Zhang, W., Gross, J. and Hayne, H. (2018)

14678624, 2018, 4, Downloaded from https://srcd.onlinelibrary.wiley.com/doi/10.1111/cdev.12890 by Universidad Nacional Autonoma De Mexico, Wiley Online Library on [08/11/2022].
See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Child Development, July/August 2018, Volume 89, Number 4, Pages e332–e341
If You’re Happy and You Know It: Positive Moods Reduce Age-Related
Differences in False Memory
Weiwei Zhang, Julien Gross , and Harlene Hayne
University of Otago
When tested in the Deese–Roediger–McDermott paradigm, children typically exhibit fewer false memories
than do adolescents or adults. Here, participants’ moods and the valence of word lists were manipulated to
explore the mechanism responsible for this developmental reversal in memory performance. Children (7- to
8-year-olds), adolescents (11- to 12-year-olds), and young adults (18- to 22-year-olds; N = 270) were assigned
to one of three induced mood conditions and were presented with emotional word lists. In negative moods,
adolescents and adults falsely recalled more negative information than did children, showing the typical
developmental reversal effect. This effect, however, was eliminated when participants were in positive moods.
The findings provide support for associative-activation theory and have important implications for our
understanding of the development of emotional false memories.
Although there is ample evidence that many of our which helps to reduce false memories. Gist traces,
memory skills improve between early childhood on the other hand, promote false memories by
and young adulthood (Brainerd & Reyna, 2005; emphasizing meaning-based information (i.e., infor-
Brainerd, Reyna, & Ceci, 2008; Gallo, 2013), there mation that is shared by the list words and the crit-
are also conditions under which they actually get ical lures). According to FTT, the age-related
worse. For example, in the Deese–Roediger–McDer- increase in false memories occurs because children’s
mott (DRM) paradigm, participants are presented ability to encode and retrieve gist information is
with a series of word lists in which all of the words inferior to that of adults (Brainerd & Reyna, 2005).
on each list (e.g., bed, rest, and pillow) are semanti- In essence, young children’s immature gist memory
cally related to a critical lure (e.g., sleep; Deese, protects them from false memory errors in the
1959; Roediger & McDermott, 1995). Adults typi- DRM paradigm.
cally exhibit a higher rate of true memory for the The second theory that has been put forward to
studied words than do children, but they also exhi- explain the developmental reversal effect is associa-
bit a higher rate of false memory for the unpre- tive-activation theory (AAT; Howe, 2005, 2006,
sented, critical lures; this finding has been referred 2008). According to AAT, children’s knowledge
to as a developmental reversal effect (Brainerd, base develops over time, resulting in more efficient
Reyna, & Forrest, 2002; Brainerd, Reyna, et al., and automatic activation of related concepts.
2008; Howe, 2005, 2006). According to AAT, children’s lower level of false
The counterintuitive developmental reversal in memory in the DRM paradigm is due to the fact
the rate of false memory has puzzled researchers that they do not activate the associative links
for decades; at least two theories have been put for- between the words on the DRM lists and critical
ward to explain it. According to fuzzy-trace theory lures, a process that occurs relatively automatically
(FTT; Brainerd & Reyna, 2002, 2005), there are two in adults (Howe, 2005, 2006; Howe, Candel, Otgaar,
distinct traces involved in memory. Verbatim traces Malone, & Wimmer, 2010).
focus on item-specific information (e.g., position, The key question is whether developmental
pronunciation, or word length of each list word), reversals of DRM false memories are due to age-
related changes in gist processing or to age-related
The research was funded by a Marsden grant (15-UOO-061) changes in the automaticity of associative activa-
from the Royal Society of New Zealand. We are grateful to Jackie tion. To date, no single developmental study has
Clearwater, Carrie Clifford, and Huan Wang for assistance with
data collection, as well as the mothers, children, and university
students who participated in our research.
Correspondence concerning this article should be addressed to © 2017 The Authors
Harlene Hayne, Department of Psychology, University of Otago, Child Development © 2017 Society for Research in Child Development, Inc.
PO Box 56, Dunedin, New Zealand 9054. Electronic mail may be All rights reserved. 0009-3920/2018/8904-0029
sent to hayne@psy.otago.ac.nz. DOI: 10.1111/cdev.12890
14678624, 2018, 4, Downloaded from https://srcd.onlinelibrary.wiley.com/doi/10.1111/cdev.12890 by Universidad Nacional Autonoma De Mexico, Wiley Online Library on [08/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Moods Influence Age-Related Differences in False Memory e333
allowed us to differentiate between FTT and AAT; impair relational processing and foster item-specific
in fact, the bulk of the data are completely compati- processing (Arndt & Reder, 2003; Hege & Dodson,
ble with both (e.g., Brainerd, Forrest, Karibian, & 2004; Storbeck & Clore, 2005, 2011). Within this
Reyna, 2006; Brainerd & Reyna, 2005; Howe, 2005, interpretive framework, FTT predicts that false
2006). Here, we took a novel approach, using what memories will be lower when participants are in a
we know about the combined effect of emotional negative mood and higher when they are in a posi-
word lists and participants’ extant mood on false tive mood. Furthermore, developmental studies
memory to generate a testable hypothesis that have shown that older children and adults have
might allow us to differentiate between these theo- greater item-specific (i.e., verbatim memory) and
retical explanations of the developmental reversal relational processing ability (i.e., gist memory) than
effect. do younger children (for review, see Brainerd,
In terms of the word lists used in the DRM para- Reyna, et al., 2008). Thus, the opportunity for nega-
digm, negative information has greater semantic tive moods to increase item-specific processing is
density compared to positive or neutral information potentially greater in older children and adults than
