2012. Vol.3, No.1, 30-35
Published Online January 2012 in SciRes (http://www.SciRP.org/journal/psych) http://dx.doi.org/10.4236/psych.2012.31005
Copyright © 2012 SciRes.
Pre-Schoolers’ Reports of Conflicting Points Secretly Inserted into
a Co-Witnessed Event: An Experimental Investigation Using
the MORI Technique
Kazuo Mori1, Ryuta Takahashi2
1Institute of Engineering, Tokyo University of Agriculture and Technology, Tokyo, Japan
2Osaka Family Court, Sakai, Japan
Received November 5th, 2011; revised December 6th, 2011; accepted December 31st, 2011
Thirteen pre-school and ten undergraduate pairs participated as eyewitnesses to a simulated criminal event
presented through animated cartoons using a presentation trick (MORI technique). Although there were
two different versions, the MORI technique had participants observe only one version without being
aware of the other. In three reporting sessions, participants recalled what they presumed they had jointly
observed; individually immediately after the presentation, collaboratively after the individual recall, and
again individually one week later. The main results were: pre-schoolers, as well as undergraduates,
showed better recall in the collaborative tests, though the former generally showed poorer recall than the
latter, pre-schoolers tended to conform more frequently than undergraduates in the week-later tests, and
both pre-school and undergraduate pairs conformed more often for amendment than distortion.
Keywords: Eyewitness Testimony; Pre-School Child Pairs; MORI Technique; Conformity
There are various reasons for the bias of information in wit-
ness testimony, but that of co-witnesses has been reported to be
crucially influential (Kanematsu, Mori, & Mori, 1996/2003;
Gabbert, Memon, & Allan, 2003; for a recent review, see
Wright, Memon, Skagerberg, & Gabbert, 2009). In Kanematsu,
et al. (1996/2003), co-witness pairs observed the same event,
but with hidden discrepancies secretly inserted using a presen-
tation trick (the MORI technique, Mori, 2003). Researchers
found that, in subsequent discussion of what they had witnessed
together, participants tended to modify their own memory of
what they had observed in order to conform to that of their
co-witnesses. Mori and his collaborators have conducted a se-
ries of experiments to investigate the memory distortions of
co-witnesses under various conditions: male vs. female pairs
(Matsuno, Mori, Hirokawa, & Ukita, 2005), couples vs. unac-
quainted pairs (French, Garry, & Mori, 2008), one-vs-two wit-
nesses and two-vs-two witnesses (Mori & Mori, 2008), as well
as mother-child pairs (Mori & Kitabayashi, 2009).
Gabbert, Memon, and Allan (2003) also investigated the
co-witness effect on memory recollection utilizing a different
paradigm. They showed pairs of participants a simulated
criminal event videotaped from two different angles such that
each participant viewed one of the two video versions which
included some details that were not visible in the other and vice
versa. They found a significant proportion (71%) of participants
who had discussed the event recalled details that had been ac-
quired during the discussion rather than through direct observa-
Co-witness conformity research utilizing the Gabbert method
(Gabbert, et al., 2003), in which the co-witness pairs observed a
slightly different version of an event separately, and that utiliz-
ing the MORI technique, in which the co-witness pairs ob-
served the same event together without being aware of the du-
ality, showed similar research findings. As stated above, fun-
damentally both sets of studies revealed that the recollections of
independent witnesses tend to be strongly influenced by those
of their co-witnesses.
