J Police Crim Psych (2012) 27:55–62

DOI 10.1007/s11896-011-9089-8

The Culprit in Target-Absent Lineups: Understanding

Young Children’s False Positive Responding
Joanna D. Pozzulo & Julie Dempsey & Kaila Bruer &
Chelsea Sheahan

Published online: 26 March 2011

# Springer Science+Business Media, LLC 2011

Abstract Young children (4- to 7-years old; N=59) and Mistaken eyewitness identification has been found to be
adults (N=53) were shown a series of targets that were the leading cause of wrongful conviction (www.innocence
either familiar (i.e., popular cartoon characters) or unfamiliar project.org). Research has shown that under some condi-
(unknown human faces) to assess whether children’s false tions, child and adult eyewitnesses differ with their identi-
positive responding with target-absent lineups is driven by fication accuracy (Pozzulo and Warren 2003; Pozzulo and
social factors to a greater degree than cognitive factors. Lindsay 1998). Specifically, children and adults produce
Although children were able to produce correct identification comparable correct identification rates when shown a target-
rates with virtually 100% accuracy for the cartoon characters, present lineup (i.e., the culprit is in the photo array; Pozzulo
they produced a significantly lower correct rejection rate and Lindsay 1998). The problem occurs when the culprit is
compared to adults. Children also produced a significantly absent from the lineup; children are more likely to pick out
lower correct rejection rate for the human faces compared to an innocent person than adults (Pozzulo and Lindsay 1998;
adults. These data are discussed for understanding children’s Zajac and Karageorge 2009). The cause or explanation for
identification evidence. this differential has yet to be fully delineated, however a
combination of social and cognitive factors are likely at play.
Keywords Child . Eyewitness . Lineup . Identification . It is assumed that social factors (in addition to cognitive
Face . Recognition factors) drive false positive responding with target-absent
lineups (e.g., Parker and Ryan 1993; Pozzulo and Lindsay
1998). The present study examined whether the pattern in
Many crimes, such as thefts, vandalisms, abductions, and
correct identification and false positive responding would
sexual assaults, can occur where the only eyewitness (or
remain when manipulating the familiarity of the target; hence
victim) is a child (Gross and Hayne 1996; Zajac and
the cognitive and social demands of the lineup task.
Karageorge 2009). While some of these cases involve a
culprit that is familiar to the eyewitness, there also are cases
where a child is asked to identify a person that is
Children’s Versus Adults’ Identification Abilities
unfamiliar, perhaps only seen during the commission of
the crime. Identification evidence can be extremely persua-
In a meta-analysis examining children’s and adult’s identi-
sive, however it may be inaccurate (Wells et al. 1998).
fication abilities with lineups, Pozzulo and Lindsay (1998)
found that children as young as 5-years produced compa-
rable correct identification rates to adults, provided the
This research was supported by a grant from the Social Sciences and target is among the photos presented. Although very few
Humanities Research Council of Canada to the first author.
lineup studies examined the abilities of children under
J. D. Pozzulo (*) : J. Dempsey : K. Bruer : C. Sheahan 5-years, this age group seemed to have greater difficulty
Department of Psychology, Carleton University,
than adults at correct identification (Pozzulo and Lindsay
1125 Colonel by Drive,
Ottawa, Ontario K1S 5B6, Canada 1998). With target-absent lineups, younger and older
e-mail: Joanna_Pozzulo@carleton.ca children were more likely than adults to produce a higher
56 J Police Crim Psych (2012) 27:55–62

