A Full Night’s Sleep at Home Improves Memory Performance in an Associative and Relational Learning Task
Nicolas Ribeiro, Yannick Gounden, and Veronique Quaglino
University of Picardy Jules Verne
This study evaluated the effect of sleep in mediating episodic memory performances
in a recognition task of visually associated and related elements. It also considered links
between memory performances, dreams mentation, and emotional salience of items. Two
groups of participants were studied; in the wake group, they stayed awake on a normal
day, and in the sleep group, they slept at home according to their usual sleep schedule and
context. Compared with the wake group, participants in the sleep group performed better
in the delayed associative and relational tests. Negative and positive emotional images
were better recognized by the participants than neutral ones with no group specificity.
Examination of dream reports suggested that reporting a dream related to the present
study was associated with an increase in recognition performances. Overall, our research
demonstrated how new associative and relational links can be consolidated in memory
following a full night’s sleep.
Keywords: sleep, memory, dreams, relational, associative
Memory is highly associative, and our everyday life experiences are integrated
into a dynamic memory network composed of an intertwinement of past event
traces and more recent ones (Nadel, Hupbach, Gomez, & Newman-Smith, 2014: 2024 – Essay Writing Service. Custom Essay Services Cheap;
Versace et al., 2014: 2024 – Essay Writing Service | Write My Essay For Me Without Delay). This study evaluated the effect of sleep in mediating episodic
memory performances in a recognition task of visually associated elements. More
specifically, we examined how visually associated images create new relationships in
memory during a learning process and the influence of these relations on
subsequent recognition performance after a full night’s sleep. In addition, we
considered the influence of emotional salience of items on memory. Finally, we
evaluated the effect of the incorporation of the studied material into dream
mentation on memory performance.
This article was published Online First April 16, 2020.
X Nicolas Ribeiro, Yannick Gounden, and Veronique Quaglino, Department of Psychology,
Research Center in Psychology: Cognition, Psychism and Organizations EA 7273, University of Picardy
Jules Verne.
We thank Cheyenne Delcourt who participated in this study as an undergraduate student.
We declare that there are no conflicts of interest for this research and that the study received no
financial support from any organization.
Correspondence concerning this article should be addressed to Nicolas Ribeiro, Department of
Psychology, Research Center in Psychology: Cognition, Psychism and Organizations EA 7273, Université de Picardie Jules Verne, Chemin du Thil, 80000, Amiens, France. E-mail: nicolas.ribeiro@upicardie.fr
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
171
Dreaming
© 2020 American Psychological Association 2020, Vol. 30, No. 2, 171–188
ISSN: 1053-0797 http://dx.doi.org/10.1037/drm0000130
Associative Processes in Memory
Associative memory refers to links that can be created in memory when
different perceptive elements are presented at the same time in our environment.
For instance, when two different images are presented at the same moment on a
screen. Relational memory refers to the links that exist in memory between
different elements that were not presented at the same time but share content
similarity. Such a relation can be based on two associations that have a similar
element in common. As an example of such associative and relational memories,
Alger and Payne (2016: 2024 – Do my homework – Help write my assignment online) have proposed that if a person knows how to get from
Place A to Place B and also from Place B to Place C, then he or she is also able to
get from Place A to Place C.
The scientific investigation of how associated items are memorized is a
long-standing transversal interest in psychology, and researchers have postulated
that associative mechanism plays, at least, a key role in human memory, if not to be
the core mechanism of memory itself (Anderson & Bower, 2014: 2024 – Essay Writing Service | Write My Essay For Me Without Delay; Chalfonte &
Johnson, 1996; Fuster, 1997; Hebb, 1949; Hutchison, 2003; Thomson & Tulving,
1970; Wheeler & Treisman, 2002). This interest for associative links in memory can
be traced back to the emergence of experimental psychology (Baldwin, 1890; Davis,
2008 – Affordable Custom Essay Writing Service | Write My Essay from Pro Writers). MacDougall (1904) specified how the presentation of one part of an
association allows memory to be reactivated for the other part. Baldwin (1890)
mentioned how images in the present, such as the face of a friend, recalls older
memories, including the context of the previous meeting with this friend. Interestingly, Baldwin (1890) also indicated how these associative phenomena can emerge
from the nature of the physiological support of cognition. He described it as “an
intricate network of fibrous and cellular tissue.” He also pointed out the way
information is related into the memory system by highlighting “the wide capacity it
affords for varied and related representations” (p. 193). Within the scope of this
study, relations refer to how associations that share similarities are related to
memory. Nadel et al. (2014: 2024 – Essay Writing Service. Custom Essay Services Cheap) suggested that consolidation is not about an isolated
memory trace but is a dynamic process that leads to changes in the memory system.
