Applied Cognitive Psychology, Appl. Cognit. Psychol.

28: 3038 (2014)

Published online 24 September 2013 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/acp.2951

Age-related Differences in the Free-recall Accounts of Child, Adolescent, and

Adult Witnesses
FIONA JACK, JESSICA LEOV and RACHEL ZAJAC
Department of Psychology, University of Otago, Dunedin, New Zealand
Summary: Many researchers have examined the factors that affect childrens ability to provide eyewitness evidence, leading to
signicant reform in policy and practice. In stark contrast, there has been virtually no eyewitness research conducted with
adolescents, even though adolescents are still undergoing developmental changes that are likely to affect eyewitness performance.
We compared the eyewitness performance of children, adolescents, and adults by showing them a brief lm clip depicting a
simulated crime and using cognitive interview instructions to elicit free-recall accounts. Adolescents provided more information
than children, but less information than adults. Accuracy did not differ with age. These data suggest that, like children, adolescents
could benet from specialised interview techniques that help them provide more complete accounts. Across all three age groups,
individual differences in cognitive functioning contributed to variation in eyewitness performance, and eliciting a second freerecall account increased the amount of information reported. Copyright 2013 John Wiley & Sons, Ltd.

Eyewitness testimony is one of the most critical forms of

evidence in criminal investigations (Engelhardt, 1999; Fisher
& Schreiber, 2007). Inaccurate or incomplete testimony can
result in wasted police time, failure to solve a case, wrongful
conviction, or victims remaining in unsafe situations
(McMurtrie, 2007; Milne & Bull, 1999). Given the high
social and nancial costs of these outcomes, an important
research goal is to understand how various factors inuence
witnesses ability to recall and recount events.
One of the most studied variables in the eyewitness testimony literature is age. In particular, although a full and
accurate description of an event that may have been sudden,
unexpected, and stressful is likely to be challenging for any
witness, it is particularly difcult for children. The ability
to encode, retain, and retrieve memories for events emerges
during the preschool years and continues to improve across
childhood (Baker-Ward & Ornstein, 2002; Fivush, 2002;
Pipe, Thierry, & Lamb, 2007). The development of language
is also protracted (Saywitz, 2002). Consequently, relative to
adults, children negotiating investigative interviews may nd
it difcult to understand what is being asked of them, to
access their memory for the event, and to put what they
remember into words (see Lamb, Orbach, Warren, Esplin,
& Hershkowitz, 2007, for a review).
An abundance of research has been conducted to examine
the eyewitness memory abilities of preschoolers and young
school-aged children (see Lamb et al., 2007; Melnyk,
Crossman, & Scullin, 2007 for reviews). Under free-recall
conditions, children typically report less information than
adolescents and adults, although they are no less accurate
(Chapman & Perry, 1995; Eisen, Goodman, Qin, Davis, &
Crayton, 2007; Marin, Holmes, Guth, & Kovac, 1979;
Sutherland & Hayne, 2001; Tustin & Hayne, 2010). Childrens
accuracy does tend to suffer, however, when more specic
questions are asked, as children are often more suggestible than
older individuals (e.g. Oates & Shrimpton, 1991; Sutherland
& Hayne, 2001; see Melnyk et al., 2007, for a review).
*Correspondence to: Fiona Jack, Department of Psychology, University of
Otago, PO Box 56, Dunedin 9054, New Zealand.
E-mail: ona@psy.otago.ac.nz

Copyright 2013 John Wiley & Sons, Ltd.

Although the difference between childrens and adults

accounts is well established, we know almost nothing about
what happens in between. Although adolescents have received
substantially less attention than children, they deserve special
consideration for at least two reasons. First, the functional
maturation of the human brain continues throughout adolescence and extends into the early 20s (Giedd et al., 1999;
Hudspeth & Pribram, 1990; Paus, 2005; Sowell, Thompson,
Holmes, Jernigan, & Toga, 1999). Signicant changes occur
between childhood, adolescence, and adulthood across a
number of brain regions, including the prefrontal cortex
(Giedd et al., 1999; Sowell et al., 1999), leading to more
efcient functioning within and between brain regions and
enhanced cognitive processing (Paus, 2005; Sowell et al.,
1999). Consequently, many of the complex cognitive skills
required in legal contexts might pose particular problems
for both children and adolescents. For example, given that
the frontal lobes play an important role in working memory,
executive functioning, and response inhibition (Paus, 2005;
Smith & Jonides, 1999), young witnesses might have more
difculty than adults inhibiting conicting or irrelevant
information when providing reports.
The second reason adolescents deserve special consideration is that adolescence represents a unique stage in social
development. Adolescents are more likely than children or
adults to change their behaviour to t in with others
(Costanzo & Shaw, 1966; Santor, Messervey, & Kusumakar,
2000). For example, adolescents are more likely than adults
to comply with authority gures in legal contexts (Grisso
et al., 2003; Gudjonsson, 2003). These data suggest that
adolescents might be particularly vulnerable to the inuence
of any demand characteristics of the interview situation.
Consistent with these predictions, adolescents, like children,
are typically less accurate or more suggestible than adults
when tested in paradigms designed to measure suggestibility
or misinformation effects (Gudjonsson & Singh, 1984; Loftus,
Levidow, & Duensing, 1992; McGuire, London, & Wright,
2011; Richardson, Gudjonsson, & Kelly, 1995; Singh &
Gudjonsson, 1992; but see Redlich, Silverman, & Steiner,
2003). But how do adolescents perform relative to other age
groups when questioned using the open-ended techniques that