Recently, Gabbert, Memon, and Wright (2007) found that
participants/witnesses who believed their partners had observed
the event longer tended to conform to them. French, Garry, and
Mori (2011) found similar results with the Mori method. They
manipulated participants’ expectations by leading them to be-
lieve that they had either “higher or lower visual acuity” than
their partners when together they watched a video event in
which several discrepancies had been inserted secretly. The
results showed that, although the actual visual acuity was equal
in reality, participants who believed that they had “lower visual
acuity” conformed more than their counterparts with “higher
It is noteworthy that these two studies independently re-
vealed the same fact, that co-witnesses tended to take their
partners’ credibility into account when deciding whether they
would conform to them or not. The co-witness conformity ef-
fects are crucial in the application of research results to actual
investigation and judicial practices. Therefore, it is worth em-
phasizing that the same research findings need to be corrobo-
rated by various independent researchers using different ex-
Relatively few experimental studies have been done on
co-witness effects in children. Child witness testimony has been
a major research topic in applied cognitive and forensic psy-
chology. Child witnesses have been studied because they are
often the victims in child abuse cases, and the credibility of
their memory and overall cognitive ability is a major concern in
K. MORI ET AL.
forensic proceedings. Techniques for interviewing young chil-
dren and suitable procedures for retrieving reliable reports from
them have also been studied rigorously (See a recent review by
Davies & Pezdek, 2010). However, the co-witness effect among
children has not yet been studied in detail.
Roos af Hjelmsäter, Granhag, Strömwall, and Memon (2008)
investigated the co-witness effect on children’s memory per-
formance using a traditional misinformation paradigm. They
interviewed children aged 7 - 13 regarding an event they had
experienced two weeks earlier. Prior to the interviews, the chil-
dren were exposed to an account of the same event that con-
tained some misinformation from a fictional adult co-witness.
Although the children’s reports were generally quite accurate,
the results showed that the information provided by the adult
co-witnesses influenced the children to make errors when they
reported about the event.
Mori and Kitabayashi (2009) studied the co-witness effect of
mother-child pairs using the Kanematsu-Mori-Mori paradigm
and found that elementary school children (between 6 to 12
years old) conformed to their co-witness mothers but that their
conformity frequency was almost the same as that of the moth-
ers to their co-witness children. Out of 92 incidences of con-
formity, mothers conformed to their children’s accounts 41
times whereas children conformed to their mothers’ versions 51
Candel, Memon, and Al-Harazi (2007) examined the mem-
ory conformity of child co-witnesses using the same paradigm
as Gabbert, et al. (2003). To our knowledge, it is the only ex-
perimental study that has examined the interaction between
child co-witnesses. In it, they showed a video to younger (aged
6 - 7) and older (aged 11 - 12) children in pairs, leading them to
believe that they had seen the same video as the other child of
the pair although they had actually seen different versions.
Then the participants were led to discuss what they had viewed
together, and finally they completed a recall task individually.
Candel, et al. (2007) found that more than 60% of the children
“recalled” at least one detail from the video that they had not
actually observed themselves. The memory conformity ten-
dency was stronger for older than for younger children.
All these preceding studies showed basically the same re-
sult—that children tend to conform to their co-witnesses. How-
ever, the first two studies, Roos af Hjelmsäter, et al. (2008) and
Mori and Kitabayashi (2009) used adult-child co-witness pairs.
The adult co-witness in Roos af Hjelmsäter, et al. (2008) was a
confederate, not a real co-witness. In Candel, et al. (2007), the
co-witness pairs were both children, but they watched a video
alone, although they were told that they had watched the same
video as their partners. Thus, those child co-witnesses might
not have had a genuine feeling of having co-witnessed the same
As stated above, co-witness conformity research has been
conducted in two different experimental modes, one utilizing
the Gabbert method, in which co-witnesses separately watched
visual stimuli, and one utilizing the MORI technique, in which
co-witness pairs watched the same presentation together. The
research groups of these two different experimental methods
produced similar experimental results, thus corroborating the
findings of the authors of the two studies. Therefore, the present
study aimed to re-examine the co-witness interactional effect
among children, found in Candel, et al. (2007) using the Gab-
bert method, by utilizing the MORI technique in a setting in
which the participants witnessed the same event simultaneously,
sitting side by side.
Thirteen pre-school pairs (5 - 6 year olds; 12 boys and 14
girls) and ten undergraduate pairs (19 - 20 year olds; six men
and 14 women) participated as eyewitnesses in the experiment.
