Suggestive questions reduce the accuracy of adults’ reports about one episode of a repeated event

ABSTRACT

Witnesses often need to describe individual episodes of repeated crimes, such as family violence. Suggestive questions containing incorrect information reduce the accuracy of adults’ reports of single events; in the current experiment, we examined the effects of suggestive questions on adults’ reports of one episode of a repeated event. Over two weeks, 134 participants completed four activity sessions containing variable details that changed each session, and new details that were not repeated across the series. One week later, they were interviewed and described one (self-nominated) target episode. Next, participants were asked four suggestive questions that varied according to whether the suggested details had been experienced in a non-target episode or not experienced, and detail-type (variable or new). As research with children indicated that question-type might be important, half our participants were asked open suggestive questions; the other half were asked closed suggestive questions. Participants accepted more suggestions about experienced (non-target) than not-experienced details, and about variable than new details. They also accepted more details in open than closed questions, but only for experienced non-target new details. Our results demonstrate the ease with which participants accepted interviewer-suggested details when reporting on an episode of a repeated event.

KEYWORDS:

• : Repeated events
• suggestive questions
• source monitoring
• question-type
• misinformation

Data availability statement

The dataset analysed during the current study is available in the OSF repository, https://osf.io/3cxq2/?view_only=23421ecb8c114c8ba6f51205818dc75d

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 As part of a larger project investigating adults’ memories for repeated events, some participants were randomly assigned to receive one of four additional interview supports during their narrative reports. These supports had no effect on the number of details that participants reported, their accuracy, or the proportion of suggestive questions that participants agreed with, Fs(3, 130) < 2.39, ps > .071, η_p² < .053, and will not be discussed further.

2 Although ‘don't know’ responses are not as strong a rejection as disagreeing or correcting the suggestive information, they also do not indicate that a person has accepted the suggestion. Indeed, Scoboria et al.'s (2008) examination of ‘don't know’ responses indicates that there are at least three underlying meanings: (1) true uncertainty about what was presented in the original event, (2) an inability to access information from the original event because it was never presented, and (3) an awareness that the information was presented in the original event, but the interviewee cannot recall it in sufficient detail. The first and third explanations are most relevant to our study. In both of these, ‘don't know’ responses suggest that participants have low confidence about particular details and are not willing to commit to a response (see also Koriat & Goldsmith, 1996). Therefore, we considered ‘don't know’ responses to be more aligned with correct (reject) responses than incorrect (accept) responses.