User involvement in research is widely valued, but evidence of its impact in quantitative research is lacking. We investigate whether survey responses are affected by interviewers' experience of mental health problems. We hypothesized that when the interviewer has experienced mental health problems, participants would be more likely to consent to participate, leave fewer items unanswered, report higher levels of discrimination and express greater satisfaction with the interview. The study was nested within a telephone survey of service users' experiences of discrimination. Participants were randomly assigned to three groups: peer disclosing, peer non-disclosing and non-peer interviewers, where 'peer interviewers' have personal experience of mental health problems. Analyses explored the impact on response rate, number of unanswered questions, reported discrimination and interview experience. No difference was found in prevalence of discrimination reported to interviewers. A significantly lower response rate was attained in the peer-disclosing group (5% compared to 6.5%, p = 0.005). Significantly fewer questions were left unanswered in the peer-disclosing group (Mean = 0.07 compared to 0.4, p = 0.004). Findings suggest that interviewers' experiences of mental health problems broadly do not impact on quantitative data collected in structured interviews about discrimination. Disclosure of peer status prior to consent may have affected recruitment.