The animated video clips of a simulated criminal event:
Two different versions of the Kanematsu et al. movie (1996/
2003) were converted into an animation by means of frame-by-
frame drawings done with Adobe Photoshop software. Each
frame was 360 × 360 pixels in size with a resolution of 72 dpi.
The corresponding frames of the two video versions were
placed side-by-side to produce a combined new frame of 720
pixels wide × 360 pixels high. The animated clip was 50 sec-
onds long and consisted of 150 combined frames, presented
three frames per second using the mpeg animation software
QuickTime Movie on an Apple iBook. In this way, two differ-
ent versions of the animation were presented simultaneously on
a frame-by-frame basis.
The two versions depicted basically the same event—a theft
from the bag of a female pedestrian by a man and woman cou-
ple driving a car. However, there were three differing points:
the color of the criminals’ car, a dark car (Version A) vs a white
car (Version B); (b) the clothes of the criminal driver, a parka
with lateral stripes (Version A) vs a white shirt (Version B);
and (c) the direction the pedestrian walked after the theft, up
toward the screen (Version A) vs down away from the screen
The experimental apparatuses used for the MORI technique
were equivalent to those used in the Kanematsu et al. study, but
modified for the child participants, that is, 1) the video clips
were animated cartoons rather than regular movies, and 2) sets
of circular polarizers were used instead of linear polarizers so
that the hidden images would not be seen if the observers tilted
their heads. (See, Mori, 2007 for details of the MORI tech-
nique.) The experimental setting is depicted in Figure 1.
Video projectors: Two DLP video projectors (Yokogawa
D-1200X) were used. One each of two types of circular polar-
izing filters (R- and L-circular polarizers) was placed in front of
the projection lenses of each projector. The projected light,
having passed through the polarizer, was polarized circularly
with either right- or left-hand rotation. Since the circularly po-
larized light can pass through another circular polarizing filter
with the same rotational direction while being blocked by a
circular polarizer of a different direction, the two images pro-
jected onto the same screen can be seen separately by the view-
ers, each wearing a pair of suitable polarizing sunglasses. The
two projectors were placed side by side approximately 60 cm
behind a half-transparent screen to project two different images
polarized counter-wise to each other.
Half-transparent rear screen: A 20 cm × 20 cm plain
ground glass pane 5 mm thick was used as a half-transparent
screen. It was mounted on a 180 cm (h) × 90 cm (w) × 0.5 cm
(thickness) wooden panel, which contained a 20 cm × 20 cm
Copyright © 2012 SciRes. 31
K. MORI ET AL.
Copyright © 2012 SciRes.
The experimental setting for presenting two different animated movies on the same screen simultane-
ously. The actual animation was presented in full color.
window in the middle for the screen. It was the same screen as
used in Mori and Kitabayashi (2009). The screen was set about
1 m away from the viewers.
Polarizing sunglasses: Two types of circular polarizing sun-
glasses suitable for viewing one or the other of the video im-
ages were prepared. They looked similar to ordinary sunglasses
and were identical to each other. They were made using ordi-
nary sunglass frames and cutouts from R- and L-circular polar-
izing filter sheets of 0.8 mm thick (MeCan Imaging Inc., Japan).
Testing sessions: The witnesses’ recall was assessed three
times using “Immediate Tests,” just after the video presentation
(individually), “Collaborative Tests,” during discussion (in
pairs), and “Week-later Tests,” a week after the discussion
Cued recall test sheets: The assessment of witness memory
was done by means of cued recall tests. Female undergraduates
pre-trained in conducting child interviews assisted the young
participants in taking the tests. The test consisted of 20 ques-
tions designed to assess the participants’ memory of the ani-
mated event. The question items were selected from both cen-
tral and peripheral information in the event. Of those, three
questions concerned the three critical items: #3 on the color of
the car, #8 on the driver’s clothing, and #20 on the direction the
pedestrian walked. The same test sheets were used in all test
periods and for all participants.
The experimental procedure was basically a replication of a
previous mother-child pair witness experiment (Mori & Kita-
Video Presentation Phase
In the first phase, the video presentation, Immediate Test,
and Collaborative Test were administered to the child partici-
pants in a room at their nursery school and to the undergraduate
participants in a psychology laboratory.
