An uncertainty monitoring account of false memories that result from forced fabrication
One potential explanation for the false memories that result from forced fabrication is that people are prone to forgetting the source of their fabrications (for reviews of research and theory on source monitoring see Johnson, Hashtroudi, & Lindsay, 1993; Lindsay, 2008). The evidence suggests, however, that the memory deficit that underlies the forced fabrication effect is much more selective than a general “source amnesia” hypothesis would suggest. In studies where we have directly assessed participants’ memory for the source of their fabricated events, we find that an overwhelming majority of participants accurately remember having provided the fabricated information during the postevent interview. For example, Ackil and Zaragoza (2011) gave participants a source recognition test and found that participants tested after a two-week retention interval correctly attributed their fabricated items to the postevent interview 81% of the time. Even more impressive was the finding by Chrobak and Zaragoza (2013) that after a six- to eight- week retention interval, 89% of participants were able to reproduce their fabricated account when asked to recall what answer they had given earlier in response to the false-event question. Also interesting was the finding that confirmatory interviewer feedback (which increases false memory) also has the seemingly paradoxical effect of improving participants’ memory for the fact that the fabricated event was from the postevent interview (Zaragoza et al.,2001).
How is it possible that the very same participants who incorrectly claim to remember witnessing their forcibly fabricated events in the video nevertheless also accurately remember providing these items/events as responses during the postevent interview? It is important to remember that the witnessed event and postevent interview are not mutually exclusive sources of information. To the contrary, because the interview is about the witnessed event, most of the information the participant encounters during the postevent interview is an accurate description of the events he or she witnessed. Hence, even if participants know that they were interviewed about things that never happened, the mere knowledge that an item was provided as a response during the interview is not diagnostic with regard to its validity.
What appears to underlie participants’ claims that they remember their fabricated event from both the video and the interview is that, although they remember providing the fabricated response, they selectively forget (or fail to retrieve) that the response they provided was fabricated under duress. This failure to remember having fabricated the response renders it susceptible to misattribu- tion for a number of reasons. First, as in all eyewitness suggestibility situations, there is a great deal of overlap between the witnessed event and subsequent interviews about the event, thus rendering the two sources highly confusable. In addition, pressing participants to describe fictitious events forces participants to create concrete, perceptually detailed memory representations that have characteristics typical of witnessed events (e.g., Johnson et al., 1993). Moreover, given the well-documented mnemonic advantage enjoyed by self-generated information (Slamecka & Graf, 1978), the content of participants’self-generated fabrications is likely well remembered and highly familiar at the time of the test, thus rendering it confusable for a real event. Finally, the fabrications that participants generated are likely constrained by their own idiosyncratic knowledge, making them particularly plausible.
In summary, many factors render a self-generated fabrication about the witnessed event confusable with a witnessed memory.The one factor that should prevent false memory is the knowledge that the item/event was fabricated under duress. As we have seen, memory for having fabricated these events is particularly vulnerable to forgetting/retrieval failure. Moreover, our review of the available evidence shows that situations that promote retrieval of participants’ uncertainty in their fabricated responses (overt verbal resistance, warnings, short retention intervals, strong memory for the witnessed event) are associated with reduced false memory development, and, conversely, situations that discourage retrieval of this uncertainty (protracted retention intervals, complex question formats, and confirmatory interviewer feedback) all serve to promote false memory development. Unfortunately, however, as we have shown, the factors that promote false memory, such as long retention intervals and confirmatory feedback, can override the protective factors.