(e.g., Kensinger & Schacter, 2008; Talmi & Moscov- in younger children. As such, the decrease in false
itch, 2004). According to FTT, the greater semantic memories should be greater in older children and
density of negative information renders its gist adults than in younger children. In other words,
easier to encode and retrieve, thus the developmen- developmental reversals should be reduced or elim-
tal reversal effect should be more pronounced for inated in negative moods.
negative information relative to positive or neutral Likewise, FTT also predicts that the opportunity
information. Similarly, according to AAT, the for positive moods to promote relational processing
greater semantic density of negative information is potentially greater in older children and adults
allows the associative links among negative words than in younger children and, in turn, should
on the list to spread more easily to the negative increase false memories more in the older age
critical lures. Overall, this might increase false groups. That is, according to FTT, developmental
memory for negative word lists in all age groups, reversals should be maintained or even increased
but given that adults have a more connected when participants are in a positive mood. Summing
knowledge base than do children, AAT would also up, assuming that the standard theoretical interpre-
lead to the prediction that the developmental rever- tation of negative and positive moods is correct,
sal effect should be more pronounced for negative FTT makes two predictions: (a) False memories will
information. Consistent with both theories, partici- be lower when participants are in a negative mood
pants of all ages generally exhibit higher false mem- compared to when they are in a neutral mood and
ories for negative information than they do for higher when they are in a positive mood compared
positive or neutral information, and the age-related to a neutral mood, and (b) developmental reversals
increase in false memories is more pronounced for will be smaller when participants are in a negative
negative information than that for positive and neu- mood and will be maintained or even larger when
tral information, particularly when arousal of the they are in a positive mood relative to a neutral
to-be-remembered information is well controlled mood.
(Brainerd, Holliday, Reyna, Yang, & Toglia, 2010; Like FTT, AAT also yields predictions about the
Brainerd, Stein, Silveira, Rohenkohl, & Reyna, 2008; relation between mood and false memory. For
Howe, 2007; Howe et al., 2010). Thus, manipulating example, we know that moods influence the level
information content alone does not allow us to dif- of automaticity needed to process information
ferentiate between FTT and AAT, because both the- (Bless et al., 1996; Bolte, Goschke, & Kuhl, 2003;
ories make exactly the same prediction. Corson, 2002; Isen, Daubman, & Nowicki, 1987;
In addition to the words on the list, another Mackie & Worth, 1989). Positive moods are thought
important factor that helps to determine what par- to lead to a heuristic processing mode, which
ticipants remember is their mood. Mood has long requires fewer cognitive resources and less effort. In
been thought to change the way that we process contrast, negative moods are thought to lead to a
information, but FTT and AAT may lead to differ- systematic processing mode, which requires more
ent predictions about the effect of mood on false cognitive resources and more effort. In terms of
memories. For example, the standard theoretical false memory, the heuristic processing mode of pos-
interpretation of the memory effects of positive and itive moods is characterized by a broad activation
negative moods is that positive moods promote of remote and weakly associated information (e.g.,
relational processing, whereas negative moods both positive and nonpositive information),
e334 Zhang, Gross, and Hayne
whereas the systematic processing mode of nega- recruited through a database of parents who had
tive moods is characterized by a relatively restricted agreed to participate in studies of child development
associative activation of close and strongly associ- in Dunedin, a medium-sized city in New Zealand.
ated information (e.g., negative information; Zhang, The sample included participants from throughout
Gross, & Hayne, 2017). According to AAT, the typi- the city and from a range of socioeconomic
cal age-related increase in false memories should be backgrounds. Children were predominantly of Cau-
reduced or even eliminated when participants are casian descent and had given their approval to par-
in positive moods, whereas this increase should be ticipate in the study, along with having their parents’
maintained when participants are in negative written consent. In addition, 90 undergraduate uni-
moods. That is, when they are in positive moods, versity students (Mage = 20.37 years, SD = 1.27)
children’s false memories are more likely to be were recruited through a research participation pool
influenced by the increased automaticity with and were paid $15 for their participation. Half of the
which associative activation occurs. Furthermore, participants in each age group were female. The sam-
according to AAT, the effect of positive moods ple size was specified prior to data collection, based
would be obtained regardless of the emotional on typical sample sizes in the false memory literature
valence of the information that participants are try- (e.g., Howe, 2007; Storbeck & Clore, 2005; Zhang
ing to remember, whereas the effect of negative et al., 2017). All of the participants were native
moods would be obtained for at least negative English speakers with normal or corrected to normal
information. In contrast, FTT makes no specific pre- vision and had no history of mood disorders. They
dictions about the role of information valence. had not participated in any similar studies before.
Taken together, given that AAT and FTT make The research was reviewed and approved by the
different predictions about the impact of mood on University of Otagos’ Human Ethics Committee,
the developmental reversal of false memory for which is approved by the New Zealand Health
emotional word lists, we took the opportunity to Research Council and whose guidelines are consis-
