ArticlePDF Available
Misinformation Effects
and the Suggestibility of
Eyewitness Memory
Maria S. Zaragoza, Robert F. Belli, and Kristie E. Payment
Social scientists and legal practitioners have long suspected that suggestive foren-
sic interview practices are a major cause of inaccuracies in eyewitness testimony.
However, it wasn’t until Elizabeth Loftus published a highly influential series of
studies on eyewitness suggestibility in the 1970s that a systematic body of scien
tific literature on this topic started to emerge. Since then, hundreds of empirical
studies on eyewitness suggestibility have been published, all of them variants of
the basic experimental paradigm that Loftus developed.
In the early 1970s, research and theorizing about memory was based almost ex
clusively on studies of memory for lists of words or sentences (see, e.g., Crowder,
1976). By studying memory for complex, fast-moving, and forensically relevant
events (typically depicted in film clips or slide shows), Loftus demonstrated that it
was possible to conduct well-controlled experiments that were high in ecological
validity (Banaji & Crowder, 1989). Her studies provided clear evidence that sug
gestive interviews can lead to profound errors in eyewitness testimony, thus rais
ing serious questions about the reliability of memory and eyewitness testimony.
Her work established that scientific research on memory and suggestibility can
and should inform the courts. In addition, her findings inspired many theoretical
debates about the constructive nature of memory, mechanisms of forgetting, and
the permanence of memory.
In this chapter, we review the empirical evidence and theoretical proposals that
have been put forward to account for the misinformation effect—the finding that
exposure to misleading postevent information can lead eyewitnesses to report
items and events they never actually saw.
In the experimental paradigm introduced by Loftus, participants view a slide se
quence depicting a complex and forensically relevant event, such as a traffic acci
dent or theft. Immediately thereafter, participants are questioned about the event
they witnessed. The critical manipulation is that the questioning includes leading
or misleading information. Finally, participants are tested on their memory for the
witnessed event. The dependent variable of interest is the extent to which misled
participants incorporate the misleading suggestions into their eyewitness reports
(as compared to control participants that were not misled).
Early demonstrations of the effects of leading questions revealed several ways in
which eyewitness reports could be influenced. For example, for participants who
had seen films of automobile accidents, the question, About how fast were the cars
going when they smashed into each other?” elicited higher estimates of speed, and
more false claims of having seen broken glass on a later test, than questions that used
verbs such as bumped or hit instead of smashed (Loftus & Palmer, 1974). Subse-
quent studies showed that misleading postevent questions could also cause a variety
of other distortions in eyewitness reports. For example, Loftus (1977) had partici-
pants view an accident involving a green car and later exposed them to misleading
questions that presupposed the car was blue. When later asked to select the color of
the witnessed car from a color wheel, misled participants showed a marked tendency
to shift their color responses in the direction of the misinformation by selecting a
“blue-green” color, a tendency that was not observed in control participants. Thus,
most misled participants reported a color that was a blend of the original and
postevent information (see also Belli, 1988, for similar evidence of color blends fol
lowing postevent information). Finally, other studies showed that participants could
be led to report entire objects that were not present in the originally witnessed event.
In Loftus (1975), participants who were asked, “How fast was the white sports car
going when it passed the barn while traveling along the country road?” (when, in
fact, no barn appeared in the scene) were much more likely to later claim they had
seen a barn than were control participants who had not been misled (Loftus, 1975).
Important to note, this study also showed that misinformation was more likely
to influence later testimony when the false information appeared as a presupposi
tion, rather than the direct focus of the question. For example, participants who
were directly asked “Did you see a barn in the film?” were much less likely to later
claim they had seen a barn than those who’d answered the previous question where
the presence of the barn was presupposed (Loftus, 1975).
In 1978, Loftus, Miller, and Burns published a study demonstrating that eye
witness testimony was much more malleable than previously thought. The experi
mental procedure was nearly identical to that described in the studies reported
earlier, with the crucial exception that the misleading postevent information di
rectly contradicted some aspect of the originally witnessed event. In Loftus et al.,
for example, participants who witnessed an auto–pedestrian accident involving a
stop sign were subsequently asked “Did another car pass the red Datsun when it
was stopped at the yield sign?” In a later test of their memory for the witnessed
events, participants were given a forced choice between the slide depicting the
stop sign and a nearly identical slide depicting a yield sign. The finding was that
75% of the control participants (who had not been misled) correctly chose the slide
they had seen depicting the stop sign, whereas only 41% of the misled participants
did so—a 34% difference in accuracy. Additional studies showed that when partic
ipants who selected the misinformation were later asked to give a second guess
(e.g., “Was it a ‘stop sign’or a ‘no parking’sign?”), their likelihood of selecting the
original item was not greater than chance. On the basis of these findings, Loftus et
al. claimed that misleading postevent information could not only supplement eye-
witness memories, it could also transform them.
The “misinformation effect” documented by Loftus et al. (1978) is one of the
best-known and most influential findings in psychology. Demonstrations of the
surprising ease with which people could be led to report objects and events they
had not seen challenged prevailing views about the validity of memory and raised
serious concerns about the reliability of eyewitness testimony. Since its publica-
tion, countless studies have replicated and extended these findings: Misinforma-
tion effects have been demonstrated in participants of all ages (from preschoolers
to older adults), for a variety of different types of events (live events, emotional
events, naturally occurring events), types of misinformation (about people, places,
and things), methods of delivering the misinformation (narratives, questionnaires,
and face-to-face interviews), and all manner of methods for assessing memory for
the witnessed event (e.g., free recall, cued recall, and recognition).
Early studies of the misinformation effect also identified factors that influence
the magnitude of these effects. For example, Loftus et al. (1978) showed that mis
information effects increase as a function of the delay between the witnessed event
and exposure to misinformation, presumably because memory for the original
event becomes weaker over time. Social factors, such as the credibility of the
postevent source, are also an important variable. Whereas participants are easily
influenced by misinformation that is provided by a credible source, they will effec
tively resist suggestion that is provided by a source who lacks credibility or whom
they perceive as having intentions to mislead (Dodd & Bradshaw, 1980; Smith &
Ellsworth, 1987; Underwood & Pezdek, 1998). Indeed, even young children are
less influenced by suggestion when it is provided by a peer rather than an authori
tative adult (Ceci, Ross, & Toglia, 1987; Lampinen & Smith, 1995). A more re
cent, and related, finding is that the magnitude of the misinformation effect is also
influenced by more subtle social cues, such as the perceived power and social at
tractiveness conveyed by the accent of the person providing the misinformation
(Vornik, Sharman, & Garry, 2003). Finally, the extent to which participants detect
a discrepancy between their memories of the witnessed event and the postevent ac
count also reduces misinformation effects (Tousignant, Hall, & Loftus, 1986), and
in the extreme case where participants are given blatantly contradictory sugges
tions, they are sometimes not misled at all (Loftus, 1979c). In sum, although stud
ies have identified some of the boundary conditions of the misinformation effect,
the high reliability and robustness of these effects is well established.
One of the most influential aspects of Loftus’s work on misinformation phenom
ena were the bold theoretical proposals that she initially advanced to account for
these effects (Loftus, 1979a, 1979b; Loftus & Loftus, 1980; Loftus et al., 1978).
Specifically, Loftus proposed a “destructive updating” process whereby contra-
dictory misleading postevent information replaces the original information and, as
a consequence, permanently erases the original information from memory. Her
claim that misinformation could permanently erase information from memory
generated tremendous interest, and many researchers set out to show that the origi-
nal details were not gone from memory but merely rendered less accessible by the
misinformation (e.g., Bekerian & Bowers, 1983; Christiaansen & Ochalek, 1983).
In 1985, McCloskey and Zaragoza published an article in which they argued
that the traditional misinformation effects (e.g., Loftus et al., 1978) could not be
taken as evidence that misleading postevent information caused impairment of
original event details. The arguments advanced by McCloskey and Zaragoza
(1985a) began with the observation that in the typical eyewitness suggestibility ex
periment (as in real-world eyewitness testimony situations) participants’ memory
for the witnessed event is far from perfect, even before they are exposed to misin
formation (as evidenced by the finding that control performance is well below
ceiling). Hence, according to McCloskey and Zaragoza, the finding that partici
pants given a choice between stop sign and yield sign report the misleading item
(yield sign) does not necessarily imply that the participant once had a memory for
a stop sign that was now impaired. Rather, for those participants who fail to re
member the stop sign (e.g., because they failed to encode it), the misleading sug
gestion “yield sign” does not conflict with a stored memory representation, it
merely fills a gap in memory. McCloskey and Zaragoza argued further that misled
participants who fail to encode the originally seen detail (stop sign) are likely to
systematically select the more recently presented misleading detail (yield sign)
because “yield sign” is all they remember and they have no reason to distrust the
postevent source. In contrast, control participants who fail to encode the original
stop sign detail, and were never exposed to misinformation about a yield sign, will
have no reason to favor the incorrect misleading response (yield sign) on the test.
Rather, they will be forced to guess, thus leading them to select the correct alterna
tive (stop sign) 50% of the time on a two-alternative forced-choice recognition
test. McCloskey and Zaragoza concluded that for these reasons, misled partici
pants are likely to perform more poorly than control participants on the
stop-sign-versus yield-sign test even when their original memory has not been im
paired by the misinformation. Finally, McCloskey and Zaragoza also noted that
the demand characteristics of the experiment (where the postevent information is
presented by an authoritative experimenter as truth) may lead some misled partici
pants to report the suggested detail on the test even if they can remember the origi
nal event detail (see also Lindsay, 1990).
To assess whether contradictory misinformation erases or impairs access to
original event details, McCloskey and Zaragoza (1985a) developed a Modified
Test procedure in which the misleading detail was not an option. Rather, partici
pants had to choose between the original detail and a new detail. In their experi
ments, participants who witnessed an office theft involving a handyman holding a
hammer were later given the misleading suggestion that he was holding a screw
driver. In contrast to the Standard Test where participants are given a choice be-
tween the original and misleading item (hammer vs. screwdriver), on the Modified
Test, participants were given a choice between the originally seen item and a new
item (hammer vs. wrench).
In six experiments using the Modified Test procedure,
they found that misled participants selected the original item as often as control
participants (grand mean was 72% vs. 75% correct for misled and control condi-
tions, respectively, when collapsing across experiments), thus providing no evi-
dence of memory impairment. In the same six experiments, McCloskey and
Zaragoza tested a second group of participants with the standard misinformation
test (a choice between the original [e.g., hammer] and misleading [e.g., screw-
driver] items) and showed that under these circumstances there were robust mis
led–control performance differences (collapsing across the six experiments mean
performance was 37% vs. 72% for misled and control conditions, respectively, a
difference of 35%). Collectively, the McCloskey and Zaragoza results showed
that, when given the opportunity, misled participants had an overwhelming ten
dency to select the misleading alternative over the item they originally saw on the
final memory test, thus replicating the misinformation effect first demonstrated by
Loftus et al. (1978). However, when the misleading alternative was not an option
on the test, participants evidenced the ability to remember the original event detail
as well as participants who had not been misled. These results support the conclu
sion that factors other than memory impairment contribute to these dramatic errors
in participants’ performance.
