Empirical evidence that forced fabrication leads to false memories
Intuitively, it would seem that the experience of being coerced into providing descriptions of fictitious items or events would be salient and memorable to participants, and that they would remember that their fabricated responses were mere speculations that they had been forced to provide. To the contrary, there is now considerable evidence that after retention intervals of just one week, participants are prone to developing false memories of having witnessed their forced fabrications even when warned before taking the test (Ackil & Zaragoza, 1998, 2011; Zaragoza et al., 2001; Frost et al., 2003; Hanba & Zaragoza, 2007; Memon, Zaragoza, Clifford, & Kidd, 2010). In what follows, we review the factors that have been shown to mitigate false memory development in the forced fabrication paradigm, and the factors that have been shown to promote these false memories.
Factors that mitigate false memory development
As we have shown, one of the unique features of suggestive interviews involving forced fabrication is that participants are pressed to provide responses to unanswerable questions (and hence must fabricate a response). Not surprisingly, participants resist doing so. In several studies we have sought to assess whether there is a relationship between participants’ resistance to providing a fabricated response and later false memory for that fabricated response. We have found that in some cases, and for some types of resistance, there is. Specifically, several studies have shown that overt, verbal resistance (e.g.,“I didn’t see that”, or “His knee wasn’t bleeding”) is associated with lower false memory development, but more passive forms of resistance (long latency or refusals to respond) are not (Zaragoza et al., 2001; Ackil & Zaragoza, 2011). Relative to passive resistance, we have found that in those fabrications that were generated following overt resistance, false memory development is substantially reduced (Ackil & Zaragoza, 2011) and in some cases not evident at all (Zaragoza et al., 2001).
It should be noted, however, that this apparent protective benefit of overt resistance occurs under restricted circumstances only. Ackil and Zaragoza (2011) manipulated whether or not participants received a warning before the final test and manipulated whether the final test was recognition or free recall. Relative to fabrications provided without verbal resistance, false memory for verbally resisted fabrications was significantly lower, but this reduction in false memory was observed only if two conditions were met jointly: participants received a pretest warning, and the final test was a recognition test. If only one of these conditions was met, overt resistance did not confer any benefit. That is, neither participants who were unwarned and given a recognition test nor those who were warned but tested with free recall evidenced any reduction in false memory when fabrications were generated following verbal resistance (Ackil & Zaragoza, 2011). Finally, as we will show later in this chapter, the association between overt resistance and reduced false memory disappears when participants receive confirmatory interviewer feedback following their fabricated responses, as well as when they are tested after very long retention intervals (Chrobak & Zaragoza, 2008).
Collectively, the data reviewed here are consistent with the conclusion that resistance is most likely to offer protection against false memory development under conditions that encourage participants to remember their uncertainty in their fabricated response. We interpret these findings as evidence that publicly expressing resistance (as opposed to keeping such thoughts internal) enhances memory for the resistance and for having fabricated the response. We further propose that pretest warnings prompt participants to retrieve their uncertainty in their fabricated response in cases where they do not spontaneously do so. Finally, we propose that because tests of narrative free recall are more cognitively demanding than yes/no recognition, participants are less likely to spontaneously retrieve memories of resistance when providing free recall.