pit one theory against the other. Assuming that the tent with those of the American Psychological Asso-
standard view of memory effects of positive and ciation. Data collection took place from April 2015 to
negative moods is correct, if developmental rever- May 2016.
sals are maintained or enhanced when participants
are in a positive mood, but developmental reversals
Materials and Procedure
are reduced or eliminated in a negative mood, this
outcome would provide empirical support for FTT. Within each age group, participants were ran-
In contrast, if developmental reversals are reduced domly assigned to one of three mood conditions
or eliminated for all information when participants (positive, negative, or neutral; n = 30 participants of
are in a positive mood, but developmental reversals each age in each condition). Mood was induced
are maintained or enhanced for negative informa- using a video clip that had a positive, negative, or
tion in a negative mood, this outcome would pro- neutral valence. This technique has been shown to
vide empirical support for AAT. However, it be the effective method of mood induction with
should be noted that if the standard view of mem- both children and young adults (for review, see
ory effects of positive and negative moods was not Brenner, 2000). To ensure the reliability and validity
supported in the present study, then FTT’s predic- of the video clips that we used, we chose emotional
tions regarding the effect of mood on developmen- video clips from classic movies and children’s cur-
tal reversals cannot be tested directly. In the current rent television programs and then asked indepen-
experiment, we examined these predictions by dent groups of 7- to 8-year-old children (N = 12, 6
assessing age-related changes in the effect of partici- female) and 18- to 23-year-olds (N = 12; 6 female)
pant mood on true and false recall for negative, to use a 5-point graphic scale (1 = happy smiling
positive, and neutral DRM lists. face, 5 = sad frowning face) to rate how happy or sad
each clip made them feel. We then selected the
video clips that produced the most consistent and
Method clear-cut mood effects. The negative and positive
video clips were from the movie “The Lion King.”
Participants
Participants in the positive mood group watched
Ninety 7- to 8-year-old children (Mage = the scene where the characters sing Hakuna Matata,
7.50 years, SD = 0.46) and ninety 11- to 12-year-old while participants in the negative mood group
adolescents (Mage = 11.40 years, SD = 0.49) were watched the scene about the death of Mufasa.
Participants in the neutral mood group watched a there was no significant difference in connectivity
National Geographic nature documentary about as a function of emotional word list, F(2, 11) = 2.19,
lions. Each video clip was approximately 4 min in p = .168. The word frequencies of both studied
duration. words and critical lures were also controlled across
Immediately following the mood-induction the emotional word lists, respectively (both
phase, we asked participants to assess their mood Fs < 1.11, ns; Kucera & Francis, 1967).
using a self-report measure called A Nonverbal The presentation order of word lists was partially
Scale of Emotion (ANSE; Lay, Waters, & Park, counterbalanced across participants; that is, half of
1989). The ANSE consists of nine different faces the participants in each age group and each mood
that cross sad, neutral, and happy facial expressions condition learned four negative word lists, followed
with calm, neutral, and excited facial expressions. by four neutral word lists, followed by four positive
Participants were shown all possible pairing of the word lists, and the other half learned the word lists
faces (N = 36) and asked to pick one face from each in the reverse order (see also, Howe, 2007; Howe
pair to indicate how they were feeling. Participants et al., 2010; Zhang et al., 2017). Each word on each
received 1 point if a chosen face showed a higher list was presented at a 3-s rate in descending order
level of pleasure than did the other face in the pair, of their BAS values. After each list had been pre-
and/or received 1 point if a chosen face showed a sented, participants completed a 30-s distractor task
higher level of excitement than did the other face in and were then asked to recall as many of the words
the pair. In this way, every participant received one from the list as they could remember.
score for valence and one score for arousal, respec-
tively. Nine pairs of faces showed the same level of
pleasure and excitement so the range of possible Results
scores was from 0 to 27. Higher scores indicate that
Mood Manipulation Check
participants felt more positive or excited, whereas
lower scores indicate that participants felt more Table 1 shows the mean valence and arousal
negative or calm. scores on the ANSE and 95% CIs as a function of
Participants then listened to audio recordings of mood. Initially, we conducted separate 3 (mood:
12 emotional DRM word lists (4 lists containing negative, positive, neutral) 9 3 (age: adults, adoles-
words with negative valence, 4 positive, and 4 neu- cents, children) analysis of variance across valence
tral) chosen from Zhang et al. (2017). Each list was and arousal scores. We then used a series of one-
10 words long and the emotional valence of the way analyses of variance (ANOVAs) to compare
words in a given list all matched and were associ- across mood and across age, respectively (alpha set
ated with a critical lure. The emotional word lists at .05).
were also matched in terms of arousal, backward For valence, there was a main effect of mood, F
associative strength (BAS; Zhang et al., 2017), and (2, 261) = 296.09, p < .001, g2p ¼ :694; participants in
interitem associative strength (connectivity). To esti- the positive mood condition felt more positive
mate the associative strength between studied (M = 24.29) than those in the neutral mood condi-
words of these word lists, an independent group of tion (M = 15.92), who in turn felt more positive
30 young adults (18 female; Mage = 19.97 years, than participants in the negative mood condition
SD = 1.07) were recruited. The results showed that (M = 4.11). There was no effect of age, F(2,
Table 1
The Mean Valence and Arousal Scores and 95% Confidence Intervals as a Function of Mood for Each Age Group
Positive mood Neutral mood Negative mood

