Accessible Online Assessments: A Practical Employer Checklist

An online assessment can appear simple to its designer yet present a significant barrier to a candidate. A countdown may be announced repeatedly by a screen reader. A drag-and-drop question may be impossible to complete using a keyboard. Instructions may rely on colour alone, while a timed session may end before someone using assistive technology can navigate its controls. These are not merely cosmetic defects. They can change whose knowledge and skills the assessment is capable of measuring.

Accessible online assessments give candidates a reasonable opportunity to demonstrate the competencies relevant to the decision. Accessibility begins before a candidate opens the test and continues through invitations, identity checks, instructions, questions, submissions, score reports and support. It is both a product-design issue and an employer operating responsibility.

This guide provides a practical employer checklist, an example review process and guidance on making defensible decisions when access needs and assessment conditions interact. It does not certify any platform as accessible or guarantee compliance with a particular jurisdiction's requirements.

Accessibility is part of assessment quality

An assessment may be intended to measure numerical reasoning but inadvertently measure a candidate's ability to operate an inaccessible chart. A workplace judgement exercise may test reading comprehension as much as decision-making because its language is unnecessarily complex. A task intended to measure technical knowledge may instead reward familiarity with a particular input device.

The central question is: Does an avoidable barrier interfere with demonstrating the skill that the employer actually needs to measure?

This question requires two kinds of review. Accessibility specialists examine how people interact with the assessment. Job experts and assessment specialists examine whether changing that interaction would alter the intended construct—the knowledge, ability or behaviour being measured. Both perspectives matter.

Accessible design does not mean that every candidate must receive an identical presentation or that all job requirements can be removed. It means the employer must distinguish relevant requirements from incidental barriers and consider appropriate adjustments under applicable law and policy.

Start with the full candidate journey

An accessibility review limited to the question screen will miss important obstacles. Map the process from the first invitation to the final communication.

Employers should test the entire journey, including the exceptional paths: a forgotten password, interrupted session, failed upload, adjustment request and retest invitation.

1. Make candidate invitations and instructions accessible

Candidates should know what the assessment involves before they begin. State the approximate duration, format, required equipment, whether timing applies, what materials are permitted, the technical support route and how to request an adjustment. Avoid requiring candidates to disclose sensitive health details to a general recruitment inbox.

Provide key information as selectable text rather than only in an image. Use descriptive link text, logical headings and clear instructions. Do not depend exclusively on phrases such as “click the green button” or “see the diagram on the right.” The button or diagram may not appear in that location for every user.

Example: An improved invitation

Instead of: “Complete the test by Friday. Click the blue link. No exceptions.”

Use: “Please complete the assessment by the stated deadline. The assessment includes 40 questions and normally takes about 45 minutes. Open the assessment using the link below. If you need an accessibility adjustment or encounter a technical barrier, contact the assessment support team before starting where possible. We will explain available arrangements and how requests are reviewed.”

This is illustrative copy, not a promise that every adjustment can be provided or that every Talentrenza test has the stated configuration. Employers should replace the figures and arrangements with the actual assessment details.

2. Test keyboard access and focus behaviour

Many candidates use a keyboard or alternative input device rather than a mouse. A basic review should confirm that a person can enter the assessment, move through controls, select responses, review answers and submit without requiring pointer-only gestures.

Check that:

• The visible focus indicator is clear enough to locate the active control.
• Focus moves in an order that makes sense with the page's content.
• Dialogues and overlays do not trap focus unexpectedly.
• Every interactive control has an understandable name and state.
• A candidate can recover from errors without restarting the whole test.
• Question navigation does not move focus unpredictably or discard an answer.

A keyboard-only pass is useful, but it is not a substitute for testing with assistive technologies and people with relevant access needs. The exact test methods should follow current accessibility standards and the technologies used by the intended population.

3. Check compatibility with screen readers and other assistive technology

A screen reader communicates page structure, labels, controls and status messages through speech or braille output. An assessment can look correct on screen while presenting confusing or missing information to a screen-reader user.

Review whether question numbers, instructions, answer choices and required fields are announced in a meaningful sequence. Confirm that feedback such as “answer saved” is available without repeatedly interrupting the candidate. Verify that a countdown does not generate excessive announcements and that important time warnings are understandable.

For tables, use appropriate row and column associations. For charts and diagrams, provide alternatives that convey the information necessary to answer the question. A generic label such as “chart image” is rarely sufficient for a question requiring comparison of plotted values.

