Investigating Interviewer Effects and Confounds in Survey‐Based Experimentation*