M [95% CI] M [95% CI] M [95% CI]
Valence
Children 23.00 [21.18, 24.82] 18.20 [15.47, 20.93] 5.30 [3.86, 6.74]
Adolescents 24.43 [23.36, 25.51] 15.63 [12.37, 18.90] 4.70 [3.10, 6.30]
Adults 25.43 [24.86, 26.00] 13.93 [10.62, 17.24] 2.33 [1.47, 3.20]
Arousal
Children 13.90 [12.49, 15.31] 13.47 [12.48, 14.46] 13.00 [12.10, 13.90]
Adolescents 13.30 [11.91, 14.69] 12.87 [11.27, 14.46] 12.70 [11.55, 13.85]
Adults 13.00 [11.82, 14.18] 11.97 [10.30, 13.63] 12.27 [11.23, 13.30]
261) = 1.89, p = .153, g2p ¼ :014, power = .391, but
.77 [.74, .80]

.89 [.85, .92]
.91 [.88, .95]
Neutral SW
there was a Mood 9 Age interaction, F(4,
261) = 3.17, p = .014, g2p ¼ :046. Adults in the posi-
tive mood condition felt more positive than did
children in the positive mood condition, F(2,
87) = 3.93, p = .023, g2p ¼ :083, and adults in the
Negative mood
Positive SW
.75 [.72, .78]

.89 [.86, .92]
.91 [.88, .94]
negative mood condition felt more negative than
did adolescents and children in the negative mood
condition, F(2, 87) = 5.74, p = .005, g2p ¼ :117.
There was no reliable difference in valence scores
as a function of age for participants in the neutral
mood condition, F(2, 87) = 1.99, p = .142,
Negative SW
.80 [.77, .83]

.90 [.87, .93]
.92 [.89, .95]
g2p ¼ :044, power = .402.
For arousal, there was no effect of mood
(p > .250), age (p = .127), nor a Mood 9 Age inter-
action (p > .250); as shown in Table 1, regardless
of age or mood, all participants reported similarly
.76 [.74, .79]

.89 [.87, .92]
.94 [.91, .97]
Neutral SW
moderate levels of arousal.
Recall
Given that the critical test of FTT and AAT
The Means and 95% Confidence Intervals for True Recall as a Function of Age, Word Valence, and Mood Condition
relies on the impact of each mood on the age-
Positive mood
Positive SW
.75 [.72, .78]

.86 [.83, .89]
.93 [.90, .96]
related patterns of false memory, we analyzed true
recall and false recall using separate 3 (age: chil-
dren, adolescents, adults) 9 3 (word valence: neg-
ative, positive, neutral) mixed factor ANOVAs
with repeated measures across word valence Negative SW
(Greenhouse-Geisser correction factor) for each
.77 [.74, .80]
.91 [.87, .94]
.93 [.89, .96]
mood condition. We then used planned compar-
isons to explore age-related differences in true and
false memories as a function of word valence in
each mood condition. For all analyses, statistical
significance was accepted at a Bonferroni-adjusted
.78 [.74, .81]
.90 [.86, .93]
.93 [.90, .97]
Neutral SW
alpha level of .017.
True Recall of Studied Words

An age-related increase in participants’ true
Neutral mood
Positive SW
.76 [.73, .79]

.86 [.83, .88]
.92 [.90, .95]
recall of studied words was found for each mood

condition (see Table 2). Adults recalled more stud-
ied words than did adolescents, who in turn
recalled more studied words than did children,
smallest F(2, 87) = 96.77, p < .001, g2p ¼ :690. In
addition, neutral studied words were recalled bet-
Negative SW
.78 [.75, .81]

.91 [.88, .94]
.93 [.90, .96]
ter than were negative studied words, which in

Note. SW = studied words.
turn were recalled better than were positive stud-

ied words, smallest F(2, 174) = 132.18, p < .001,
g2p ¼ :603.
Adolescents
False Recall of Critical Lures