false positive rate (i.e., lower correct rejection rate). All these this response option was made salient. Thus, children’s
data involve the identification of a “stranger” or someone higher false positive rates compared to adults may occur
unfamiliar to the participant/witness. It is likely that both because of a greater sense to make a selection/identification
social and cognitive factors are responsible for lineup when shown a lineup (e.g., Parker and Ryan 1993; Pozzulo
identification accuracy. It has been suggested that cognitive and Lindsay 1998).
factors are more likely to drive decision processes with a A feeling of pressure may make children less likely than
target-present lineups (i.e., the need to match a previously adults to reject a lineup (Beal et al. 1995). Although adults,
formed memory to a current image), whereas, social factors too, may perceive pressure to make an identification,
are more likely to drive decision processes with a target- children are likely to perceive greater pressure than adults,
absent lineup (i.e., rejecting the images when the task resulting in a lower rejection rate. For example, Pozzulo
requires the witness to make a selection; Parker and Ryan and Dempsey (2006) found that with biased lineup
1993; Pozzulo and Lindsay 1998, 1999). It should be noted instructions, (i.e., instructions that do not explicitly state
that both cognitive and social factors are at play with a an option to reject the lineup) children had a higher rate of
lineup task, regardless of whether the target is present. false positives compared to adults. In the same study,
However, as stated above, social or cognitive factors may children also had a higher rate of false positive responding
exert a greater influence depending on the presence of the compared to adults when neutral/non-biased lineup instruc-
target. tions were presented. Thus, the researchers found that a
manipulation known to increase false positives in adults
(e.g., Malpass and Devine 1981; Steblay 1997), also
Are Social Factors Likely Driving Children’s increased false positives in children. Moreover, the propor-
Target-Absent Responding? tion increase in false positive responding between children
and adults remained constant across neutral and biased
Beal et al. (1995) suggested that identification errors made instructions. In contrast, correct identification rates seem
by children may not be primarily a result of poor memory unaffected by pressure (e.g., Malpass and Devine 1981).
of the events but rather could reflect other factors such as Correct identification rates (target-present lineups) are not
social pressure. Raskin and Yuille (as cited in Ricci et al. influenced negatively by pressure because children feel the
1996) suggested that the mere presentation of a lineup may need to make a selection, see the target, and choose the
suggest to the child that the presenter expects the child to target. Correct rejection rates (target-absent lineups) are
make a choice (i.e., why else would a lineup be shown?). influenced because the child feels pressure to make a
Further, Wells and Luus (1990) likened the lineup task to a selection but does not see the target so another individual is
social psychology experiment. Just as there are social selected. If children’s false-positive responding with a
demands that are experienced by the participant in an target-absent lineup is driven by the social demands of the
experimental task so too does the witness experience task to a greater degree than cognitive demands, even with
similar demands when examining a lineup. The mere a “lower cognitive demand” lineup task (i.e., familiar
presentation of a lineup suggests a “selection” is being target), children should produce a higher false positive rate
requested. Making no selection (or rejecting the lineup) compared to adults.
may be viewed as a “non response” and a participant not A caveat is necessary in that we are not suggesting
willing to complete the task. For example, both participant that eyewitness identification can be neatly split into two
and witness may want to please the experimenter/police crude categories, “social” and “cognitive”. Certainly,
officer by choosing the “right person”. The participant may identification involves both social and cognitive processes
guess at the experimenter’s hypothesis. The witness may and this interplay will vary depending on the various factors
guess who the police suspect and who the officer wants the present at the time of task, factors present at encoding,
witness to choose. Moreover, the social demands associated factors inherent in the individual eyewitness, etc. Rather,
with a lineup task may be more pronounced for the younger we have used the terms “social” and “cognitive” as labels
child. For example, the experimenter/police officer is an for groups of factors and processes where one group of
authority figure who is older than the child. There is an factors and processes may be more prevalent under certain
implicit demand to make a selection when shown a lineup; conditions.
why else would you be shown a lineup if not to pick
someone out. The child may worry about getting into
trouble if no selection is made because this behaviour Current Study
would suggest non-compliance. For example, Pozzulo and
Lindsay (1997) found that children were less likely to use We define a “lower cognitive demand lineup task” as one
an “I don’t know” response compared to adults even when where the correct identification rate would be at approxi-
J Police Crim Psych (2012) 27:55–62 57

