Don’t Make Me Think, Revisited a common Sense Approach to Web Usability Steve Krug

Use a camera pointed at the screen instead of mirroring.
Mirroring
is the same as screen sharing: It displays what’s on the screen. You can
do it with software (like Apple’s Airplay) or hardware (using the same
kind of cable you use to play a video from your phone or tablet on a
monitor or TV).
But mirroring isn’t a good way to watch tests done on touch screen
devices, because you can’t see the gestures and taps the participant is
making. Watching a test without seeing the participant’s fingers is a
little like watching a player piano: It moves very fast and can be hard
to follow. Seeing the hand and the screen is much more engaging.
If you’re going to capture fingers, there’s going to be a camera
involved. (Some mirroring software will shows dots and streaks on the
screen, but it’s not the same thing.)
Attach the camera to the device so the user can hold it naturally.
In some setups, the device sits on a table or desk and can’t be moved.
In others, the participant can hold the device, but they’re told to keep it
inside an area marked with tape. The only reason for restricting
movement of the device is to make it easier to point a camera at it and
keep it in view.
If you attach the camera to the device, the participant can move it
freely and the screen will stay in view and in focus.
Don’t bother with a camera pointed at the participant.
I’m really
not a fan of the face camera. Some observers like seeing the
participant’s face, but I think it’s actually a distraction. I’d much rather
have observers focus on what’s happening on the screen, and they can
almost always tell what the user is feeling from their tone of voice
anyway.
Adding a second camera inevitably makes the configuration much more
complicated, and I don’t think it’s worth the extra complexity. Of course, if
your boss insists on seeing faces, show faces.