Age-related difference in free-recall account

are appropriate for use during investigative interviews?
Open-ended questions have been used in only three eyewitness
memory studies that included adolescent participants
(Chapman & Perry, 1995; Eisen et al., 2007; Memon, Wark,
Holley, Bull, & Koehnken, 1997). In two of these, adolescents performance was compared with that of children (as
cited earlier), but researchers have yet to compare adolescents performance with that of older witnesses. We address
this issue in the present experiment.
We also examine the extent to which free-recall reports
might differ as a function of individual difference variables
other than age. Although several researchers have investigated
how cognitive variables might contribute to eyewitness performance, most of this research has been conducted with children,
and much of it has focused on suggestibility (see Bruck &
Melnyk, 2004; Clarke-Stewart, Malloy, & Allhusen, 2004;
Lee, 2004, for reviews). However, a few researchers have
examined the extent to which scores on standardised memory, language, and intelligence tests might predict eyewitness
performance under neutral interview conditions.
Given that event recall depends on memory for the target
event, it seems likely that individual differences in general
memory skills would play a role in eyewitness performance.
In the few studies where researchers have tested this hypothesis,
they have found positive relations. For example, Eisen, Qin,
Goodman, and Davis (2002) found that maltreated childrens
digit span and memory for sentences were negatively related
to the number of errors made when recalling a medical examination (see also Eisen et al., 2007).
Language underlies witnesses ability to understand interviewers questions and to express what they remember. It
also plays an important role in the encoding, storage, and
retrieval of event memories (Hayne & Jack, 2011). However,
aside from verbal intelligence, tests of language have not
demonstrated a consistent relation with eyewitness performance. For example, some researchers have found no relation between childrens event recall and their vocabulary
scores (Greenhoot, Ornstein, Gordon, & Baker-Ward,
1999; Gross & Hayne, 1999), whereas others have found
positive relations (Gordon et al., 1993; Salmon, Roncolato,
& Gleitzman, 2003; see also Brown & Pipe, 2003). There
is some evidence that the relation between language and
eyewitness recall strengthens with age across childhood
(Gordon et al., 1993).
General intellectual functioning might contribute to eyewitness memory through, for example, the organisation of event
representations or the use of memory strategies to aid
retrieval (Brown & Pipe, 2003; Geddie, Fradin, & Beer,
2000). Indeed, childrens scores on intelligence tests correlate
positively with the amount of information recalled about
classroom events (Brown & Pipe, 2003; Geddie et al., 2000).
Verbal intelligence tends to be more closely related to childrens eyewitness performance than non-verbal intelligence
(Brown & Pipe, 2003; Henry & Gudjonsson, 2007), although
both measures are positively related (e.g. Eisen et al., 2002;
Elischberger & Roebers, 2001; Roebers & Schneider, 2001).
There is some evidence that, as for language, the relation
between eyewitness recall and intelligence (both verbal and
non-verbal) increases with age across childhood (Elischberger
& Roebers, 2001; Roebers & Schneider, 2001).
Copyright 2013 John Wiley & Sons, Ltd.

31

It therefore appears that childrens general memory, linguistic, and intellectual abilities all contribute to their eyewitness performance to some extent. For both theoretical
and applied reasons, it is important to know whether the role
of these variables is consistent across the lifespan. From a
theoretical perspective, this will tell us whether the reasons
for individual variation in the quality of older witnesses
reports are the same as the reasons for individual variation
in the quality of childrens reports. From an applied perspective, knowing the extent to which these cognitive variables
contribute to eyewitness performance will be helpful for
experts who are required to evaluate witnesses reports
(Geddie et al., 2000).
Interviewers obviously cannot manipulate a witnesss age
or cognitive skills. There are, however, a number of other
variables that inuence eyewitness performance and are
under the interviewers control. Perhaps the most inuential
of these is questioning style. As such, a large amount of
research attention has been devoted to developing interview
techniques that maximise the volume and accuracy of
witnesses reports (e.g. Aschermann, Mantwill, & Kohnken,
1991; Larsson, Granhag, & Spjut, 2003; Lipton, 1977;
Roberts & Higham, 2002; for reviews, see Fisher &
Schreiber, 2007; Milne & Bull, 1999). Free-recall prompts
(such as Please tell me everything that happened that
morning) have been identied as particularly useful, with
information obtained during free recall generally highly
accurate (e.g. Lipton, 1977) and forensically relevant (e.g.
Roberts & Higham, 2002).
The main drawback to free-recall accounts is that they are
often relatively brief. In order to elicit additional information,
investigators often use increasingly specic questions.
Unfortunately, when asking specic questions, investigators
often introduce newand potentially inaccuratedetails
(Milne & Bull, 1999). For example, investigators might ask
leading questions (e.g. Was he wearing a blue hat?) or
forced-choice questions (e.g. Was the hat blue or black?). It is
well established that, relative to free-recall prompts, these types
of questions impact negatively on accuracy (e.g. Aschermann
et al., 1991; Larsson et al., 2003; Lipton, 1977; see Milne &
Bull, 1999, for a review).
A crucial step in reducing the need for specic questioning
is to develop techniques that maximise the amount of information that witnesses provide during free recall. The cognitive interview (CI), which is recommended by scientists and
widely used by police, includes a number of instructions
designed to achieve this (Fisher & Schreiber, 2007; Milne
& Bull, 1999). Chiey, witnesses are asked to report everything they can remember, even seemingly minor details,
and to mentally reinstate the context in which the event took
place. Used together, these instructions signicantly increase
the amount of information reported by both children and
adults, with no effect on accuracy (Milne & Bull, 2002).
In its full version, the CI also includes a number of mnemonics
that are variations on the typical free-narrative technique (e.g.
recalling the event in reverse chronological order or describing
the event from a different point of view). However, researchers
have voiced concern that some witnesses might nd these
latter techniques confusing (Memon, 2006; Saywitz, Geiselman,
& Bornstein, 1992) and that evidence obtained using the
Appl. Cognit. Psychol. 28: 3038 (2014)

32

F. Jack et al.

change perspective mnemonic could be deemed speculation or

hearsay (Boon & Noon, 1994). Furthermore, two studies
have shown that simply asking participants to try again
was just as effective at eliciting additional correct information
as using these special mnemonics (Boon & Noon, 1994;
Milne & Bull, 2002).
The reminiscence literature offers further evidence that
simply asking witnesses to provide a second free-recall
account might effectively increase the amount of information
reported (see La Rooy, Pipe, & Murray, 2005, for a review).
Reminiscence is the tendency for the cumulative number of
unique details reported to increase across multiple retrieval
attempts. Both adults (e.g. Scrivner & Safer, 1988) and
children (e.g. Bluck, Levine, & Laulhere, 1999; La Rooy
et al., 2005; Peterson, Moores, &White, 2001; Salmon & Pipe,
2000) have shown reminiscence for witnessed or personally
experienced events.
The implications of these results for applied settings are
unclear, however, as participants did not receive CI instructions prior to the rst retrieval attempt. Given that the CI
instructions were specically designed to maximise the
amount of information that is reported, they might create
a ceiling effect such that subsequent retrieval attempts will
fail to elicit additional details. Even in the Boon and Noon
(1994) and Milne and Bull (2002) experiments, in which CI
mnemonics were used, participants were not given both the
report everything and context reinstatement instructions
prior to the rst retrieval attempt, both of which are standard features of the CI (Fisher & Schreiber, 2007; Milne
& Bull, 1999) and which have been shown in tandem to
be crucial to its success (Milne & Bull, 2002). It remains
to be seen how effective a second free-recall attempt will
be when these standard instructions are given prior to the
rst attempt. One of the goals of the present experiment
was to answer this question.
In light of the research discussed above, we had three
main aims in conducting the present experiment. The rst
was to provide the rst empirical evaluation of adolescents
free-recall accounts relative to those of both children and
adults. Consistent with New Zealands legal denitions of
a child, young person, and adult (Children, Young Persons,
and Their Families Act, 1989), we recruited children aged
911 years and adolescents aged 1416 years. Given that
brain maturation extends into early adulthood (Giedd et al.,
1999; Hudspeth & Pribram, 1990; Sowell et al., 1999) and
older adults exhibit a decline in eyewitness performance
(Moulin, Thompson, Wright, & Conway, 2007), we restricted
the age of our adult participants to 2560 years. Whereas the
adult samples in many memory studies comprise college
students, who may differ systematically from the general
population (Loftus et al., 1992), we recruited all participants
from the wider community. Our second aim was to examine
the extent to which childrens, adolescents, and adults performance on standardised cognitive tests are related to their
eyewitness performance. Our nal aim was to clarify previous research ndings (Boon & Noon, 1994; Milne & Bull,
2002) by establishing whether a second free-recall attempt
is successful at eliciting additional details when the standard
report everything and context reinstatement instructions are
given prior to the rst attempt.
Copyright 2013 John Wiley & Sons, Ltd.

