West, R.L., & Stone, K.R. (2013). Age differences in eyewitness memory for a realistic event. Journals of Gerontology, Series B: Psychological Sciences and Social Sciences,
69(3), 338–347, doi:10.1093/geronb/gbt014. Advance Access publication March 26, 2013
Age Differences in Eyewitness Memory for a
Realistic Event
Robin L. West and Kevin R. Stone
Department of Psychology, University of Florida, Gainesville, Florida.
Objectives. To better understand the effects of misinformation on eyewitnesses of different ages, older and younger
adults experienced an event under intentional and incidental learning conditions in a naturalistic experiment using multiple memory tests.
Method. Following exposure to the event, which was a brief interruption of a group testing session, participants
completed several memory tests. For half of the participants, misinformation was embedded in the first cued recall test.
On subsequent free recall and cued recall tests, basic scores and misinformation-based memory errors were examined.
Results. As expected, younger adults had higher recall scores than older adults. Older and younger adults made the
same number of misinformation errors in free recall and in cued recall with intentional learning. However, in the incidental condition, younger adults made more misinformation errors likely due to the information processing strategies they
employed after incidental learning.
Discussion. Misinformation effects were quite strong, even with a realistic scene and intentional learning. Older adult
suggestibility was no worse than that of younger adults. When misinformation was combined with incidental learning,
younger adults may have used strategic processing to encode misinformation to their detriment.
Key Words: Aging—Eyewitness—Incidental learning—Memory.
P
eople of all ages are involved in the criminal justice
system as witnesses. Thus, it is important to understand
how witness skills and limitations may be affected by age.
Past research clearly shows that older adults are poorer at
lineup identification than young people, showing a tendency toward more false alarms (Memon, Gabbert, & Hope,
2004). Extensive work has also shown that older adult witnesses are viewed as less believable than younger witnesses
(Allison, Brimacombe, Hunter, & Kadlec, 2006). But the
research on aging is mixed with respect to the impact of
misinformation at different ages (cf. Dodson & Krueger,
2006; Roediger & Geraci, 2007). It “remains unclear if
older witnesses are generally more vulnerable to suggestions than young witnesses” (Mueller-Johnson, 2009,
p. 163). The purpose of this research was to address some
critical gaps in the eyewitness aging literature, by examining misinformation under incidental and intentional encoding conditions, using a naturalistic experiment.
Aging and Misinformation
It is well established that the accuracy of memory
can be significantly compromised by misleading postevent suggestion (Loftus & Palmer, 1974; Loftus, Miller,
& Burns, 1978). When this happens, participants recall
or recognize details they never saw (the misinformation
effect). Brief descriptions of false details or imagination
of fictional events are sufficient to cause later memory distortion (Nash, Wade, & Lindsay, 2009). This is likely due
to weak initial encoding (Loftus, Levidow, & Duensing,
338
1992), reconstruction of the memory trace to include new
information (Reyna & Brainerd, 1995), and/or failure to
encode or retain the correct source of the information
(Mitchell, Johnson, & Mather, 2003). All of these hypothesized influences have some support in the literature
(Loftus, 2005).
Regardless of the reason, the impact of misinformation
represents a powerful false memory effect that occurs
across all ages. Traditional laboratory studies of false
memory with word lists clearly show age-related declines,
with more false memory errors on the part of older
participants (Jacoby, 1999), and source memory tests reveal
age declines as well (Johnson, Hashtroudi, & Lindsay,
1993; Dodson, Bawa, & Slotnick, 2007; Vakil, Hornik,
& Levy, 2008), but such studies are not focused on the
more complex task of recalling a witnessed scene. Using
more complex scenes, aging is sometimes associated with
significantly stronger misinformation effects (cf. Bulevich
& Thomas, 2012; Cohen & Faulkner, 1989; Loftus et al.,
1992; Schacter, Koutstaal, Johnson, Gross, & Angell,
1997). Older adults have sometimes been more suggestible
than younger adults, and confident of their false memories,
with both short and long delays between the event and the
critical misinformation test (Karpel, Hoyer, & Toglia, 2001;
Mueller-Johnson & Ceci, 2004; Roediger & Geraci, 2007).
Mitchell and colleagues (2003) specifically argued that
these effects were related to source memory difficulties;
older adults, more so than young adults, indicated that they
“saw” things that were merely “suggested.”
© The Author 2013. Published by Oxford University Press on behalf of The Gerontological Society of America.
All rights reserved. For permissions, please e-mail: [email protected].
Received August 13, 2012; Accepted February 4, 2013
Decision Editor: Robert West, PhD
AGE AND EYEWITNESS MEMORY
Other scholars have not reported age differences in suggestibility after planting misinformation (Bornstein, Witt,
Cherry, & Greene, 2000; Chan, Thomas, & Bulevich,
2009; Coxon & Valentine, 1997; Dodson & Krueger, 2006;
Searcy, Bartlett, & Memon, 2000). In a study by Gabbert,
Memon, and Allan (2003), a significant proportion of both
old and young participants reported seeing things that they
had only heard about in a postvideo discussion session, but
older adults were no more prone to such memory errors than
younger adults (Gabbert et al., 2003). In other research using
a legal testimony format, when witnesses received a direct
examination (concerning the main details seen in a video)
and a cross-examination (designed to get them to endorse
misinformation), there was no interaction of age with type
of examination—all age groups were less accurate on crossexamination (Brimacombe, Jung, Garrioch, & Allison,
2003). In some cases, younger adults have even reported
more misinformation than older adults (Gabbert, Memon,
Allan, & Wright, 2004; Marche, Jordan, & Owre, 2002).