McCloskey and Zaragoza’s (1985a, 1985b) claim that contradictory misinfor
mation does not impair memory for originally seen details was highly controver
In all of these experiments, the materials are completely counterbalanced so that across par
ticipants, hammer, screwdriver, and wrench serve as the original, misleading, and new informa
tion equally often.
sial and generated considerable debate, a debate in which the authors of this
chapter held contrasting points of view (e.g., Belli, 1989; Loftus & Hoffman,
1989; Loftus, Schooler, & Wagenaar, 1985; McCloskey & Zaragoza, 1985b;
Metccalfe, 1990; Tversky & Tuchin, 1989; Zaragoza & McCloskey, 1989; see also
Ayers & Reder, 1998, for a review). McCloskey and Zaragoza’s “no impairment”
claim seemed to fly in the face of decades of research on retroactive interference
effects, and considerable research was devoted to obtaining unambiguous evi
dence of memory impairment caused by contradictory misinformation. Although
McCloskey and Zaragoza’s (1985a) results with the Modified Test have been rep
licated many times and under a variety of conditions (e.g., Belli, 1993; Bowman &
Zaragoza, 1989; Loftus, Donders, Hoffman, & Schooler, 1989) clear evidence that
misinformation can impair memory for original event details has now been ob
tained with the Modified Test procedure (Belli, Windschitl, McCarthy, & Winfrey,
1992; Eakin, Schreiber, & Sergent-Marshall, 2003; Schooler, Foster, & Loftus,
1988; Schreiber & Sergent, 1998), though it has been difficult to identify the cir-
cumstances under which it consistently does so. Payne, Toglia, and Anastasi
(1994) conducted a meta-analysis of 44 published experiments that used modified
recognition tests and showed that, individually, only 14 of the 44 experiments
yielded significantly poorer misled than control performance. Collectively, how-
ever, the overall memory impairment effect across the 44 experiments was statisti-
cally significant.
What determines whether memory impairment effects are detected? On studies
using the Modified Test, evidence of memory impairment has most consistently
been obtained when participants are given an interpolated test that forces them to
overtly commit to misinformation (by not providing the correct item as an option)
prior to taking the Modified Test (Eakin et al., 2003; Foster, Schooler, & Loftus,
1988; Schreiber & Sergeant, 1998) but not when participants freely commit to the
misinformation prior to taking the Modified Test (see Belli, 1993, for evidence
that freely committing to the misinformation does not lead to such impairment).
Presumably, forcing participants to overtly commit to the misinformation further
boosts activation of the misleading item, thus inhibiting access to the originally
seen details. Another factor that appears related to the detection of memory im
pairment effects is overall memory performance. Memory impairment effects
have more readily been observed when overall memory performance is relatively
high as opposed to when it is low (Chandler, 1989; Payne et al.. 1994), and when
participants are misled about centrally presented items and tested after a lengthy
retention interval (Belli, Windschitl, McCarthy, & Winfrey, 1992; although see
Windschitl, 1996, for contradictory evidence regarding the effects of retention in
terval). There is some evidence that memory impairment effects are more likely to
be observed when participants are preschoolers rather than adults (Ceci, Ross, &
Toglia, 1987; although see Zaragoza, 1991, and Zaragoza, Dahlgren, & Muench,
1992, for failures to replicate these findings under nearly identical conditions).
Evidence for memory impairment has been more readily observed when the fi
nal test requires recall rather than recognition (for evidence of memory impair
ment with cued-recall tests, see Belli, Lindsay, Gales, & McCarthy, 1994; Lindsay,
1990). An important condition for obtaining impairment effects in recall tests is
that the misinformation is permitted as a potential response. When cued-recall
tests disallow reporting of the misleading information (e.g., by providing a cue to
which the misleading item does not apply), no impairment is observed (Zaragoza,
McCloskey, & Jamis, 1987). Whether these impairment effects are at times due to
destructive updating of memory traces or solely due to retrieval failure is still not
completely resolved (e.g., Belli & Loftus, 1996), but what is clear is that the bulk
of the evidence favors a role for the retrieval failure hypothesis in most of the con
ditions that have been observed so far (for recent treatments of this issue, see
Chandler, Gargano, & Holt, 2001; Eakin et al., 2003). In summary, research on the
memory impairment issue shows that one potential consequence of exposure to
contradictory postevent information is impaired access to the originally seen
The paper by McCloskey and Zaragoza (1985a) and those that followed in re
sponse (Belli, 1989; Loftus & Hoffman, 1989; McCloskey & Zaragoza, 1985b;
Tversky & Tuchin, 1989; Zaragoza & McCloskey, 1989) marked a turning point in
the history of research on the misinformation effect. It soon became clear that re-
search on the memory impairment issue—which focuses on the “fate” of the origi-
nal memory following exposure to suggestion—did not address a fundamental
aspect of misinformation phenomena, namely, participants’ tendency to incorpo-
rate misleading postevent suggestions into their eyewitness reports. Whether or
not misinformation impairs memory for originally seen details, the fact remains
that participants can be easily led to report misinformation that has only been sug-
gested to them. Hence, a critically important issue of both theoretical and practical
concern is understanding the nature and extent of this misleading influence. Does
misleading questioning simply influence what participants say, or might such
questioning lead to the development of false beliefs about the witnessed event? Al
ternatively, is it possible that exposure to misinformation might lead participants
to create genuine false memories of having witnessed the suggested events? Much
of the research that followed was concerned with addressing these issues
From her earliest writings, Loftus was a strong proponent of the view that mislead
ing postevent suggestions led to distortions in eyewitness memory rather than sim
ply influencing what eyewitnesses report (see, e.g., Loftus, 1979a, 1979b; Loftus
& Loftus, 1980). However, as alluded to in the preceding section, the misinforma
tion effect does not provide definitive evidence for this conclusion. It wasn’t until
researchers started employing methods that probed participants more extensively
about the basis for their eyewitness reports that clearer evidence bearing on the
false-memory hypothesis started to emerge. Although Loftus’s claim that mislead
ing suggestions can lead to false memories was ultimately proved correct, this re
search has also shown that misled participants sometimes report misinformation
for other reasons.
In the standard eyewitness suggestibility experiment, misinformation is pre
sented to participants as an accurate description of the events they witnessed by an
experimenter whom they are likely to view as knowledgeable and credible. As
many have noted (see, e.g., Lindsay, 1990) this experimental situation is imbued
with substantial demand. Participants may feel pressured to report the suggestion
whether or not they believe the suggested information or misremember seeing it in
the original event. Hence, to rule out the possibility that participants report misin
formation simply because they are playing along, it is necessary to make every ef
fort to eliminate this demand. At a minimum, participants need to be alerted to the
possibility that the information provided by the postevent source may not corre
spond to the events they witnessed.
Although it is a relatively straightforward matter to change the demand charac
teristics of the experiment, it is somewhat more difficult to discriminate between
situations where a participant-witness has developed a false belief in the suggested
information as opposed to a false memory of having witnessed the suggested infor-
mation. Even a high-confidence endorsement of the suggested details may simply
reflect a strong belief that the suggested events transpired. As mentioned previ-
ously, in cases where participants have no memory that contradicts the misleading
suggestions, they have little reason to distrust the experimenter and may therefore
come to believe that the suggested information is true. In an attempt to be helpful,
participant-witnesses are likely to report everything they know about the event
without regard to whether they specifically recollect witnessing it at the original
event or whether they learned it from another source
One method that investigators have used to more directly assess whether partici-
pants misremember witnessing the suggested information is to give them a
source-monitoring test, which forces them to discriminate between possible sources
of information in memory. In the typical study, participants are asked to identify the
source of the suggested item by choosing among multiple possible sources (e.g., the
witnessed event, the postevent questions, both, or neither). Note that source-moni
toring test procedures inform participants prior to the test that the postevent narrative
and questions contain information that was not in the witnessed event, thus reducing
any perceived demand to go along with the suggested information.
Studies have shown that when misled participants are given a source-monitor
ing test, rather than a traditional recognition test, their tendency to claim they re
member witnessing the suggested items is substantially reduced (Zaragoza &
Lane, 1994) and in some cases eliminated (Lindsay & Johnson, 1989; Zaragoza &
Koshmider, 1989). Nevertheless, a great deal of evidence supports the conclusion
that misled participants do claim to remember witnessing the suggested details,
even when given a source-monitoring test (Belli, Lindsay, Gales, & McCarthy,
1994; Lindsay, 1990; Chambers & Zaragoza, 2001; Drivdahl & Zaragoza, 2001;
Frost, Ingraham, & Wilson, 2002; Hekkanen & McEvoy, 2002; Lane, Mather,
Villa, & Morita, 2001; Mitchell & Zaragoza, 1996, 2001; Zaragoza & Lane, 1994;
Zaragoza & Mitchell, 1996; see Zaragoza, Lane, Ackil, & Chambers, 1997, for a
review). Moreover, if, in addition to the mild warning provided by the
source-monitoring test, participants are told very directly and explicitly that they
were misinformed (e.g., by telling them that the misleading source contained inac
curacies [Zaragoza & Lane, 1994] or telling them that the experimenter was trying
to trick them [Chambers & Zaragoza, 2001]), they still persist in claiming they re
member witnessing the suggested items on the source-monitoring test (see also
Lindsay, 1990, for evidence that the very strong warning not to report any informa
tion from the postevent source on the test does not always eliminate false reports
on a cued-recall test).
It might be argued that participants’tendency to claim they remember witnessing
the suggested items on the source-monitoring test is a reflection of a false belief that
they saw the suggested item rather than a genuine false memory of having witnessed
the suggested event. To address this possibility, several studies have also assessed
the phenomenological experience that accompanies participants’ “memory” of wit
nessing the suggested item or event (cf. Schooler, Gerhard, & Loftus, 1986). One
method that has been used to assess the phenomenological experience of false mem-
ories is Tulving’s (1985) remember/know procedure, a technique that has been used
quite extensively in other domains (see Gardiner & Java, 1993, for a review). Fol-
lowing recall or recognition of a test item, participants are asked to indicate whether
they remember seeing it during the original event or they just know it occurred, but
cannot actually remember the specific episode (see also Zaragoza & Mitchell, 1996,
for a related measure where participants are asked to distinguish between “remem-
bering” and “believing”). The distinction between “remembering” and “knowing”
is carefully explained to participants (see, e.g., Gardiner & Java, 1990; Rajaram,
1993) and it is emphasized that one can be quite confident that something happened
without being able to recollect the specific experience. The question of interest is
whether misled participants given remember/know instructions would indicate they
“remember” witnessing suggested details, and several studies have now shown that
they do (Drivdahl & Zaragoza, 2001; Frost, 2000; Roediger, Jacoby, & McDermott,
1996; Zaragoza & Mitchell, 1996).