Participants individually viewed a brief lm clip depicting

a simulated crime. After a 45-minute delay, during which the
interviewer administered standardised cognitive tests, participants were asked to provide two free-recall accounts about
what they had witnessed. The interviews were then transcribed
and coded, allowing us to systematically assess the type, quantity, and accuracy of the information that participants reported.
METHOD
Participants
Participants were recruited via a newspaper advertisement,
invitations distributed to local schools, a research participant
database, and word of mouth. The nal sample comprised 48
children aged 911 years (M = 10.6, SD = 0.9; 24 boys), 48
adolescents aged 1416 years (M = 15.4, SD = 0.8; 24 boys),
and 48 adults aged 2560 years (M = 42.8, SD = 9.3; 24
men). An additional ve participants were excluded because
of insufcient uency in English (one adult and one child) or
experimenter error (two adults and one adolescent). All participants, and a parent/guardian of each child and adolescent
participant, gave written informed consent.
Across the three age groups, 86% of participants identied
as NZ European, 6% as NZ Maori, 9% as Asian or Indian,
and 9% as other ethnicities (percentages do not sum to 100
because some participants reported multiple ethnicities).
Child and adolescent participants provided their parents
occupations, and adult participants provided their own and,
if applicable, their partners occupation. On the basis of the
Elley-Irving (2003) socio-economic index, in which scores
range from 1 (e.g. lawyer, university professor) to 6 (e.g.
cook or cleaner), the average household occupation level
across the sample was 2.7 (SE = 0.11); scores did not differ
as a function of the age group, F(2, 131) = 1.98, p > .10.
Film clips
In order to increase the internal and external validity of our
ndings (Wells & Windschitl, 1999), we used three different
lm clips as the target events. The clips were shown on a
24-in. computer monitor, with participants seated approximately 1.5 m from the screen. The lm clips were created
by the New Zealand Police for use in their interview training
workshops. The use of these clips ensured that our target
events were representative of events that are often the focus
of investigative interviews.
The rst lm clip (54 seconds) shows a man walking on a
suburban footpath, looking over a fence, looking in a letterbox, and entering a property. The second (31 seconds) shows
two people walking into a shop and, a few seconds later,
running out of the shop and around the corner out of sight.
The third clip (14 seconds) shows a young woman shoplifting
in a convenience store.
Memory interview
The interview was conducted using a standardised script,
based on the standard CI techniques used by the New
Zealand Police. Each session was conducted by one of two
trained female interviewers. At the beginning of the memory
Appl. Cognit. Psychol. 28: 3038 (2014)

Age-related difference in free-recall account

interview, participants were instructed to take their time, to
concentrate hard, and to report everything they could recall.
They were told to make sure they only reported the information they could actually recall, and not to guess. Finally,
participants were asked to reinstate the context of the event
by thinking about what they had seen, heard, thought, and
felt when watching the clip.
While participants described what they remembered, the
interviewer used minimal encouragers (e.g. M-hm), but did
not interrupt. Once the participant stopped giving information,
the interviewer prompted once, Is there anything else you can
remember?
Next, the interviewer asked the participant to talk her
through the event again. The interviewer explained that this
did not mean she thought the participant had got anything
wrong the rst time; it was just in case he or she could
remember any more. The participant was reminded not to
guess or to try to ll in any gaps in his or her memory. Once
the participant began to give information, the interview
proceeded as for the rst recall phase.
Procedure
Participants were tested individually. Sessions were balanced
such that each interviewer saw half of the participants from
each age group and half of each gender within each age
group. After consent procedures were completed, the participant was shown one of the three lm clips. Clips were randomly assigned, with the restriction that each was viewed
equally often by participants from each combination of age
group, gender, and interviewer. Before showing the clip, the
interviewer asked the participant to watch and listen carefully
and to try to imagine he or she was actually there, seeing it
happen.
After viewing the clip, the participant was administered
the two-subtest form of the Wechsler Abbreviated Scale of
Intelligence (WASI), which provides an acceptable general
summary of cognitive ability (Wechsler, 1999). It comprises
one test of verbal intelligence (vocabulary) and one test of
non-verbal intelligence (matrix reasoning). The interviewer
also administered the four subtests of the Wide Range
Assessment of Memory and Learning (WRAML; Sheslow
& Adams, 2003) that contribute to the screening index (story
memory, design memory, verbal learning, and picture
memory). Both the WASI and the WRAML were able to
be administered across all three of our age groups. To keep
the session length reasonable for participants, rather than
administering an additional test of language, we used scores
on the WASI vocabulary subtest as a measure of language
ability.
The participant was then interviewed about the lm clip
using the interview protocol described earlier. The delay
between the lm clip and the interview was approximately
45 minutes. Upon completion of the interview, participants
were thanked and given $15.
Coding
All interviews were audio-recorded and transcribed verbatim.
Any identifying information was removed from the transcripts
prior to coding. A coding scheme was developed to quantify
Copyright 2013 John Wiley & Sons, Ltd.