Children
Table 2
Adults
Participants in the neutral mood condition (see

Figure 1, top panel) exhibited the typical
Neutral Mood
0.6
0.5
False Recall Rate
0.4 Negative Critical Lures
0.3 Positive Critical Lures
0.2 Neutral Critical Lures
0.1
0
Children Adolescents Adults
Positive Mood
0.6
0.5
False Recall Rate

Neutral Critical Lures
0.2
0.1
0
Negative Mood
0.6
0.5
False Recall Rate

Neutral Critical Lures
0.2
0.1
0
Age
Figure 1. The false recall rate of critical lures as a function of word valence and age for participants in the neutral mood condition (top
panel), positive mood condition (middle panel), and negative mood condition (bottom panel). Error bars show 1 SEM.
developmental reversal effect in false memory; information tends to increase false memory com-
there was a main effect of word valence, F(2, pared to positive and neutral information and
174) = 13.96, p < .001, g2p ¼ :138, and an developmental reversals should be more pro-
Age 9 Word Valence interaction, F(4, 174) = 5.31, nounced for negative information than for positive
p < .001, g2p ¼ :109, but no effect of age, F(2, and neutral information. This prediction has been
87) = 4.06, p = .021, g2p ¼ :085. Consistent with supported by prior research (Brainerd, Stein, et al.,
prior research, children falsely recalled fewer nega- 2008; Brainerd et al., 2010; Howe, 2007; Howe et al.,
tive critical lures than did adults, t(87) = 4.11, 2010) and again by the current study, but it does
p < .001, but there were no age-related differences not provide a lynchpin for one theory over the
in false recall of positive or neutral critical lures (|t| other.
s ≤ 1.50, ps > .05). The primary goal of the present study was to
For participants in the positive mood condition differentiate between the two theoretical explana-
(see Figure 1, middle panel), there was no effect of tions of developmental reversals in false memories.
age, F(2, 87) = 2.08, p = .131, g2p ¼ :046, The current findings directly support AAT. Accord-
power = .417, and no Age 9 Word Valence interac- ing to AAT, the lower level of children’s false mem-
tion, F(4, 174) = .50, p > .250, g2p ¼ :011, ories relative to adults’ is due to the fact that
power = .168. There was, however, a main effect of children do not activate the associative links
word valence, F(2, 174) = 15.93, p < .001, g2p ¼ :155; between words and concepts as automatically as
participants in the positive mood condition falsely adults. In other words, if a manipulation enhances
recalled more negative critical lures (M = 0.31) than the level of automaticity with which children pro-
neutral (M = 0.19) or positive critical lures cess information, then according to AAT, develop-
(M = 0.17). Planned comparisons indicated that, mental reversals should be reduced. Drawing on
regardless of word valence, there were no reliable previous research, inducing positive moods is such
age-related differences in false recall (|t|s ≤ 1.91, a manipulation because positive moods lead to a
ps > .05). heuristic processing mode which is relatively auto-
For participants in the negative mood condition matic (Bolte et al., 2003; Corson, 2002; Zhang et al.,
(see Figure 1, bottom panel), there were main 2017). In contrast, negative moods lead to a system-
effects of age, F(2, 87) = 6.02, p = .004, g2p ¼ :122, atic processing mode, which requires more cogni-
and word valence, F(2, 174) = 6.11, tive resources and effort. Consistent with AAT’s
p = .003,g2p ¼ :066, but no Age 9 Word Valence predictions, we found that in a positive mood,
interaction, F(4, 174) = 2.53, p = .042, g2p ¼ :055. developmental reversals were eliminated for all
Planned comparisons indicated that, for negative kinds of information, whereas in a negative mood,
word lists, there was a developmental reversal developmental reversals still occurred for negative
effect: Children falsely recalled fewer negative criti- information. In addition, the present findings sug-
cal lures than did adolescents and adults, t gested that with increasing knowledge and experi-
(87) = 3.92, p < .001, and t(87) = 2.70, p < .01, ence, adolescents’ associative links among negative
respectively. There were no age-related differences concepts may be as integrated as those of adults,
in false recall of positive or neutral critical lures (|t| increasing their false recall to an adult level.
s ≤ 1.93, ps > .05). Although the current findings provide direct
support for AAT’s explanation for the developmen-
tal reversal effect, they do not directly refute FTT’s
explanation. Recall that FTT’s predictions regarding
Discussion
the effect of negative and positive moods on devel-
Although adults typically exhibit superior true opmental reversals are primarily based on the stan-
memories in the DRM paradigm, they also exhibit dard theoretical interpretation of negative and
more false memories. Here, we examined whether positive moods (e.g., Storbeck & Clore, 2005). How-
this effect is due primarily to adults’ superior gist- ever, the standard interpretation of the effects of
processing ability, as proposed by FTT, or due to positive moods was not supported in the present
their superior automatic associative activation, as study: We found that participants in positive
