A long description can sit in nearby visible text or on a linked page
Aliases: nearby description · described-by page · visible chart summary
What it is
A precise long description that only assistive technology can find inside an attribute is unusable to sighted people who struggle with charts, to magnifier users, and to anyone who wants to cite the prose. The description must live in visible text near the figure, or behind an explicit link to a separate page. Placement is part of access: an invisible description is not provided.
Both placements are legitimate. Nearby text is immediately available to everyone; a separate page suits a long note that will be revisited. Hiding the prose in an invisible node serves neither.
Why it happens
“Nearby” uses the same visual group and the same reading order: figure then description (or the reverse). People who can see the figure can check it; people who cannot reach the description at the next stop in the structure. A linked page pulls length out of the main prose so readers who do not need the detail are not interrupted, at the cost of one extra navigation. The link must be visible beside the figure, and its name must say “this is the data note for this figure,” not a collision-prone “details.”
The historical pattern of stuffing the note into a long-description attribute that only assistive technology reads failed twice: support was uneven, and sighted users never saw it. An invisible aria-describedby pointing at display:none text is read by some screen readers, skipped by others, and never seen under magnification. Visible is the stable channel across technologies.
Studying it
Place the same figure three ways: a visible paragraph under the figure, a beside-the-figure “text description of this chart” link to a page, and an invisible attribute only. Ask screen-reader users, magnifier users, and people who use no assistive technology but report difficulty reading charts to answer the same data questions, and time how long they take to find the note.
Independent variables: placement, whether the link name identifies the object, whether the note is structurally adjacent to the figure. Dependent variables: success and time to find the note, how often people follow an unrelated “details” link, variance across assistive-technology combinations.
Sampling should include mainstream reader–browser combinations, because invisible attributes split most there. Do not replace a finding task with a walkthrough in which the author already knows where the link is.
Where it stops holding
When the figure already illustrates adjacent prose that states the same trend, repeating a long description under the figure causes redundant announcement; keep a short name on the figure and let the prose carry the note. Interactive dashboards with many figures on one screen cannot paste an essay under each; they need one description link per figure plus a summary table. Print has no links, so the separate-page path dies and the note must be printed beside the figure. Social and content platforms that strip HTML under the image force the placement into whatever visible caption field the platform allows; do not assume your DOM survived.
Applying it
- Prefer a visible paragraph immediately above or below the figure. When the note is long, put a visible link beside the figure to a page that discusses only that figure.
- Name the link with the figure’s object (“data notes for the quarterly revenue chart”), not “details,” “more,” or “click here.”
- Do not use an invisible attribute or a
display:nonenode as the only channel. - How to check: starting from the figure, find the note with a screen reader, with a magnifier, and with ordinary browsing and assistive technology off. If any of those groups cannot find it, or land on an unrelated page, placement has failed.
Related
- Same group: J2.12.1 Long descriptions of charts should capture data trends, not only visual appearance · J2.12.3 Interactive charts need an equivalent data table as an alternative presentation · J2.12.4 Writing a long description requires understanding the data behind the graphic, not a post-hoc caption
- Nearby: J2.04 Alternative Text · J2.07 Low Vision and Screen Magnification · J2.08 Reading Order
- Search terms:
long description placement·visible description·described-by page