33

the amount, type, and accuracy of the information that participants reported. First, each new relevant detail participants
reported was assigned to one of the ve categories described
in the following. Repetitions and off-topic utterances were
ignored. Any word or group of words that gave embellishment
to the main idea (e.g. adjectives and adverbs) scored a further
code of the same type. In the following examples, each
separately underlined word or phrase received unique credit
for that particular information type:
Person: Information describing a persons physical
appearance, including his or her clothing (e.g. she had
brown hair or he was wearing a brown striped hoodie)
Behaviour: Information describing the behaviour, actions,
or manner of the people in the clip (e.g. he looked into a
letterbox or the girl looked annoyed)
Surroundings: Information describing physical surroundings
and objects (e.g. he looked into a letterbox or the girl
took a packet of noodles from the shelf)
Time: Information describing the sequence of events (e.g.
then, while, or rst), the duration of something (e.g.
They were in the shop for about 10 seconds), or the time
of day (e.g. It must have been the middle of the day
because the sun was very bright)
Sound: Information describing any sounds (e.g. the girl
shouted Sam quite loudly or I could hear birds chirping)
Next, each detail was further coded as correct, possible, or
incorrect. The possible code covered subjective or evaluative statements (e.g. It looked like a nice neighborhood).
This category also included reasonable speculation about
things that were not shown in the clip but could have been
correct (e.g. the shopkeeper might have been out the back).
Thirty randomly selected transcripts (10 from each age
group) were coded independently by two coders. Agreement between coders was good; Pearson correlations ranged
from r = 0.89 to r = 0.98 for the ve information-type codes
and from r = 0.91 to r = 0.99 for the three accuracy codes
(all ps < .001). Disagreements were resolved through discussion; one coder coded the remaining transcripts.

RESULTS
Prior to analysis, we adjusted for outliers by replacing any
values that were 3 SD from the mean with the next-highest
value plus 1 (Tabachnick & Fidell, 2007). Across all variables
analysed, we replaced 7 child, 9 adolescent, and 11 adult data
points. For analyses of variance (ANOVAs), where the
homogeneity-of-variance assumption was violated, we report
Welchs F-ratio and Tamhanes post-hoc comparisons.
Elsewhere, Tukeys post-hoc test was used. Unless stated
otherwise, alpha = .05 throughout.
Preliminary analyses showed no main effects of gender,
and gender did not enter into any signicant interactions,
so we collapsed across gender for subsequent analyses.
Given the wide age range of our adult participants, we tested
for correlations between age and our eyewitness measures
within this group. None were signicant (ps > .10). There were
effects of lm clip on the volume of participants reports,
F(2, 141) = 13.37, and the proportion of information that
Appl. Cognit. Psychol. 28: 3038 (2014)

34

F. Jack et al.

was correct, possible, and incorrect, F(2, 141) = 3.72, 11.10,

and 5.70, respectively. The interactions with age were nonsignicant (ps > .08). Because lm clip was a nominal variable,
it was not possible to include it as a covariate (Brace, Kemp, &
Snelgar, 2009). Given our thorough counterbalancing of lm
clip, however, we are condent that the variation in performance due to lm clip will not have affected our other results.
Amount and type of information reported
We included only correct details in these analyses. Table 1
shows the mean number of correct details reported as a
function of age group, recall phase, and information type.
Because details about sound were rare, we omitted this category from analyses. We subjected these data to a 3 (age: children, adolescents, and adults) 5(information type: person,
behaviour, surroundings, and time) multivariate ANOVA,
with repeated measures across information type. The dependent variables were the numbers of unique details reported
during the rst and second recall phases. The main effects of
age, F(4, 280) = 17.43, Wilks lambda = 0.64, 2p 0:20, and
information type, F(6, 136) = 96.38, Wilks lambda = 0.19,
2p 0:81 , and their interaction, F(12, 272) = 7.83, Wilks
lambda = 0.55, 2p 0:26, were all signicant for the combined dependent variable amount recalled.
Next, we analysed the amount recalled during the two
individual recall phases, using a Bonferroni-adjusted alpha
level of .025. Mauchlys test showed that for both recall
phases, the assumption of sphericity had been violated for
information type, so the degrees of freedom were corrected
using GreenhouseGeisser estimates. The main effect of
age was signicant for both recall phases, F(2, 141) = 36.70,
2p 0:34 , and F(2, 141) = 4.76, 2p 0:06 , respectively.
During the rst recall phase, adults reported more details than
adolescents, who in turn reported more details than children.
During the second recall phase, the same pattern emerged, although adolescents did not differ signicantly from either of
the other groups. The effect of information type was also signicant for both recall phases, F(2.45, 345.29) = 131.06, 2p

0:48, and F(2.37, 334.46) = 46.23, 2p 0:25, respectively. As

shown in Table 1, most of the details participants reported described people, behaviour, or surroundings; relatively few described time. For the rst recall phase only, these main effects
were qualied by a signicant interaction, F(4.90,
345.29) = 14.68, 2p 0:17 . A series of one-way ANOVAs
showed that this interaction was due to signicant effects of
age on the amount of information that was reported about people, Welchs F(2, 84.26) = 40.23, behaviour, Welchs F(2,
90.98) = 12.88,
and
surroundings,
Welchs
F(2,
76.46) = 34.22, but not time. Tamhanes post-hoc tests
revealed that children reported less information about people
than adolescents, who in turn reported less than adults;
children and adolescents reported less than adults about
behaviour; and children reported less than adolescents and
adults about surroundings.

Accuracy
Recall that each detail reported was categorised as correct,
possible, or incorrect. For each participant, these data were
converted to proportions by dividing the total number of
details falling into each category by the total number of
details reported, separately for each recall phase. We analysed
these proportions using a 3 (age: children, adolescents, and
adults) 2 (recall phase: rst and second) ANOVA with
repeated measures across recall phase. There were no signicant effects of age or interactions. There were, however, significant effects of recall phase on the proportion of information
that was correct, F(1, 141) = 55.80, 2p 0:28 , possible,
F(1, 141) = 9.89, 2p 0:07, and incorrect F(1, 141) = 48.27,
2p 0:26. Relative to the rst recall phase, the proportion of
correct information decreased during the second recall phase,
whereas the proportion of possibly correct information and
the proportion of incorrect information both increased
(Table 2). Because participants did not always report details
for each information type, we were unable to examine accuracy as a function of this variable.

Table 1. Mean (and SE) number of correct details reported as a function of age group, recall phase, and information type
Age group
Information type

Children

People
Behaviour
Surroundings
Time
Sound
Total

5.06 (0.38)
5.52 (0.40)
5.88 (0.40)
1.98 (0.28)
0.38 (0.08)
18.92 (1.06)

First recall phase

7.25 (0.69)
7.00 (0.45)
11.69 (1.00)
2.52 (0.24)
0.44 (0.10)
28.69 (1.79)

12.92
9.35
14.29
2.40
0.44
38.96

(0.79)
(0.65)
(1.15)
(0.24)
(0.13)
(1.94)

8.41 (0.46)
7.29 (0.32)
10.62 (0.60)
2.30 (0.15)
0.42 (0.06)
28.85 (1.16)

People
Behaviour
Surroundings
Time
Sound
Total

1.02 (0.17)
2.13 (0.29)
2.15 (0.34)
0.85 (0.15)
0.02 (0.14)
6.06 (0.64)

Second recall phase

1.10 (0.25)
1.83 (0.25)
3.17 (0.43)
0.90 (0.17)
0.10 (0.04)
6.94 (0.77)

1.81
3.17
3.42
1.06
0.17
9.60

(0.23)
(0.38)
(0.37)
(0.18)
(0.07)
(0.83)

1.31 (0.13)
2.38 (0.18)
2.91 (0.22)
0.94 (0.10)
0.10 (0.03)
7.53 (0.45)

Copyright 2013 John Wiley & Sons, Ltd.