mately 100%. In such a case, false positives would have to Predictions

be driven by “social factors” to a greater degree than
“cognitive factors”, i.e., accurate identification is at ceiling Children and adults were predicted to have a similar rate of
with the target-present lineup so an error in the target- correct identification for cartoon faces at approximately
absent lineup must be a result of social pressure more so 100% accuracy. Given the age of our participants and the
than cognitive factors. Using a within-subjects design limited research available with this population, we pre-
would be powerful to eliminate explaining away results as dicted that children would produce a lower correct
a function of individual differences. identification rate for human faces compared to adults
It is predicted that young children and adults will (see Pozzulo and Lindsay 1998). Children were predicted to
produce comparable correct identification rates (with a have a lower correct rejection rate for both the cartoon faces
lower cognitive demand lineup). Assuming that errors, i.e., and the human faces compared to adults.
false positive responding, are more so driven by social
factors than cognitive factors in the target-absent condition,
it is predicted that young children will produce a higher Method
false positive rate than adults, regardless of the cognitive
demands of the lineup task. Participants
According to the Media Awareness Network (1999), in
the United States, it was estimated that in 1998, pre- Young children (N=59; age range 4- to 7-years, M=4.98
schoolers watched an average of 2.6 hours of television a years, SD=0.82; 21 females and 38 males) were recruited
day. In addition, children, ages 2- to 11-years, watch on from pre-kindergarten/kindergarten classes from three private
average 17.7 hours of television per week. Similarly, in schools in Eastern Ontario, Canada.
Canada, children watch, on average, 16.8 hours of Adult participants (N=53; age range, 17- to 30-years,
television a week. A study of children’s television habits M=20.54 years, SD=3.34; 36 females and 17 males) were
revealed that the majority of children’s programming airing recruited from the Introductory Psychology Participant
were animated programs (Atwal et al. 2003). In addition, Pool from an Eastern Ontario university.
this study found that the number of hours devoted to
animation and preschool programming has dramatically Design
increased from 1996 to 2001 (Atwal et al. 2003).
Using cartoon characters that young children are familiar A 2 (age; young children vs. adult) × 2 (target; cartoon vs.
with and could correctly identify at approximately 100% human) × 2 (lineup type; target-present vs. target-absent)
would indicate a familiar target and a relatively “easy” (i.e., mixed factorial design was used.
lower) cognitive task. If children produce near perfect
correct identification rates but go on to produce signifi-
cantly lower correct rejection rates (target-absent lineups) Materials
than adults, these data would provide strong evidence that
social demands to make a selection drives children’s higher Demographic and Cartoon Watching Form Each partici-
false positive rate compared to adults. It is important to pant (for child participants this form was provided to the
compare within the same study and across the same parent/guardian) was provided with a response form for
participants the identification patterns using human faces demographics and to assess level of familiarity with the
as well as cartoon faces. Using human face stimuli provides target cartoons used in this study. Eight questions were
a “baseline” of sorts to better understand the influence of asked requesting the participant’s age, gender, primary
unfamiliar (i.e., human faces) versus familiar (i.e., cartoon language, ethnicity, number of children in the household
faces) stimuli. Moreover, these data could be compared to and their ages, amount of time spent watching cartoons per
other literature that uses human face stimuli. week, and how much time spent watching the two target
In the current study, young children (4- to 7-years-old) cartoons used in this study (i.e., Dora the Explorer and Go
and adults were shown a set of clips from two popular Diego Go). Adult participants completed the form them-
cartoon shows (Dora the Explorer and Go Diego Go) and selves. For child participants, the form was completed by
two clips of human actors engaged in everyday tasks (a their parent/guardian.
female combing her hair and a male putting on his coat).
All participants viewed all targets, however, presence/ Human Face Targets One female and one male Caucasian
absence of the target in the lineup task was randomized university student (each 22-years-old) were used as targets.
across participants. Correct identification and correct Each human target was filmed completing an everyday task
rejection was measured for each target. for a six-second video clip; a female brushing her hair in
58 J Police Crim Psych (2012) 27:55–62

