I recall a time I was conducting an exit poll and without much thought to it I assigned two Muslim girls to a precinct in a Catholic school. They were very sweet young ladies and approachable so I didn’t much consider any interviewer effects when assigning them. A few hours into the election I got a call asking to be reassigned because they were not having much luck with the respondents and the folks there were making them feel uncomfortable. Both were wearing hijabs. I was a little embarrassed and saddened that I had to reassign them. I was raised Catholic. Anyways, here is some empirical research on the topic.