Testing my 1st interview before release, when recording several takes (alternating between audio and video to test what happens then ; it happens normally, my choice is free at each new take), I notice that I have no sign that some of the previous takes are video, and even if they are active (or re-activated by me as a respondent), if I play these video takes... I only hear the sound, but there is no image displayed. In effect, the respondent cannot check his/her previous video takes’ pictures, it seems. In the attached screenshot, the 3rd take (playing) has been recorded as a video. But no picture showing. It took me some tests to understand the following, that respondents will clearly never take the time to search for : To get back access to this previous take’s pictures, I am forced to delete the 4th (last) video take altogether... or I can only "disable" the last take, and then the displayed video corresponds to the 3rd take. Also, if takes 3 and 4 are active, if take 3 was played to the end, and I click on "play" for take 3 ... it plays take 4 , instead of restarting take 3 from beginning. Also, if take 3 is active AND take 4 disabled, then I play take 3, but the video player does not open directly under take 3, it remains at the very same place than take 4. Yet, since we are in the situation of recording several takes of the same answer, so most probably in the very same setting and position, even on a 22’’ screen I don’t see obvious signs that the picture corresponds take 3 rather than take 4, until really it begins to reach the point where I say something clearly different. So, moving the player directly under the take being watched would be best to give a very immediate and unambiguous sign to the respondent about what take he’s watching. Or, alternatively, it could just be displaying "take n°[X]" just right beyond the player. But in the case (probably rare, granted ? depending on use case) where the respondent takes quite a lot of takes, if the player remains at the end of the list of takes, then the respondent may have to scroll repeatedly between the takes and the player at the end.