One possible explanation for this variability across studies is that researchers have used a number of different methodologies to examine misinformation effects—incidental or
intentional memory during initial exposure to a scene, different memory assessments (recognition, free recall, cued
recall), and variations in misinformation presentation, including participant discussion (Gabbert et al., 2003), cued recall
(Coxon & Valentine, 1997), and written narratives (Cohen &
Faulkner, 1989). Few researchers have systematically investigated these methodological differences. In one exception,
Gabbert and colleagues (2004) compared misinformation
presented in discussion with misinformation presented in a
written narrative. They found that the discussion format led
to more memory errors for both older and younger adults, but
no age interaction occurred (Gabbert et al., 2004). Systematic
examination of methodological variations is essential to discover the factors that predict suggestibility across age. Here,
we compared incidental and intentional conditions for initial
exposure to the to-be-remembered event and assessed misinformation errors in free recall and cued recall.
The aging literature indicates that learning conditions often
make a difference (Hess & Ennis, 2012). Younger adults
typically have an advantage with intentional learning (Block,
2009), due to their greater ability to activate and benefit
from skilled memory strategies (Hertzog & Hultsch, 2000),
including successful use of perceptual details (Koutstaal,
2003). In contrast, age differences can be reduced with incidental learning, depending on the encoding and retrieval context (Crook, Larrabee, & Youngjohn, 1993; Naveh-Benjamin
et al., 2009). Nevertheless, no one to our knowledge has
directly compared incidental and intentional encoding for a
witnessed event in an aging study. Scholars have described
intentional or incidental learning conditions or have simply
instructed participants to watch a video (cf. Chan et al., 2009;
Gabbert et al., 2003; Karpel et al., 2001; Mueller-Johnson &
Ceci, 2004; Searcy, Bartlett, & Memon, 1999).
339
It is not clear, with older adults, whether misinformation will vary as a function of encoding conditions. In a
typical paradigm, the witness views a brief scene and then
reads faulty information from a narrative or questionnaire.
To be affected by misinformation, the witness must encode
and retrieve briefly presented misinformation. Given older
adults’ memory deficits, one could argue that they might
have difficulty encoding details from the to-be-remembered
event (and would thus be susceptible to misinformation)
or they might have difficulty encoding the misinformation
(which would then have a weaker impact). At the same
time, we know that older adults make errors in identifying sources, and many scholars have suggested that source
memory is the basis for the misinformation effect (Mitchell
et al., 2003; Roediger & Geraci, 2007). However, there is
also evidence that adult misinformation errors are reduced
when attention is focused on the memory task (Roediger &
Geraci, 2007), which might occur with intentional learning.
Thus, it is important to explicitly test the extent to which
older and younger adults are susceptible to misinformation
errors under different learning conditions.
Present Study
To our knowledge, no studies of aging and eyewitness
memory have compared intentional and incidental learning.
To make a believable scenario for incidental learning, we
employed a realistic stimulus—early in the testing session,
a research assistant interrupted the experimenter to discuss
a problem. Half of the participants experienced this event
after instructions to remember it and half observed it without knowing about the memory test. Free recall, cued recall,
and face recognition tests were administered, with confidence assessed for the last two measures. For half of the
participants, misinformation was embedded in the first cued
recall test. Scores on a free recall test and a second cued
recall test revealed the suggestibility of participants.
Based on past research, we expected misinformed participants to have lower overall memory scores and to make
more misinformation-based errors (on free recall and the
second cued recall test) than those not receiving misinformation. Due to weaker encoding of the initial event, we
expected the misinformation effect to be more prominent
in the incidental condition than the intentional condition.
Given age-related memory deficits, we expected that the
younger adults would have higher scores on all memory tests.
It was also expected that intentional memory conditions
would result in higher recall than incidental memory
conditions; however, age differences were expected to be
reduced with incidental learning, as is often the case. With
respect to aging and misinformation, no predictions were
made. Source memory deficits might increase age effects.
On the other hand, as is typical in eyewitness research,
the misinformation was not strongly emphasized. If older
adults do not retain this misinformation, it might have little
or no impact on their memory.
340
WEST AND STONE
Method
Participants
Data were collected from 99 undergraduates and 77 community-dwelling older adults (M = 72.30, SD = 5.17). Older
adults (M =16.4 years, SD =2.8) had more education than
younger adults
(M = 13.0, SD = 1.1), F(1,173) = 118.92,
2
p < .001, ηp = .41. Both age groups reported above average self-rated health (10 = Excellent Health; 0 = Very Poor
Health; younger: M = 8.77, SD = 1.10; older: M = 8.29,
SD = 1.81). Participants were excluded for potential cognitive problems (anticholinergic medications, dementia
diagnosis, stroke, or difficulty following instructions) or
significant hearing or vision problems; only one younger
person was excluded (for a significant hearing deficit).
Procedures and Materials
To begin the testing session, all participants, regardless of encoding condition, were given a story to read and
remember, as a filler task. Half of the participants (intentional condition only) were then warned about an upcoming interruption and were told to remember that event. After
the event, a demographic questionnaire (questions about
age, gender, race, health, and education) was followed by
administration of memory tests. Finally, health questions
and control questions were presented (see Figure 1).
All instructions were provided in writing. Participants
completed memory tests and questionnaires at their own
pace. Each session lasted 20–30 min. Participants were
tested in groups of 1–7 people (younger: M = 4.10, SD =
1.40; older: M = 4.77 SD = 1.36); participants in each group
were of same age and were all assigned to the same condition for encoding and misinformation (assigned in advance,
using counterbalancing).
Filler task.—For the filler task, all participants read a story
of 24 sentences, written at a seventh-grade reading level
(Dixon, Hultsch, & Hertzog, 1989). All participants were
asked to study and remember the story, but no such memory
test was administered. Participants spent approximately 20 s
reading the filler story before the interruption and then spent
another 20 s reading after the confederate departed.
To-be-remembered event.—Shortly after the filler task
story was distributed, a research confederate knocked on the
door and entered the testing room. The experimenter and
research confederate engaged in a 1-min scripted conversation developed for this experiment. In general, the interrupter
reported that she was on the phone conducting an interview
and the respondent was having difficulty. She wanted to
know what to do. The experimenter offered two possible
solutions to the problem, and then the interrupter left.
