Responsivity to Interviewer During Interview-Based Assessment of Physical Intimate Partner Violence

Objective: Interview assessments of intimate partner violence (IPV) may provide more accurate behavior frequency estimates than self-report questionnaires. However, concerns have been raised about whether participants underreport IPV during interviews due to an emotional response to the interviewer. Method: Participants were 42 mixed-gender community couples (83 individuals) in which at least one partner endorsed physical IPV perpetration or victimization in their relationship. We examined whether participants were emotionally responsive to the interviewer during an interview about physical IPV. Responsivity was defined as the extent to which participants' emotional arousal, indexed by vocal fundamental frequency (f(0)), was predicted by interviewers' emotional arousal at the previous talk turn on a moment-by-moment basis. We then examined whether participants' responsivity predicted interview-based reporting of IPV relative to their own self-report on an IPV measure and to the highest other available report (including partner report). Results: Repeated measures actor-partner interdependence models conducted in a multilevel modeling framework indicated that, on average, participants were responsive to interviewers' emotional arousal, even when controlling for responsivity to their own arousal, and that responsivity varied across participants. However, participants' responsivity to interviewer arousal did not significantly predict reporting of IPV perpetration or victimization during the interview relative to their own self-report or to the highest other available report. Conclusions: Participants are emotionally responsive to interviewer arousal, but this responsivity does not appear to reduce interview-based reporting of IPV relative to self-report, supporting the utility of IPV interviews in clinical and research settings.