Presentation: Participants entered an experimental room in
pairs and sat on chairs placed side-by-side facing the screen ap-
proximately 1 m apart. Participants chose a chair at will. Two
types of circular polarizing sunglasses, one pair for Version A
and another for Version B, were placed on the chairs. As a ruse,
the participants were instructed to wear the sunglasses to pro-
tect their eyes from glare.
Immediate Tests: After watching the animation, the par-
ticipants’ recollection of the event was assessed individually
by means of cued recall tests. No time limits were set, but it
took about five to ten minutes to complete each test. For
pre-school participants, female undergraduates majoring in
educational psychology assisted the youngsters in taking the
test. The assistants were trained beforehand how to handle
young children and how to elicit their responses without bi-
Collaborative Tests: After completion of the Immediate
Tests, participants were asked to report collaboratively on what
they had observed on a new test sheet. It took about ten minutes
to complete it and the same female assistants helped the pre-
school pairs to answer the test questions.
Week-Later Test Phase
Approximately a week after the first phase, participants were
asked to take a memory test individually again. Following these
Week-later Tests, a questionnaire inquiring whether they had
noticed any anomaly in the video presentation was administered.
The same female assistants helped the pre-school pairs during
The questionnaire conducted at the week-later phase revealed
that no participants noticed any anomaly during the video
presentations. The effectiveness of the MORI technique has
been shown in various application experiments (French, Gerrie,
Garry, & Mori, 2009). It was noteworthy that the present
study demonstrated its validity even for pre-school age par-
Memory Test Scores
The memory test scores were converted into percentage
scores; that is, the number of items correctly reported divided
by the total number of items (= 20) and multiplied by 100. Each
test contained three questions concerning the three critical items
that had different correct answers depending on which version
K. MORI ET AL.
the participant had observed. No significant differences were
found between the memory test scores, Immediate or Week-
later, of participants who had observed different video versions;
F(1,42) = 0.05, ns. Therefore, we combined the data of the two
video versions in the following analyses.
Figure 2 shows the average memory test scores of pre-
school and undergraduate participants for the three test sessions.
As expected, the average memory scores of the pre-schoolers
were much lower than those of the undergraduates; 53.2 and
84.3, respectively (F(1,88) = 125.56, p < .01; Cohen’s d = 3.16).
The results also showed that collaboration had improved the
memory test scores for both participant groups (F(2,88) = 7.92, p
< .05; d = .61). Tukey’s post hoc test (MSe = 53.5, p < .05)
revealed that only the Collaborative Test scores were statisti-
cally higher than the other two test scores. The interaction was
not significant (F(2,88) = 0.73, ns).
Conformity on Week-Later Tests
We adopted the same definition of “conformity” as Mori and
Kitabayashi (2009). That is, “conformity” was defined as a
change in the answers on the Week-later Tests towards the
group decision on the Collaborative Tests. A typical example of
conformity was when one witness who had answered “white
car” on the Immediate Test changed his/her answer to “dark
car” on the Week-later Test, after having acceded to their group
decision that it was a “dark car” on the Collaborative Test. It
was also regarded as conformity when a witness who had no
answer on the Immediate Test acceded to the group decision on
the Week-later Test.
We had intended to examine the conformity frequencies only
on the three target items that would cause conflicts between the
co-witness pairs. However, we incidentally obtained several
conformity responses on some non-target items when one
member of a pair overlooked or misperceived information.
Some participants made errors or left unanswered some non-
target items on the Immediate Tests, and provided on the
Week-later Tests answers that had come from the Collaborative
Tests. These incidents fit the criteria for conformity as we had
defined it. Therefore, we included such items in our analyses of
conformity occurrences. The results showed that pre-schoolers
tended to conform more than undergraduates on both the target
Average memory test scores at the three test periods (max. = 100; ver-
tical lines represent standard deviations).
and non-target items, as shown in Figure 3. One pre-school
pair showed conformity on 2.7 items on average out of 20 (0.7
of the 3 target items) while one undergraduate pair conformed
on 1.5 items (0.5 of 3 target items). The Chi-square test per-
formed on the total conformity frequency in the two groups
showed a significant difference (χ2
(1) = 4.15, p < .05; Cohen’s w
= .10). The probability calculation on the conformity frequen-
cies based only on the target items showed no significance (p
= .42, ns.)
