Accessibility statement

How accessible our interview pages are, and what we know still falls short.

Causal Map Ltd wants everyone who is invited to a QualiaInterviews interview to be able to take part. This statement covers the interview pages that respondents use: the consent screen, the conversation with the AI interviewer, structured questions, and the closing screens. It does not cover the editor that researchers use to set up interviews and read the results.

Compliance status

The respondent interview pages are partially compliant with the Web Content Accessibility Guidelines (WCAG) 2.2 at level AA. The known exceptions are listed below.

How we assessed the pages

We reviewed the respondent pages ourselves against WCAG 2.2 AA in September 2026, and on 30 September 2026 put the fixes that review produced into service. We ran the axe-core accessibility checker in the Chrome browser over each stage of an interview (the consent screen, the conversation, the restart confirmation and the closing screen) in both the light and dark themes, and measured the contrast of every piece of visible text. We then checked by hand: using the pages with the keyboard alone, the order and visibility of keyboard focus, the pages at 320 pixels wide and at 200% zoom, increased text spacing, the size of touch targets, reduced motion, and the messages announced to screen readers. After the fixes, the checker reported no failures at any of those stages.

The pages have not been audited by an independent organisation. We have not yet tested them with a screen reader such as NVDA or VoiceOver, or with disabled respondents, and we reviewed the spoken-answer features by reading their code rather than by testing them in use.

What respondents can expect

  • No installation or account. The interview runs in any modern browser on a phone, tablet or computer, and text can be enlarged with the browser’s own zoom.
  • Keyboard use. Controls show a visible focus outline when reached with the keyboard, and the conversation can be scrolled from the keyboard. Dialogs (consent, restart and closing) keep focus inside until they are closed and then return it to where it was, and Escape cancels a restart or a correction to an earlier answer.
  • Screen readers. The page has a main landmark and a heading, and the conversation is a labelled log. Each interviewer reply is announced once it is complete, not while it is still being written, and waiting, transcription and error messages are announced too. Icons are hidden from screen readers, and the answer box and conversation have names in the interview’s language.
  • Structured questions. Question sets are labelled groups, selected answers are announced as selected, and required questions are announced as “required”. If a respondent tries to send without answering a required question, the page says so in words and moves focus to the first unanswered question.
  • Progress. The progress bar reports how much of the interview is complete to assistive technology as well as visually.
  • Colour and contrast. In our standard colours, text, including secondary text and button labels, has a contrast ratio of at least 4.5:1 in both the light and dark themes.
  • Motion. If the device is set to reduce motion, animations such as the waiting indicator and fades are switched off.
  • Language and direction. The page declares the interview’s language so that screen readers pronounce it correctly, and right-to-left scripts such as Arabic are laid out right to left.
  • Voice as an alternative to typing. Where the researcher enables it, respondents can answer by speaking; their speech is transcribed and shown as text.
  • No time limit on answering. Respondents can take as long as they need over each answer, and an interview does not expire while it is left open.
  • Stopping and returning. Respondents can stop at any point, and on the same device and browser can come back later and continue where they left off.

Known limitations

  • The researcher editor has not been reviewed. It is designed for laptop and desktop screens and may present barriers to researchers who use assistive technology.
  • Interviewer wording is generated by AI. How plain the questions are depends partly on the instructions the researcher writes. We advise researchers to ask for short, plain questions, and they can test every interview with an automated test respondent before launch.
  • Some interface text is machine translated into the interview’s language on first use. Researchers can check and edit these translations before launch; unchecked translations may be imperfect.
  • Voice answers need a microphone and a browser that permits microphone access. Text remains available in every interview. A single spoken answer stops recording after 55 seconds, and screen readers are not told when recording starts or stops.
  • The closing screen moves on by itself. When the research team has set a page to go to after the interview, the closing screen opens it after a few seconds, or after a minute when the respondent is offered a copy of their transcript, and there is no control to pause this (WCAG 2.2.1).
  • A research team’s own brand colour is chosen by the researcher, and we do not yet check it for contrast.
  • Text written by the research team, such as the interview description on the consent screen, can include images, and we cannot check that the team has given them text alternatives.
  • Some warnings appear in the browser’s own dialog boxes, whose appearance we cannot control.
  • Resuming on a different device is not supported, for privacy reasons.

Feedback and contact

If you meet a barrier when taking part in an interview, or need the interview in another format, please tell us:

If you were invited by a research team, you can also contact them using the details on the interview’s opening screen.

About this statement

This statement was prepared on 30 September 2026, based on our own review of the respondent pages that month. We will review it whenever the respondent pages change significantly, and at least once a year.

Causal Map Ltd, 39 Apsley Road, Bath BA1 3LP, United Kingdom. Registered in England and Wales, company number 12926630.