Adolescents

Adults

Full sample

Appl. Cognit. Psychol. 28: 3038 (2014)

Age-related difference in free-recall account

35

Table 2. Mean (and SE) proportion of information that was correct, possible, and incorrect, as a function of recall phase and age group
Age group
Accuracy

Children

Adolescents

Correct
Possible
Incorrect

0.75 (0.02)
0.12 (0.01)
0.13 (0.01)

Correct
Possible
Incorrect

0.61 (0.03)
0.17 (0.02)
0.22 (0.02)

Adults

Full sample

First recall phase

0.79 (0.01)
0.12 (0.01)
0.09 (0.01)

0.78 (0.01)
0.12 (0.01)
0.10 (0.01)

0.77 (0.01)
0.12 (0.01)
0.11 (0.01)

Second recall phase

0.66 (0.04)
0.14 (0.02)
0.19 (0.03)

0.67 (0.03)
0.15 (0.02)
0.18 (0.02)

0.64 (0.02)
0.15 (0.01)
0.20 (0.01)

Individual difference measures

On the basis of the WASI and WRAML subtests, we were
able to obtain six standardised scores for each participant:
WASI vocabulary T score (verbal intelligence), WASI matrix
reasoning T score (non-verbal intelligence), WASI Full-2
score (general intelligence), WRAML verbal memory index
(verbal memory), WRAML visual memory index (visual
memory), and WRAML screening memory index (general
memory). We used linear regression to examine whether
these six individual difference measures predicted two
dependent measures: the number of correct details reported
in our eyewitness paradigm and, separately, the proportion
of correct information. Entering all six predictor variables
into the regression analyses simultaneously resulted in
unacceptable levels of multicollinearity among predictors
(tolerance = 0.0010.012). To avoid multicollinearity and to
maintain a good ratio of participants to predictor variables,
we conducted a separate analysis for each individual difference measure. Because lm clip affected both dependent
variables, we controlled for lm clip in the models using a
set of two dummy predictors. We also examined whether
the relations between our individual difference measures

and eyewitness performance varied with age by including

two dummy variables representing the three age groups and
terms for their interaction with the nominated individual
difference measure. We used BonferroniHolm-adjusted
alpha levels when interpreting the results of these analyses.
All of the 12 models were signicant (Table 3). In each
case, the nominated individual difference measure and the
dummy variables for clip were the only signicant predictors,
with the exceptions that verbal intelligence, verbal memory,
and visual memory scores were not signicant predictors of
proportion correct. In all cases, the association between the
individual difference measure and the eyewitness memory
measure was in a positive direction. None of the interactions
with age were signicant (ps > .10)

DISCUSSION
Age-related differences in recall
Children and adolescents provided less information about
the lms than adults but were no less accurate. These data
are consistent with existing literature on the effects of age

Table 3. Statistics for regression models predicting the amount of correct information reported and proportion correct and for the individual
difference measures included as predictor variables
Statistics for the individual difference
predictor variable

Statistics for the model

Individual difference measure

Adjusted R2

Verbal intelligence
Non-verbal intelligence
General intelligence
Verbal memory
Visual memory
General memory

0.482
0.465
0.492
0.471
0.491
0.499

Verbal intelligence
Non-verbal intelligence
General intelligence
Verbal memory
Visual memory
General memory

0.090
0.160
0.143
0.105
0.084
0.116

.225
.208
.259
.201
.187
.210

3.65
3.12
4.07
2.91
2.88
3.19

<.001
.002
<.001
.004
.005
.002

Models predicting proportion correct

3.02
.005
.195
4.88
<.001
.343
4.41
<.001
.313
3.40
.002
.209
2.87
.008
.146
3.66
.001
.242

2.39
4.10
3.78
2.34
1.68
2.77

.018
<.001
<.001
.021
.096
.006

Models predicting amount correct

20.00
<.001
18.76
<.001
20.78
<.001
19.18
<.001
20.57
<.001
21.17
<.001

Note: Each model included two dummy variables representing clip on the rst step and one individual difference measure, two dummy variables representing
age group, and terms for their interaction on the second step. For the models including visual memory and general memory, model df = 7 and error df = 135; for
all other models, model df = 7 and error df = 136. Verbal intelligence = Wechsler Abbreviated Scale of Intelligence (WASI) vocabulary T score; non-verbal
intelligence = WASI matrix reasoning T score; general intelligence = WASI Full-2 score; verbal memory = Wide Range Assessment of Memory and Learning
(WRAML) verbal memory index; visual memory = WRAML visual memory index; general memory = WRAML screening memory index.
Copyright 2013 John Wiley & Sons, Ltd.

Appl. Cognit. Psychol. 28: 3038 (2014)

36

F. Jack et al.

on free recall in both eyewitness and autobiographical

memory paradigms. Typically, the amount of information
that individuals report increases with age in childhood and
between childhood and adulthood, although accuracy remains
high across age groups (e.g. Marin et al., 1979; Oates &
Shrimpton, 1991; Sutherland & Hayne, 2001; Tustin &
Hayne, 2010). Although our ndings are highly consistent
with past research, they make a unique contribution by
allowing us to evaluate adolescents performance relative to
that of both younger and older witnesses. The amount of
information reported by adolescents in response to free-recall
prompts was intermediate between the amounts reported by
children and adults, suggesting that eyewitness memory
improves gradually across this developmental window.
Most of the information that participants reported
described the people in the lm clips, their behaviour, and
the surrounding objects and scenery. These were also the
categories in which age differences emerged. Children
reported less information than adults across all three of these
categories. Adolescents reported less than adults about
people and behaviour. These are the categories of information that are likely to be vital in eyewitness situations; it is
therefore important for interviewers and researchers to
consider how we can maximise the amount of information
that young witnesses provide about these aspects of events.
Across all three age groups, few details were reported about
time or sound. This could have been due to the nature of our
target events. It is possible that age differences would have
emerged had there been more scope for describing these
aspects of the events.
What factors might be responsible for the age differences
we observed in the volume of information recalled? From a
social/motivational perspective, older participants might
have been more sensitive to a perceived requirement to perform well in the laboratory or might have felt more at ease
with the adult interviewer than their younger counterparts.
Alternatively, cognitive or linguistic factors could have
played the major role. For example, older participants might
have used more effective encoding or retrieval strategies
than younger participants or been better able to communicate
their recollections (Bjorklund & Douglas, 1997; Lamb et al.,
2007; Saywitz, 2002).
Individual differences
Our analyses indicated that all six of our individual difference
variables were positively related to the amount of correct
information participants reported and that non-verbal intelligence, general intelligence, and general memory were
positively related to the proportion correct. The failure of
the interaction terms to make a signicant contribution to
the regression models indicates that the relations between the
individual difference variables and the eyewitness recall
variables did not differ meaningfully across our three age
groups. Past research indicates that the relations between
eyewitness memory performance and scores on tests of
language and intelligence increase with age in childhood
(Elishberger & Roebers, 2001; Gordon et al., 1993; Roebers
& Schneider, 2001). If this trend continues beyond childhood,
it was not strong enough to be detected in our analyses.
Copyright 2013 John Wiley & Sons, Ltd.