Three young women with similar appearance served as
interrupters. All wore the same color and type of clothing.
Counterbalancing of confederate assignment to conditions
ensured that each one had an equal chance of serving as
interrupter in each condition, and confederates were blind
to condition. Seats were arranged in the testing room so that
all participants had a clear view of the door (no more than
6 feet away) and the interrupter, who stepped fully into the
room. Clear speech techniques were implemented to ensure
that the event was heard and understood; specifically, the
interrupter and experimenter enunciated each word clearly,
especially consonants, paused before key phrases, and
spoke at a slightly lower pitch (see Kricos, 2003).
Cued recall (CR1 and CR2).—In the two cued recall tests
(11 items each, with new items for each test), participants
were asked questions about the interruption (e.g., After the
experimenter greeted the interrupter, what problem did
she report? What was the name of the interrupter?), and
these questions were presented in the same order that the
details occurred in the event. The participants assigned to
misinformation had a false detail embedded into 5 of the
11 questions on the first cued recall test (CR1), following
procedures used by Zaragoza and Lane (1994). All participants received the same 11 new questions on the second
cued recall test (CR2), regardless of condition assignment;
5 of these 11 items were critical target items asking about
details related to the misinformation presented in CR1. For
CR1 and CR2, the total number of correct answers was the
primary dependent measure. Analyses of the five critical
items were conducted separately to examine errors based
on misinformation.
For example, a question in CR1 for participants who
were not misinformed was When the interrupter opened
the door, did she appear anxious? That same question was
reworded for the misinformation group: When the interrupter opened the door and looked at her watch, did she
appear anxious? (This information was false because the
interrupter wore no watch.) All participants later received
this item on CR2: Was the interrupter wearing a watch? If a
participant answered yes, it was scored as a misinformation
error, regardless of condition assignment (participants with
no misinformation sometimes guessed incorrectly).
Free recall.—Free recall was tested by asking participants
to write a detailed description of the interruption, with no
time limit. Recall protocols were transcribed and compared
to a template. To develop this template, the event script
was used to generate details about the conversation. In
addition, the to-be-remembered details were recorded by an
independent group of 22 coders who described the event in
writing as the scripted event occurred. The event was acted
out three times for the coders (using all three interrupters),
and a new list of details was generated by each coder for
each re-enactment to create a complete list of all possible
to-be-remembered details about the event (details about
AGE AND EYEWITNESS MEMORY
341
Figure 1. Participant flow chart showing the order of procedures used in this research.
the conversation, the appearance of the interrupter, etc.).
Each correctly reported detail, out of a possible score of
39 details, was assigned one point. For example, for the
recalled statement, The interrupter knocked twice on the
door and then opened it, the participant received three
points; one point each for knocked on door, twice, and
opened it (following the methods of Sutherland & Hayne,
2001). Text analysis software (SPSS Text Analysis for
Surveys 2.0) was utilized to compare the participants’
recalled text to the template developed by coders. The
number of details correctly recalled was the primary
dependent measure. Intrusions based on misinformation
were also recorded.
Control questions.—All participants were asked whether
they (a) were expecting a memory test on the interruption,
(b) were trying to remember the interruption, (c) noticed any
incorrect information while answering questions, (d) could
see the interruption clearly, and (e) could hear the interruption clearly. Participants responded on a 7-point scale from
1 = No, not at all to 7 = Yes, definitely for the first three
items and from 1 = Could not hear/see at all to 7 = Could
hear/see clearly for the last two items. Participants were
also asked to report the number of seconds they spent paying attention to the event.
Results
To determine whether group size had an effect on
memory performance, an Encoding condition (incidental or intentional memory) × Misinformation (misleading
or no misleading information) × Group size analysis of
variance (ANOVA) was examined for all memory tests
with group size entered as a random effect. This preliminary analysis showed no significant main effects or
342
WEST AND STONE
interactions—group size did not affect free recall or cued
recall. Preliminary analyses also indicated that there were
no age differences for reported ability to clearly see or
hear the interruption or for time spent attending to the
interruption. Results for confidence and face recognition
measures are not reported here and are available from
the authors. Primary hypotheses were tested using Age
(younger, older) × Encoding condition (incidental or
intentional memory) × Misinformation (misleading or no
misleading information) analyses, with multivariate tests
conducted where appropriate.
Cued Recall
Age, encoding, and misinformation effects were examined in a multivariate analysis of variance using scores
from CR1 and CR2. This multivariate analysis allowed
us to examine main effects across both tests, and univariate follow-up testing revealed the results for each individual test. There was a significant multivariate effect
of
2
encoding condition, F(2,167) = 24.07, p < .001, ηp = .22,
with participants in the intentional condition scoring significantly higher than those in the incidental condition,
as expected. There was a significant multivariate effect
of age, as older adults scored 2lower than younger adults,
F(2,167) = 22.8, p < .001, ηp = .22, and a main
effect
2
of misinformation, F(2,167) = 7.75, p = .01, ηp = .09.
Participants who received misinformation scored significantly lower overall than those who did not receive
misinformation.
In univariate follow-up tests, the effects of encoding
condition and age were significant on both CR1 and CR2
(see Table 1), as expected. The misinformation groups
showed no difference, as expected, for CR1 (p > .25), but
the misinformed group performed worse than the nonmisinformed
group on CR2, F(1,168) = 14.4, p < .001,
2
h p = .08. Univariate tests also revealed an Encoding condition × Misinformation interaction
significant for CR2 only,
2
F(1,168) = 3.70, p = .05, ηp = .02; the intentional notmisinformed participants performed significantly higher on
CR2 than the intentional misinformed group. There were no
other significant interactions.
