I'm working on the improvement of Voice Over support for the application which delivers language tests to end-users. The app asks a user to provide a verbal response to test questions that are recorded by the app. Tests have a form of timed assessment so the time span allocated for response providing is absolutely critical (cannot be extended).
When a test is taken with a VO enabled there is a chance that a user will activate an element and VO readout will be recorded by the app as a part of a response which will impact the user's score.
As we didn't find any option to silence/disable VO during the recording we minimized the number of active elements during the test (there is only one left). However, there is still a chance that a user will activate it or refresh a focus and the response will be impacted.
Do any of you guys have some experience with similar cases? Is there anything we missed? All the accommodation-related recommendations suggest letting the user start the recording/playback whenever he/she is ready but this is not an option for us due to the nature of the test.
Any help appreciated.