Researchers have questioned the accuracy of participants’ reports of the frequency of intimate partner violence (IPV). We used an experimental design to probe the adequacy of a 1-month recall period. Representatively sampled 18- to 19-year-old participants in romantic relationships (N = 435) were randomly assigned to self-report on the presence (yes/no) and frequency of physical and psychological IPV perpetration and victimization on a monthly versus semimonthly basis. It logically follows that monthly IPV reports should be double the IPV of semimonthly reports. Findings differed for IPV rate (i.e., binary scores) and frequency (i.e., continuous scores). Risk ratio (monthly:semimonthly schedule) confidence intervals (CIs) for binary IPV variables each included 2 (consistent with the logical supposition above), except for physical IPV victimization, although the point estimates fell short of this mark. All mean ratio (monthly:semimonthly schedule) CIs for IPV frequency variables included 1, and the point estimates were adjacent to 1 (not consistent with the logical supposition above). Results suggest that, for young adults, IPV recall windows shorter than 1 month are preferable, particularly if act frequency is of interest.
(Publisher abstract provided.)