Hover exists only on pointer devices; touch and keyboard have no hover state
Aliases: hover state · pointer device · touch interaction · keyboard interaction
What it is
Hover modality dependence means that hover is created by a continuously locatable pointer entering a hit region, so it exists only for mice, trackpads, hovering pens, and similar devices. A touch tap normally activates directly; keyboard movement changes focus. Neither has the same “over it, not yet committed” state. Treating hover as a universal current-item marker leaves touch and keyboard users without information or causes focus, press, and selection to be misread as hover.
Why it happens
Hover affords low-cost preview: people explore clickability, hints, or secondary commands before activation. Touch lacks this consequence-free intermediate distance—contact can obscure and initiate action—while keyboard exploration follows discrete focus. When essential explanation, available command, or risk appears only on hover, nonpointer users either cannot discover it or must guess through long press, accidental action, or external help. Different device capabilities mean the same visual component cannot presume the same state semantics.
Studying it
For the same task, test mouse, touch, keyboard, and assistive-technology routes for discovery, supplementary information, current-state interpretation, and completion. Do not test only responsive width: a narrow laptop has hover, while a large touch display does not. Record whether touch users trigger actions to seek information, keyboard users lack an equivalent route, and mixed devices retain erroneous hover appearance after switching input.
Where it stops holding
Some pens and proximity sensors support hover-like capability, but cannot be assumed universally; input availability can also change dynamically. Long press may supplement discovery but is not equivalent because it costs time, conflicts with platforms, and is less discoverable. Decorative hover animation need not be reproduced everywhere; only content carrying understanding, decision, or necessary commands needs an equivalent route.
Applying it
- Inventory every hover-triggered information item and command, marking whether it changes task understanding or outcome; consequential items need touch and keyboard paths.
- Use focus style and explicit controls for keyboard, and persistent summary, details, or context menus for touch instead of mechanically porting hover visuals.
- Adapt hints to actual input capability on mixed devices, without making transient pointer state the only information source.
- Validate first-use tasks with no mouse or hover: people should find equivalent information and complete work without accidentally triggering the primary action.