Research Guides
Verify the Keyboard Exit from a Research Editor
A research editor may be easy to enter with a keyboard yet difficult to leave. If every press of Tab changes indentation instead of moving to the next control, a reviewer needs a discoverable exit method. Testing only how the editor opens misses the rest of the research journey.
This guide supplies an exit-path worksheet for embedded research components. It does not claim a particular Colquhoun.AI editor implementation or certify a whole website's accessibility.
1. Ask whether focus can leave
W3C's No Keyboard Trap guidance requires a keyboard route away from a component that can receive keyboard focus. If leaving needs more than ordinary arrow or Tab keys or another standard exit method, the user must be advised of that method. The guidance distinguishes an inescapable trap from an appropriate modal focus boundary that the user can dismiss.
Do not mark every repeated focus cycle as a defect. First identify the interaction pattern and its intended exit. Equally, do not accept “click somewhere else” as the keyboard solution.
2. Follow a fictional note-editor journey
Imagine a fictional comparison workspace with a source link, an embedded note editor and a save control. A reviewer tabs into the editor, writes a harmless test note and presses Tab. The editor inserts indentation. Further Tab presses continue editing, and no instruction explains how to reach the toolbar or leave.
The defect is not that an editor supports indentation. It is that the tested keyboard user cannot discover and use a route out. A proposed repair may provide an appropriate exit command with accessible instructions, or change the interaction to a suitable standard pattern.
The retest should begin before entering the editor. Verify that the instructions are available when needed, that the command works and that the next relevant control is reachable without a pointer. A developer knowing an undocumented shortcut is not evidence that a reader can discover it.
3. Record an entry-and-exit path
- Starting point: identify the control before the component.
- Entry: record how focus arrives and what receives it.
- Inside behavior: note what Tab, reverse navigation and relevant documented commands do.
- Exit method: identify the actual keyboard route and where its instructions appear if needed.
- Destination: record where focus lands after leaving.
- Continuation: confirm the research task can proceed.
Use a short, reproducible sequence rather than “keyboard navigation is broken.” Include the browser, operating system and assistive technology used, if any. A different environment should not inherit the result without testing.
4. Review dialogs as a separate case
If an export dialog intentionally keeps focus inside while open, test its available dismissal route and the return to the surrounding task. Do not remove the focus boundary simply to make Tab reach the background page; that can create a different interaction problem.
Use a harmless local fixture for actions that could publish, delete or spend credits. A test of escaping a dialog should not require completing a real paid export or changing production research records.
5. Verify the exit did not damage the work
Check the note after leaving and returning. An exit command that silently clears text creates a separate data-loss issue, even if focus now moves. Record preservation and focus movement independently so the repair does not hide one problem behind another.
Include repeated entry and exit, empty and populated content, and any alternate editing mode the application exposes. A fix for the first focus cycle can still fail after the component rerenders or switches its toolbar.
The completion note should name the tested path and remaining coverage gaps. Source inspection, a scripted key sequence and an observed assistive-technology interaction are different forms of evidence. Being precise about which one was performed makes this small website review useful to the next person reproducing it.