Not every diagram can be reduced to a short text alternative without changing the task. When visual interpretation is itself relevant to the role, the employer may need a specialist review to determine a suitable assessment method or adjustment. Do not silently substitute a different construct and then assume scores remain comparable.

4. Support readable text and flexible presentation

Candidates may need enlarged text, increased spacing, high contrast or a different presentation mode. Check that content remains usable when text is resized and when the viewport is narrow. Avoid fixed layouts that hide answer options or submission controls.

Use plain, precise language where advanced reading is not part of the target competency. Avoid unnecessary idioms, culturally narrow examples and dense instructions that create additional processing demands. However, do not rewrite a verbal reasoning assessment so extensively that it no longer measures the intended reading skill; review changes against the assessment blueprint.

Information should not rely on colour alone. A red error border should be accompanied by text explaining the problem. A chart using two coloured lines may also need line styles, labels or a data table where appropriate.

5. Review time limits carefully

Time limits can create barriers, but removing a limit can also change what a test measures. The correct approach begins with the assessment's intended purpose.

Ask whether speed is a genuine, job-relevant requirement or merely a convenient administrative setting. If the test is intended to measure accuracy and reasoning, a restrictive timer may introduce avoidable speed pressure. If response speed is genuinely essential, the employer should document why and consider appropriate ways to assess the requirement fairly.

Worked example: A timed assessment

A fictional employer uses a 40-question test with a 45-minute standard duration. That allows an average of 67.5 seconds per question: 45 × 60 ÷ 40. The average does not mean each question needs the same amount of time; some may require substantially longer.

A candidate requests additional time because the standard interface presents an access barrier. The employer should not simply apply an arbitrary percentage or reject the request on the ground that “everyone receives 45 minutes.” Instead, it should review the task's purpose, the candidate's functional access needs, available alternatives, applicable obligations and the likely effect on score interpretation. The decision and rationale should be recorded with appropriate privacy safeguards.

The calculation illustrates the time budget only. It does not validate the test's duration or establish a universal adjustment rule.

6. Design an adjustment request process

An employer should offer a clear, confidential way to request reasonable adjustments. Candidates should not have to navigate an inaccessible test to discover how to request help.

A practical process has six stages:

1. Inform: Explain that adjustment requests can be made and identify the contact route.
2. Receive: Gather only information necessary to understand the access barrier and proposed support.
3. Review: Consider the assessment purpose, available options, candidate needs and applicable requirements.
4. Decide: Communicate the arrangement in time for the candidate to prepare.
5. Implement: Confirm that the agreed configuration or alternative actually works.
6. Record: Document the decision and operational outcome while restricting sensitive information.

Examples may include an accessible format, assistive technology compatibility, modified presentation, scheduled breaks, additional time where appropriate or an alternative assessment method. These are possibilities, not automatic entitlements or universally equivalent substitutes.

A decision-maker should avoid assuming that a diagnosis alone dictates a particular adjustment. The practical barrier, the assessment's purpose and the candidate's circumstances all matter.

7. Avoid collecting unnecessary sensitive information

Adjustment requests may involve disability or health information. Employers should avoid circulating this information to hiring managers when only the operational arrangement is relevant. An assessment administrator may need to know that an extended session is authorised; they may not need to know the candidate's medical history.

Define who receives requests, who approves adjustments, how decisions are communicated, how long records are retained and when records are deleted. Use applicable privacy and employment requirements rather than adopting an unsupported universal retention period.

Keep the assessment score separate from the adjustment request record where practical. Do not automatically mark a candidate's report with sensitive adjustment details. If an adjustment materially affects how scores should be interpreted, obtain competent assessment advice and communicate only what decision-makers legitimately need.

8. Test charts, mathematical content and interactive tasks

Different item types present different accessibility challenges.

Charts and data interpretation: Ensure labels, units and relevant relationships can be perceived. Where a data table provides an equivalent route to the target reasoning task, consider offering it. Confirm that the alternative does not supply an answer that the original task required the candidate to infer.

Mathematical expressions: Use representations that can be interpreted by relevant assistive technologies. Test how fractions, exponents, symbols and units are announced rather than assuming a visual formula will be understood.

Drag-and-drop matching: Provide a keyboard-operable alternative if the interaction itself is not what the test intends to measure. A candidate should not fail a knowledge question merely because a mouse gesture is required.