Free Recall
A three-way ANOVA examined the effects of encoding
condition, misinformation, and age on the number of correct details recalled. Encoding condition had 2a significant
effect on recall, F(1,176) = 23.15, p < .001, ηp = .12, with
intentional memory better than incidental memory (see
Table 2). Younger adults provided significantly more2 details
than older adults, F(1,176) = 32.71, p < .001, ηp = .16.
The presence of misinformation had no significant effect
on recall, F(1,176) = 2.08, p > .10. No interactions were
significant.
Table 1. Cued Recall Scores as a Function of Experimental
Condition and Age CR1
Group
Older adults
Intentional
Not misinformed
Misinformed
Incidental
Not misinformed
Misinformed
Younger adults
Intentional
Not misinformed
Misinformed
Incidental
Not misinformed
Misinformed
All adults
Intentional
Not misinformed
Misinformed
Incidental
Not misinformed
Misinformed
CR2
M
SD
M
SD
5.78
6.00
5.58
4.31
4.60
4.00
2.18
2.03
2.34
1.73
1.60
1.86
5.42
6.32
4.53
3.85
4.15
3.53
2.01
2.16
1.39
1.62
1.69
1.62
6.69
6.82
6.50
5.41
5.23
5.60
1.68
1.59
1.82
1.70
1.63
1.78
6.98
7.29
6.55
5.47
5.62
5.32
1.39
1.38
1.32
1.54
1.70
1.38
6.28
6.47
6.05
4.93
4.96
4.91
1.95
1.80
2.11
1.79
1.63
1.96
6.29
6.89
5.56
4.77
4.98
4.55
1.85
1.78
1.68
1.76
1.83
1.68
Notes. The maximum score on each cued recall test was 11. The table
provides accuracy (number correct) scores in relation to age and condition for
each test across all 11 items (CR1 = the first cued recall test; CR2 = the second
cued recall test).
Misinformation-Based Errors
Cued recall.—We examined the five critical items on
CR2 to see whether specific misleading details had been
provided by the misinformed participants. Compared with
the not-misinformed group, those who were misinformed
responded incorrectly more often to 2the critical items (see
Table 3), F(1,175) = 58.91, p < .001, ηp = .26. Younger adults
made more specific misinformation
errors than older adults,
2
F(1,175) = 12.70, p < .001, ηp = .07, which was qualified
by a significant age by learning condition interaction.
Younger adults made more misinformation errors than
older adults 2in the incidental condition, F(1,168) = 15.40,
p < .001, ηp = .08, but both age groups made an equal
number of misinformation errors with intentional encoding
(see Table 3). As expected, the lowest accuracy scores on the
five critical items were obtained by the misinformed older
adults in the incidental condition (M = 0.79, SD = 0.22), and
the highest accuracy scores were obtained by younger adults
in the intentional condition who were not misinformed
(M = 2.75, SD = 0.18). Out of concern for the possibility that
older adults were simply saying I don’t know when unsure,
while younger adults were providing a response, we also
examined an output-bound measure (Pansky, Goldsmith,
Koriat, & Pearlman-Avnion, 2009). Using an output-bound
measure—number of specific misinformation errors/number
of specific responses given (not including don’t know or
blank responses)—younger adults still made significantly
343
AGE AND EYEWITNESS MEMORY
Table 2. Free Recall Scores as a Function of Experimental Condition
and Age Table 3. Misinformation Errors on Cued Recall 2 as a Function of
Experimental Condition and Age Misinformation errors
Free recall
Group
Older adults
Intentional
Not misinformed
Misinformed
Incidental
Not misinformed
Misinformed
Younger adults
Intentional
Not misinformed
Misinformed
Incidental
Not misinformed
Misinformed
All adults
Intentional
Not misinformed
Misinformed
Incidental
Not misinformed
Misinformed
M
SD
7.05
7.26
6.84
4.21
3.55
4.89
3.57
3.48
3.75
3.07
2.70
3.35
9.46
9.32
9.65
7.49
6.69
8.32
3.16
3.31
3.01
3.34
3.06
3.49
8.40
8.49
8.28
6.07
5.33
6.84
3.54
3.50
3.63
3.60
3.28
3.80
Group
Older adults
Intentional
Not misinformed
Misinformed
Incidental
Not misinformed
Misinformed
Younger adults
Intentional
Not Misinformed
Misinformed
Incidental
Not misinformed
Misinformed
All adults
Intentional
Not misinformed
Misinformed
Incidental
Not misinformed
Misinformed
M
SD
1.58
0.79
2.37
1.49
1.00
2.00
1.13
0.79
0.83
1.45
1.03
1.67
1.75
1.00
2.80
2.55
1.78
3.36
1.52
1.22
1.28
1.68
1.31
1.66
1.67
0.91
2.59
2.09
1.43
2.77
1.36
1.06
1.09
1.66
1.24
1.78
Note. The table indicates the number of objective details recalled in relation
to age and condition.
Note. The table indicates the number of misinformation details incorrectly
produced on the second cued recall test as responses to six critical cued recall items.
more errors in the incidental condition (M = 0.486) than
older adults in that condition (M = 0.396).
p < .001, ηp = .08. As expected, those in the intentional
condition noticed the misinformation more than those
in the
2
incidental condition, F(1,168) = 6.40, p < .01, ηp = .04. In
addition, 65% of the misinformed younger adults compared
with 37% of misinformed older
adults reported noticing it,
2
F(1,168) = 11.80, p < .001, ηp = .07.
A two-way ANOVA using these two independent variables (a) expectation of a memory test and (b) detection
of false information, run only on those participants who
were misinformed, revealed that expectation of a memory
test and detection of false information did not significantly
predict the number of misinformation-based errors. That is,
those who noticed false information were just as likely to
report it as those who did not notice it, and those who were
expecting a memory test were also just as likely to report
misinformation as those who were not expecting a memory
test. These results did not vary with age.