the bathroom and a male putting on his coat and exiting his target along with three other foils were presented. In the
home. Each video provided a 2- to 3-second close-up of the target-absent condition, the target was replaced with a
individual’s face. The target videos were filmed in colour. similar foil. The target and target replacement was placed in
the same position. Each target’s lineup position however
Human Face Foils Each human target was photographed in was randomized. Each lineup also included a silhouette to
a different outfit than what was worn during the video clip. represent the possibility of an absent target.
The foils were selected from a pool of 90 female faces and Each participant saw four videos in a random order.
90 male faces. The foil photographs were selected based on Participants were shown one photoarray, following each
similar appearance to the intended target. Similarity was video, where the position of the target/replacement was
measured in terms of general facial structure, hair length, counterbalanced across photoarrays. Videos and photoarrays
and colour. Three raters selected the 4 foils for each target. were displayed on 13-inch laptop screens using the Microsoft
Targets and foils were closely cropped such that their face, PowerPoint program.
neck and the tops of their shoulders were photographed
(similar to the cartoon foils). Target-present lineups Instructions for Lineup Identification The following instruc-
contained the target and three foils. Target-absent lineups tions were provided prior to the display of each lineup:
contained four foils. All photos were in black and white. “Please look at the photos. The person/cartoon from the
video may or may not be here. If you see the person/cartoon
Cartoon Targets One female and one male cartoon charac- please point to the photo. If you do not see the person/
ter were used as the targets. Six-second clips of each of the cartoon, please point to this box (indicating)”.
following were used; Dora the Explorer talking to the
audience and Go Diego Go putting on a pair of gloves for Lineup Administrators Three female experimenters showed
safety. Each video provided a 2- to 3-second close-up of the children the video clips and photoarrays. As “dress” has
target character’s face and involved no other characters. been found to influence children’s choosing behaviour in
The video clips were in colour. The sound on the videos lineup tasks (see Lowenstein et al. 2010) possibly through
was muted as there was no sound with the human face social pressure, the experimenters wore professional-casual
videos. clothing (e.g., sweater/blouse and dress-pants) that would
reduce external visual cues of authority. More specifically,
Cartoon Foils The foils were selected from a vast number experimenters were “neat” in appearance but not overly
of readily available cartoon images on the internet. The foil formal (e.g., no uniforms or lab coats).
photographs were selected based on similar appearance to
the intended target. Similarity was measured in terms of Free Recall Descriptions All participants were asked an
general facial structure, hair length, and colour. Three raters open-ended question to describe everything they could
judged approximately 10 photographs for each target. The remember about each video clip. The researchers recorded
four cartoons receiving the highest similarity ratings (most each child participant’s responses, while the adult partic-
similar to least similar) were selected. Most cartoon ipants recorded their own responses. This task was used as
characters were displayed in similar clothes across different a brief filler between exposure of the video and presentation
videos. For this reason, cartoons depicted in the photoarrays of the lineup. Approximately 2 minutes lapsed between
were displayed closely cropped to the target’s face (from each video exposure and lineup presentation.
the top of the shoulders) to reduce the appearance of any
clothing worn. In order to compensate for the strong and Procedure
often vibrant colours unique to the cartoon images, all the
photographs were displayed in black and white. Also, black Young Children Parents/guardians of the children attending
and white images reduced the possibility that the bright and private schools in the community were supplied with a
vibrant colours would be the focus of recognition rather written consent form, as well as a demographics sheet. The
than the identity of the target. Target-present lineups demographic sheet was to be completed by the parent/
contained the target and three foils. Target-absent lineups guardian in order to ensure children were familiar with the
contained four foils. target cartoons. Upon receiving written consent forms and
completed demographic forms, three female experimenters
Lineup Presentation For each target, a lineup was pre- and one female facilitator arrived at each private school.
sented. A simultaneous procedure was used to present the Only children with consent were invited to participate. The
lineup. That is, for each lineup, all pictures were shown at researchers were introduced to the students as a group from
once. For the target-present conditions, a photograph of the the university doing a project on TV shows and computer
J Police Crim Psych (2012) 27:55–62 59

