate version, provided it is equivalent and reachable (source 23), and some faculties would have built one for the learners who needed it and kept the platform’s version for the rest. This path retires the original, on a judgement rather than a rule: two versions of a safety module diverge over time, the one fewer people use is the one that stops being maintained, and a learner sent to the alternate version is sent to a different module. One version, built to work for everyone, is the design the duty points to. A decision reasoned this way is an argument that links the performance to the inference drawn from it (source 10), and it is written down so that the next leader does not have to make it again.

Stop 6: Making it real

Where the request is now

September. The module is live; the practical is in week two. Now every piece has to exist on the day.

The question

The module has been rebuilt and the arrangements agreed. What has to occur, and who has to do it, for each to exist when the learner reaches it?

What to read

• Interpreting, Live Captioning, Note-Taking and Alternate Formats — this time for the booking: lead time, the materials the interpreter needs beforehand, and who confirms.
• Adapting Test Papers to Alternate Formats — the quiz is a secure, time-critical document; its accessible form is produced and checked before term, not on the morning.
• Delivering Accommodated Exams in a Learning Platform — the timer setting, the attempt settings, and the assistive technology check, confirmed for the live module rather than the test copy.
• Training and Supervising Student Access Staff — the brief as a document with a walk-through, for the technicians and for any assistant in the room.
• Accessible Physical and Hybrid Learning Environments — the laboratory itself. An alarm that is only audible, a bench the interpreter cannot stand beside, a spill station behind a step: the part of the problem that belongs to facilities, and must be handed to them in writing.
• Working with Accommodation Offices and Access Advisors — the implementation gap, in which an approved arrangement is recorded and then not delivered, and what a referral has to carry to be usable.

A good decision looks like

Every line has an owner, a date, and a check, and the associate dean holds the list. The captions are reviewed against the transcript by a person, not only generated, because a caption track with the wrong hazard class is worse than none. The timer is confirmed off in the live module by the unit’s designer, with a screenshot in the file. The rebuilt activity is retested with a screen reader after the platform’s September update. The interpreter is booked through the office for the practical and for the safety briefing that precedes it, with the procedure sheet sent a week ahead. The technicians have the brief, have walked through it with the unit’s designer, and know who to call if something in the room does not work. The associate dean reports the laboratory’s audible-only alarm to facilities as a barrier with a date, and an interim arrangement for the practical is agreed with the technicians in the meantime.

In week one the associate dean’s office checks with the access office that each learner has what was agreed, in one message that asks about the arrangements and nothing about the people. An arrangement that is recorded and not delivered has discharged nothing (source 11), and the professional standards for a disability service expect the service and the units it works with to evaluate whether what was approved occurred (source 12).

Stop 7: Changing the institution

Where the request is now

December. The module ran for the whole cohort. The three learners completed it and the practical without a further request.

The question

What did this one request show about how the institution buys, builds, and assesses, and what changes so that the next leader does not sign a report nobody read?

What to read

• Accessible Procurement and Vendor Accountability — read again, now as the proposal: what procurement asks for before signing, who reads the report, and what the contract says about exceptions and remediation.
• Institutional Accessibility Planning as a Change Program — where the inventory goes, and what turns a published plan into changed practice.
• Closing the Knowing–Doing Gap — the distance between what an institution has agreed and what it does, and the leader’s part in closing it: a decision is not a change until the next licence is read before it is signed.
• Making the Case for Resourcing Accessibility Work — the argument after the numbers, and its limits: what a rebuild cost, what it would have cost at procurement, and what the duty requires regardless.
• Disability Data and the Limits of Prevalence Figures — how to use a national figure honestly when making the case for a module everyone must complete.
• Evaluating Whether Design Reduced Barriers — measuring whether the rebuilt module lowered barriers rather than assuming it did.
• Making Accessibility More Than Compliance — the module as a template: access as a dimension of design quality for every mandatory module the faculty builds next.

A good decision looks like

Three things leave the associate dean’s desk. For the faculty: the rebuilt module becomes, by the associate dean’s direction, the template for every mandatory module it builds, with captions, transcript, keyboard operation, named controls, and no time limit as defaults rather than fixes; the essential-requirements statement written in June and the technician’s brief written in August are filed with the module and reviewed each year. The note is about the design of the module; it is not an evaluation of the people who built the first one.

For procurement: the conformance report and the unit’s test results go to whoever signs the next licence, with a proposal that every report be read against the national procurement standard before signing, that “partially supports” require the exceptions in writing, and that the contract name who remediates and by when (sources 24, 22). The renewal in two years is the first test of the proposal, and the date is in the plan.

For the institution: the inventory is counted, and the count goes to the accessibility plan with the associate dean as its named sponsor. The provincial legislation requires the plan, a committee, and a feedback mechanism (source 7); ownership and sequencing are what the proposal adds. The count rests on a figure used carefully: in the most recent national survey, twenty-seven per cent of Canadians aged fifteen and over reported a disability (source 25); a module every learner must complete cannot assume otherwise of its cohort. Next year the faculty measures two things, not one: whether the modules built since were built to the template, and whether any learner had to ask. A request answered well becomes evidence, and evidence counted across modules becomes the case for the next change.

After the seventh stop

The same seven stops, followed from the other three positions, are the other paths: measuring what matters (the practitioner’s), working from the account (the participant’s), and acting before disclosure (the expert’s). The rest of the landscape is reached from the same place:

• The map, where resources sit by group and by the sources and themes they share.
• The full list, in reading order, with no diagram or scripting.

Until each resource is published, its link opens the group it belongs to.