Video or audio scenarios: Consider captions, transcripts and appropriate alternatives. The right accommodation depends on whether understanding the audio or visual signal is itself a relevant competency.

Code editors and technical simulations: Test focus behaviour, keyboard shortcuts, screen-reader support and error feedback. A visually polished editor may still be unusable with certain assistive technologies.

A single accessible question template does not prove that all content entered into it will be accessible. Every new item format and significant content change needs review.

9. Make technical support part of the assessment design

A candidate who experiences an inaccessible control or unexpected disconnection needs a clear route to help. The support team should know how to distinguish an access problem from suspected misconduct and how to preserve fairness across candidates.

Document whether the timer continues during an interruption, what happens to saved answers, how a session can be resumed and when a retest is permitted. Communicate the policy in advance where possible.

Scenario: A screen reader cannot announce the answer options

A candidate reports that Question 17's choices are not announced. The support team should record the issue, avoid telling the candidate to guess, and determine whether the problem is isolated or affects others. Depending on the evidence and assessment design, an appropriate response may involve pausing the attempt, offering a corrected accessible version or arranging a fair retest.

The employer should investigate whether the item was defective, whether a platform change introduced the problem and whether other candidates were affected. Any score correction must follow an approved process; simply awarding full marks to one candidate may create new comparability issues.

10. Validate accessibility claims through evidence

An accessibility statement is a useful starting point, not a substitute for testing. Procurement teams should ask vendors which product versions, interfaces and item formats were reviewed, what methods were used, which issues remain open and how remediation is managed.

The Web Content Accessibility Guidelines (WCAG) provide a widely used technical reference for web accessibility. Employers should confirm the applicable version and conformance expectations rather than assuming a product's marketing claim covers every workflow. A documented review might combine automated checks, manual keyboard testing, assistive technology testing and user research. Automated tools alone cannot identify every barrier.

Ask whether the review covered candidate registration, invitations, authentication, questions, timers, submission, results and support. Verify whether custom employer-uploaded questions are included in the scope. Distinguish technical accessibility evidence from psychometric evidence about what scores mean after an adjustment.

11. Create an accessibility acceptance checklist

Use this checklist before launching an assessment and after significant changes.

Candidate communication

• Invitations contain selectable text and descriptive links.
• Candidates know the format, timing, equipment needs and deadline.
• An accessible adjustment-request route is available before the test.
• Support contact details are visible and understandable.

Interface and question delivery

• Core tasks can be completed without a mouse where appropriate.
• Focus order and visible focus are usable.
• Controls, labels and errors are meaningfully announced.
• Text remains readable when resized or reflowed.
• Colour is not the sole means of conveying information.
• Charts, tables, equations and media have suitable accessible representations.
• Timer announcements do not overwhelm or mislead candidates.
• Candidates can review and submit answers without unexpected loss of work.

Employer operations

• The adjustment process has an owner and clear response route.
• Decisions are recorded without unnecessary disclosure of sensitive information.
• Technical incidents and retests follow a documented policy.
• Scoring consequences of material changes are reviewed.
• Vendor evidence and unresolved accessibility issues are tracked.
• Results and candidate communications are accessible.

A checked box should represent actual evidence, not an assumption. Record the tester, test date, product version, methods used and unresolved limitations.

12. Prioritise fixes using risk, not convenience

An employer may discover more issues than can be corrected immediately. Prioritisation should consider the severity of the barrier, number of affected candidates, stakes of the assessment, availability of a reliable alternative and time until the next administration.

This table illustrates triage decisions, not an automatic severity rating. A defect that blocks a high-stakes assessment should not be deferred merely because it affects a small number of users.

13. Review fairness without assuming identical scores

Accessibility, fairness and psychometric validity are related but distinct. An accessible interface removes avoidable barriers. A fair process evaluates candidates against relevant criteria and handles requests consistently. Validity concerns whether evidence supports the intended interpretation and use of assessment scores.

An adjustment can improve access while also affecting the conditions under which a score was obtained. Whether that matters depends on the construct, the adjustment and the decision. Employers should not assume either that all adjusted scores are incomparable or that every adjustment leaves scores unchanged.

Where high-stakes decisions are involved, consult qualified assessment professionals. Examine whether alternative forms or conditions measure the same competencies to a sufficiently comparable standard. Document uncertainty rather than claiming exact equivalence without evidence.