Prior research with children has typically shown that

scores on verbal intelligence tests play a more important role
in predicting eyewitness performance than scores on tests of
non-verbal intelligence. As Table 3 shows, we found no
evidence to support this conclusion.
One reason researchers have examined the relations
between childrens eyewitness performance and individual
differences in cognitive functioning is to inform professionals
who might be asked to evaluate a witnesss evidence in a legal setting (Geddie et al., 2000). Our results suggest that,
across all of the individual difference variables we measured,
standardised test scores may be just as predictive for older
witnesses as they are for children.
Utility of second free recall
In order to elicit as much information as possible under freerecall conditions, we asked participants to provide a second
free-recall account. How successful was this strategy?
Participants of all ages volunteered some new information
during this phase, with children, adolescents, and adults
reporting approximately 6, 7, and 10 correct new details on
average, respectively. As these means indicate, eliciting a
second free-recall account did nothing to decrease the agerelated differences in the overall amount of information
reported; as in the rst free-recall phase, the amount of
new information reported increased with age.
These data suggest that even when witnesses are provided
with both the context reinstatement and report everything
instructions from the outset, they might not report all
available information during their rst attempt. Our results
add to existing evidence that simply asking witnesses to try
again can increase the amount of information obtained under
free-recall conditions.
An important caveat to this conclusion is that the new
information might come at the cost of accuracy. Across all
three of our age groups, participants accuracy decreased
between the rst and second recall phases. This aspect of
our data is consistent with past research. Whether the
subsequent account is given within the same interview
(Boon & Noon, 1994; La Rooy et al., 2005, Experiment 2),
24 hours after the rst (La Rooy et al., 2005, Experiment
1), or over 6 months later (La Rooy et al., 2005, Experiment
3; Peterson et al., 2001; Salmon & Pipe, 2000), the error rate
(as a proportion of new information reported) is typically at
least double that of the original account (although see Bluck
et al., 1999). This trend of decreasing accuracy suggests that
asking participants to provide a second free-recall account
might have an effect comparable with that of asking more
specic questions. Researchers generally nd that as the line
of questioning becomes more specic, the accuracy of
witnesses reports decreases (e.g. Aschermann et al., 1991;
Larsson et al., 2003; Lipton, 1977; see Milne & Bull,
1999, for review). Given that details that witnesses have less
condence in are also less likely to be accurate (Roberts &
Higham, 2002), one possible explanation for this trend of
decreasing accuracy is that when interviewers encourage
witnesses to provide more specic or simply more information, the witnesses might begin to report details that they
had initially refrained from mentioning because they were
Appl. Cognit. Psychol. 28: 3038 (2014)

Age-related difference in free-recall account

less sure of them. Supplementary analyses revealed a small
but signicant trend consistent with this interpretation. Participants indicated uncertainty (e.g. I think) more frequently during the second recall phase relative to the rst,
suggesting that they had less condence in the new information that they provided.
Limitations and future directions
Like all laboratory experiments, our experiment has some
limitations in terms of its ecological validity. First, we asked
participants to watch carefully as we showed them a brief
lmed event in a quiet laboratory setting, and we interviewed
them after a 45-minute delay. Actual witnesses would view
live events with no prior warning, might have their view
obstructed or their attention diverted, and would be
interviewed after a delay that could range from a few minutes
to a week or more, during which they might be exposed to
post-event information from a range of sources. The task of
remembering and describing what happened might therefore
be considerably more difcult for an actual witness than it
was for our participants. Another difference is that witnesses
to an actual transgression might be more or less willing to
provide information depending on factors such as their role in
the situation, whether they know the perpetrator (or suspect),
and whether they fear recrimination. These issues might be
particularly relevant for adolescent witnesses, who tend to
offend in groups (Steinberg, 2004). Finally, our interviewers
asked only free-recall questions; investigators are likely to
ask many different types of questions, including more
specic open-ended or closed questions (Evans & Webb,
1993; Fisher, Geiselman, & Raymond, 1987). Accuracy tends
to suffer as questions become more specic, particularly for
younger witnesses.
With these issues in mind, the present results are likely to
represent a best-case scenario of eyewitness performance.
Further laboratory and eld research is required to assess
the relative effect that the aforementioned factors have on
the eyewitness performance of children, adolescents, and
adults. In our own laboratory, we are currently investigating
age-related differences in eyewitness reports elicited under
less than ideal interview conditions.
ACKNOWLEDGEMENTS
This research was funded by a New Zealand Science &
Technology Postdoctoral Fellowship awarded to Fiona Jack
by the Foundation for Research, Science, & Technology,
and by a grant awarded to Rachel Zajac by the Marsden
Fund Council, from Government funding administered by
the Royal Society of New Zealand.
A portion of the data reported in this article was presented
at the 4th International Congress on Psychology and Law,
Miami FL, 2011. We thank The New Zealand Police for
supplying the lm clips used in this research; Senior Sergeant Nina Westera and Detective Inspector Ross Grantham
for their advice on the preparation of our interview script;
Nicola Davis, Celia Wright, and Karen Quarrell for their assistance with data collection and coding; Mele Taumoepeau,
Dione Healey, and Tamlin Conner for statistical advice;
Copyright 2013 John Wiley & Sons, Ltd.