In summary, even when participants are warned about the misinformation and
are given a source-monitoring test that forces them to discriminate between differ
ent sources of information in memory, they continue to claim they remember wit
nessing suggested details. Moreover, on measures of their phenomenological
experience, participants indicate that they recollect seeing the suggested items,
much like they recollect memories derived from perceptual experiences. Collec
tively, these studies provide clear evidence that participants sometimes develop
genuine false memories for items and events that were only suggested to them.
The belief that one remembers witnessing an item that was only suggested is an ex
ample of a situation where a memory derived from one source (e.g., misleading
suggestions provided by an experimenter) is misattributed to another source (e.g.,
the witnessed event), an error we refer to as a source misattribution error. Marcia
Johnson and colleagues (see Johnson, Hashtroudi, & Lindsay, 1993; Johnson &
Raye, 1981; Lindsay, 1994) have developed a general theoretical framework, the
source-monitoring framework (SMF), that provides some insight into how such
errors come about.
According to the SMF, memory for source is an attribution (see also Jacoby,
Kelley, & Dywan, 1989, for a similar approach) that is the product of both con
scious and nonconscious judgment processes. From this view, information about
the source of a memory is not stored directly, but is based on an evaluation of the
characteristics of the memory representation. The SMF assumes that memory rep
resentations are records of the processing that occurred at encoding and thus con
tain features or characteristics that reflect the conditions under which the memory
was acquired (where and when each piece of information was acquired, modality
of presentation, emotional reactions, records of reflective processes, etc). So, for
example, if a memory contains a great deal of visual detail, an individual would
likely attribute this memory to an event he or she saw. People can, and often do, ac-
curately attribute the source of their memories because memories from different
sources tend to differ on average in the quantity and quality of the characteristics
associated with them (e.g., memories of perceived events typically have more
vivid perceptual, temporal, and spatial information than memories of imagined
events; Johnson, Foley, Suengas, & Raye, 1988). Nevertheless, because there can
be overlap in the distributions of the features associated with memories from dif-
ferent sources, errors can occur. For example, imagining words spoken in another
person’s voice increases people’s tendency to confuse what they imagined they
heard the person say with what they actually heard the person say, presumably be-
cause it increases the overlap between the characteristics of the two sources of in-
formation (Johnson, Foley, & Leach, 1988).
In situations where eyewitness suggestibility is a concern, the overlap between
the witnessed event and postevent interviews is extensive. First, the two episodes
are intimately related because they share a common referent—the witnessed
event. Note that the common referent factor is inherent in every eyewitness inter
rogation because, by definition, the postevent interview is always about the wit
nessed event (Mitchell & Zaragoza, 2001). As a consequence, with the exception
of several misleading details in the postevent interview, the content of the original
and postevent episodes is nearly identical. Second, in attempting to answer ques
tions about the witnessed event, participants are likely to actively retrieve and re
construct the originally witnessed events in their minds. This process of activating
the original memory while processing the postevent misinformation likely in
creases the overlap between the original event and postevent questioning even fur
ther. Hence, from the perspective of the SMF, it is not surprising that participants
sometimes confuse suggested items for items they witnessed firsthand.
The SMF also assumes that the accuracy of source-monitoring judgments is
heavily influenced by the circumstances at the time of retrieval (i.e., the appropri
ateness and stringency of the decision-making processes and criteria used). A
good illustration of this is Lindsay and Johnson’s (1989) finding that a suggestibil
ity effect is obtained with a yes–no recognition test but not on a source-monitoring
test. They proposed that yes–no recognition tests may encourage participants to
use a familiarity criterion when responding on a test of memory for the witnessed
event. Because the test list consists primarily of witnessed items interspersed with
novel foils, responding on the basis of familiarity will in most cases lead to a cor
rect response. For this reason, participants may slip into a tendency of using high
familiarity as the basis for deciding whether or not a test item was seen. Of course,
the suggested items are familiar not because they were witnessed but because par
ticipants had been exposed to them recently in the context of a postevent narrative.
In contrast to the yes–no test, the source-monitoring tests direct participants to re
trieve and use source-specifying information, thereby enhancing participants’
ability to discriminate between memories of the witnessed event and memories of
the postevent narrative.
Just as the criteria for deciding whether something was “seen” might change as
a function of test demands, the SMF also posits that the criteria by which people
judge a memory as “real” might change over time. Specifically, Johnson et al.
(1993) posit that the amount of perceptual detail needed to accept a remembered
experience as a real memory (and not imagined or suggested) is much greater for
recent events than for events from the distant past (see also Belli & Loftus, 1994).
In support of these predictions, Frost (2000) has shown that when misled partici-
pants are asked whether the suggested details they report are “remembered” or
“known, participants are more likely to claim they “remember” seeing the misin-
formation on delayed tests than on immediate tests.
An important aspect of the source-monitoring account is that people can mis-
take the origin of some item in memory even if memory for the item itself is very
good. Consistent with this idea, several of the studies reviewed in the following
sections show that people can have very strong memories of the content of mis-
leading suggestions yet misattribute their source.
Inspired in part by the SMF, some progress has been made in identifying factors
that, in combination with misleading suggestion, influence the creation of false
memories for suggested events. We review some of this research in the following
subsections. Because our primary concern in this section is false memories, we re
strict this review to studies that have used methods that attempt to differentiate be
tween false reports that reflect false memories and false reports that occur for other
reasons (e.g., source tests, warnings, measures of phenomenological experience).
The Role of Processing Resources
When attentional resources are limited, memory for an item’s source is more likely
to be disrupted than is the familiarity of the memory’s contents. This is because the
encoding and retrieval of source-relevant information are highly effortful, atten
tion-demanding processes, whereas familiarity is a relatively automatic conse
quence of exposure to an item (Johnson, Kounios, & Reeder, 1994). Thus, limiting
attentional resources can cause a relatively selective impairment of source-speci
fying information that renders the memory highly susceptible to misattribution
(cf. Jacoby, Woloshyn, & Kelley, 1989). A study by Zaragoza and Lane (1998)
verified these predictions. In one experiment, participants encountered the misin
formation under conditions of either divided or full attention, and in a second ex
periment participants were either given ample time to make the source judgment or
were forced to provide source judgments very quickly. The results showed that a
scarcity of attentional resources—either when encoding misinformation or when
retrieving misinformation—led to impoverished memory for the suggested infor
mation’s true source but no impairment in memory for the content of the suggested
item. This, in turn, led participants to misremember the suggestion as part of the
witnessed event. These results are consistent with the finding that forgetting of
source information that occurs over long retention intervals is accompanied by in-
creased suggestibility (see, e.g., Lindsay, 1990; Zaragoza & Mitchell, 1996), al-
though the possibility that other factors related to delay (e.g., weaker memory for
the witnessed event over time) might contribute to these latter results cannot be
ruled out. In sum, the finding that attentional resources influences suggestibility is
highly relevant to assessing and predicting suggestibility in real-world contexts,
where multiple environmental and internal stimuli (e.g., distraction due to height-
ened arousal) compete for attentional resources.
Repeated Suggestion and False Memory
Whereas the foregoing study shows that poor memory for the suggested item’s
source increases suggestibility, it is also the case that “source amnesia” is not a pre
condition for misattribution errors. Paradoxically, there are some variables, such
as repetition, that simultaneously increase source misattributions and improve
memory for the suggestions’ source.
Understanding the cognitive consequences of repeated suggestive interviews
has considerable practical, as well as theoretical implications. For example, re
peated suggestive questioning is not uncommon in eyewitness interrogation pro
cedures. Moreover, one of the reasons the therapeutic process is thought to be
potentially conducive to the formation of false memories is because suggestions
encountered in the course of therapy are likely to be repeated over time. Given the
current controversy surrounding allegedly false memories induced by therapy (see
chap. 8, this volume), the need for scientific evidence on the relationship between
repeated suggestion and false memory seems especially acute.
Several studies that have attempted to mimic the complexity of real-world in
terview situations involving repeated suggestion have demonstrated striking ex
amples of false memory in which participants claim to remember entire fictitious
events, such as getting lost in a mall (e.g., Ceci, Hoffman, Smith, & Loftus, 1994;
Ceci, Loftus, Leichtman, & Bruck, 1994; Hyman & Billings, 1998; Hyman, Hus
band, & Billings, 1995; Hyman & Pentland, 1996; Loftus & Ketcham, 1994; Por
ter, Yuille, & Lehman, 1999). However, in the interest of ecological validity, in all
of these studies repeated suggestion was confounded with several other variables
including the passage of time, experimenter demand, and generation of elabora
tive details. Hence, the role that repetition alone played in the creation of these
memories is difficult to discern.
Zaragoza and Mitchell (1996; see also Mitchell & Zaragoza, 1996) conducted
several studies whose goal was to isolate the effects of repeated exposure to sug
gestion on false-memory creation. Zaragoza and Mitchell found that, relative to a
single exposure, repeatedly exposing participants to suggestions increased the in
cidence of false memories for the suggested items, even when controlling for dif
ferences in recognition of the suggested items. Using a variant of the
remember/know paradigm, they also showed that repeated suggestions increased
participants’ claims that they specifically recollected witnessing the suggested
item in the video. In a follow-up study, Mitchell and Zaragoza showed that increas
ing the contextual variability of the repeated exposures (i.e., each repetition oc-
curred in a different modality) further exacerbated the repeated-exposures effect.
In both these studies, the deleterious effect of repeated exposures could not be at-
tributed to better memory for the suggested information following repetition, be-
cause the repetition effects remained the same when the data were conditionalized
on old–new recognition (see Mitchell & Zaragoza, 1996; Zaragoza & Mitchell,
1996). Hence, it is not the case that this false-memory effect is merely a function of
the greater familiarity of the repeated suggestion. It was also the case that in both
these studies, one consequence of repeated suggestion was an increase in partici-
pants’ claims that the suggested item came from two sources: both the witnessed
event and the postevent questionnaire. Hence, rather than impair participants’
memory for the suggestions’ actual source (the postevent questionnaire) repeated
suggestion had the opposite effect. What became more difficult with repeated sug
gestion was discriminating between the original and postevent episodes, as evi
denced by the increase in “both” responses. Recall that in the typical eyewitness
suggestibility situation, simply knowing that some piece of information came
from a postevent source is not very diagnostic with regard to its accuracy, because
much of the information provided by the postevent source is highly accurate.
Hence, the difficulty for participants was discriminating between information that
was only in the postevent questionnaire as opposed to both the postevent question
naire and the originally witnessed event.
How might repeated exposure to suggestion lead participants to
misremember witnessing suggested events? In the studies by Zaragoza and
Mitchell (1996; Mitchell & Zaragoza, 1996) the misleading suggestions were
embedded in questions about the video event that participants were required to
answer. They proposed that in answering such questions participants were likely
to retrieve and reflect upon the events they had witnessed. When the questions
contained misleading suggestions, it is likely that participants implicitly incor
porated the suggested information into their imagined reconstructions of the wit
nessed event. With repetition, these images of suggested events probably
became more elaborate and detailed (see Suengas & Johnson, 1988, for evidence
that rehearsing imagined events serves to preserve and embellish them), thus in
creasing their similarity to records of actually witnessed events. In addition, it is
likely that repetition increased the speed or fluency with which images of the
suggested events could be generated by the subjects, thereby increasing partici
pants’ confidence that the memories were real (cf. Kelley & Lindsay, 1993).
