ACCESSIBILITY
What works here and what does not yet
This page states how usable this website is with a keyboard, a screen reader, magnification and no mouse, what we have tested and what is still missing. We write nothing here that we have not measured.
Status
Status of this statement: 9 October 2026. We tested the version of the website of that same day.
Our aim
We follow the Web Content Accessibility Guidelines 2.2, level AA. For body text we aim for the contrast of level AAA. Every page is readable without JavaScript, every graphic has a data table, every control measures at least 44 by 44 pixels (links within body text excepted, source marks among them), and zoom is never locked.
How we test
On 9 October 2026 we tested 21 pages: the home page, pieces with graphics and tables, polls, every page under “About us”, contact, legal notice, privacy and the error page. Each page at 375 and at 1280 pixels wide, in light and in dark mode, automatically with axe-core against WCAG 2.0, 2.1 and 2.2, levels A and AA.
- Zoom is locked on no page, and no page overflows sideways at 375 pixels wide.
- The first step with the tab key leads to “Skip to content”; the focus is visibly outlined.
- The home page, one piece and one poll are fully readable without JavaScript.
- We have recalculated the contrast of our colours: body text 17.1 to 1 in light and 16.3 to 1 in dark mode (level AAA), secondary text at least 5.9 to 1 (level AA).
We found one violation: wide tables on the method and source standard pages could be moved sideways on narrow screens, but not reached with the keyboard. It has been fixed since 9 October 2026, as has the size of the links in the breadcrumb navigation, which were only 24 pixels high on the desktop. Automated tools do not find everything; that is why what remains to be tested by hand is set out below.
Known gaps
Each gap states what is missing, since when it has been known and by when we intend to close it.
- Not yet tested: a pass with a screen reader, operating the site by hand at 200 per cent magnification, a simulation of colour vision deficiency and a measurement with Lighthouse. Known since 9 October 2026; we will catch up by 31 December 2026.
- For some 390 pieces of text per view the automated test could not judge the contrast itself, because our colours for light and dark mode are derived from one shared definition. We have recalculated the colour pairs behind them (see above); a test for each individual place is outstanding. Known since 9 October 2026; deadline 31 December 2026.
Beyond this, as at the status given above, we know of no gaps. If you find one, please tell us.
Contact
If you come across a barrier or need a piece of content in another form, please write to info@mbgglobal.com or use the contact form with the subject “A question for the editors”. We answer within two working days.
Next test
By 31 December 2026 at the latest. We also test after every larger change to the design or the technology.