proposed by AAT. Manipulating only the valence moods exhibited fewer, rather than more, false
of the to-be-remembered information does not memories than did participants in neutral moods.
allow us to differentiate between these two theoreti- In other words, positive mood, just like negative
cal explanations because both theories predict the mood, appears to reduce rather than to enhance
same outcome; according to both theories, negative gist processing. From this perspective, according to
FTT, false memories produced in a positive mood associative activation is better in adults than in chil-
should be qualitatively similar to those produced in dren, which in turn leaves more room for negative
a negative mood. In addition, the age-related information to show the age-related increase in
increase in false memories should be less for partici- false memory than for positive or neutral informa-
pants in a positive mood than for participants in a tion.
neutral mood. This pattern was exactly what we The second finding from the present research
observed in the present study, especially for nega- that appears to be inconsistent with some prior
tive information. research is that, according to the standard theoreti-
On the surface, two findings from the present cal interpretation of negative and positive moods,
research appear to be inconsistent with some prior negative moods are thought to promote item-speci-
research. First, it was somewhat surprising that fic processing, which reduces false memories,
developmental reversals were not obtained in the whereas positive moods are thought to enhance gist
neutral mood condition for neutral and positive processing, which increases false memories (Arndt
word lists considering that developmental reversals & Reder, 2003; Hege & Dodson, 2004; Storbeck &
have been found when participants are in their natu- Clore, 2005, 2011). In contrast, as described earlier,
ral (nonmanipulated) mood, which is often assumed our positive mood manipulation did not enhance
to be neutral. The difference between the present false memories. Instead, participants who were in a
findings and prior research is likely due to differ- positive mood exhibited fewer false memories than
ences in arousal levels, which we measured here, those who were in a neutral mood. Again, this dif-
but is often only assumed in other research. In the ference is likely due to differences in arousal levels
present study, participants in the neutral mood con- across studies. Prior research has shown that
dition were moderately aroused, but we have no valence and arousal have different effects on false
idea whether the arousal level of participants in their memories (Brainerd et al., 2010; Bookbinder &
natural mood in prior research was also moderate or Brainerd, 2016; Corson & Verrier, 2007). In the pre-
whether it was consistent across age groups. This sent study, we strictly controlled arousal across
uncertainty makes direct comparison between the moods, whereas in prior research, arousal has often
present findings and prior research impossible. In been confounded with valence (e.g., Storbeck &
addition, the difference between the present findings Clore, 2005). Therefore, it is likely that the effect of
and prior research might also be due to arousal positive moods on false memories in prior research
caused by the emotional content of the word lists. was due to the combined effects of valence and
Arousal of emotional word lists was well controlled arousal rather than to valence per se.
in the present study; however, in some prior studies, Overall, we found that children who were in a
arousal has often been confounded with valence (for positive mood exhibited the same level of false
review, see Bookbinder & Brainerd, 2016). Prior recall irrespective of the valence of the list words,
research has indicated that once the arousal level of but when they were in a negative mood, the level
emotional word lists is controlled across list valence, of false recall for negative information was lower
developmental reversals for positive or neutral word than that exhibited by adolescents and adults.
lists tend to be less pronounced than for negative These findings cannot be explained solely on the
word lists (Brainerd et al., 2010), a finding that is basis of age-related differences in mood strength.
consistent with the results of the present study. Although adults reported that they felt more
The mechanism of the age-related changes in positive than did children and that they felt more
false memories as a function of valence of to-be- negative than did adolescents and children, if age-
remembered information may be due to semantic related differences in mood strength played a role
density of information. Compared to positive or in the present results, we would have expected dif-
neutral information, negative information has ferent levels of false memories between adults and
greater semantic density (e.g., Kensinger & Schac- children in a positive mood and different levels of
ter, 2008; Talmi & Moscovitch, 2004). The greater false memories among adults, adolescents, and chil-
semantic density of negative information renders dren in a negative mood. This was not the case.
gist information easier to retrieve (as suggested by Additional research will be required to clearly
