I am currently reading a draft research document - and was reminded of the difficulties in the field when trying to audio- and video-record young children's speech so that reliable speech transcription can be made and accurate data entry can be assured:
"The speech sample collection was not carried out if the assessment room had too much noise such as drilling machines, saws, wall demolition, continuous honking, etc. in the surrounding area."
It reminded me of conducting speech assessments in Vietnam in a room that was next to the parking area for parents' and teachers' motorbikes. We kept having to stop as people came and went.