games. During the introduction and invitation to participate, completed, the experimenter displayed the corresponding
the researchers made it clear to the children that they could lineup (via PowerPoint) on a laptop to the participant. The
change their minds at any time and not get into trouble. In experimenter asked the participant to identify the cartoon/
order to create a level of comfort with the children, the person they saw in the video if he/she was present by
researchers worked with the children to make some crafts indicating their selection on a matching sheet. The
prior to engaging the children in the experimental task. experimenter informed the participant that the person they
Experimenters tested children individually. Children were saw may not be there and demonstrated that, in this case,
monitored for fatigue, anxiety, and stress. Each child was the participants should select the option that corresponds to
told that he/she would be watching some videos of people the silhouetted photograph in each lineup. Following the
doing different things. The child was told to pay attention identification, the procedure was repeated for the additional
because, following the video, they would be asked some three videos, each time reminding the participants that the
questions and shown some pictures. person they are looking for may not be in the lineup.
Once the child was comfortable, the experimenter played Following completion of the videos and lineups, the partic-
the first video (human or cartoon). Following the video ipants were given a demographic questionnaire assessing their
clip, the experimenter asked the child one free recall familiarity with the cartoons shown. Finally, the participants
question about what they remembered about the video, were debriefed and thanked for their participation.
i.e., “What did the cartoon character/person look like?”.
Following the child’s response, the experimenter asked, a
non-specific, probing question twice, i.e., “Do you remember Results
anything else?”. If the child offered no response to the initial
question, the experimenter, again, asked, “Do you remember Data were divided into target-present versus target-absent
anything from the video?”. After recording the information lineups given the identification decision differs for each.
provided by the child, the experimenter displayed the Specifically, correct identifications (i.e., selecting a face)
corresponding lineup (in PowerPoint) on a laptop to the child. versus correct rejections (i.e., not selecting a face) may be
The experimenter asked the child to identify the cartoon/ driven by different processes. In order to test the study’s
person they saw in the video by pointing. The experimenter hypotheses it was necessary to separate the data. Pozzulo
instructed the child that the person they saw may or may not and Lindsay (1998) have suggested that target-present
be there and demonstrated that, if the correct person was not lineup decisions are driven by cognitive processes more
there, they should point to the silhouetted box. The so whereas target-absent decisions are driven by cognitive
experimenter recorded the child’s response. Following the as well as social processes.
identification, the procedure was repeated for the additional
three videos, each time reminding the children that the Target-Present Lineups
cartoon/person they are looking for may not be in the lineup.
Following the end of the study, the children were thanked and Young Children
given a small token (i.e., crayons and colouring book). The
facilitator was responsible for entertaining the children while Human Faces Versus Cartoon Faces Our goal was to
they waited to complete the experimental task. average the correct identification rate for the human faces
per child in order to produce a stabilized correct identifi-
Adults Upon entering the laboratory, each participant was cation rate per child. We wanted to eliminate target specific
given a short introduction to the study and provided with a peculiarities. We calculated a mean correct identification
consent form that explained they would be participating in rate for human faces per child. Following this same logic,
a study about memory. Following the signing of the we calculated a mean correct identification rate for cartoon
consent, the participants were told they would be watching faces per child. See Table 1 for individual targets’ correct
some short video clips. The participants also were asked to identification rates (target-present lineups) as a function of
pay attention because following the video they would be age and stimuli.
asked some questions and shown some pictures.
After the first video, the participants were provided with Human faces. Correct identification rates were aver-
a sheet asking a free recall question, “What did the cartoon aged to produce an overall correct identification rate
character/person look like?”. This question was followed up for human faces of .23.
with, “Do you remember anything else about the cartoon Cartoon faces. Correct identification rates were aver-
character/person.” The participant then wrote down all they aged to produce an overall correct identification rate
could remember about what they saw on the video. Once for cartoon faces of .99.
60 J Police Crim Psych (2012) 27:55–62

Table 1 Identification rates for target-present lineups (n) as a function Human versus cartoon faces. Adults were significantly
of age and stimuli
more accurate with cartoon faces (.95) versus human
Target-present lineups Correct identification rate (n) faces (.66), X2 (1, N=103)=11.25, p=.001.