FTT) and also allows the associative links among understand any potential interaction between age
negative studied words to spread more easily to and mood strength on false memories.
negative critical lures (as suggested by AAT; Brain- To the best of our knowledge, the current study is
erd, Stein, et al., 2008; Brainerd et al., 2010; Howe, the first to investigate the effect of moods on devel-
2007; Howe et al., 2010). Moreover, gist retrieval or opmental reversals in false memories for emotional
information. Although the explanatory power of FTT Bolte, A., Goschke, T., & Kuhl, J. (2003). Emotion and intu-
for the developmental reversal effect was not directly ition effects of positive and negative mood on implicit
tested in the present study, our data provides the judgments of semantic coherence. Psychological
first empirical support of AAT’s explanation for the Science, 14, 416–421. https://doi.org/10.1111/1467-9280.
01456
effect. In future studies, this explanation could be
Bookbinder, S. H., & Brainerd, C. J. (2016). Emotion and
further strengthened by conducting neuroimaging
false memory: The context-content paradox. Psychologi-
studies. For example, the anterolateral temporal lobe cal Bulletin, 12, 1315–1351. https://doi.org/10.1037/
is thought to be involved in semantic gist memory, bul0000077
whereas the orbitofrontal cortex is thought to be Brainerd, C., Forrest, T., Karibian, D., & Reyna, V. (2006).
involved in associative memory (Lewis, Critchley, Development of the false-memory illusion. Developmen-
Smith, & Dolan, 2005; Simons et al., 2005). In the tal Psychology, 42, 962–979. https://doi.org/10.1037/
generation of DRM false memories, both regions 0012-1649.42.5.962
should be activated. In addition, individuals exhibit Brainerd, C., Holliday, R., Reyna, V., Yang, Y., & Toglia,
stronger amygdala activity during sad moods than M. (2010). Developmental reversals in false memory:
during happy moods (Fitzgerald et al., 2011). Thus, Effects of emotional valence and arousal. Journal of
Experimental Child Psychology, 107, 137–154. https://doi.
we predict that children and adults might exhibit a
org/10.1016/j.jecp.2010.04.013
similar degree of interaction among the amygdala,
Brainerd, C. J., & Reyna, V. F. (2002). Fuzzy-trace theory
the orbitofrontal cortex, and the anterolateral tempo- and false memory. Current Directions in Psychological
ral lobe in a DRM task when the activation of amyg- Science, 11, 164–169. https://doi.org/10.1111/1467-
dala is relatively low (e.g., happy moods) but that 8721.00192
there would be an age-related increase in the interac- Brainerd, C. J., & Reyna, V. F. (2005). The science of false
tion among these areas when the activation of the memory. New York, NY: Oxford University Press.
amygdala is relatively strong (e.g., negative moods). Brainerd, C. J., Reyna, V. F., & Ceci, S. J. (2008). Develop-
Finally, there is also an important practical impli- mental reversals in false memory: A review of data and
cation of our findings. When children are required theory. Psychological Bulletin, 134, 343–382. https://doi.
to deploy their memory in stressful situations like in org/10.1037/0033-2909.134.3.343
Brainerd, C., Reyna, V., & Forrest, T. (2002). Are young
a courtroom, they are often attempting to recall
children susceptible to the false–memory illusion? Child
emotionally charged information in a negative
Development, 73, 1363–1377. https://doi.org/10.1111/
mood. How does the nature of the information and 1467-8624.00477
the child’s mood at the time influence the accuracy Brainerd, C. J., Stein, L. M., Silveira, R., Rohenkohl, G., &
of his or her recall? The current study is the first to Reyna, V. F. (2008). How does negative emotion cause
address this question. We found that adults and false memories? Psychological Science, 19, 919–925.
adolescents are more likely to exhibit false memories https://doi.org/10.1111/j.1467-9280.2008.02177.x
than are children in general and that mood has a Brenner, E. (2000). Mood induction in children: Method-
significant effect on this age-related pattern. Coun- ological issues and clinical implications. Review of Gen-
terintuitively, children who are in a negative mood eral Psychology, 4, 264–283. https://doi.org/10.1037/
may be more likely to make fewer negatively 1089-2680.4.3.264
Corson, Y. (2002). Effects of positive, negative and neutral
charged errors than adolescents and adults. In con-
moods on associative and conceptual priming. Current
trast, children who are in a positive mood are likely
Psychology of Cognition, 21(1), 33–62.
to make a similar number of errors regardless of the Corson, Y., & Verrier, N. (2007). Emotions and false
valence of information they are attempting to recall. memories. Valence or arousal? Psychological Science, 18,
208–211. https://doi.org/10.1111/j.1467-9280.2007.01874.x
Deese, J. (1959). On the prediction of occurrence of partic-
References ular verbal intrusions in immediate recall. Journal of
Experimental Psychology, 58, 17–22. https://doi.org/10.
Arndt, J., & Reder, L. M. (2003). The effect of distinctive 1037/h0046671
visual information on false recognition. Journal of Mem- Fitzgerald, D. A., Arnold, J. F., Becker, E. S., Speckens, A.
