>>> Does anyone have an example of an ONS research project using interviewers?
The annual Crime Survey for England and Wales is conducted by ONS, who select 50,000 households at random. Their protocol doesn't allow any substitution of addresses, so everyone who refuses to take part reduces the size of the database upon which the survey is based.
I've taken part in the survey myself. The interviewer turns up at a pre-agreed time, asks his/her questions and enters the answers directly onto his/her laptop. Dead easy. No problems.