Free recall.—The number of misinformation-based
errors made during free recall was relatively low (0–3)
and was not normally distributed, so Kruskall–Wallis tests
were employed. The misinformation group made significantly more errors (M = 0.70, SD = 0.92) than those who
did not receive misinformation (M = 0.23, SD = 0.55), χ2 (1,
N = 176) = 16.68, p < .001. The presence of misinformationbased errors in free recall did not vary as a function of age
group or encoding condition. Potential interactions of age,
encoding, and misinformation were also examined with
Kruskall–Wallis tests, and none were found to be significant.
Control questions.—Further analyses examined the
misinformation effect in relation to two questions: (a) Were
you expecting a memory test on the interruption? and (b)
Did you notice any incorrect information…? Those who
expected a memory test or noticed the misinformation
were defined as having a score of 5 or higher (with 7 being
Yes, a lot). Only 14% of the participants in the incidental
condition expected a memory test compared with 78% of
participants in
the intentional condition, F(1,168) = 164.80,
2
p < .001, ηp = .50. About half of the participants (52%)
who received misinformation reported noticing some
incorrect information compared with 29% of participants
who did not receive misinformation, F(1,168) = 15.17,
2
Discussion
There is a large literature on aging and recognition for
faces in lineups (Goodsell, Neuschatz, & Gronlund, 2009)
and on the credibility of older adult witnesses (Kwong See,
Hoffman, & Wood, 2001), but research on aging and suggestibility to misinformation shows mixed findings. This
study filled gaps in the literature by examining two methodological factors that might shed light on past research:
learning condition (intentional and incidental learning)
and type of misinformation test (free and cued recall). The
basic findings of this research were consistent with previous
344
WEST AND STONE
evidence: participants in the intentional encoding condition
had higher scores on most memory measures than those
in the incidental condition; participants who were misinformed made more memory errors than those who were not
misinformed; and older adults recalled less than younger
adults. These are all well-established findings from past
research.
Going beyond past research, this study provided some
of the first evidence on misinformation effects in relation
to different encoding conditions. For cued recall, overall
score varied with encoding condition but number of misinformation errors did not. Misinformed participants were
successfully misled, regardless of encoding condition.
These results emphasize the power of the misinformation
effect, validating past research with videos using a similar paradigm, where false information was presented in a
cued recall context (Brimacombe et al., 2003; Coxon &
Valentine, 1997; Sutherland & Hayne, 2001; Zaragoza &
Lane, 1994). Additionally, younger adults in the incidental condition reported more misinformation than younger
adults in the intentional condition, providing some support
for the hypothesis that misinformation could be more problematic in incidental conditions. On free recall, however,
the number of remembered details did not vary as a function
of misinformation and few incorrect details were reported
overall, suggesting some resistance to misinformation for
that testing format. Given that the free recall test occurred
after the critical target items were presented on CR2, it
is not surprising that some nonmisinformed individuals
included erroneous details; these participants may have
thought, for example—they asked me about a watch, so
maybe I should mention a watch. Considering these points
together, police interviewers may want to assess free recall
first to avoid the potential for planting misinformation
during direct questioning. When later questioning begins,
interviewers should be careful to avoid misinformation,
especially when a scene was initially viewed without any
intent to remember (Gabbert, Hope, Fisher, & Jamieson,
2011). Useful aids may be the Cognitive Interview, which
uses open-ended items that encourage free recall (cf. Mello
& Fisher, 1996; Searcy, Bartlett, Memon, & Swanson,
2001), or the Self-Administered Interview, a standardized
protocol that gathers detailed witness accounts before that
information becomes lost or distorted (Gabbert, Hope, &
Fisher, 2009).
Aging and Eyewitnessing
Consistent with previous research (Backman, Small,
& Wahlin, 2001), older adults scored lower than younger
adults on all tasks, under incidental and intentional conditions, providing fewer details on free and cued recall tests.
Our key question was whether there would be age differences in the number of misinformation errors under different learning conditions and with different test formats.
Some studies have reported that older adults are more prone
to misinformation (Cohen & Faulkner, 1989; Karpel et al.,
2001; Loftus et al., 1992; Mueller-Johnson & Ceci, 2007;
Roediger & Geraci, 2007), whereas others found no age
increase in the misinformation effect or mixed results for
different measures (Bornstein et al., 2000; Brimacombe
et al., 2003; Bulevich & Thomas, 2012; Coxon & Valentine,
1997; Dodson & Krueger, 2006; Gabbert et al., 2003, 2004;
Marche et al., 2002; Schacter et al., 1997; Searcy et al.,
2000). Our results—compared with older adults, younger
adults were just as susceptible to misinformation on free
recall and cued recall—were consistent with much of this
past research.
It has been suggested that older adults are prone to
misinformation because they recall the main event more
poorly (Mueller-Johnson & Ceci, 2004). In this case, older
adults were clearly less accurate overall in free and cued
recall but did not make more misinformation errors. That
finding suggests that poor recall of the initial event cannot
be the only factor influencing misinformation errors. It
is also unlikely that our findings were due to greater
suggestibility for the young, given the extensive evidence
for greater susceptibility to false memory among older
adults (Backman et al., 2001). Instead, our results may be
related to the production of specific misinformation details.
Older adults may be less prone to spontaneously report false
information on a cued recall or free recall test if they do not
encode or retrieve the misinformation and younger adults,
causing its impact to be lessened (Roediger & Geraci,
2007). In fact, these older adults were much less likely
than the younger adults to even notice the misinformation,
suggesting that they were not fully processing the details
on CR1.
In contrast, the younger adults may have paid special
attention to the CR1 questionnaire for strategic reasons,
treating it as a postevent review and hoping to pick up
details to aid their memory (see Chan & Langley, 2010).