A less studied but related issue is the effect of repeatedly reporting suggested
information on false-memory creation. Using a cued-recall paradigm, Roediger,
Jacoby, and McDermott (1996) induced participants to report misinformation on a
first occasion, and assessed whether this would alter participants’ performance on
a later recall test (relative to a condition with no inducing manipulation). The clear
finding was that producing the misinformation on a first test increased the likeli
hood that participants misrecalled the misinformation on a later test (where they
were explicitly warned to ignore the postevent information). Moreover, repeated
reporting of suggested details also made participants more likely to claim they “re-
membered” witnessing the misinformation when instructed to use the remem-
ber/know procedure. Hence, Roediger et al. showed that repeated reporting of
suggested information, like repeated exposure, increases false memories for
suggested information.
Several different lines of evidence have shown that mentally elaborating on sug-
gested events increases false memory for these events. By “mental elaboration” we
mean any type of mental processing that embellishes the suggested event with de-
tails or other characteristics that render it confusable with a memory for a “real”
Imagery Instructions
There is considerable evidence that imagery is a catalyst for false-memory forma
tion. For example, there is evidence that both imagery ability (e.g., Dobson &
Markham, 1993) and preference for an imagic cognitive style (e.g., Labelle,
Laurence, Naden, & Perry, 1990) are related to susceptibility to false-memory cre
ation. In addition, studies have shown that instructing people to imagine fictitious
childhood experiences increases their belief that these fictitious events actually
occurred (Garry, Manning, Loftus, & Sherman, 1996; Paddock et al., 1998) and
can lead to false memories for fictitious childhood events (Hyman & Pentland,
1996). For example, Hyman and Pentland instructed participants to reminisce
about a number of events from their childhood, including one that had never actu
ally happened but had only been suggested by the experimenter (e.g., that they had
tipped over a punch bowl at a wedding onto the parents of the bride). Those partici
pants who were repeatedly instructed to imagine these events over three sessions
were later more likely to claim they remembered them (both fictitious and real)
than participants who were repeatedly instructed to merely think about them. Sim
ilar findings have been obtained with a somewhat different paradigm where partic
ipants are repeatedly instructed to imagine performing simple actions (Goff &
Roediger, 1998; Thomas, Bulevich, & Loftus, 2003).
Why might imagery be associated with false-memory creation? It is well docu
mented that discriminating between imagination and reality can be difficult, espe
cially when imagined events contain large numbers of features or characteristics
that are typical of actually experienced events (e.g., Johnson, Foley, & Leach,
1988; Johnson & Raye, 1981). In the case of false memories, imagining a fictitious
event likely results in a mental representation that closely resembles a real event,
because the act of imagination involves creating a specific instantiation of the ficti
tious event in one’s mind. To use an example from a study by Hyman and Pentland
(1996), imagining that as a child I once spilled a punch bowl at a wedding probably
involves creating a mental version of this incident that specifies such things as who
was there, what they looked like, where the wedding took place, how I spilled the
punch bowl, what the consequences were, how I felt about it, and so forth. In other
words, imagining and visualizing an event (in this example, a fictitious one) in-
volves reflectively elaborating on a hypothetical idea in a variety of ways so as to
produce a more concrete, specific, and perceptually and semantically detailed ver-
sion of the incident in one’s mind. At a broader level, imagining a fictitious event
likely renders the imagined event more familiar, available, and plausible, thus in-
creasing the likelihood that one would accept the imagined event as a real one (see,
e.g., Garry et al., 1996). Thus, there are probably multiple dimensions, both gen-
eral (e.g., familiarity) and specific (e.g., sensory-perceptual detail), on which
imagined events resemble real ones.
Sensory-Perceptual Elaboration
Although the association between imagery and false-memory creation is well
documented, the mechanisms by which imagery induces false-memory creation
are not yet well understood (cf. Thomas et al., 2003). Given that imagining a fic
titious event likely involves a great deal of perceptual elaboration (as well as re
flective elaboration on other dimensions) Drivdahl and Zaragoza (2001) set out
to assess whether leading participants to reflectively elaborate on the perceptual
characteristics of misleading suggestions would increase false memory for the
suggested events. To this end, they used an eyewitness suggestibility paradigm
where participants viewed a videotape depicting a burglary and were later ex
posed to misleading suggestions (e.g., they were misinformed that the thief stole
a ring, when in fact he did not steal any jewelry). The perceptual elaboration ma
nipulation was implemented by asking participants yes-or-no follow-up ques
tions about the perceptual characteristics (e.g., location, physical appearance,
etc.) of a previously mentioned item. For example, some participants who had
been misinformed that the thief stole a ring were asked the follow-up question,
“Did he find the ring in the top drawer?” When a suggestion was repeated, partic
ipants answered a different question about the suggested item each time they en
countered it (e.g., “Was the ring in a box?” “Did the ring have a gemstone?”). In
this way, the follow-up questioning resembled interview situations where wit
nesses are pressed to provide specific details of fictitious or poorly remembered
events. Two days later, participants were given a test of their memory for the
source of the suggested items. The results showed that participants who an
swered questions that encouraged them to elaborate on the perceptual character
istics of the suggested items, such as their location and physical appearance,
were much more likely to later claim they “definitely” remembered seeing them
than participants in a no-elaboration group, who were exposed to the same sug
gestions but answered follow-up questions about relatively superficial aspects of
the suggested items (e.g., its rhyming characteristics). Drivdahl and Zaragoza
proposed that answering perceptual-elaboration questions increased false mem-
ory because it induced participants to form a more perceptually detailed, spe-
cific, and embellished representation of the suggested events than they would
have otherwise done (see Thomas et al., 2003, for the related finding that in-
structing participants to include sensory details in their imaginings increased
source misattributions). Consistent with this proposal, there is considerable evi-
dence that the greater the vividness, clarity, and detail associated with imagined
events, the greater the likelihood that they will be confused for actually experi-
enced events (see Johnson et al., 1993, for a review). Note that one important
way in which the Drivdahl and Zaragoza study differs from previous imagina-
tion studies is that participants were not explicitly instructed to imagine the sug-
gested events, but were asked follow-up questions that encouraged the implicit
generation of visualized detailed images of the suggested events.
Semantic Elaboration
Although the foregoing results show that the sensory-perceptual characteristics
of imagined suggested events contribute to false-memory creation, it is also pos
sible that the effects of imagery are also due in part to the meaningful elaborative
processing that imaging entails. Note that attempts to imagine fictitious
items/events are also likely to involve more abstract sorts of reasoning about the
meaning and implications of the fictitious events, and a consideration of the
plausible scenarios within which they might have transpired. This sort of elabo
rative processing may serve to establish stronger and more numerous connec
tions between the misleading suggestions and other information in memory. To
the extent that suggested memories that are embedded in a coherent network of
relations are likely to be confused for “real memories, it is possible that simply
elaborating on the meaning and implications of suggested events might serve to
increase false memory.
In a follow-up to the Drivdahl and Zaragoza (2001) study, Zaragoza, Mitch
ell, Payment, and Drivdahl (2004) examined the relative contributions of percep
tual and meaningful elaboration on false-memory creation by manipulating the
type of follow-up questions participants received about the suggested informa
tion. Participants in the semantic-elaboration group answered follow-up ques
tions that led them to think about the meaning and implications of the suggested
details, but not their perceptual characteristics. For example, some participants
in the semantic-elaboration condition who received the misleading suggestion
that the “thief stole a ring” were asked the follow-up question “How incriminat
ing would a jury find it if they were told that the thief was found with a stolen ring
in his possession?” If the suggestion was repeated, participants answered a dif
ferent type of follow-up question every time they encountered it (e.g., “Was the
fact that the thief stole a ring central to the plot?” “Do you think the thief was
disappointed that he did not find other jewelry besides the ring?”). As in the
Drivdahl and Zaragoza study, participants in the perceptual-elaboration group
were asked follow-up questions about the perceptual characteristics of the
suggested items. Relative to a repetition-control group that had repeated expo-
sure to the suggested information but no elaboration, participants in both elabo-
ration groups evinced higher levels of false memory for the suggested detail. The
proportion of times participants claimed they “definitely” remembered seeing
suggested details was .28, .41, and .59 in the repetition-control, perceptual-elab-
oration, and semantic-elaboration groups, respectively. Hence, a novel and un-
expected finding was that semantic elaboration led to greater increases in false
memory than perceptual elaboration (even though type of elaboration had no ef-
fect on any other dependent measure). Zaragoza et al. proposed that semantic
elaboration promotes false memory by establishing stronger and more numerous
connections between the misleading suggestions and other well-developed
knowledge structures in memory (e.g., schemas regarding juries, what consti-
tutes incriminating evidence, etc). In sum, it appears that misinformation that is
embedded in a coherent network of relations is more likely to be confused for a
“real” memory. (See also Drivdahl, 2001, for the related finding that inducing
participants to think about the emotional consequences of suggested events also
increases false memory.)
The foregoing studies on perceptual and semantic elaboration induced partici
pants to mentally elaborate on suggested items that pertained to recently experi
enced events. When participants are misled about events from their remote
past—such as their childhood—their relatively impoverished memory of their dis
tant past may make it more difficult for them to construct a perceptually and se
mantically detailed image of the suggested childhood event. On the other hand, if
participants are given cues, such as photographs, that help them retrieve aspects of
their childhood, they may be better equipped to form a compelling and plausible
image of the suggested (i.e., fictitious) childhood event. This, in turn, should in
crease their susceptibility to developing a false memory for the suggested event. A
fascinating study by Lindsay, Hagen, Read, Wade, and Garry (2004) provides evi
dence consistent with this prediction.
In the Lindsay, Hagen et al. (2004) study, participants were asked to remem
ber three school-related childhood events: two true events provided by parents,
and a third, fictitious event (putting slime in the teacher’s desk in Grade 1 or 2).
For participants in the photo group, the experimenter provided a color photo of
their class picture at the time of suggested the event, and participants in all
groups were encouraged to recall as much as possible using mental-context re
instatement and guided-imagery exercises over a period of several days. In a fi
nal test session, participants in the photo condition were twice as likely as
participants in as participants in a no-photo condition to experience false mem
ories, and their ratings of the extent to which remembering the event was like
reliving it was comparable to the ratings they gave for true events. The authors
proposed that although the photo did not depict the slime prank, the photo may
have enabled participants to speculate about plausible scenarios involving the
suggested event (e.g., “Who would my collaborator in the slime prank have
been?”). Participants in the no-photo condition may have had difficulty con-
structing such scenarios because of their inability to recall relevant elements,
such as who was in their classroom and the appearance of the teacher. The au-
thors further proposed that participants may have used perceptual details from
the photo (e.g., the teacher’s appearance) to produce vivid images of the ficti-
tious event, thus producing especially compelling false memories (see Lindsay
et al., 1994).
