What a Fair Personality Assessment Should Provide for Disabled Test Takers

Start with the decision, not the disability label

Suppose a person is asked to complete a personality questionnaire before a coaching conversation. The questionnaire uses a small, crowded interface, requires a mouse, times every page, and sends a graph that a screen reader cannot describe. The immediate question is not whether the person has a particular diagnosis. It is whether the testing conditions let the person answer the intended questions.

A fair process asks what the assessment is for and what the person must do. Self-reflection and coaching normally call for a usable route to thoughtful answers and a report the reader can understand. Hiring or selection adds a higher-stakes decision, where the test user must show that the assessment is relevant to the job and that the same score interpretation is defensible for the intended population. A general personality report is not a clinical evaluation, and difficulty using a test is not evidence of a disorder.

The International Test Commission's test-use guidance says information about a person's disability should not be gathered more broadly than needed to understand the activities required by the test. It also says there is no single accommodation rule that works for every disability. That is a useful starting point: describe the access task, the support, and the effect on interpretation rather than treating a label as a prediction of how someone will respond.

Fairness means protecting the construct

In assessment language, a construct is the characteristic the instrument claims to measure. For a personality questionnaire, it might be a defined trait or facet. Construct-irrelevant variance is the part of a score produced by something outside that target. The 2025 ITC/ATP Guidelines for Technology-Based Assessment give a simple kind of example: a test taker's fluency with a computer interface or the effort required to scroll can affect performance even when those abilities are not the target.

This distinction changes the meaning of accommodation. Keyboard navigation can remove an irrelevant motor barrier when the item asks the person to choose a response. Text enlargement can remove a visual access barrier. A pause may reduce fatigue or give assistive technology time to work. Those supports can make the response better reflect the person's answers to the personality items.

But a support is not automatically neutral. If an instrument deliberately measures speed, spontaneous spoken performance, or a response mode that is part of its documented construct, changing that feature may alter what the score means. Many personality questionnaires do not make speed the target, but that cannot be assumed for every instrument. The provider should state which features are construct-relevant and which are merely delivery choices.

Accessibility should cover the whole assessment journey

A fair assessment is more than an accessible question page. The ITC/ATP guidance names the assessment lifecycle: test information, registration, accommodation-request forms, login, sample tests, the assessment interface, and test results. A barrier at any of these points can prevent a person from reaching the score or understanding it.

Look for plain instructions, labelled controls, visible keyboard focus, sensible heading structure, and enough contrast. A person who uses a screen reader should be able to move through the item, identify the response options, change an answer, and know when the page has saved or advanced. A person who cannot use a mouse should not be required to drag an answer unless dragging is genuinely part of the construct. If images, audio, or video carry meaning, equivalent alternatives should be provided where that would preserve the item.

WCAG 2.2 is a technical web standard, not a complete fairness study for a personality instrument. It covers many needs, including visual, auditory, physical, cognitive, learning, language, and neurological access, while warning that it cannot address every combination of disability. The ITC/ATP guidance makes the same practical point: technical conformance does not by itself prove that an assessment is usable. Human testing with people who use different access methods is still needed.

Compare a barrier with a proportionate support

A short comparison helps make the decision concrete. Imagine a self-report item asking a person to select how accurately a statement describes them. The person can read the statement with enlarged text but cannot reliably operate a pointer. Keyboard selection is a proportionate support because it changes the input method, not the statement or the available answers. The report can still describe the response as an answer to the same item, provided the instrument was designed to support that route.

Now change the case. The page automatically advances after a few seconds, and the provider argues that extra time would make scores incomparable. First ask what the timer measures. If it is only an interface setting, it may add irrelevant speed pressure. If speed is a documented part of the construct, a different instrument or a separately studied accommodation may be needed. The correct answer cannot come from a universal list of “fair” supports.

A third case concerns the report itself. The assessment may be completed successfully, but the result is delivered only as a colour-coded chart with no text explanation. The score has not necessarily been changed, yet the reader cannot inspect its meaning. An accessible report should identify the scale, score type, comparison group if one is used, uncertainty, and limits in a format the recipient can use.

Ask what happens when the standard format changes

Changing the format can affect both access and comparability. The ITC notes that modified tests are rarely normed on large enough groups of people with equivalent disabilities to establish a simple conversion to the standard version. That does not mean accommodations should be avoided. It means the provider should not quietly present a modified result as perfectly interchangeable with every other result.

