We're experiencing a problem where participants are allowed to create blank/empty consent recordings and then move on with the study. This problem is caused by the video-consent frame's 'start recording' button being enabled when the frame first loads, rather than being initially disabled and only becoming enabled with the recorder is ready to start.
This problem is more common when a consent frame is presented as the first frame that uses the video recorder, because in that case the recorder takes longer to start up so there's a higher chance that the user will click 'start recording' before the recorder is actually ready.
The fix for this is to disable the video-consent frame's 'start recording' button until the recorder is actually ready to start recording.
Things to do/consider
[x] Are there other frames that use similar logic to enable/disable participant-controlled recordings?
video-assent: this frame has the same problem as video-consent, the change is included in the PR
I checked the following frames, and these all correctly check that the recorder is ready before allowing the recording to start:
observation
start-recording
stimuli-preview
webcam-display
video-config-quality
[x] Should we add a message about the recorder initializing while the button is disabled, in case it takes a few seconds and participants get confused/annoyed? -- I've done this for the video-assent frame, since the delayed trial start will be more confusing in that case, but I don't think it's needed for the video-consent frame since it has a webcam display and the 'not recording yet' message.
[x] Do we need to handle the case where the recorder takes a super long time to initialize or never becomes ready to record? -- I don't think this is needed at this point, and we should wait to see the current solution causes problems/confusion.
Summary
We're experiencing a problem where participants are allowed to create blank/empty consent recordings and then move on with the study. This problem is caused by the
video-consentframe's 'start recording' button being enabled when the frame first loads, rather than being initially disabled and only becoming enabled with the recorder is ready to start.This problem is more common when a consent frame is presented as the first frame that uses the video recorder, because in that case the recorder takes longer to start up so there's a higher chance that the user will click 'start recording' before the recorder is actually ready.
The fix for this is to disable the
video-consentframe's 'start recording' button until the recorder is actually ready to start recording.Things to do/consider
video-assent: this frame has the same problem as video-consent, the change is included in the PR I checked the following frames, and these all correctly check that the recorder is ready before allowing the recording to start:observationstart-recordingstimuli-previewwebcam-displayvideo-config-quality