Do not use observed differences between demographic groups as automatic proof of discrimination or as a reason to dismiss concerns. Investigate job relevance, accessibility, sampling, administration and other plausible explanations using lawful and appropriate methods.

14. Build an accessible assessment review cycle

Accessibility is not a one-time project. A platform update, new question type or employer-uploaded item can introduce barriers after a previous review passed.

Establish three review triggers:

Scheduled review: Periodically test representative candidate journeys and review unresolved defects.

Change-triggered review: Recheck relevant workflows after interface changes, authentication updates, new question formats or significant content revisions.

Incident-triggered review: Investigate reports of inaccessible controls, failed adjustments or candidates unable to complete tasks.

Maintain an issue register showing the affected workflow, observed behaviour, reproduction steps, impact, owner, temporary mitigation, remediation status and verification evidence. Do not record more personal information than needed.

A practical 30-day employer implementation plan

Week 1: Inventory the journey

List the assessments currently in use, candidate groups, delivery devices, item types and known support issues. Identify who owns accessibility decisions and who can pause a compromised assessment.

Week 2: Test critical paths

Run keyboard-only and relevant assistive technology checks across invitation, login, instructions, representative questions, submission and results. Include actual candidate-facing content, not only a demonstration environment.

Week 3: Fix barriers and formalise adjustments

Prioritise blockers, test workarounds, establish a confidential adjustment process and train support staff. Review whether any changes affect scoring or interpretation.

Week 4: Verify and document

Retest corrected workflows, record remaining limitations, approve launch decisions and set the next review date. Where a critical barrier remains without a suitable alternative, defer the affected assessment rather than treating the issue as acceptable by default.

The schedule is illustrative. Complex platforms, specialist assessments and unresolved high-impact issues may require more time.

How Talentrenza Assess fits into an accessible employer workflow

Employers using Talentrenza Assess should consider accessibility alongside competency selection, assessment design, candidate communications and interpretation of results. Before relying on a particular feature or question format, verify the actual deployed behaviour and the evidence available for that workflow. Do not assume that the existence of an online test builder establishes accessibility conformance, that every assistive technology is supported or that all accommodations are automated.

Where the platform does not support a candidate's access needs adequately, employers should identify a fair alternative process and obtain specialist advice when necessary. Employers retain responsibility for their employment decisions, privacy obligations and the appropriateness of assessment conditions.

Frequently asked questions

Are accessible assessments easier assessments?

Not necessarily. Accessibility aims to remove barriers unrelated to the skill being measured. It does not require lowering a justified performance standard. Any change that may affect the construct or score interpretation should be reviewed carefully.

Is a keyboard-accessible test automatically accessible?

No. Keyboard access is important, but accessibility also includes meaningful labels, readable presentation, assistive technology compatibility, clear instructions, accessible media, error handling and operational support.

Must every candidate receive the same time limit?

Not automatically. Employers should consider the assessment's purpose, relevant job requirements, individual access needs and applicable obligations. Timing changes may need psychometric review where speed is integral to the construct.

Can automated accessibility scans certify a hiring assessment?

No. Automated checks can identify certain technical problems, but they do not replace manual review, assistive technology testing, user testing or analysis of assessment validity and fairness.

Should hiring managers see candidates' medical information?

Generally, access should be restricted to what is necessary for an authorised purpose and consistent with applicable law. Hiring managers often need to know the assessment result and its legitimate interpretation, not detailed sensitive information behind an adjustment request.

What if an inaccessible question is discovered after candidates have completed the test?

Preserve evidence, assess who was affected, review score consequences and determine a fair correction or retest policy. Avoid ad hoc adjustments that create inconsistent treatment. Record the rationale and communicate appropriately with affected candidates.

Does compliance with a web accessibility standard prove an assessment is valid?

No. Technical accessibility and validity address different questions. An accessible interface can still deliver irrelevant or poorly designed items, and a psychometrically reviewed test can still contain access barriers.

Conclusion

Accessible online assessments require more than an accessible-looking interface. Employers need a candidate journey that can be navigated, understood and completed; a confidential route for adjustments; tested question formats; responsive support; and documented decisions when access needs interact with score interpretation.

The strongest approach is to design accessibility into the assessment programme from the beginning, verify it with evidence and review it whenever the technology or content changes. That helps employers assess relevant capabilities rather than avoidable barriers—and gives candidates a clearer, fairer opportunity to demonstrate what they can do.