Age
Children
Young Children Versus Adults As predicted, young children
Cartoons
and adults produced a comparable correct identification
Dora-Target 1 (29)
rate for cartoon characters (.99 vs. 95), X2 (1, N=110)=.39,
p=.53.
Foil 0 (0)
As predicted, young children compared to adults
False rejection 0 (0)
produced a significantly lower rate of correct identification
Diego-Target .97 (28)
for human faces (.23 vs. .66). X 2 (1, N = 168) = 18.83,
Foil 0 (0)
p = .001.
False rejection .03(1)
Human Faces
Target-Abesent Lineups
Female-Target .24 (7)
Foil .38 (11)
Young Children
False rejection .38 (11)
Male-Target .21 (6)
Human Faces Versus Cartoon Faces The same logic and
Foil .45 (13)
analyses were followed for target-absent data as was
False rejection .34 (10) followed for target-present data. See Table 2 for individual
Adults targets’ correct rejection rates (target-absent lineups) as a
Cartoons function of age and stimuli.
Dora-Target 1 (25)
Foil 0 (0) Human faces. Correct rejection rates were averaged to
False rejection 0 (0) produce an overall correct rejection rate for human
Diego-Target .89 (24)
faces of .45.
Foil 0 (0)
Cartoon faces. Correct rejection rates were averaged to
False rejection .11 (3)
produce an overall correct rejection rate for cartoon
faces of .74.
Human Faces
Human versus cartoon faces. Young children were
Female-Target .46 (13)
significantly more accurate with a higher correct
Foil 0 (0)
rejection rate with cartoon faces (.74) versus human
False rejection .54 (15)
faces (.45), X2 (1, N=114)=7.66, p=.01.
Male-Target .85 (22)
Foil .15 (4)
Adults
False rejection 0 (0)

Human faces. Correct rejection rates were averaged to

produce an overall correct rejection rate for human
Human versus cartoon faces. Young children were faces of .70.
significantly more accurate with a higher correct Cartoon faces. Correct rejection rates were averaged to
identification rate for cartoon faces (.99) versus human produce an overall correct rejection rate for cartoon
faces (.23), X2 (1, N=116)=66.10, p=.001. faces of .94.
Human versus cartoon faces. Adults were significantly
Adults more accurate with a higher correct rejection rate
for cartoon faces (.94) versus human faces (.70), X2
The same logic and analyses used with child participants (1, N=106)=9.80, p=.01.
was used for adult participants.
Human faces. Correct identification rates were aver- Young Children Versus Adults As predicted, young children
aged to produce an overall correct identification rate produced a significantly lower correct rejection rate than
for human faces of .66. adults for cartoon faces (.74 vs. .94), X2 (1, N=114)=7.66,
Cartoon faces. Correct identification rates were aver- p=.01. Also as predicted, young children produced a
aged to produce an overall correct identification rate significantly lower correct rejection rate than adults for
for cartoon faces of .95. human faces (.45 vs. .70). X2 (1, N=112)=5.70, p=.02.
J Police Crim Psych (2012) 27:55–62 61