The misinformation paradigm provides a laboratory analogue of real-world situa
tions involving eyewitnesses to forensically relevant events who are interviewed
suggestively about the events they have witnessed. For this reason, most studies of
misinformation effects have presented misinformation in the context of postevent
questions or postevent narratives about the witnessed event. Recent studies have
begun to document that eyewitness memories can also be contaminated by misin
formation encountered outside the postevent interview context.
Studies of the social-contagion effect (Meade & Roediger, 2002; Roediger,
Meade & Bergman, 2001) have shown that when two people witness the same
event, a person’s memory for the witnessed information can be contaminated by
false information provided by the cowitness. In these studies, two participants
(one of whom is a confederate) view a series of scenes and then engage in a collab
orative-recall task where they each recall several items from each scene. However
some of the items “recalled” by the confederate are false in that they suggested ob
jects that had not actually appeared in the scene (and thus serve as misinforma
tion). The finding of interest is that when participants are later tested individually
they recall these suggested items even when told to recall only those items they re
membered from the scenes. Moreover, neither warning participants that others’re
sponses might influence their recall, nor giving them a source-monitoring test
(which oriented them to discriminate between items they saw and those the
cowitness reported) eliminates the social-contagion effect (Meade & Roediger,
2002). These findings extend research on the misinformation effect by showing
that even information provided by an unfamiliar peer can affect participants’mem
ories for events they have witnessed.
Perhaps even more surprising than the social-contagion effects is the recent
finding that participants will sometimes intrude details from one event into their
memory of a different event (Allen & Lindsay, 1998; Lindsay, Allen, Chan, &
Dahl, 2004).
In these studies, participants viewed a movie clip depicting a museum burglary
and read a narrative description of either (a) the same event (i.e., the museum bur-
glary they had seen), (b) a different, but thematically related, event (a palace bur-
glary), or (c) a less closely related event (a school field trip to a palace). In every
case, the narratives contained information that was not in the original movie clip.
The finding of interest was that when participants were given a cued-recall test of
their memory for the video, participants in all three conditions falsely recalled in-
formation from the narratives, though false recall was greater when the narrative
described the witnessed event rather than a different event. A striking finding was
that these intrusions occurred even when the narrative described an event (a school
field trip to a palace) that was in many ways quite different from the event depicted
in the movie clip participants had seen (a burglary in a museum).
To date, most studies in the eyewitness suggestibility literature have focused al
most exclusively on suggestive interviews involving false-memory implanta
tion. In the implantation paradigm, the witness is given misinformation about a
witnessed event, and suggestibility is measured as the extent to which the wit
ness then (or later) assents to the misinformation provided by the interviewer.
However, in real-world forensic and therapeutic settings, suggestive interview
practices are not restricted to situations involving the explicit provision of misin
formation. Rather, in some cases interviewers attempt to elicit from witnesses
accounts that support their beliefs about what transpired (cf. Bruck, Ceci, &
Hembrooke, 1998). To this end, interviewers may forcibly press witnesses to de
scribe those events they believe transpired, even when witnesses cannot remem
ber or never witnessed the events they are pressed to testify about (cf.
Gudjonssson, 1992; Leo, 1996).
Might witnesses eventually develop false memories for events they had ear
lier been forced to confabulate? Intuitively, it seems unlikely they would do so.
Presumably, events that are confabulated deliberately and under duress will be
remembered as mere fabrications, even over the long term. However, contrary to
this intuition, a recent study (Zaragoza, Payment, Ackil, Drivdahl, & Beck,
2001) showed that participants who were pressed to confabulate information
about a witnessed event later evidenced false memories for some of the events
they had earlier confabulated knowingly (see also Ackil & Zaragoza, 1998, for
similar evidence with children), a phenomenon they call the “forced confabula
tion effect. In this study, participants viewed a movie clip and were then asked
specific questions about both true and blatantly false events. In order to answer
these false-event questions, the participants were required to confabulate, or
make something up. For example, in going over a scene from the video, the ex
perimenter said, “It [the chair] broke, and Delaney fell on the floor. Where was
Delaney bleeding?” This question required a confabulated response because al-
though Delaney did fall off a chair in the video, he clearly did not bleed nor hurt
himself in any way.
As one might expect, participants firmly resisted answering such questions, but
were repeatedly pressed to guess until they eventually acquiesced. To illustrate,
the following is a transcript from a forced-confabulation interview:
Interviewer: After he fell, where was Delaney Bleeding?
Participant: He wasn’t. He was? I didn’t see any blood.
Interviewer: What’s your best guess?
Participant: Where was he bleeding?
Interviewer: Yeah.
Participant: But he wasn’t bleeding. Oh, I don’t have a best guess. I didn’t
think he was bleeding. His knee?
Interviewer: Okay, his knee.
Participant: Its not his knee! (Zaragoza et al., 2001, p. 476)
Throughout the interview, the interviewer selectively reinforced some of the
participants’ confabulated responses by providing confirmatory feedback (e.g.,
“Knee is the correct answer!”) and provided neutral (uninformative feedback for
the remaining confabulated responses (e.g., “OK, knee”—see earlier transcript)
(see Wells & Bradfield, 1998, for other evidence that confirmatory feedback
leads to distortions in eyewitness testimony). One week later, a large proportion
of participants misremembered witnessing the events they had earlier confabu
lated knowingly (even though they were warned in advance that they had been
questioned about fictitious events). Moreover, confirmatory feedback increased
false memory for forcibly confabulated events, increased confidence in those
memories, and increased the likelihood that participants would freely report the
confabulated events 1 to 2 months later. Important to note, the authors were able
to show that these false-memory effects were not dependent on memory for the
In an ingenious experiment, Kassin and Kiechel (1996) demonstrated that
false incriminating evidence can lead people to believe—and even remem
ber—that they committed a crime they did not commit. Research on this topic is
of great practical importance as the presentation of false incriminating evidence
is a common interrogation technique, and studies have shown that coerced con
fessions increase the conviction rate even when participants recognize they are
coerced, even when it is stricken from the record, and even when jurors claim it
had no influence on their verdict (Kassin & Sukel, 1997). In their study, partici
pants engaged in a computer reaction-time task and were incorrectly accused of
hitting the wrong key and damaging the computer. All of the participants initially
denied hitting the key, and indeed none of them had done so. Two variables were
manipulated: First, in order to influence participants’ certainty in their own inno
cence, the pace of the computer task was varied (either very fast paced or slow
paced). Second, they varied whether or not participants received the false in
criminating evidence (the “misinformation”). Specifically, for participants in
the false-witness condition, an experimental confederate claimed she saw the
participant hit the key, whereas for participants in the no-witness condition the
same confederate said she had not seen what happened. There were three out-
come measures: participants’ willingness to sign a confession (a measure of
compliance), participants’ tendency to freely report to another participant that
they had hit the key/damaged the computer (a measure of internalization), and
participants’ tendency to “recall” specific details to fit the allegation, such as “I
hit it with the side of my right hand after you called out the A’” (Kassin &
Kiechel, 1996, p. 127). The latter served as a measure of confabulation (or false
memory). The results showed that both variables influenced participants’perfor-
mance on all measures: Compared to participants in the slow-pace/no-witness
group, participants in the fast-pace/false-witness group were more likely to sign
the confession, internalize guilt for the event, and confabulate details in memory
consistent with the false belief. Interestingly, the false incriminating evidence
was a critical ingredient in producing confabulations. No participants in the
no-witness groups (whether slow or fast paced) produced confabulations. In
summary, these findings extend studies of misinformation phenomena by show
ing that people can be misled about their own actions, not just events they have
observed. Moreover, like the forced-confabulation studies, these studies show
that coercing witnesses to say things against their will (describing events they
did not see, providing self-incriminating statements), can seriously distort their
memories over the long term.
Most of the research on misinformation phenomena reviewed here has docu
mented false memories for selected aspects of a witnessed event (e.g., a false
memory that the thief had a gun, when in fact he had no weapon). In recent
years, the controversy surrounding allegedly false recovered memories of
childhood abuse has raised important questions about the extent of people’s
susceptibility to memory illusions. In particular, a question of central concern
in this debate is whether people can be led to develop false memories for entire
events that never actually transpired (see chap. 8, this volume). Once again,
Elizabeth Loftus has been at the forefront of the effort to provide scientific evi
dence that bears on this issue of pressing social concern by developing innova
tive methods for addressing such questions in the laboratory (e.g., Loftus,
Coan, & Pickrell, 1979/1996)
Although the recovered/false-memory debate has captured a great deal of
attention in recent years, research on this problem is in many ways a natural
outgrowth of research on misinformation phenomena. Although the latter fo
cuses on participants’ memories for details of recent eyewitness events, many
of the factors that appear to promote the creation of false memories for sug
gested childhood events (e.g., imagery, repetition, forced generation, mental
elaboration) were discovered in the context of basic research on the misinfor-
mation effect. Our hope is that the ongoing recovered/false-memory debate
may resolve itself as the once lively debates on the misinformation effect have
largely been resolved, with scientists who once took adversarial positions be-
ginning to see that their points of agreement are more numerous than those
points on which they disagree. Finally, as this chapter has shown, research on
the misinformation effect is an ongoing energetic direction of investigation
that continues to provide unique insights into the nature of memory and its sus-
ceptibility to error.
Ackil, J. K., & Zaragoza, M. S. (1998). Memorial consequences of forced con-
fabulations: Age differences in susceptibility to false memories. Developmental
Psychology, 34, 1358–1372.
Allen, B. P., & Lindsay, D. S. (1998). Amalgamations of memories: Intrusions
of information from one event into reports of another. Applied Cognitive Psychol
ogy, 12, 277–285.
Ayers, M. S., & Reder, L. M. (1998). A theoretical review of the misinforma
tion effect: Predictions from an activation-based memory model. Psychonomic
Bulletin & Review, 5, 1–21.
Banaji, M. R., & Crowder, R. G. (1989). The bankruptcy of everyday memory.
American Psychologist, 44, 1185–1193.
Bekerian, D. A., & Bowers, J. M. (1983). Eyewitness testimony: Were we mis
led? Journal of Experimental Psychology: Learning, Memory, and Cognition, 9,
139–145.Belli, R. F. (1988). Color blend retrievals: Compromise memories or de
liberate compromise responses? Memory & Cognition, 16, 314–326.
Belli, R. F. (1989). Influences of misleading postevent information: Misinfor
mation interference and acceptance. Journal of Experimental Psychology: Gen
eral, 118, 72–85.
Belli, R. F. (1993). Failure of interpolated tests in inducing memory impair
ment with final modified tests: Evidence unfavorable to the blocking hypothesis.
American Journal of Psychology, 106, 407–427.
Belli, R. F., Lindsay, D. S., Gales, M. S., & McCarthy, T. T. (1994). Memory
impairment and source misattribution in postevent misinformation experiments
with short retention intervals. Memory & Cognition, 22, 40–54.