Most misinformation studies showing equal or greater
suggestibility for young adults were studies in which misinformation was embedded in a cued recall or narrative
format, and the subsequent questions were cued recall
items requiring participants to report remembered details
(Chan et al., 2009; Coxon & Valentine, 1997; Gabbert
et al., 2004; Marche et al., 2002; Mitchell et al., 2003),
rather than a forced choice testing format where general familiarity or gist-based responses might lead older
adults to respond yes to misinformation (cf. Karpel et al.,
2001; Mitchell et al., 2003; Schacter et al., 1997). In situations such as this experiment, where initial exposure to
the event was relatively brief and specific responses were
required on a cued recall test, younger adults (especially
after incidental encoding) might have searched through
any available information to bolster a weak memory trace.
Ordinarily, a single postevent review would aid recall
for younger adults more than older adults (Koutstaal,
AGE AND EYEWITNESS MEMORY
Schacter, Johnson, Angell, & Gross, 1998) because older
adults have difficulty encoding and using details related
to the context of to-be-remembered information. In this
case, however, when the details encoded in their review
contained misinformation, younger adults were misled.
Future studies could use a think-aloud protocol to determine whether particular participants focus carefully on
postevent narratives or test questions to identify informative details about a to-be-remembered event. Participants
who act strategically in this way may be more likely to
report the specific details that they noticed in those misleading narratives or questions. Suggestibility may then
be influenced by the strength of the memory traces for the
misinformation as much, or more so, than memory traces
for the initial to-be-remembered event (also see Bulevich
& Thomas, 2012; Roediger & Geraci, 2007).
Limitations
Using a live event provided a unique perspective for
examining age differences in the misinformation effect.
Although the use of a live event served to make the experience more familiar to participants and more externally
valid than most previous research, naturalistic experiments have the potential to reduce internal reliability.
Precautions were taken to ensure that this was not an
issue. All participants viewed the same event (e.g., the
interruption script was rehearsed extensively, condition assignments were counterbalanced and preassigned,
and the interrupter was blind to experimental condition).
Nevertheless, it is possible that an unconscious bias on
the part of the experimenter (who was necessarily aware
of condition assignment) affected how the event unfolded.
Furthermore, participants in the incidental condition
were not instructed to attend to the event (similar to the
approach of Loftus et al., 1992), which may have disadvantaged that group more than in other studies of incidental memory where researchers directed participants to
notice some specific aspect of a to-be-remembered video
(Brimacombe et al., 2003; Schacter et al., 1997).
Conclusions
Consistent with previous research (Loftus & Palmer,
1974; Sutherland & Hayne, 2001; Zaragoza & Lane,
1994), the current study has demonstrated how easily
memory can be distorted. By simply suggesting that a
false detail was present, participants believed they heard
and saw details that did not exist. That result is not new.
However, this is one of the first studies to examine the
impact of misinformation under both intentional and
incidental learning conditions (see Davies & Hine, 2007;
Yarmey, 2004). Even when participants were paying attention to the event and trying to remember everything about
345
it, they were still susceptible to believing false information. These results emphasize the power of misinformation to distort memory at all ages, even with a realistic
stimulus. At the same time, free recall of basic details was
not compromised by encoding condition. These data suggest that a free recall format or special interview process
before detailed question-answering (with no opportunity
for misinformation to be planted in people’s minds) might
yield valuable information from witnesses.
With respect to aging, these results confirmed past
research showing that older adults have more memory difficulties than young adults. At the same time, this research
demonstrated that misinformation has a strong impact on
memory for all adults, even under optimal encoding conditions. In fact, younger adult misinformation errors were
comparable to older adult errors on free recall, and younger
adults in the incidental condition showed a stronger misinformation effect on cued recall than the older adults. These
findings are likely related to information processing skills
in younger adults—either their ease of encoding the misinformation planted in the first cued recall questionnaire
or strategic efforts they made to learn that information to
aid subsequent retrieval. Further research in this area will
help psychologists learn more about the processing of real
world events, with the long-term goal of identifying ways
to improve the accuracy of everyday memory for younger
and older adults alike.
Acknowledgments
The authors would like to thank Erin Hastings and the research
assistants in the Adult Development and Aging Laboratory, especially
Nicole Barroso, Margarita Corredor, Viktoria Sisto, Michelle Melanson,
and Tamara Murray.
Correspondence
Correspondence should be addressed to Robin L. West, PhD, Department
of Psychology, University of Florida, P. O. Box 112250, Gainesville, FL
32611. E-mail: [email protected].
References
Allison, M., Brimacombe, C. A. E., Hunter, M. A., & Kadlec, H. (2006).
Young and older adult eyewitnesses’ use of narrative features
in testimony. Discourse Processes, 41, 289–314. doi:10.1207/
s15326950dp4103_3
Backman, L., Small, B. J., & Wahlin, A. (2001). Aging and memory:
Cognitive and biological perspectives. In J. Birren & K. W. Schaie
(Eds.), Handbook of the psychology of aging (5th ed., pp. 349–377).
San Diego: Academic Press.
Block, R. A. (2009). Intent to remember briefly presented human faces
and other pictorial stimuli enhances recognition memory. Memory &
Cognition, 37, 667–678. doi:10.3758/MC.37.5.667
Bornstein, B. H., Witt, C. J., Cherry, K. E., & Greene, E. (2000). The suggestibility of older witnesses. In M. B. Rothman, B. D. Dunlop, & P.
Entzel (Eds.), Elders, crime, and the criminal justice system: Myth,
perception and reality in the 21st century (pp. 149–161). New York:
Springer.
Brimacombe, C. A., Jung, S., Garrioch, L., & Allison, M. (2003).
Perceptions of older adult eyewitnesses: Will you believe me
when I’m 64? Law and Human Behavior, 27, 507–522. doi:10.102
3/A:1025486006769
346
WEST AND STONE
Bulevich, J. B., & Thomas, A. K. (2012). Retrieval effort improves memory
and metamemory in the face of misinformation. Journal of Memory
and Language, 67, 45–58. doi:10.1016/j.jml.2011.12.012
Chan, J. C., Thomas, A. K., & Bulevich, J. B. (2009). Recalling a witnessed event increases eyewitness suggestibility: The reversed testing
effect. Psychological Science, 20, 66–73. doi:10.1111/j.1467-9280.