Good Conformities vs Bad Conformities
As Mori and Kitabayashi (2009) noted in their mother-child
pair co-witness conformity experiment, there were good and
bad conformities. Conformity may well be a source of distor-
tion, as most of the post-event information (PEI) eyewitness
literature has demonstrated. However, conformity has positive
aspects as well. Witnesses having misperceived the original
event may correct their memory through conforming to the
correct information (Mori & Kitabayashi, 2009).
We observed 50 cases of conformity in total that could be
classified either good or bad; “good” if the conformity occurred
in the correcting direction, and “bad” if it resulted in a distor-
tion. There were ten cases that did not fall into either category,
and those were excluded from the analysis. They were mostly
for the target items, because there were two correct answers for
those. Table 1 shows the frequencies of good and bad confor-
mity for child and undergraduate pairs. It showed that good
conformity occurred three times more often than bad as a whole
(p = .0006, direct probability calculation). It should be noted
that the good-bad conformity ratios were similar for children
and undergraduates (χ2 = .06, ns).
Average number of conformity responses per pair on week-later tests
(Each pair reported on the 3 target and 17 non-target items).
Frequencies of good and bad conformity for pre-school and undergra-
Good Bad Unclassified Total
Pre-schoolers 22 6 7 35
Undergraduates9 3 3 15
Total 31 9 10 50
Copyright © 2012 SciRes. 33
K. MORI ET AL.
Do Children Really Conform More Often than
The present experiment showed that child co-witnesses con-
formed to each other more often than did the undergraduate
co-witnesses. This result at first seems a matter of simple com-
mon sense. However, we should be cautious about taking these
results at face value, because the literature on co-witness re-
search has shown conflicting results. For example, Gabbert, et al.
(2003) reported that 71% of their undergraduate co-witnesses
recalled details that had been acquired during the discussion,
whereas in Candel, et al. (2007) only 60% of the child co-wit-
nesses (11 - 12 years old) did. A simple comparison of the re-
sults of these two studies is inadequate because the conformity
indices were different, as were the experimental procedures. As
for a genuine comparison under the same conditions, Candel, et
al. (2007) concluded that the memory conformity effect was
stronger for older children (11 - 12 years old) than for younger
(6 - 7 years old). Even this result from a direct comparison
within a single study was not consistent with a conflicting re-
port by other researchers. Roos af Hjelmsäter, et al. (2008)
reported that their 7- and 12-year old children showed no age
group difference in their susceptibility to suggestion from adult
Then why did the children in the present experiment conform
more? A probable interpretation is that the higher frequency of
conformity in children stemmed from the higher rate of mem-
ory errors in general. Conformity would only occur when there
were some discrepancies among co-witnesses. The more errors
there were, the more possibilities for conformity to occur. To
state it another way, there might be a “ceiling effect” in the
memory performance of the undergraduates. They correctly
recalled about 80% on the 20-item cued recall tests, leaving
only a small margin for discrepancies among their original
answers. As shown in Table 1, three-quarters of conformities
occurred in the correcting direction. Our undergraduate co-
witnesses did not conform nearly as frequently because they
remembered very well what they had observed, leaving little
room for amendment. That is why conformity occurred less
frequently among undergraduates than among children in the
We have some supporting data obtained from a series of dif-
ferent, but closely related, recent studies in the Asch conformity
experiments. Mori and Arai (2010) replicated the Asch original
line judgment experiment without using a confederate by util-
izing the MORI technique with foursomes of undergraduate
participants. In each foursome of participants, only the third
responders wore a different type of polarizing sunglasses than
the other three, so that only they observed the lines differently,
therefore becoming a minority among the group of four. They
found minority female undergraduates made conformity errors
in 28.6% of their responses. Using exactly the same experi-
mental procedure, Hanayama and Mori (2011) found the corre-
sponding error rate of six-year-olds was 50.0%. However, the
overall error rate of the child participants was higher than that
of the undergraduate participants. Then they estimated the net
conformity rates by extracting the error rates of the majority
groups from those of the minority groups, obtaining almost
equivalent rates for children (50.0 – 21.3 = 28.7%) and under-
graduates (28.6 – 8.6 = 20.0%).