Belli, R. F., & Loftus, E. F. (1994). Recovered memories of childhood
abuse: A source monitoring perspective. In S. J. Lynn & J. Rhue (Eds.), Disso
ciation: Theory, clinical, and reseach perspectives (pp. 415–433). New York:
Belli, R. F., & Loftus, E. F. (1996). The pliability of autobiographical memory:
Misinformation and the false memory problem. In D. C. Rubin (Ed.), Remember
ing our past: Studies in autobiographical memory (pp. 157–179). New York:
Cambridge University Press.
Belli, R. F., Windschitl, P. D., McCarthy, T. T., & Winfrey, S. E. (1992). Detect-
ing memory impairment with a modified test procedure: Manipulating retention
interval with centrally presented event items. Journal of Experimental Psychol-
ogy: Learning, Memory, and Cognition, 18, 356–367.
Bowman, L. L., & Zaragoza, M. S. (1989). Similarity of encoding context does
not influence resistance to memory impairment following misinformation. Ameri-
can Journal of Psychology, 102, 249–264.
Bruck, M., Ceci, S. J., & Hembrooke, H. (1998). Reliability and credibility of
young children’s reports: From research to policy and practice. American Psychol-
ogist, 53, 136–151.
Ceci, S. J., Huffman, M. L. C., Smith, E., & Loftus, E. F. (1994). Repeatedly
thinking about a non-event: Source misattributions among preschoolers. Con-
sciousness & Cognition: An International Journal, 3(3–4), 388–407.
Ceci, S. J., Loftus, E. F., Leichtman, M. D., & Bruck, M. (1994). The possible
role of source misattributions in the creation of false beliefs among preschoolers.
International Journal of Clinical and Experimental Hypnosis, 42, 304–320.
Ceci, S. J., Ross, D. F., & Toglia, M. P. (1987). Suggestibility of children’s
memory: Psycholegal implications. Journal of Experimental Psychology: Gen
eral, 116, 38–49.
Chambers, K. L., & Zaragoza, M. S. (2001). Intended and unintended effects of
explicit warnings on eyewitness suggestibility: Evidence from source identifica
tion tests. Memory & Cognition, 29, 1120–1129.
Chandler, C. C. (1989). Specific retroactive interference in modified recogni
tion tests: Evidence for an unknown cause of interference. Journal of Experimen
tal Psychology: Learning, Memory, and Cognition, 15, 256–265.
Chandler, C. C., Gargano, G. J., & Holt, B. C. (2001). Witnessing postevents
does not change memory traces, but can affect their retrieval. Applied Cognitive
Psychology, 15, 3–22.
Christiaansen, R. E., & Ochalek, K. (1983). Editing misleading information
from memory: Evidence for the co-existence of original and postevent informa
tion. Memory & Cognition, 11, 467–475.
Crowder, R. G. (1976). Principles of learning and memory. Oxford, England:
Lawrence Erlbaum Associates.
Dobson, M., & Markham, R. (1993). Imagery ability and source monitoring:
Implications for eyewitness memory. British Journal of Psychology, 84,
Dodd, D. H., & Bradshaw, J. M. (1980). Leading questions and memory:
Pragmatic constraints. Journal of Verbal Learning & Verbal Behavior, 19(6),
Drivdahl, S. B. (2001). The role of emotion and self-reference in the creation of
false memories for suggested events. Dissertation Abstracts, 61, 6156.
Drivdahl, S. B., & Zaragoza, M. S. (2001). The role of perceptual elaboration
and individual differences in the creation of false memories for suggested events.
Applied Cognitive Psychology, 15, 265–281.
Eakin, D. K., Schreiber, T. A., & Sergent-Marshall, S. (2003). Misinformation
effects in eyewitness memory: The presence and absence of memory impairment
as a function of warning and misinformation accessibility. Journal of Experimen-
tal Psychology: Learning, Memory, and Cognition, 29, 813–825.
Frost, P. (2000). The quality of false memory over time: Is misinformation “re-
membered” or “known”? Psychonomic Bulletin & Review, 7(3), 531–536.
Frost, P., Ingraham, M., & Wilson, B. (2002). Why misinformation is more
likely to be recognised over time. A source monitoring account. Memory, 10(3),
Gardiner, J. M., & Java, R. I. (1990). Recollective experience in word and
nonword recognition. Memory & Cognition, 18, 23–30.
Gardiner, J. M., & Java, R. I. (1993). Recognition memory and awareness: An
experiential approach. European Journal of Cognitive Psychology, 5, 337–346.
Garry, M., Manning, C. G., Loftus, E. F., & Sherman, S. J. (1996). Imagination
inflation: Imagining a childhood event inflates confidence that it occurred.
Psychonomic Bulletin & Review, 3, 208–214.
Goff, L. M., & Roediger, H. L. (1998). Imagination inflation for action events:
Repeated imaginings lead to illusory recollections. Memory & Cognition, 26,
Gudjonsson, G. H. (1992). The psychology of interrogations, confessions and
testimony. Oxford, England: Wiley.
Hekkanen, S. T., & McEvoy, C. (2002). False memories and source-monitoring
problems: Criterion differences. Applied Cognitive Psychology, 16, 73–85.
Hyman, I. E., & Billings, F. J. (1998). Individual differences and the creation of
false childhood memories. Memory, 6, 1–20.
Hyman, I. E., Husband, T. H., & Billings, F. J. (1995). False memories of child
hood experiences. Applied Cognitive Psychology, 9, 181–197.
Hyman, I. E., & Pentland, J. (1996). The role of mental imagery in the creation
of false childhood memories. Journal of Memory and Language, 35, 101–117.
Jacoby, L. L., Kelley, C. M., & Dywan, J. (1989). Memory attributions. In H. L.
Roediger&F.I.M.Craik(Eds.),Varieties of memory and consciousness: Essays
in honor of Endel Tulving (pp. 391–422). Hillsdale, NJ: Lawrence Erlbaum Asso
Jacoby, L. L., Woloshyn, B., & Kelley, C. M. (1989). Becoming famous with
out being recognized: Unconscious influences of memory produced by divided at
tention. Journal of Experimental Psychology: General, 118, 72–85.
Johnson, M. K., Foley, M. A., & Leach, K. (1988). The consequences for mem
ory of imagining another person’s voice. Memory & Cognition, 16, 337–342.
Johnson, M. K., Foley, M. A., Suengas, A. G., & Raye, C. L. (1988). Phenome
nal characteristics of memories for perceived and imagined autobiographical
events. Journal of Experimental Psychology: General, 117, 371–376.
Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring.
Psychological Bulletin, 114, 3–28.
Johnson, M. K., Kounios, J., & Reeder, J. A. (1994). Time-course studies of re-
ality monitoring and recognition. Journal of Experimental Psychology: Learning,
Memory, and Cognition, 20, 1409–1419.
Johnson, M. K., & Raye, C. L. (1981). Reality monitoring. Psychological Re-
view, 88, 67–85.
Kassin, S. M., & Kiechel, K. L. (1996). The social psychology of false confes-
sions: Compliance, internalization, and confabulation. Psychological Science, 7,
Kassin, S. M., & Sukel, H. (1997). Coerced confessions and the jury: An exper-
imental test of the “harmless error” rule. Law & Human Behavior, 21, 27–46.
Kelley, C. M., & Lindsay, S. D. (1993). Remembering mistaken for knowing:
Ease of retrieval as a basis for confidence in answers to general knowledge ques
tions. Journal of Memory and Language, 32, 1–24.
Labelle, L., Laurence, J., Nadon, R., & Perry, C. (1990). Hypnotizability, pref
erence for an imagic cognitive style, and memory creation in hypnosis. Journal of
Abnormal Psychology, 99, 222–228.
Lampinen, J. M., & Smith, V. L. (1995). The incredible (and sometimes incred
ulous) child witness: Child eyewitnesses’ sensitivity to source credibility cues.
Journal of Applied Psychology, 80, 621–627.
Lane, S. M., Mather, M., Villa, D., & Morita, S. K. (2001). How events are re
viewed matters: Effects of varied focus on eyewitness suggestibility. Memory &
Cognition, 29, 940–947.Leo, R. A. (1996). Miranda’s revenge: Police interroga
tion as a confidence game. Law & Society Review, 30, 259–288.
Lindsay, D. S. (1990). Misleading suggestions can impair eyewitnesses’ ability
to remember event details. Journal of Experimental Psychology: Learning, Mem
ory, and Cognition, 16, 1077–1083.
Lindsay, D. S. (1994). Memory source monitoring and eyewitness testimony.
In D. F. Ross, J. D. Read, & M. P. Toglia (Eds.), Adult eyewitness testimony: Cur
rent trends and developments (pp. 27–55). New York: Springer-Verlag.
Lindsay, D. S., Allen, B. P., Chan, J. C. K., & Dahl, L. C. (2004). Eyewitness
suggestibility and source similarity: Intrusions of details from one event into
memory reports of another event. Journal of Memory and Language, 50, 96–111.
Lindsay, D. S., Hagen, L., Read, J. D., Wade, K. A., & Garry, M. (2004). True
photographs and false memories. Psychological Science, 15, 149–154.
Lindsay, D. S., & Johnson, M. K. (1989). The eyewitness suggestibility effect
and memory for source. Memory & Cognition, 17, 349–358.Loftus, E. F. (1975).
Leading questions and the eyewitness report. Cognitive Psychology, 7, 560–572.
Loftus, E. F. (1977). Shirting human color memory. Memory & Cognition, 5,
696–699.Loftus, E. F. (1979a). Eyewitness testimony. Cambridge, MA: Harvard
University Press.
Loftus, E. F. (1979b). The malleability of human memory. American Scientist,
67, 312–320.Loftus, E. F. (1979c). Reactions to blatantly contradictory informa-
tion. Memory & Cognition, 7, 368–374.
Loftus, E. F., Coan, J. A., & Pickrell, J. E. (1996). Manufacturing false memo-
ries using bits of reality. In L. M. Reder (Ed.), Implicit memory and metacognition
(pp. 195–220). Mahwah, NJ: Lawrence Erlbaum Associates. (Original work pub-
lished 1979)
Loftus, E. F., Donders, K., Hoffman, H. G., & Schooler, J.W. (1989). Creating
new memories that are quickly accessed and confidently held. Memory & Cogni-
tion, 17, 607–616.
Loftus, E. F., & Hoffman, H. G. (1989). Misinformation and memory: The cre-
ation of new memories. Journal of Experimental Psychology: General, 118,
Loftus, E. F., & Ketcham, K. (1994). The myth of repressed memory.NewYork:
St. Martin’s Press.Loftus, E. F., & Loftus, G. R. (1980). On the permanence of
stored information in the brain. American Psychologist, 35, 409–420.
Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of ver
bal information into a visual memory. Journal of Experimental Psychology: Hu
man Learning and Memory, 4, 19–31.
Loftus, E. F., & Palmer, J. E. (1974). Reconstruction of automobile destruction:
An example of the interaction between language and memory. Journal of Verbal
Learning and Verbal Behavior, 13, 585–589.