2008.02245.x
Chan, J. C. K., & Langley, M. M. (2010). Paradoxical effects of testing:
Retrieval enhances both accurate recall and suggestibility in eyewitnesses. Journal of Experimental Psychology: Learning, Memory, and
Cognition, 38, 248–255. doi:10.1037/a0021204
Cohen, G., & Faulkner, D. (1989). Age differences in source forgetting:
Effects on reality monitoring and on eyewitness testimony. Psychology
and Aging, 4, 10–17. doi:10.1037/0882-7974.4.1.10
Coxon, P., & Valentine, T. (1997). The effects of the age of eyewitnesses on
the accuracy and suggestibility of their testimony. Applied Cognitive
Psychology, 11, 415–430. doi:10.1002/(SICI)1099-0720(199710)11:
5<415::AID-ACP469>3.0.CO;2-A
Crook, T. H., 3rd, Larrabee, G. J., & Youngjohn, J. R. (1993). Age and
incidental recall for a simulated everyday memory task. Journal of
Gerontology, 48, P45–P47.
Davies, G., & Hine, S. (2007). Change blindness and eyewitness testimony. The Journal of Psychology, 141, 423–434. doi:10.3200/
JRLP.141.4.423-434
Dixon, R. A., Hultsch, D. F., & Hertzog, C. (1989). A manual of twentyfive three-tiered structurally equivalent texts for use in aging research.
Collaborative Research Group on Cognitive Aging Technical Report
No. 2. University of Victoria and Georgia Institute of Technology.
Dodson, C. S., Bawa, S., & Slotnick, S. D. (2007). Aging, source
memory, and misrecollections. Journal of Experimental
Psychology. Learning, Memory, and Cognition, 33, 169–181.
doi:10.1037/0278-7393.33.1.169
Dodson, C. S., & Krueger, L. E. (2006). I misremember it well: Why older
adults are unreliable eyewitnesses. Psychonomic Bulletin & Review,
13, 770–775. doi:10.3758/BF03193995
Gabbert, F., Hope, L., & Fisher, R. P. (2009). Protecting eyewitness evidence:
Examining the efficacy of a self-administered interview tool. Law
and Human Behavior, 33, 298–307. doi:10.1007/s10979-008-9146-8
Gabbert, F., Hope, L., Fisher, R. P., & Jamieson, K. (2011, June). Does an
early recall opportunity make witnesses more susceptible to misinformation? Paper presented at the meeting of the Society for Applied
Research in Memory and Cognition, New York, NY.
Gabbert, F., Memon, A., & Allan, K. (2003). Memory conformity: Can
eyewitnesses influence each other’s memories for an event? Applied
Cognitive Psychology, 17, 533–543. doi:10.1002/acp.885
Gabbert, F., Memon, A., Allan, K., & Wright, D. (2004). Say it to
my face: Examining the effects of socially encountered misinformation. Legal and Criminological Psychology, 9, 215–227.
doi:10.1348/1355325041719428
Goodsell, C. A., Neuschatz, J. S., & Gronlund, S. D. (2009). Effects of mugshot commitment on lineup performance in young and older adults.
Applied Cognitive Psychology, 23, 788–803. doi:10.1002/acp.1512
Hertzog, C., & Hultsch, D. F. (2000). Metacognition in adulthood and old
age. In F. I. M. Craik & T. A. Salthouse (Eds.), Handbook of aging
and cognition (2nd ed., pp. 417–466). Mahwah, NJ: Erlbaum.
Hess, T. M., & Ennis, G. E. (2012). Age differences in the effort and costs
associated with cognitive activity. The Journals of Gerontology.
Series B, Psychological Sciences and Social Sciences, 67, 447–455.
doi:10.1093/geronb/gbr129
Jacoby, L. L. (1999). Deceiving the elderly: Effects of accessibility bias in
cued-recall performance. Cognitive Neuropsychology, 16, 417–436.
doi:10.1080/026432999380861
Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring.
Psychological Bulletin, 114, 3–28. doi:10.1037/0033-2909.114.1.3
Karpel, M. E., Hoyer, W. J., & Toglia, M. P. (2001). Accuracy and qualities of real and suggested memories: Nonspecific age differences.
The Journals of Gerontology. Series B, Psychological Sciences and
Social Sciences, 56, 103–110. doi:10.1093/geronb/56.2.P103
Koutstaal, W. (2003). Older adults encode–but do not always use–perceptual details: Intentional versus unintentional effects of detail
on memory judgments. Psychological Science, 14, 189–193.
doi:10.1111/1467–9280.01441
Koutstaal, W., Schacter, D. L., Johnson, M. K., Angell, K. E., & Gross, M.
S. (1998). Post-event review in older and younger adults: Improving
memory accessibility of complex everyday events. Psychology and
Aging, 13, 277–296. doi:10.1037/0882-7974.13.2.277
Kricos, P. B. (2003, March). Hearing loss in elders. Lecture presented to
GEY 4001/6646: Issues and Concepts in Gerontology, University of
Florida.
Kwong See, S. T., Hoffman, H. G., & Wood, T. L. (2001). Perceptions of an
old female eyewitness: Is the older eyewitness believable? Psycho­
logy and Aging, 16, 346–350. doi:10.1037/0882-7974.16.2.346
Loftus, E. F. (2005). Planting misinformation in the human mind: A 30-year
investigation of the malleability of memory. Learning & Memory
(Cold Spring Harbor, N.Y.), 12, 361–366. doi:10.1101/lm.94705
Loftus, E. F., Levidow, B., & Duensing, S. (1992). Who remembers best?
