Research Guides
Pause a Research Display Without Mislabeling Its Freshness
A research table that refreshes while someone is reading can move the evidence underneath their cursor. A pause control can help, but a frozen display must not keep claiming that its values are current. Review reading control and data freshness together.
This is a manual test plan for an authorized interface or prototype. It does not claim that Colquhoun.shop has a live-data refresh or pause feature, and it is not a complete accessibility assessment.
1. Describe what updates automatically
Inventory the changing parts: values, row order, timestamps, status messages and any moving presentation. Record what triggers each change and whether the user intentionally starts it. A refresh icon alone does not tell a reviewer whether background updates also occur.
Keep the display behavior separate from data retrieval. Pausing visible changes may or may not stop background requests. Verify the implementation before claiming that the control saves network usage, stops processing or prevents provider charges.
2. Test the control without trapping the reader
W3C's Pause, Stop, Hide guidance covers automatically updating information presented alongside other content, with an exception for essential activity. It describes options to pause, stop, hide or control update frequency. Unlike its moving-content provision, the auto-updating provision has no five-second exception. W3C: Pause, Stop, Hide
Use a harmless test view. Pause updates, move to another part of the page and continue reading. Check whether the chosen state persists instead of depending on keeping the pointer or keyboard focus on the control. Record keyboard and touch observations separately; do not infer them from a mouse test.
3. Give the paused state an honest label
Imagine a fictional research queue that shows 12 completed reviews at 10:00. At 10:01, the underlying queue reaches 15 while a reader has paused the display. The visible 12 can be a useful reading snapshot, but it is no longer evidence of the current queue total.
Show which time the visible values describe and make the paused state distinguishable from a failed update. Avoid changing a “last updated” label while leaving the displayed numbers unchanged unless the label explicitly refers to a different event.
Record source observation time separately from retrieval time when both are available. A response fetched just now can still contain older observations. If a timestamp's meaning is unknown, leave that uncertainty visible rather than labeling it real-time.
4. Define what resume means
Decide whether resuming returns to the newest available state or continues a historical sequence. W3C discusses this distinction because presenting old information as current can mislead users of status displays. The interface label and timestamp should match the behavior actually implemented.
For the fictional queue, test whether resume shows the current completion count and whether an in-progress selection remains understandable. If rows reorder, confirm that the selected record is still the intended one. Do not let a refresh turn an action aimed at one research item into an action on another.
5. Keep a pause-and-resume acceptance record
- Trigger: what changes automatically and when.
- Control: how the reader pauses, stops or changes the update rate.
- Scope: visible presentation versus background retrieval.
- Freshness: source time, retrieval time and displayed-state time.
- Resume: current-state or historical behavior actually observed.
- Continuity: selection, focus and reading context after refresh.
- Limits: untested devices, assistive technologies or provider behavior.
A successful check establishes the specific behavior you tested, not that the whole dashboard is accessible or its source data is accurate. The useful result is a view that readers can hold still without losing track of when its evidence was true.