Ask the provider to record the material change: for example, an alternative response method, a break, a different presentation format, or a change to timing. Then ask whether research or pilot evidence exists for that change and how it affects interpretation. For a low-stakes reflective use, the result may be most useful as a qualified prompt for discussion. For a selection decision, the organization may need an alternative assessment or evidence from other job-relevant methods.

The distinction is between an access support and a change to the thing being measured. If a support removes a barrier unrelated to personality, it may improve validity. If it changes the content or response process that defines the construct, the result may require a different inference. A transparent report makes that boundary visible.

Check the report, not only the test screen

A report should let the reader answer four basic questions. What was measured? How was the response converted into the reported score? What comparison group or interpretation rule was used? What should the score not be used to conclude? These questions apply to disabled and non-disabled readers alike, but inaccessible presentation can make them harder to answer.

Do not treat a percentile as a disability-adjusted truth. A percentile is a position relative to a specified comparison group, not the percentage of a trait a person possesses. If the provider changes the norm group, scoring rule, or response conditions, it should explain the consequence rather than imply that the number is self-explanatory. A band such as “higher” or “lower” also needs a stated basis and should not be presented as good, bad, capable, or incapable.

The ITC test-use guidance recommends considering the relevant comparison group, measurement error, validity evidence, variations from standard administration, and the information available about the person. It also says test data are one source of information. That is especially important when an accommodation, fatigue, technology failure, communication method, or inaccessible item may have affected the response process.

Privacy and choice are part of fairness

A person should not have to disclose more medical information than the testing decision requires. A useful accommodation process asks what access support is needed to complete the assessment, who will receive that information, and how it will be stored. It should explain whether a provider, coach, employer, or another party can see the answers and report.

The ITC guidance asks test users to tell test takers who may receive results, the conditions for release, and how to raise a complaint or report a problem. Its technology guidance also treats privacy as an equity concern. Remote monitoring, cameras, or unusual security steps can create extra barriers or anxiety, so the organization should explain their purpose and consider an effective alternative when appropriate.

In the United States, the EEOC says an employer generally may not ask a job applicant about disability or require a medical examination before a job offer. It also says applicants may be asked whether they can perform the job and how they would do so. Employment law differs by jurisdiction and situation, so this is a prompt to check the applicable rules, not a universal legal conclusion.

What a responsible provider should disclose

A provider does not need to promise that every interface works for every person. It does need to be specific enough for a test taker to make an informed choice. Useful disclosure includes the assessment's purpose, the traits or facets measured, the item and response formats, supported browsers and assistive technologies, available supports, request timing, and a contact route that itself is accessible.

It should also describe scoring and reporting: raw or standardized scores, percentiles or bands, norm or criterion interpretation, missing-answer rules, measurement uncertainty, and the intended uses. If a timed element, spoken response, image, or interaction is central to an inference, say so. If a support changes the administration, identify whether the result is interpreted qualitatively, quantitatively, or through an alternative procedure.

The ITC/ATP guidance recommends accessible sample tests and says report documents should be piloted for accessibility, usability, and understanding before deployment. That is a stronger promise than a generic accessibility badge. It connects the feature to the actual assessment task and asks whether readers can use the final report.

A practical checklist before you rely on a result

Use this checklist before taking a personality assessment or acting on its report:

1. Purpose: What decision will the result inform, and is the assessment designed for that purpose?

2. Access: Can you complete registration, sample items, login, questions, responses, and report review with your usual access tools?

3. Support: Which features are available to everyone, and which require a request? Can you ask without disclosing unnecessary medical detail?

4. Construct: Does the provider explain whether timing, reading aloud, speech, mouse use, or another response feature is part of what is measured?

5. Record: Will changes, interruptions, technical failures, or accommodations be recorded and considered in interpretation?

6. Score: Does the report identify the score type, comparison group or rule, uncertainty, and limits?

7. Privacy: Who sees the answers and report, how long are they retained, and how can you challenge an error?

8. Alternatives: If access cannot be made valid, is there another way to gather relevant evidence?

A provider that cannot answer these questions may still offer a useful informal reflection prompt, but it has not shown that the result can carry a stronger claim.