Accessibility is not a badge a site earns once. It is a set of decisions about who can use a page, under what conditions, with what assistive technology, and what happens when the answer is "not well". This page describes the current state of the site. It is deliberately cautious: we do not claim full WCAG conformance, and we do not claim that a visual reaction test is usable by everyone.
Keyboard access
The main reaction test can be started and played with the keyboard. The play area is focusable, and the space bar or Enter activates it; the page also handles keyboard input on the document so a click is not required. The aim trainer, number memory test, and spacebar clicker have their own keyboard paths; the spacebar clicker is explicitly designed for the space bar. The CPS test accepts spacebar presses as a click-equivalent on desktop.
This is not perfect. A keyboard-only user still needs to understand a visual green signal; a blind user cannot perform the task at all, because the task is defined by a visual event. Keyboard support removes the mouse requirement; it does not remove the visual requirement. A future audio-signal mode would be a different test, not the same test with a different input, and we have not shipped one.
Focus and screen readers
The play areas are focusable and have accessible names, and the result panels are in the document rather than being injected into an inaccessible canvas. The test is designed so the important state changes are also reflected in text: the page shows "wait", "go", the attempt result, and the final summary rather than relying on colour alone.
That said, a screen reader cannot convey the exact moment a visual signal appears, and a live region that announces every state change would be too slow to serve as a reaction signal. The honest position is that the visual tests are not screen-reader playable; the surrounding explanation, the FAQ, and the result summary are readable, but the timed task itself is not. We would rather say that than pretend a stopwatch announcement is equivalent.
Colour, contrast, and flashing
The test uses a green action state on a dark background, with a light theme available. The signal is not conveyed by colour alone: the board also changes its text and state, so a colour-vision difference does not remove all information, though it still affects the experience. We avoid rapid flashing or strobing. A reaction test necessarily has sudden changes, but the signal is a steady green state rather than a flash sequence.
Contrast and focus styles are checked as part of the release process, including the keyboard focus outline. If you find a control that is hard to see or a state that is announced poorly, that is a bug worth reporting.
Motor and input differences
The reaction test accepts mouse, touch, and keyboard input. That gives people with different motor abilities more than one way to respond. The aim trainer requires moving to targets and is harder to adapt; the CPS test rewards rapid repetition; the spacebar test rewards keyboard repetition. These are different physical demands, and a person who can use one may not be able to use another.
We do not currently offer switch access, dwell clicking, or an input-remapping layer. Those would be meaningful additions for some users, but they change the task and the measurement; they should be designed as explicit alternative modes, not quietly patched into the default test.
Cognitive load and instructions
The pages try to explain the task before the timer starts, keep the first action simple, and avoid deceptive patterns. The reaction test has no account, no pop-up, and no forced tutorial; the instructions are short and in the page text, not hidden in a modal. The FAQ and guides are written in plain language where possible.
Some pages are long. A long guide is not automatically inaccessible, but it can be harder to navigate. Headings are used to structure the content, and the pages can be read linearly. If a specific section is hard to follow, tell us and we will rewrite it.
Widget and embedding
The embeddable widget uses a title attribute so the iframe has an accessible name, and the credit line is in the frame. The host page still controls the surrounding context; a site that embeds the widget is responsible for its own page accessibility. We cannot guarantee how a third-party page presents the frame.
What we do not claim
We do not claim WCAG conformance for the timed tests, and we do not claim that the tests are usable by blind users. We do not offer an audio reaction test, a switch-accessible mode, or a reduced-motion alternative that preserves the same measurement. Those are real gaps, not marketing footnotes.
What we do claim is narrower: keyboard access to the main test, text alternatives for state and results, a light/dark theme, no rapid flashing, and a process for fixing concrete accessibility bugs when they are reported.
Text size, zoom, and responsive layout
The layout is responsive down to 320px wide, and the pages can be zoomed without horizontal scrolling. The test board is sized from the viewport so the signal stays on screen on a phone. Tables that are too wide are wrapped in a scrollable region rather than forcing the whole page sideways, and the chart on the age page can be scrolled on small screens.
Zoom and text-size settings vary by browser and operating system. If a layout breaks at a specific zoom level, that is useful feedback; send the page and the zoom percentage. We would rather fix a concrete break than claim universal responsiveness.
What we have tested, and what we have not
We run automated accessibility checks on the main pages, including keyboard focus and contrast checks, and we test the layout at mobile and desktop widths. We also test the keyboard path on the reaction test and the spacebar clicker. These checks catch structural problems, but they do not replace testing with people who use assistive technology.
We have not run a formal usability study with screen-reader users, switch users, or people with low vision. That is a real limitation. The honest conclusion is that the site has had automated and manual keyboard checks, not a full accessibility audit with disabled users. We would welcome that kind of review, and we will correct what it finds.
How you can help
Specific reports are more useful than general ones. "The focus ring disappears on the CPS board after I press Tab" is actionable; "the site is inaccessible" is a starting point but harder to fix. Include the page, the browser, the assistive technology, and what you expected to happen. If you can, include a screenshot or a short recording.
We will not argue with a report about your own experience. We may ask for details, and we may explain that a limitation is inherent to a visual timed test, but the report itself is data.
What we would like to add
Several improvements would make the site more usable, but each changes the task or the engineering enough that it should be done deliberately rather than promised casually. An audio-signal reaction test would be a separate test, not the visual test with a sound; a switch or dwell-clicking mode would change the response requirements; and a formal usability study with disabled users would be more valuable than another automated scan.
We would also like clearer text announcements for state changes, a simpler mode for the long guides, and a documented accessibility statement that distinguishes supported features from unsupported ones. Those are engineering and editorial tasks, not marketing copy. Until they ship, this page describes the current state rather than the aspirational one.
Plain-language summary
- What works: the main reaction test can be played with a keyboard; the pages are readable, responsive, and available in light and dark themes; state and results are shown in text as well as colour; there is no rapid flashing.
- What does not work: the timed tests are visual and are not screen-reader playable; there is no audio-signal version, switch access, or dwell-clicking mode; the aim and CPS tests have their own motor demands.
- What we claim: keyboard access, text alternatives for state and results, and a process for correcting concrete bugs — not full WCAG conformance.
- What to do: report a specific barrier with the page, browser, and assistive technology; we will investigate and say honestly what we can and cannot fix.
Reporting an accessibility problem
If something blocks you — a control that cannot be reached, a state that is not announced, text that is too small, or a result you cannot read — email hello@reflextimer.com with the page and, if possible, the browser and assistive technology you use. We will investigate and fix what we can; where a limitation is inherent to a visual timed test, we will say so rather than promise a fix we cannot deliver.
References and further reading
- W3C. Web Content Accessibility Guidelines (WCAG) — the standard this page uses as a reference, not a conformance claim.
- Methodology — how the tests are built and scored.
- Reaction Time Test — the main test, with keyboard support.
- Contact — report a concrete accessibility problem.