No Collaborativ e Efforts Shown in the Child
As described above, our child co-witnesses seemed not to
recollect collaboratively what they had witnessed together. We
got two clues to that through observation of their behavior dur-
ing the Collaborative Test sessions. First, we found that the
children were seldom surprised at hearing different answers
from their co-viewers. On the contrary, the undergraduate par-
ticipants always showed signs of surprise when they experi-
enced disagreement on the three differing items. It is quite
natural for adult participants to have become puzzled at unex-
pected disagreements on such a simple memory test of what
they observed together. Secondly, our child co-witnesses did
not try to reconcile the discrepancies between their answers.
We expected them to discuss the conflicting items and attempt
to resolve them, but such discussion seldom occurred among
the children during the Collaborative Test sessions.
They behaved almost like children at an ice cream shop,
where one said “Chocolate” and the other said “Vanilla,” but
soon changed to “Chocolate.” These observations were not
done systematically, but they implied the interesting interpreta-
tion that pre-school children might not have a fully developed
Theory of Mind yet. Adult participants started discussion to
reconcile the discrepancy when their answers differed each
other because they believed their answers should not be differ-
ent on something they had witnessed together. It was because
they knew what their partner observed and they anticipated
certain answers. However, children aged five or so, as in our
experiment, seemed lack this sense of reasoning, or inferring
what their partners would report. They were not aware of how
their co-witness partners had observed the event or how they
would answer concerning what they had observed together.
Perhaps that is why they were not surprised at the discrepancies
they encountered. It is an interesting finding, but it should be
examined in detail elsewhere rather than in the present study.
Further well-constructed research on this interesting interpreta-
tion would be illuminating.
Is the Collaboration of Co-Witnesses Good or Bad?
Implications for Forensic Practices
It would be rather difficult to give concrete implications from
the present study for forensic practice on co-witness effects.
However, it clearly showed that collaboration of co-witnesses
may lead to better memory performance. It was repeatedly
shown in the experimental studies utilizing the MORI method
(Kanematsu, et al., 1996/2003; Matsuno, et al., 2005; Mori &
Kitabayashi, 2009). That was mostly because witnesses may
well be mutually reminded of the correct information by their
partners if they collaborate. We also found that conformity after
the collaborative recollection may occur in the correcting direc-
tion more often than in the distorting direction. These findings
showed the positive side of collaboration among co-witnesses.
However, a negative side of collaboration was found, too. If
one of the co-witnesses misperceives or miscomprehends some
detail and insists, discussion may become a two-edged sword.
The mistaken details may be corrected through discussion, but,
alternatively, they may well persist and become reinforced even
though they are incorrect. Though, it was less frequent, we did
find that some conformities resulted in wrong answers. The
problem is that, when it comes to actual co-witnesses, we never
know which may occur. Therefore, we cannot present here any
Copyright © 2012 SciRes.
K. MORI ET AL.
Copyright © 2012 SciRes. 35
definitive recommendation on whether or not co-witnesses
should be allowed to discuss their observations.
This research was done while the authors were at Shinshu
University. It was supported by a Grant-in-Aid from the Japa-
nese Ministry of Education, Culture, Sports, Science, and
Technology (Grant No.13610081) to the first author. We thank
all the children and teachers in Hiraoka Nursery School in Na-
gano Prefecture for their cooperation in the conduct of this
research. The animation was made by Akane Yamazaki for her
capstone thesis submitted to Shinshu University in 2002. We
are indebted to Rebecca Ann Marck for her superb work in
editing the English manuscript.