Loftus, E. F., Schooler, J. W., & Wagenaar, W. (1985). The fate of memory:
Comment on McCloskey and Zaragoza. Journal of Experimental Psychology:
General, 114, 375–380.
McCloskey, M., & Zaragoza, M. (1985a). Misleading postevent information
and memory for events: Arguments and evidence against memory impairment hy
potheses. Journal of Experimental Psychology: General, 114, 1–16.
McCloskey, M., & Zaragoza, M. (1985b). Postevent information and memory:
Reply to Loftus, Schooler, and Wagenaar. Journal of Experimental Psychology:
General, 114, 381–387.
Meade, M. L., & Roediger, H. L. (2002). Explorations in the social contagion of
memory. Memory & Cognition, 30, 995–1009.Metcalfe, J. (1990). Composite Ho
lographic Associative Recall Model (CHARM) and blended memories in eyewit
ness testimony. Journal of Experimental Psychology: General, 119, 145–160.
Mitchell, K. J. ,& Zaragoza, M. S. (1996). Repeated exposure to suggestion and
false memory: The role of contextual variability. Journal of Memory and Lan
guageI 246V260.
Mitchell, K. J., & Zaragoza, M. S. (2001). Contextual overlap and eyewitness
suggestibility. Memory & Cognition, 29, 616–626.
Paddock, J. R., Joseph, A. L., Chan, F. M., Terranova, S., Manning, C., & Lof
tus, E. F. (1998). When guided visualization procedures may backfire: Imagina
tion inflation and predicting individual differences in suggestibility [Special
issue]. Applied Cognitive Psychology, 12, 563–575.
Payne, D. G., Toglia, M. P., & Anastasi, J. S. (1994). Recognition performance
level and the magnitude of the misinformation effect in eyewitness memory.
Psychonomic Bulletin & Review, 1, 376–382.Porter, S., Yuille, J. C., & Lehman,
D. R. (1999). The nature of real, implanted, and fabricated memories for emo-
tional childhood events: Implications for the recovered memory debate. Law &
Human Behavior, 23, 517–537.
Rajaram, S. (1993). Remembering and knowing: Two means of access to the
personal past. Memory & Cognition, 21, 89–102.
Roediger, H. L., Jacoby, D., & McDermott, K. B. (1996). Misinformation ef-
fects in recall: Creating false memories through repeated retrieval. Journal of
Memory and Language, 35, 300–318.
Roediger, H. L., Meade, M. L., & Bergman, E. T. (2001). Social contagion of
memory. Psychonomic Bulletin & Review, 8, 365–371.
Schooler, J. W., Foster, R. A., & Loftus, E. F. (1988). Some deleterious conse
quences of the act of recollection. Memory & Cognition, 16, 243–251.
Schooler, J. W., Gerhard, D., & Loftus, E. F. (1986). Qualities of the unreal. Jour
nal of Experimental Psychology: Learning, Memory, and Cognition, 12, 171–181.
Schreiber, T. A., & Sergent, S. D. (1998). The role of commitment in producing
misinformation effects in eyewitness memory. Psychonomic Bulletin & Review, 5,
Smith, V. L., & Ellsworth, P. C. (1987). The social psychology of eyewitness
accuracy: Misleading questions and communicator expertise. Journal of Applied
Psychology, 72, 294–300.
Suengas, A. G., & Johnson, M. K. (1988). Qualitative effects of rehearsal on
memories for perceived and imagined complex events. Journal of Experimental
Psychology: General, 117, 377–389.
Thomas, A. K., Bulevich, J. B., & Loftus, E. F. (2003). Exploring the role of
repetition and sensory elaboration in the imagination inflation effect. Memory &
Cognition, 31, 630–640.
Tousignant, J. P., Hall, D., & Loftus, E. F. (1986). Discrepancy detection and
vulnerability to misleading postevent information. Memory & Cognition, 14,
Tulving, E. (1985). Memory and consciousness. Canadian Psychology, 26,
Tversky, B., & Tuchin, M. (1989). A reconciliation of the evidence on eyewit
ness testimony: Comments on McCloskey and Zaragoza. Journal of Experimental
Psychology: General, 118, 86–91.
Underwood, J., & Pezdek, K. (1998). Memory suggestibility as an example of
the sleeper effect. Psychonomic Bulletin & Review, 5, 449–453.
Vornik, L. A., Sharman, S. J., & Garry, M. (2003). The power of the spoken word:
Sociolinguistic cues influence the misinformation effect. Memory, 11, 101–109.
Wells, G. L., & Bradfield, A. L. (1998). “Good, you identified the suspect”:
Feedback to eyewitnesses distorts their reports of the witnessing experience. Jour-
nal of Applied Psychology, 83, 360–376.
Windschitl, P. D. (1996). Memory for faces: Evidence of retrieval-based im-
pairment. Journal of Experimental Psychology: Learning, Memory, and Cogni-
tion, 22, 1101–1122.
Zaragoza, M. S. (1991). Preschool children’s susceptibility to memory impair-
ment. In J. Doris (Ed.), The suggestibility of childrens recollections (pp. 27–39).
Washington, DC: American Psychological Association.
Zaragoza, M. S., Dahlgren, D., & Muench, J. (1992). The role of memory im-
pairment in children’s suggestibility. In M. L. Howe, C. J. Brainerd, & V. F. Reyna
(Eds.), The development of long-term retention (pp. 184–216). New York:
Zaragoza, M. S., & Koshmider, J. W., III (1989). Misled subjects may know
more than their performance implies. Journal of Experimental Psychology:
Learning, Memory, and Cogntion, 15, 246–255.
Zaragoza, M. S., & Lane, S. M. (1994). Source misattributions and the suggest
ibility of eyewitness memory. Journal of Experimental Psychology: Learning,
Memory, &and Cogntion, 20, 934–945.
Zaragoza, M. S., & Lane, S. M. (1998). Processing resources and eyewitness
suggestibility. Legal & Criminological Psychology, 3(Pt. 2), 305–320.
Zaragoza, M. S., Lane, S. M., Ackil, J. K., & Chambers, K. L. (1997). Confus
ing real and suggested memories: Source monitoring and eyewitness suggestibil
ity.InN.Stein,P.A.Ornstein,B.Tversky,&C.Brainerd(Eds.),Memory for
everyday and emotional events (pp. 401–428). Mahwah, NJ: Lawrence Erlbaum
Zaragoza, M. S., & McCloskey, M. (1989). Misleading postevent information
and the memory impairment hypothesis: Comment on Belli and reply to Tversky
and Tuchin. Journal of Experimental Psychology: General, 118, 92–99.
Zaragoza, M. S., McCloskey, M., & Jamis, M. (1987). Misleading postevent in
formation and recall of the original event: Further evidence against the memory
impairment hypothesis. Journal of Experimental Psychology: Learning, Memory,
and Cogntion, 13, 36–44.
Zaragoza, M. S., & Mitchell, K. J. (1996). Repeated exposure to suggestion and
the creation of false memories. Psychological Science, 7, 294–300.
Reflective elaboration and false memory: The role of relational processing.Manu
script in preparation, Kent State University, Department of Psychology.
(2001). Interviewing witnesses: Forced confabulation and confirmatory feedback
increase false memories. Psychological Science, 12, 473–477.
... Subsequently, the participants answer a series of questions about the original event, including critical questions relating to the misinformation. In almost all experiments of this kind, it has been found that participants in the experimental group perform worse on the final memory test as they usually include some of the misinformation in their answers (for a review, see Zaragoza et al., 2007). ...
Full-text available
The misinformation effect occurs when an eyewitness includes information in his or her account that is incongruent with the event he or she witnessed, and stems from being exposed to incorrect external sources. This is a serious threat to the quality of witness testimony and to the correctness of decisions reached by courts. However, few methods have been developed to reduce the vulnerability of witnesses to misinformation. This article presents such a method, namely, reinforced self-affirmation (RSA), which, by increasing memory confidence of witnesses, makes them less inclined to rely on external sources of information and more on their own memory. The effectiveness of this method was confirmed in three experiments. It was also found that memory confidence, but not general self-confidence, is a mediator of the impact of RSA on misinformation effect (ME), and that contingent self-esteem and feedback acceptance, but not sense of self-efficacy or general self-esteem, are moderators of this impact. It is concluded that RSA may be a promising basis for constructing methods, which can be used by forensic psychologists in real forensic settings.
... The misinformation effect is when memories are distorted or altered by presenting misleading information after an event [13][14][15]. For example, participants watch a video of an event followed by a statement of another witness including false information about the previously watched event [14,16]. ...
Full-text available
The aim of the present study was to examine how negative emotion and sex affect self-generated errors as in fabrication set-up and later false recognition of those errors. In total, 120 university students volunteered to take part in the study. Participants were assigned at random into two equal sized groups (N = 60) depending on the type of event they received (negative emotional or neutral). We expected that fabrication and false recognition would be enhanced for the emotional event compared to the neutral one. We further hypothesized that both the willingness to fabricate and later false recognition would be enhanced for women compared with men. The results partly confirmed the hypotheses. The results showed that emotional valence (negative) affects both the willingness to fabricate about events that never took place, and the recognition of the fabrication as true at a later point. Women and men were equally likely to fabricate but women were more likely to recognize their fabrication, particularly for the emotional event. The results are discussed in the context of prior work.
... The internal, cognitive, regulation of these social influences has recently been intensively studied to address forensic concerns about the malleability of witness memory under social pressure 30,31 . Various factors promoting or inhibiting conformity to other peoples' knowledge have consequently been examined, usually from the theoretical perspective that social influences cause malfunctions that corrupt and distort one's knowledge [32][33][34][35][36] . ...
Full-text available
If artificial intelligence (AI) is to help solve individual, societal and global problems, humans should neither underestimate nor overestimate its trustworthiness. Situated in-between these two extremes is an ideal ‘Goldilocks’ zone of credibility. But what will keep trust in this zone? We hypothesise that this role ultimately falls to the social cognition mechanisms which adaptively regulate conformity between humans. This novel hypothesis predicts that human-like functional biases in conformity should occur during interactions with AI. We examined multiple tests of this prediction using a collaborative remembering paradigm, where participants viewed household scenes for 30 s vs. 2 min, then saw 2-alternative forced-choice decisions about scene content originating either from AI- or human-sources. We manipulated the credibility of different sources (Experiment 1) and, from a single source, the estimated-likelihood (Experiment 2) and objective accuracy (Experiment 3) of specific decisions. As predicted, each manipulation produced functional biases for AI-sources mirroring those found for human-sources. Participants conformed more to higher credibility sources, and higher-likelihood or more objectively accurate decisions, becoming increasingly sensitive to source accuracy when their own capability was reduced. These findings support the hypothesised role of social cognition in regulating AI’s influence, raising important implications and new directions for research on human–AI interaction.