Individual differences in memory for events that occurred in a science
museum. Applied Cognitive Psychology, 6, 93–107. doi:10.1002/
acp.2350060202
Loftus, E. F., Miller, D. G., & Burns, H. J. (1978). Semantic integration of verbal information into a visual memory. Journal of
Experimental Psychology. Human Learning and Memory, 4, 19–31.
doi:10.1037/0278- 7393.4.1.19
Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of automobile
destruction: An example of the interaction between language and
memory. Journal of Verbal Learning and Verbal Behavior, 13, 585–
589. doi:10.1016/S0022-5371(74)80011–3
Marche, T. A., Jordan, J. J., & Owre, K. P. (2002). Younger adults can be
more suggestible than older adults: The influence of learning differences on misinformation reporting. Canadian Journal on Aging, 21,
85–93. doi:10.1017/S0714980800000660
Mello, E. W., & Fisher, R. P. (1996). Enhancing older adult eyewitness
memory with the cognitive interview. Applied Cognitive Psychology,
10, 403–417. http://psycnet.apa.org.lp.hscl.ufl.edu/index.cfm?fa=
search.displayRecord&id=19946F93-969D-E0E5-284E-029E54E
2C886&resultID=1&page=1&dbTab=all doi:10.1002/(SICI)10990720(199610)10:5<403::AID-ACP395>3.0.CO;2-X
Memon, A., Gabbert, F., & Hope, L. (2004). The ageing eyewitness. In J.
R. Adler (Ed.), Forensic psychology: Concepts, debates and practice
(pp. 96–112). Devon, United Kingdom: Willan.
Mitchell, K. J., Johnson, M. K., & Mather, M. (2003). Source monitoring
and suggestibility to misinformation: Adult age-related differences.
Applied Cognitive Psychology, 16, 1–13. doi:10.1002/acp.857
Mueller-Johnson, K. (2009). Ältere menschen als zeugen vor gericht.
Forensische Psychiatre, Psychologie, Kriminologie, 3, 163–170.
Mueller-Johnson, K., & Ceci, S. J. (2004). Memory and suggestibility in
older adults: Live event participation and repeated interview. Applied
Cognitive Psychology, 18, 1109–1127. doi:10.1002/acp.1078
Mueller-Johnson, K., & Ceci, S. J. (2007). The elderly eyewitness:
A review and prospectus. In M. P. Toglia, J. D. Read, D. F. Ross, &
R. C. L. Lindsay (Eds.), Handbook of eyewitness psychology (Vol. 1,
pp. 577–603). Mahwah, NJ: Erlbaum.
Nash, R. A., Wade, K. A., & Lindsay, D. S. (2009). Digitally manipulating memory: Effects of doctored videos and imagination in distorting beliefs and memories. Memory & Cognition, 37, 414–424.
doi:10.3758/MC.37.4.414
Naveh-Benjamin, M., Shing, Y. L., Kilb, A., Werkle-Bergner, M.,
Lindenberger, U., Li, S. (2009). Adult age differences in memory
AGE AND EYEWITNESS MEMORY
for name-face associations: The effects of intentional and incidental
learning. Memory, 17, 220–232. doi:10.1080/09658210802222183
Pansky, A., Goldsmith, M., Koriat, A., & Pearlman-Avnion, S. (2009).
Memory accuracy in old age: Cognitive, metacognitive, and neurocognitive determinants. European Journal of Cognitive Psychology,
21, 303–329. doi:10.1080/09541440802281183
Reyna, V. F., & Brainerd, C. J. (1995). Fuzzy-trace theory: An
interim synthesis. Learning and Individual Differences, 7, 1–75.
doi:10.1016/1041–6080(95)90031–4
Roediger, H. L., 3rd, & Geraci, L. (2007). Aging and the misinformation effect: A neuropsychological analysis. Journal of Experimental
Psychology. Learning, Memory, and Cognition, 33, 321–334.
doi:10.1037/0278-7393.33.2.321
Schacter, D. L., Koutstaal, W., Johnson, M. K., Gross, M. S., & Angell, K.
E. (1997). False recollection induced by photographs: A comparison
of older and younger adults. Psychology and Aging, 12, 203–215.
doi:10.1037/0882-7974.12.2.203
Searcy, J. H., Bartlett, J. C., & Memon, A. (1999). Age differences in accuracy and choosing in eyewitness identification and face recognition.
Memory & Cognition, 27, 538–552.
Searcy, J. H., Bartlett, J. C., & Memon, A. (2000). Influence of post-event
narratives, line-up conditions and individual differences on false
347
identification of young and older eyewitnesses. Legal and
Criminological Psychology, 5, 219–235. doi:10.1348/1355325
00168100
Searcy, J. H., Bartlett, J. C., Memon, A., & Swanson, K. (2001). Aging and
lineup performance at long retention intervals: Effects of metamemory and context reinstatement. The Journal of Applied Psychology,
86, 207–214. doi:10.1037/0021-9010.86.2.207
Sutherland, R., & Hayne, H. (2001). Age-related changes in the misinformation effect. Journal of Experimental Child Psychology, 79,
388–404. doi:10.1006/jecp.2000.2610
Vakil, E., Hornik, C., & Levy, D. A. (2008). Conceptual and perceptual
similarity between encoding and retrieval contexts and recognition
memory context effects in older and younger adults. The Journals of
Gerontology. Series B, Psychological Sciences and Social Sciences,
63, 171–175. doi:10.1093/geronb/63.3.P171
Yarmey, D. A. (2004). Eyewitness recall and photo identification: A field
experiment. Psychology, Crime & Law, 10, 53–68. doi:10.1080/
1068316021000058379
Zaragoza, M. S., & Lane, S. M. (1994). Source misattributions and the suggestibility of eyewitness memory. Journal of Experimental Psychology.
Learning, Memory, and Cognition, 20, 934–945. doi:10.1037/02787393.20.4.934