Table 2 Correct rejection rates for target-absent lineups (n) as a When examining children’s performance with the human
function of age and stimuli
targets, the predicted pattern was observed. Young children
Target-absent lineups Correct rejection rate (n) (M=4.98 years) produced a lower correct identification rate
as well as a lower correct rejection rate compared to adults.
Age This pattern is consistent with previous work comparing
Children children’s and adults’ correct identification and correct
Cartoons rejection rates (Pozzulo and Lindsay 1998). Pozzulo and
Dora (correct rejection) .80 (24) Lindsay (1998) found that children under 5 years of age
Incorrect identification .20 (6) produced a lower correct identification rate than adults. As
Diego (correct rejection) .67 (20) well, Pozzulo and Lindsay (1998) found that children under
Incorrect identification .33 (10) 5 years of age produced a lower correct rejection rate than
Human Faces adults. In the current study, the correct identification and
Female-Target (correct rejection) .47 (14) rejection rate between children and adults was predicted
Incorrect identification .53 (16) and found (for unfamiliar human faces).
Male-Target (correct rejection) .43 (13) The main objective of the study was to better
Incorrect identification .57 (17) understand children’s false positive responding. By pro-
Adults viding a target(s) that children would be familiar with and
Cartoons could identify at approximately 100%, errors (i.e., a lower
Dora (correct rejection) .96 (27) correct rejection rate) in the target-absent condition would
Incorrect identification .04 (1) more likely be a result of social factors (e.g., an
Diego (correct rejection) .92 (24) expectation to make a selection) than cognitive factors,
Incorrect rejection .08 (2) of course cognitive factors also are necessary for correct
Human Faces rejection. Indeed, children and adults were able to
Female-Target (correct rejection) .72 (18) correctly identify the familiar targets (i.e., cartoon characters)
Incorrect rejection .28 (7) at approximately 100%. In the target-absent conditions
Male-Target (correct rejection) .67 (18) however, children produced a significantly lower correct
Incorrect rejection .33 (9) rejection rate for the familiar cartoon targets compared to
adults. Children were more likely to choose an incorrect
cartoon character than to reject the lineup compared to
adults. This same pattern was observed for the human
faces as well. These data are suggestive that children are
Discussion likely to make an error in the target-absent condition due
to an expectation to the “social demands” to make a
The present study examined whether young children’s false selection rather than due to faulty memory (e.g., memory
positive responding in target-absent lineups was driven was approximately 100%).
more so by the social versus cognitive demands of the task.
Correct identifications with target-present lineups also were Limitation
considered between young children and adults. Young
children were given a set of familiar targets (i.e. cartoon It is important to recognize that the two levels of the
characters) where identity (i.e., cognitive factors) was not at experiment, that is cartoon faces versus human faces,
issue. If the correct identification rate (target-present line- present, in an absolute sense a confound. We must
ups) for these characters was at approximately 100%, a acknowledge the possible influence of unknown factors
lower correct rejection rate (target-absent lineups) would inherent in the processing of the cartoon faces versus
suggest a greater reliance on social factors driving false human faces. Several studies demonstrate that the process-
responding than cognitive factors. These “familiar” targets ing of drawings from that of photographs differ in
were contrasted to “unfamiliar” targets, that is, human faces numerous ways (Davies et al. 1978; Stevenage 1995;
who were unknown to all the participants. Adults were Tversky and Baratz 1985). For the current study, it was
shown the same targets as the children to compare correct important to use a set of stimuli that were known to be
identification and rejection rates. It is important to include easily recognized by children (drawings) and then to
both types of targets in the same study and to use a within- contrast these stimuli with photographs. The results
subjects design so that a “baseline” could be established to presented in this paper are likely not influenced by any
interpret differences in identification patterns and across putative factors delineated in the literature demonstrating
other literature. processing differences for drawings versus photographs.
62 J Police Crim Psych (2012) 27:55–62