Full-text available
Like scientists, investigators and decision-makers in criminal cases both explain known evidence and use the resulting explanations to make novel predictions. Philosophers of science have made much of this distinction, arguing that hypotheses which lead to successful predictions are—all else being equal—epistemically superior to those that merely explain known data. Their ideas also offer important lessons for criminal evidence scholarship. This article distinguishes three values of prediction over explaining known facts in criminal cases. First, witnesses who predict are—all else being equal—more reliable than those who do not because they are less likely to be biased or lying. Second, investigators who only explain known facts run the risk of ‘fudging’ the scenarios that they formulate. Predictions can protect us against this danger. Third, carefully constructed predictions may help investigators to avoid confirmation bias. This article ends with a case study of the murder of Hae Min Lee.
Long before psychology, bias has existed in science. From the beginning, concerns have been raised about the reliability, validity, and accuracy of social science research (Meehl, 1954). In this chapter, we define and discuss the origins of bias and how it can erode the scientific method. We focus specifically on bias in psychological research, theory, assessment, and treatment. We discuss the range of common misconceptions and misinformation that permeates the female offender literature. Finally, we conclude with ten myths about female offenders and offer guidelines for identifying bias and how to avoid it.
Full-text available
Several studies have shown that executing a series of visually guided saccades immediately before taking a memory test can help people recall details of mock crimes. The nature of the memory test has varied across studies, reflecting the multiple forms of questioning in investigative interviewing. Here, we examined mnemonic effects of saccade execution when people answered two important question types: specific-open, which called for open-ended descriptions of multifaceted event elements (e.g., a person’s appearance), and specific-closed, which probed memory for isolated features (e.g., the color of a particular article of clothing). Saccade execution enhanced recall in response to open questions, with non-trivial effect sizes, but had no effect when questions were closed. As in some prior research, enhancement was moderated by handedness: only individuals who reported highly consistent hand preference benefited from saccade execution. Although inconsistent individuals did not benefit from saccade execution, they exhibited several natural memory advantages when tested with specific-open questions. The theory that saccade execution potentiates the contribution of top-down attention to memory retrieval is discussed. These findings provide new insights regarding the circumstances under which saccade execution might be used to enhance eyewitness recall in real-world settings.
Modern technology allows people to easily doctor their own photos (e.g., cropping out unwanted objects, applying color-changing filters). We examined whether editing and reviewing photos alters people’s memory of their experiences. In three experiments, participants photographed scenes, then edited their photos. In Experiment 1 (n = 54), editing photos by cropping out objects or applying a grayscale filter did not impact memory for the original experience. In Experiment 2 (n = 55), cropping objects from photos focused attention on the remaining objects, thereby increasing memory for the intact objects regardless of whether they were reviewed or not. Reviewing unedited photos where details were intact increased memory. In Experiment 3 (n = 39), applying a grayscale filter to photos did not impact memory for the color and content of the original scene, but reviewing photos improved recognition memory and memory for the scene’s color and content. Depending on the type of edit and cognitive demands, reviewing or editing photos can shape what is remembered.
While major depressive disorder has been associated with increased veridical memory for negative information, prior false memory literature has linked high depressive symptoms to increased false memory for negative information. We tested whether these contradictory findings may be due to semantic and emotional cohesion inflating false alarm rates in the most commonly used false memory paradigm – the Deese-Roediger-McDermott (DRM). In Experiment 1, healthy controls and depressive participants completed two emotional false memory tasks: the DRM, using semantically associated words, and an adaptation using orthographically associated words. Participants encoded words associated with neutral or negative critical lures. During subsequent recognition memory testing, errors were greater for negative than neutral semantic critical lures, but the opposite pattern was observed for orthographic critical lures. These findings support that orthographically associated stimuli provide a non-confounded method to test emotional false memory. In Experiment 2, we deployed the orthographic paradigm with simultaneous electroencephalography (EEG) and a state measure of negative mood (Positive and Negative Affective Scale) to better understand neurocognitive components of emotional false memory. Replicating Experiment 1, we found fewer false alarms for negative than neutral novel orthographically associated words. The EEG data showed significant differences in two memory-related event-related potentials: the familiarity-related N400, and the recollection-related late positive component (LPC). N400 amplitudes were larger for true versus false endorsements. For LPC, increasing depressive symptoms predicted greater amplitude differences for true versus false negative items and reduced differences for neutral items. However, increasingly negative mood predicted reduced amplitude differences for negative items, and greater amplitude differences for neutral items. Although depressive symptoms and negative mood state are highly correlated, they may inform differentially influence shifts in memory retrieval strategies, with depressive symptoms predicting reduced false alarms for negative information, and negative mood state predicting increased recollection errors for negative information.
Full-text available
Fact-checking and warnings of misinformation are increasingly salient and prevalent components of modern news media and political communications. While many warnings about political misinformation are valid and enable people to reject misleading information, the quality and validity of misinformation warnings can vary widely. Replicating and extending research from the fields of social cognition and forensic psychology, we find evidence that valid retrospective warnings of misleading news can help individuals discard erroneous information, although the corrections are weak. However, when informative news is wrongly labeled as inaccurate, these false warnings reduce the news’ credibility. Invalid misinformation warnings taint the truth, lead individuals to discard authentic information, and impede political memory. As only a few studies on the tainted truth effect exist, our research helps to illuminate the less explored dark side of misinformation warnings. Our findings suggest general warnings of misinformation should be avoided as indiscriminate use can reduce the credibility of valid news sources and lead individuals to discard useful information.
Full-text available
A central issue in the study of children’s long-term retention is an understanding of children’s susceptibility to memory failures. It has long been recognized that an important cause of memory failures is interference caused by new learning. Recently, interest in memory failures caused by subsequent learning has been revived in the context of studies on suggestibility and eyewitness memory. These studies have shown that, for subjects of all ages, exposure to misinformation (i.e., false information presented as truth) after viewing an event can lead to profound decrements in performance on later tests of memory for the originally seen event.
Full-text available
People viewed a security video and tried to identify the gunman from a photospread. The actual gunman was not in the photospread and all eyewitnesses made false identifications (n = 352). Following the identification, witnesses were given confirming feedback ("Good, you identified the actual suspect"), disconfirming feedback ("Actually, the suspect is number _"), or no feedback. The manipulations produced strong effects on the witnesses' retrospective reports of (a) their certainty, (b) the quality of view they had. (c) the clarity of their memory, (d) the speed with which they identified the person, and (e) several other measures. Eyewitnesses who were asked about their certainty prior to the feedback manipulation (Experiment 2) were less influenced, but large effects still emerged on some measures. The magnitude of the effect was as strong for those who denied that the feedback influenced them as it was for those who admitted to the influence.
Full-text available
Three experiments explored the effects of rehearsal and the passage of time on qualitative characteristics of memories for perceived and imagined complex events. Subjects thought or talked about events, focusing on either the perceptual (e.g., colors, sounds) or apperceptive (e.g., thoughts, feelings) aspects of the events (Experiment 1). Thinking about apperceptive aspects of events decreased the salience of context and sensory characteristics of memories and made memories for perceived and imagined events seem more similar in the subjective amounts of thoughts and feelings included in the memories. When the aspects of events subjects thought about were unspecified, thinking about events primarily affected rated clarity (Experiment 2). The clarity of imagined events was more affected than was the clarity of perceived events by whether the memories had been rated previously (Experiments 1 & 3). Over 24 hrs, clarity and sensory ratings decreased more for imagined than for perceived events (Experiment 3). Implications for reality monitoring (M. K. Johnson and C. L. Raye [see PA, Vol 65:6694]) are discussed. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Because of the biasing nature of retrieval tests, evidence that the introduction of misleading postevent information will impair the memory for an original event has recently been in dispute. In two experiments, a retrieval test sensitive to both biasing effects of misinformation (misinformation acceptance) and influences of the misinformation on memory (misinformation interference) was used. Both experiments demonstrated misinformation acceptance, and one of the experiments suggested that misinformation interferes with the ability to remember the original event. Two misinformation interference hypotheses are evaluated; they suggest that the misinformation may have either impaired memory or led to confusion regarding what had occurred during the event.
In two studies we examined the effect of questioner expertise on the error rates of subjects who were asked misleading versus unbiased questions. A total of 105 introductory psychology students watched a videotaped clip of a bank robbery and were then questioned about the crime. The questioner was represented to subjects as either highly knowledgeable or completely naive about the events the subject witnessed. One half of the subjects in each expertise condition were asked misleading questions, and the other half were asked unbiased questions. In the knowledgeable questioner conditions, misleading questions were associated with error rates significantly higher than those obtained with the unbiased questions (p < .05). In the naive questioner conditions, equivalent error rates for both types of questions were obtained (ns). These results indicate that misleading questions decrease witness accuracy when the questioner is assumed to be knowledgeable about the crime, but have no effect on accuracy when the questioner is assumed to be naive.
Three experiments explored the effects of rehearsal and the passage of time on qualitative characteristics of memories for perceived and imagined complex events. Subjects thought or talked about events, focusing on either the perceptual (e.g., colors, sounds) or apperceptive (e.g., thoughts, feelings) aspects of the events (Experiment 1). Thinking about apperceptive aspects of events decreased the salience of context and sensory characteristics of memories and made memories for perceived and imagined events seem more similar in the subjective amounts of thoughts and feelings included in the memories. When the aspects of events subjects thought about were unspecified, thinking about events primarily affected rated clarity (Experiment 2). The clarity of imagined events was more affected than was the clarity of perceived events by whether the memories had been rated previously (Experiments 1 & 3). Over 24 hr, clarity and sensory ratings decreased more for imagined than for perceived events (Experiment 3). Implications for reality monitoring (Johnson & Raye, 1981) are discussed.
A new approach to the study of memory has emerged recently, characterized by a preoccupation with natural settings and with the immediate applicability of research findings. In contrast, the laboratory study of memory relies on experimental techniques for theory testing and is concerned with the discovery of generalizable principles. Although both approaches share the goal of generalizability, they differ sharply in the evaluation of how that goal is best accomplished. In this article, we criticize the everyday memory approach, arguing that ecologically valid methods do not ensure generalizability of findings. We discuss studies high in ecological validity of method but low in generalizability, and others low in ecological validity of method but high in generalizability. We solidly endorse the latter approach, believing that an obsession with ecological validity of method can compromise genuine accomplishments.
Purpose. Two experiments were conducted to test the hypothesis that encountering or retrieving suggested information under conditions of limited, rather than full, attentional resources is likely to increase false memory for suggested events. Methods. A typical eyewitness suggestibility paradigm was employed in which participants viewed a slide sequence depicting an office theft, answered misleading questions regarding the theft, and were later tested on their memory for the source of the suggested details. In Expt 1, participants encountered the misinformation under conditions of either divided or full attention, and in Expt 2 participants were given either ample time to make the source judgment or were forced to provide source judgments very quickly. Results. The results of both experiments showed that participants who encountered (Expt 1) or retrieved (Expt 2) misleading suggestions under conditions of limited attentional resources were more likely to misattribute the suggested items to the slides and less likely to remember having encountered the suggestions in the post-event questions. Conclusions. The results support the hypothesis that limiting attentional resources impairs participants' ability to retrieve source-specifying information and increases false memory for suggested details.