ory and Language, 48, 1–15. https://doi.org/10.1016/ E., Rinck, M., Rijpkema, M., & Tendolkar, I. (2011).
S0749-596X(02)00518-1 How mood challenges emotional memory formation:
Bless, H., Clore, G. L., Schwarz, N., Golisano, V., Rabe, An fMRI investigation. NeuroImage, 56, 1783–1790.
C., & W€ olk, M. (1996). Mood and the use of scripts: https://doi.org/10.1016/j.neuroimage.2011.02.061
Does a happy mood really lead to mindlessness? Jour- Gallo, D. (2013). Associative illusions of memory: False mem-
nal of Personality and Social Psychology, 71, 665–679. ory research in DRM and related tasks. New York, NY:
https://doi.org/10.1037/0022-3514.71.4.665 Psychology Press.
Hege, A. C., & Dodson, C. S. (2004). Why distinctive as a mediator. Child Development, 60, 1405–1411.
information reduces false memories: Evidence for both https://doi.org/10.2307/1130930
impoverished relational-encoding and distinctiveness Lewis, P., Critchley, H., Smith, A., & Dolan, R. (2005).
heuristic accounts. Journal of Experimental Psychology: Brain mechanisms for mood congruent memory facilita-
Learning, Memory, and Cognition, 30, 787–795. https://d tion. NeuroImage, 25, 1214–1223. https://doi.org/10.
oi.org/10.1037/0278-7393.30.4.787 1016/j.neuroimage.2004.11.053
Howe, M. L. (2005). Children (but not adults) can inhibit Mackie, D. M., & Worth, L. T. (1989). Processing deficits
false memories. Psychological Science, 16, 927–931. and the mediation of positive affect in persuasion. Jour-
https://doi.org/10.1111/j.1467-9280.2005.01638.x nal of Personality and Social Psychology, 57, 27–40.
Howe, M. L. (2006). Developmentally invariant dissocia- https://doi.org/10.1037/0022-3514.57.1.27
tions in children’s true and false memories: Not all relat- Roediger, H. L., & McDermott, K. B. (1995). Creating false
edness is created equal. Child Development, 77, 1112– memories: Remembering words not presented in lists.
1123. https://doi.org/10.1111/j.1467-8624.2006.00922.x Journal of Experimental Psychology: Learning, Memory, and
Howe, M. L. (2007). Children’s emotional false memories. Cognition, 21, 803–814. https://doi.org/10.1037/0278-
Psychological Science, 18, 856–860. https://doi.org/10. 7393.21.4.803
1111/j.1467-9280.2007.01991.x Simons, J. S., Lee, A. C., Graham, K. S., Verfaellie, M.,
Howe, M. L. (2008). What is false memory development Koutstaal, W., Hodges, J. R., & Budson, A. E. (2005).
the development of? Comment on Brainerd, Reyna, Failing to get the gist: Reduced false recognition of
and Ceci (2008). Psychological Bulletin, 134, 768–772. semantic associates in semantic dementia. Neuropsychol-
https://doi.org/10.1037/0033-2909.134.5.768 ogy, 19, 353–361. https://doi.org/10.1037/0894-4105.19.
Howe, M. L., Candel, I., Otgaar, H., Malone, C., & Wimmer, 3.353
M. C. (2010). Valence and the development of immediate Storbeck, J., & Clore, G. L. (2005). With sadness comes
and long-term false memory illusions. Memory, 18, 58–75. accuracy; with happiness, false memory: Mood and the
https://doi.org/10.1080/09658210903476514 false memory effect. Psychological Science, 16, 785–791.
Isen, A. M., Daubman, K. A., & Nowicki, G. P. (1987). https://doi.org/10.1111/j.1467-9280.2005.01615.x
Positive affect facilitates creative problem solving. Jour- Storbeck, J., & Clore, G. L. (2011). Affect influences false
nal of Personality and Social Psychology, 52, 1122–1131. memories at encoding: Evidence from recognition data.
https://doi.org/10.1037/0022-3514.52.6.1122 Emotion, 11, 981–989. https://doi.org/10.1037/a0022754
Kensinger, E. A., & Schacter, D. L. (2008). Memory and Talmi, D., & Moscovitch, M. (2004). Can semantic related-
emotion. In M. Lewis, J. M. Haviland-Jones, & L. F. ness explain the enhancement of memory for emotional
Barrett (Eds.), The handbook of emotions (pp. 601–617). words? Memory & Cognition, 32, 742–751. https://doi.
New York, NY: Guilford. org/10.3758/BF03195864
Kucera, H., & Francis, W. N. (1967). Computational analysis Zhang, W., Gross, J., & Hayne, H. (2017). The effect of
of present day American English. Providence, RI: Brown mood on false memory for emotional DRM word lists.
University Press. Cognition and Emotion, 31, 526–537. https://doi.org/10.
Lay, K. L., Waters, E., & Park, K. A. (1989). Maternal 1080/02699931.2016.1138930
responsiveness and child compliance: The role of mood

Zhang, W., Gross, J. and Hayne, H. (2018)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Zhang, W., Gross, J. and Hayne, H. (2018)

Uploaded by

Copyright:

Available Formats

14678624, 2018, 4, Downloaded from https://srcd.onlinelibrary.wiley.com/doi/10.1111/cdev.12890 by Universidad Nacional Autonoma De Mexico, Wiley Online Library on [08/11/2022].

Positive mood Neutral mood Negative mood

261) = 1.89, p = .153, g2p ¼ :014, power = .391, but

.77 [.74, .80]

.75 [.72, .78]

.80 [.77, .83]

.76 [.74, .79]

.75 [.72, .78]

alpha level of .017.

True Recall of Studied Words

.76 [.73, .79]

recall of studied words was found for each mood

.78 [.75, .81]

ter than were negative studied words, which in

turn were recalled better than were positive stud-

False Recall of Critical Lures

Participants in the neutral mood condition (see

0.2 Neutral Critical Lures

0.4 Negative Critical Lures

0.4 Negative Critical Lures

You might also like