Implications simultaneous and elimination lineup procedures. Am J Forensic

Psychol 26:67–81
Gross J, Hayne H (1996) Eyewitness identification by 5-to 6-year-old
The knowledge that children’s false positive responding has children. Law Hum Behav 20:359–373
a strong “social” component representing a critical finding Lowenstein JA, Blank H, Sauer JD (2010) Uniforms affect the
for understanding children’s identification evidence. As has accuracy of children’s eyewitness identification decisions. J of
Investigative Psychol and Offender Profiling 7:59–73
been echoed in much of the recall literature for children’s
Malpass RS, Devine PG (1981) Eyewitness identification: lineup
memory for events, that with non-leading and appropriate instructions and the absence of the offender. J Appl Psychol
interview techniques children can recall much that is 66:482–489
forensically relevant (Ceci and Bruck 1993), so too can Media Awareness Network, The Culture Statistics Program. (1999).
Statistics on TV viewing habits (1994-2000). Retrieved from http://
children identify accurately when the social demands of the www.media-awareness.ca/english/resources/research_documents/
lineup task are reduced. A substantial literature is now statistics/television/tv_viewing_habits.cfm
mounting that social factors, such as lineup instructions Parker JF, Ryan V (1993) An attempt to reduce guessing behavior in
(Pozzulo and Lindsay 1999), clothing worn by the children’s and adult’s eyewitness identifications. Law Hum
Behav 17:11–26
experimenter (Lowenstein et al. 2010), and type of lineup
Pozzulo JD, Balfour J (2006) The impact of change in appearance on
administration (Pozzulo and Lindsay 1999) can influence children’s eyewitness identification accuracy: comparing simul-
children’s false positive responding. To obtain the most taneous and elimination lineup procedures. Legal Criminol
accurate identification evidence from child witnesses, Psychol 11:25–34
Pozzulo JD, Dempsey J (2006) Biased lineup instructions: examining
understanding the social factors at play and then addressing
the effect of pressure on children’s and adults’ eyewitness
them effectively is necessary. Lineup administrators must identification accuracy. J Appl Soc Psychol 36:1381–1394
take care to use procedures that reduce the social pressure Pozzulo JD, Lindsay RCL (1997) Increasing correct identifications by
associated with the lineup task. children. Expert Evid 5:126–132
Pozzulo JD, Lindsay RCL (1998) Identification accuracy of children
One such lineup procedure that was created with the goal versus adults: a meta-analysis. Law Hum Behav 22:549–570
of reducing social demands with lineup presentation is the Pozzulo JD, Lindsay RCL (1999) Elimination lineups: an improved
elimination lineup by Pozzulo and Lindsay (1999). With identification procedure for child eyewitnesses. J Appl Psychol
the elimination lineup procedure, children are asked to 84:167–176
Pozzulo JD, Warren KL (2003) Descriptions and identifications of
make two judgments; first, to select the person most similar
strangers by youth and adult eyewitnesses. J Appl Psychol
to the target and second, to determine whether this most 88:315–323
similar person is in fact the target. The empirical data for Pozzulo JD, Dempsey J, Corey S, Girardi A, Lawandi A, Aston C
this procedure with different aged children (preschoolers, (2008) Can a lineup procedure designed for child witnesses work
for adults: comparing simultaneous, sequential, and elimination
older children, adults) under varying conditions (multiple
lineup procedures. J Appl Soc Psychol 38:2195–2209
perpetrators, change of appearance) is very positive and Pozzulo JD, Dempsey J, Crescini C (2009) Preschoolers’ person
robust (Dempsey and Pozzulo 2008; Pozzulo et al. 2008, description and identification accuracy: a comparison of the
2009, 2010; Pozzulo and Balfour 2006). Reliable identifi- simultaneous and elimination lineup procedures. J Appl Dev
Psychol 30:667–676
cation evidence is possible from child eyewitnesses (as
Pozzulo JD, Dempsey JL, Clarke C (2010) Can the elimination lineup
reliable as adult identification evidence) when appropriate procedure overcome lineup bias: a comparison of procedures.
procedures are used. Psychiatry, Psychology, and Law 17:32–38
Ricci CM, Beal CR, Dekle DJ (1996) The effect of parent versus
unfamiliar interviewers on children’s eyewitness memory and
References identification accuracy. Law Hum Behav 20:483–500
Steblay N (1997) Social influence in eyewitness recall: a meta-
analytic review of lineup instruction effects. Law Hum Behav
Atwal K, Millwood-Hargrave A, Sancho J, Agyeman L, Karet N 21:283–297
(2003) What children watch: an analysis of children’s program- Stevenage S (1995) Can caricatures really produce distinctiveness?
ming prevision between 1997-2001, and children’s views. The British Journal of Psychology 86:127–146
Broadcasting Standards Commission, Independent Television Tversky B, Baratz D (1985) Memory for faces: are caricatures better
Commission than photographs? Mem Cogn 13:45–49
Beal CR, Schmitt KL, Dekle DJ (1995) Eyewitness identification of Wells GL, Luus E (1990) Police lineups as experiments: social
children: effects of absolute judgments, nonverbal response methodology as a framework for properly-conducted lineups.
options, and event encoding. Law Hum Behav 19:197–216 Pers Soc Psychol Bull 16:106–117
Ceci SJ, Bruck M (1993) Suggestibility of the child witness: a Wells GL, Small M, Penrod SJ, Malpass RS, Fulero SM, Brimacombe
historical review and synthesis. Psychol Bull 113:403–439 CAE (1998) Eyewitness identification procedures: recommen-
Davies G, Ellis H, Shepherd J (1978) Face recognition accuracy as dations for lineups and photospreads. Law Hum Behav
a function of mode of representation. J Appl Psychol 63:180– 22:603–647
187 Zajac R, Karageorge A (2009) The wildcard: a simple technique for
Dempsey JL, Pozzulo JD (2008) Identification accuracy of eye- improving children’s target-absent lineup performance. Appl
witnesses for a multiple perpetrator crime: examining the Cogn Psychol 23:358–368