Skill in interviewing reduces confirmation bias

Interviewers given prior information are biassed to seek it from interviewees. We examined whether the detrimental impact of this confirmation bias in terms of leading question use was moderated by interviewers' demonstrated ability to adhere to open questions. We classified interviewers' adherence as 'good' or 'poor' in an independent interview before they interviewed children about a staged event. Half the interviewers were given biassing true and false information about the event; half were given no information. As predicted, only poor interviewers showed the effect of bias. Poor interviewers asked fewer open questions in the biassed condition than the non-biassed condition; good interviewers asked the same (high) proportion of open questions in both conditions. Poor interviewers asked more leading questions in the biassed condition than the non-biassed condition; good interviewers asked the same (low) proportion of leading questions in both conditions. These results demonstrate that interviewers' skill in adhering to open questions reduces the detrimental impact of confirmation bias on question type.