37

Julien Gross for her helpful comments on an earlier draft of

this manuscript; and all of the children, adolescents, and
adults who participated in this research.
REFERENCES
Aschermann, E., Mantwill, M., & Kohnken, G. (1991). An independent
replication of the effectiveness of the cognitive interview. Applied Cognitive
Psychology, 5, 489495.
Baker-Ward, L., & Ornstein, P. A. (2002). Cognitive underpinnings of
childrens testimony. In H. L. Westcott, G. M. Davies, & R. H. C. Bull
(Eds.), Childrens testimony: A handbook of psychological research
and forensic practice (pp. 2135). Chichester, England: John Wiley &
Sons Ltd.
Bjorklund, D. F., & Douglas, R. N. (1997). The development of memory
strategies. In N. Cowan (Ed.), The development of memory in childhood
(pp. 201246). Hove, England: Psychology Press.
Bluck, S., Levine, L. J., & Laulhere, T. M. (1999). Autobiographical
remembering and hypermnesia: A comparison of older and younger
adults. Psychology and Aging, 14, 671682.
Boon, J., & Noon, E. (1994). Changing perspectives in cognitive
interviewing. Psychology, Crime & Law, 1, 5969.
Brace, N., Kemp, R., & Snelgar, R. (2009). SPSS for psychologists (4th ed.).
Hampshire, UK: Palgrave Macmillan.
Brown, D., & Pipe, M.-E. (2003). Individual differences in childrens event
memory reports and the narrative elaboration technique. Journal of
Applied Psychology, 88, 195206.
Bruck, M., & Melnyk, L. (2004). Individual differences in childrens
suggestibility: A review and synthesis. Applied Cognitive Psychology,
18, 947996.
Chapman, A. J., & Perry, D. J. (1995). Applying the cognitive interview
procedure to child and adult eyewitnesses of road accidents. Applied
Psychology: An International Review, 44, 283294.
Clarke-Stewart, K. A., Malloy, L. C., & Allhusen, V. D. (2004). Verbal ability,
self-control, and close relationships with parents protect children against
misleading suggestions. Applied Cognitive Psychology, 18, 10371058.
Costanzo, P. R., & Shaw, M. E. (1966). Conformity as a function of age
level. Child Development, 37, 967975.
Eisen, M. L., Goodman, G. S., Qin, J., Davis, S., & Crayton, J. (2007).
Maltreated childrens memory: Accuracy, suggestibility, and psychopathology. Developmental Psychology, 43, 12751294.
Eisen, M. L., Qin, J., Goodman, G. S., & Davis, S. L. (2002). Memory and suggestibility in maltreated children: Age, stress arousal, dissociation, and psychopathology. Journal of Experimental Child Psychology, 83, 167212.
Elischberger, H. B., & Roebers, C. M. (2001). Improving young childrens free
narratives about an observed event: The effects of nonspecic verbal
prompts. International Journal of Behavioral Development, 25, 160166.
Elley, W. B., & Irving, J. C. (2003). The ElleyIrving socio-economic
index: 2001 census revision. New Zealand Journal of Educational
Studies, 38, 317.
Engelhardt, L. (1999). The problem with eyewitness testimony: Commentary
on a talk by George Fisher and Barbara Tversky. Stanford Journal of Legal
Studies, 1, 2729.
Evans, G., & Webb, M. (1993). High prolebut not that high prole:
Interviewing of young persons. In E. Shepherd (Ed.), Aspects of police
interviewing (pp. 3745). Leicester, England: British Psychological Society.
Fisher, R. P., Geiselman, R. E., & Raymond, D. S. (1987). Critical analysis of
police interviewing techniques. Journal of Police Science & Administration,
15, 177185.
Fisher, R. P., & Schreiber, N. (2007). Interview protocols to improve
eyewitness memory. In M. P. Toglia, J. D. Read, D. F. Ross, & R. C. L.
Lindsay (Eds.), Handbook of eyewitness psychology: Memory for events
(Volume 1, pp. 5380). Mahwah, NJ: Erlbaum.
Fivush, R. (2002). The development of autobiographical memory. In H. L.
Westcott, G. M. Davies, & R. H. C. Bull (Eds.), Childrens testimony:
A handbook of psychological research and forensic practice (pp. 319).
Chichester, England: John Wiley & Sons Ltd.
Geddie, L., Fradin, S., & Beer, J. (2000). Child characteristics which impact
accuracy of recall and suggestibility in preschoolers: Is age the best
predictor? Child Abuse & Neglect, 24, 223235.
Appl. Cognit. Psychol. 28: 3038 (2014)

38

F. Jack et al.

Giedd, J. N., Blumenthal, J., Jeffries, N. O., Castellanos, F. X., Liu, H.,
Zijdenbos, A., et al. (1999). Brain development during childhood and
adolescence: A longitudinal MRI study. Nature Neuroscience, 2, 861863.
Gordon, B. N., Ornstein, P. A., Nida, R. E., Follmer, A., Crenshaw, M. C.,
& Albert, G. (1993). Does the use of dolls facilitate childrens memory of
visits to the doctor? Applied Cognitive Psychology, 7, 459474.
Greenhoot, A. F., Ornstein, P. A., Gordon, B. N., & Baker-Ward, L. (1999).
Acting out the details of a pediatric check-up: The impact of interview
condition and behavioural style on childrens memory reports. Child
Development, 70, 363380.
Grisso, T., Steinberg, L., Woolard, J., Cauffman, E., Scott, E., Graham, S.,
Lexcen, F., Reppucci, N. D., & Schwartz, R. (2003). Juveniles competence to stand trial: A comparison of adolescents and adults capacities
as trial defendants. Law and Human Behavior, 27, 333363.
Gross, J., & Hayne, H. (1999). Drawing facilitates childrens verbal reports
after long delays. Journal of Experimental Psychology: Applied, 5, 265283.
Gudjonsson, G. H. (2003). The psychology of interrogations and confessions:
A handbook. Chichester, England: Wiley.
Gudjonsson, G. H., & Singh, K. K. (1984). Interrogative suggestibility and
delinquent boys: An empirical validation study. Personality and Individual
Differences, 5, 425430.
Hayne, H., & Jack, F. (2011). Childhood amnesia. Wiley Interdisciplinary
Reviews: Cognitive Science, 2, 136145.
Henry, L. A., & Gudjonsson, G. H. (2007). Individual and developmental
differences in eyewitness recall and suggestibility in children with
intellectual disabilities. Applied Cognitive Psychology, 21, 361381.
Hudspeth, W. J., & Pribram, K. H. (1990). Stages of brain and cognitive
maturation. Journal of Educational Psychology, 82, 881884.
Lamb, M. E., Orbach, Y., Warren, A. R., Esplin, P. W., & Hershkowitz,
I. (2007). Enhancing performance: Factors affecting the informativeness
of young witnesses. In M. P. Toglia, J. D. Read, D. F. Ross, & R. C. L.
Lindsay (Eds.), Handbook of eyewitness psychology: Memory for events.
(Volume 1, pp. 429451). Mahwah, NJ: Erlbaum.
La Rooy, D., Pipe, M.-E., & Murray, J. E. (2005). Reminiscence and
hypermnesia in childrens eyewitness memory. Journal of Experimental
Child Psychology, 90, 235254.
Larsson, A. S., Granhag, P. A., & Spjut, E. (2003). Childrens recall and the
cognitive interview: Do the positive effects hold over time? Applied
Cognitive Psychology, 17, 203214.
Lee, K. (2004). Age, neuropsychological, and social cognitive measures as
predictors of individual differences in susceptibility to the misinformation
effect. Applied Cognitive Psychology, 18, 9971019.
Lipton, J. P. (1977). On the psychology of eyewitness testimony. Journal of
Applied Psychology, 62, 9095.
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, 93107.
Marin, B. V., Holmes, D. L., Guth, M., & Kovac, P. (1979). The potential of
children as eyewitnesses: A comparison of children and adults on eyewitness
tasks. Law and Human Behavior, 3, 295306.
McGuire, K., London, K., & Wright, D. B. (2011). Peer inuence on event
reports among adolescents and young adults. Memory, 19, 674683.
McMurtrie, J. (2007). Incorporating Elizabeth Loftuss research on memory
into reforms to protect the innocent. In M. Garry & H. Hayne (Eds.), Do
justice and let the sky fall: Elizabeth F. Loftus and her contributions to
science, law, and academic freedom (pp. 171191). Mahwah, NJ: Erlbaum.
Melnyk, L. Crossman, A. M., & Scullin, M. H. (2007). The suggestibility of
childrens memory. In M. P. Toglia, J. D. Read, D. F. Ross, & R. C. L.
Lindsay (Eds.), Handbook of eyewitness psychology: Memory for events
(Volume 1, pp. 401427). Mahwah, NJ: Erlbaum.
Memon, A. (2006). The cognitive interview. In O. Hargie (Ed.), The handbook
of communication skills (3rd ed, pp. 531550). London: Routledge.
Memon, A., Wark, L., Holley, A., Bull, R., & Koehnken, G. (1997). Eyewitness
performance in cognitive and structured interviews. Memory, 5, 639656.
Milne, R., & Bull, R. (1999). Investigative interviewing: Psychology and
practice. Chichester, UK: Wiley.
Milne, R., & Bull, R. (2002). Back to basics: A componential analysis of the
original cognitive interview mnemonics with three age groups. Applied
Cognitive Psychology, 16, 743753.