Candel, I., Memon, A., & Al-Harazi, F. (2007). Peer discussion affects
children’s memory reports. Applied Cognitive Psychology, 21, 1191-
Davies, G., & Pezdek, K. (2010). Children as witnesses. In G. J. Towl,
& D. A. Crighton (Eds.), Forensic psychology (pp. 178-193).
Leicester, UK: BPS Blackwell.
French, L., Garry, M., & Mori, K. (2008). You say tomato? Collabora-
tive remembering between intimate couples leads to more false
memories than for strangers. Memory, 16, 262-273.
French, L., Garry, M., & Mori, K. (2011). Relative—Not absolute—
Judgments of credibility affect susceptibility to misinformation con-
veyed during discussion. Ac ta Psychologica, 136, 119-128.
French, L., Gerrie, M. P., Garry, M., & Mori, K. (2009). Evidence for
the efficacy of the MORI technique: Viewers do not notice or im-
plicitly remember details from the alternate movie version. Behavior
Research Methods, 41, 1224-1232. doi:10.3758/BRM.41.4.1224
Gabbert, F., Memon, A., & Allan, K. (2003). Memory conformity: Can
eyewitnesses influence each other’s memories for an event? Applied
Cognitive Psychology, 17, 533-543. doi:10.1002/acp.885
Gabbert, F., Memon, A., & Wright, D. B. (2007). I saw it for longer
than you: The relationship between perceived encoding duration and
memory conformity. Acta Psychologica, 124, 319-331.
Hanayama, A. & Mori, K. (2011). Conformity of six-year-old children
in the Asch experiment without using confederates. Psychology, 2,
Matsuno, E., Mori, K., Hirokawa, K., & Ukita, J. (2005). The effects of
gender differences in pairs of eyewitnesses on recall memory. Japa-
nese Journal of Cognitive P sy c h ol o g y, 3, 83-94.
Kanematsu, H., Mori, K., & Mori, H. (1996/2003). Kotonaru Jitai wo
mokugekishita futari no mokugekisha no hanashiai niyoru kioku no
hen-you [Memory distortion in eyewitness pairs who observed non-
conforming events and discussed them]. Cognitive Studies: Bulletin
of the Japanese Cognitive Science Society, 3, 29-40, 1996. (In Japa-
nese. English version published in Journal of the Faculty of Educa-
tion, Shinshu University, 10 9, 75-84, 2003).
Mori, K. (2003). Surreptitiously projecting different movies to two
subsets of viewers. Behavior Research Methods, Instruments, and
Computers, 35, 599-604. doi:10.3758/BF03195539
Mori, K. (2007). A revised method for projecting two different movies
to two groups of viewers without their noticing the duality. Behavior
Research Methods, 39, 574-578. doi:10.3758/BF03193028
Mori, K., & Arai, M. (2010). No need to fake it: Reproduction of the
Asch experiment without using confederates. International Journal
of Psychology, 45, 390-397. doi:10.1080/00207591003774485
Mori, K., & Kitabayashi, M. (2009). How child-mother pairs reported
what they had witnessed together: An experimental examination us-
ing the MORI technique. Ps ych olog y Jo urn al, 6, 60-69.
Mori, K., & Mori, H. (2008). Conformity among cowitnesses sharing
same or different information about an event in experimental col-
laborative eyewitness testimony. Perceptual and Motor Skills, 106,
Roos af Hjelmsäter, E., Granhag, P. A., Strömwall, L. A., & Memon, A.
(2008). The effects of social influence on children’s memory reports:
The omission and commission error asymmetry. Scandinavian Jour-
nal of Psychology, 49, 507-513.
Wright, D. B., Memon, A., Skagerberg, E. M., & Gabbert, F. (2009).
When eyewitnesses talk. Current Directions in Psychological Sci-
ence, 18, 174-178. doi:10.1111/j.1467-8721.2009.01631.x