An Interview Adjustment Is Not a Lower Hiring Standard.

Separate the skill from the delivery channel

A candidate is asked to solve a data-analysis problem on a video call while reading dense instructions from a small shared screen. The job requires analytical reasoning and stakeholder communication. It does not require reading tiny text under an artificial timer. If the format blocks access, the assessment may be measuring the barrier as much as the competency. ^↗

A reasonable adjustment changes how someone can demonstrate the same job-related ability. It does not automatically waive an essential function, award extra points or guarantee a pass. The U.S. EEOC guidance for job applicants and the UK government's recruitment adjustment guidance provide jurisdiction-specific starting points; actual obligations depend on local law and the facts. ^{↗ ↗}

Treat the request as a process-design question. What competency is the exercise supposed to reveal? Which parts of the exercise are essential to that competency? Which parts are merely a chosen delivery mechanism? The distinction protects both fairness and the quality of evidence. ^↗

The equivalence test for an alternative format

Write the competency in plain language before choosing the test. For example: 'Identify material risks in an unfamiliar dataset and explain the recommendation.' Then list the observable behaviors: checking assumptions, finding inconsistencies, making a defensible recommendation and acknowledging uncertainty. ^↗

Now compare two formats: a timed live screen share and a downloadable accessible spreadsheet with a scheduled discussion. Both may measure the same behaviors if the task difficulty, allowed resources, evaluator rubric and evidence expectations are aligned. But if the role genuinely requires live incident response, an untimed take-home task may not be equivalent for that specific competency. ^↗

Never improvise a completely different rubric for one candidate. Use a documented equivalence map: target competency, original task, barrier, proposed alternative, invariant scoring anchors and approver. The WCAG 2.2 standard helps with digital accessibility, while assessment validity still requires job analysis. ^↗

Four examples recruiters can actually use

Example one: a written reasoning exercise delivered as an inaccessible PDF. Offer a tagged document or accessible HTML, keeping the same case facts and scoring criteria. Example two: a candidate requests captioning for a video interview. Provide captioning or another effective communication arrangement, and avoid scoring the captioning mechanism as a communication weakness. ^{↗ ↗}

Example three: a candidate needs breaks during a lengthy assessment. Separate total task time from elapsed session time and decide whether sustained continuous performance is actually an essential job demand. Example four: a candidate requests advance instructions. Provide the same process overview to everyone where feasible; explain what preparation materials are permitted and why. ^↗

Do not require the recruiter to adjudicate a medical condition. Route sensitive information to an authorized accommodation contact, disclose only what interviewers need to administer the format, and keep medical details out of the scoring notes. ^↗

Train assessors on what must stay constant

Interviewers need to know the unchanged competency and rubric, the alternative format they will see, and what not to infer from it. They do not need the candidate's diagnosis. A score of 'strong' should mean the same observed job behavior across formats. ^↗

Keep the scoring anchors behavioral. 'Explains a decision using two independently checked data points' is observable. 'Looks confident while sharing the screen' is often a weak proxy. The latter can be distorted by disability, connectivity, language or simply the chosen medium. ^{↗ ↗}

If an alternative format materially changes what can be observed, flag the limitation and design a complementary observation for all similarly situated candidates. Do not fill the evidence gap with a personality judgment. ^↗

Build a request path candidates can find

Put a clear adjustment contact in the invitation, not only in a policy hidden in a footer. Confirm the format, logistics and timing privately before the assessment. Test the actual platform with keyboard navigation, captions and screen-reader basics. The W3C preliminary checks are useful but are not a substitute for a full accessibility evaluation. ^{↗ ↗}

Track requests by stage, response time and resolution, with strict access controls and aggregate reporting. A spike in requests for the same barrier can reveal a broken default design. Improving the default for everyone is often better than handling the same workaround repeatedly. ^↗

After the assessment, ask whether the arrangement allowed the candidate to demonstrate the target skill. Avoid using that feedback as an input to the hiring score. ^↗

Run the operational check