Copyright 2013 John Wiley & Sons, Ltd.

Moulin, C. J. A., Thompson, R. G., Wright, D. B., & Conway, M. A. (2007).

Eyewitness memory in older adults. In M. P. Toglia, J. D. Read, D. F.
Ross, & R. C. L. Lindsay (Eds.), Handbook of eyewitness psychology:
Memory for events (Volume 1, pp. 627646). Mahwah, NJ: Erlbaum.
Oates, K., & Shrimpton, S. (1991). Childrens memories for stressful and
non-stressful events. Medicine, Science & the Law, 31, 410.
Paus, T. (2005). Mapping brain maturation and cognitive development
during adolescence. Trends in Cognitive Science, 9, 6068.
Peterson, C., Moores, L., & White, G. (2001). Recounting the same events
again and again: Childrens consistency across multiple interviews.
Applied Cognitive Psychology, 15, 353371.
Pipe, M.-E., Thierry, K. L., & Lamb, M. E. (2007). The development of event
memory: Implications for child testimony. In M. P. Toglia, J. D. Read, D. F.
Ross, & R. C. L. Lindsay (Eds.), Handbook of eyewitness psychology: Memory for events (Volume 1, pp. 453478). Mahwah, NJ: Erlbaum.
Redlich, A. D., Silverman, M., & Steiner, H. (2003). Pre-adjudicative and
adjudicative competence in juveniles and young adults. Behavioral
Sciences and the Law, 21, 393410.
Richardson, G., Gudjonsson, G. H., & Kelly, T. P. (1995). Interrogative
suggestibility in an adolescent forensic population. Journal of Adolescence,
18, 211216.
Roberts, W. T., & Higham, P. A. (2002). Selecting accurate statements from
the cognitive interview using condence ratings. Journal of Experimental
Psychology: Applied, 8, 3343.
Roebers, C. M., & Schneider, W. (2001). Individual differences in childrens
eyewitness recall: The inuence of intelligence and shyness. Applied
Developmental Science, 5, 920.
Salmon, K., & Pipe, M.-E. (2000). Recalling an event one year later: The
impact of props, drawing and a prior interview. Applied Cognitive
Psychology, 14, 99120.
Salmon, K., Roncolato, W., & Gleitzman, M. (2003). Childrens reports of
emotionally laden events: Adapting the interview to the child. Applied
Cognitive Psychology, 17, 6579.
Santor, D. A., Messervey, D., & Kusumakar, V. (2000). Measuring peer
pressure, popularity, and conformity in adolescent boys and girls:
Predicting school performance, sexual attitudes, and substance abuse.
Journal of Youth and Adolescence, 29, 163182.
Saywitz, K. J. (2002). Developmental underpinnings of childrens testimony.
In H. L. Westcott, G. M. Davies, & R. H. C. Bull (Eds.), Childrens
testimony: A handbook of psychological research and forensic practice
(pp. 319). Chichester, England: John Wiley & Sons Ltd.
Saywitz, K. J., Geiselman, R. E., & Bornstein, G. K. (1992). Effects of
cognitive interviewing and practice on childrens recall performance.
Journal of Applied Psychology, 77, 744756.
Scrivner, E., & Safer, M. A. (1988). Eyewitnesses show hypermnesia for
details about a violent event. Journal of Applied Psychology, 73, 371377.
Sheslow, D., & Adams, W. (2003). Wide Range Assessment of Memory and
Learning (2nd ed.). Lutz, FL: Psychological Assessment Resources.
Singh, K., & Gudjonsson, G. (1992). The vulnerability of adolescent boys to
interrogative pressure: An experimental study. Journal of Forensic
Psychiatry, 3, 167170.
Smith, E. E., & Jonides, J. (1999). Storage and executive processes in the
frontal lobes. Science, 283, 16571661.
Sowell, E. R., Thompson, P. M., Holmes, C. J., Jernigan, T. L., & Toga,
A. W. (1999). In vivo evidence for post-adolescent brain maturation in
frontal and striatal regions. Nature Neuroscience, 2, 859861.
Steinberg, L. (2004). Risk taking in adolescence: What changes, and why?
Annals of the New York Academy of Sciences, 1021, 5158.
Sutherland, R., & Hayne, H. (2001). Age-related changes in the misinformation
effect. Journal of Experimental Child Psychology, 79, 388404.
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics
(5th ed.). Boston, MA: Pearson.
Tustin, K., & Hayne, H. (2010). Dening the boundary: Age-related changes in
childhood amnesia. Developmental Psychology, 46, 10491061.
Wechsler, D. (1999). Wechsler Abbreviated Scale of Intelligence. San
Antonio, TX: Harcourt Assessment.
Wells, G. L., & Windschitl, P. D. (1999). Stimulus sampling and social
psychological experimentation. Personality and Social Psychology Bulletin,
25, 11151125.

Appl. Cognit. Psychol. 28: 3038 (2014)