They proposed that new experiences must be integrated in the memory system to
be remembered. For this integration, memory system needs to modify its previous
state. Thus, memory is about creating new links between newer and older
experiences by the means of mechanisms that reshape previously fixed memories.
These transformations affect memory system depending on similarities between
previous experiences and the ongoing experiences. As a simplification, this dynamic
process is essentially about associating and relating experiences.
Association Interference Task
Current considerations for associative learning and for the relational dimension deriving from these associations coincided with the creation of new research
paradigms. For instance, Preston, Shrager, Dudukovic, and Gabrieli (2004) devised
a paradigm, known as Associative Inference Tasks (AIT; Lau, Tucker, & Fishbein,
2010 – Essay Writing Service: Write My Essay by Top-Notch Writer; Ragland et al., 2009). The rationales of AIT can be described as such: “If A
is associated with B and B is associated with C, then A is related to C.” In the
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
172 RIBEIRO, GOUNDEN, AND QUAGLINO
present research, AB and BC are referred to as associations and CA as relations
(Armstrong, Kose, Williams, Woolard, & Heckers, 2014: 2024 – Essay Writing Service. Custom Essay Services Cheap; Carpenter & Schacter,
2017; Ragland et al., 2009; Shohamy & Wagner, 2008 – Affordable Custom Essay Writing Service | Write My Essay from Pro Writers; Tripathi & Mishra, 2017).
AIT evaluates the performance on recognizing associations of pair that were
previously presented and on recognizing the relational inference that were created
from these associations (Bunsey & Elchenbaum, 1996; Lau et al., 2010 – Essay Writing Service: Write My Essay by Top-Notch Writer; Preston et
al., 2004). AIT has been initially used to investigate the neural structures influencing relational and associative learning (Preston & Eichenbaum, 2013). AIT was also
used to demonstrate the hippocampal contribution in mediating relational information in human participants (Preston et al., 2004). Preston and collaborators
(2004) used AIT with the aim of evaluating how individual experiences need to be
related or conjoined in new ways to deal with novel situations (Preston et al., 2004).
Later, Zeithamova, Dominick, and Preston (2014: 2024 – Essay Writing Service. Custom Essay Services Cheap) used a modified version of AIT
to assess how hippocampus and ventromedial prefrontal cortex activation correlated during learning of associations. In the present study, we also chose to use the
AIT paradigm to investigate both associative and relational dimensions of memory.
Sleep, Dream, and Memory Learning
Investigations on associative and relational dimensions of memory have been
crossed with other factors, such as sleep. For instance, Lau et al. (2010 – Essay Writing Service: Write My Essay by Top-Notch Writer) were
interested in how specific sleep stages of non-REM and REM sleep contributed to
memory learning. The authors suggested that sleep influences both associative and
relational learning. Interestingly, these dimensions have also been used to consider
the influence of additional factors (such as emotion) on sleep enhancement of
memory. For instance, Alger and Payne (2016: 2024 – Do my homework – Help write my assignment online) demonstrated that participants of a
nap group were less likely to forget association with emotionally neutral stimuli
than participants who stayed awake. Additionally to emotional valence, another
aspect of emotional salience, emotional arousal is complementary to evaluate how
sleep enhance memory (Wagner, Gais, & Born, 2001). These two dimensions of
emotional salience (valence and arousal) are considered in the present study, as
they are both relevant in their links with memory performance.
In the research proposed by Alger and Payne (2016: 2024 – Do my homework – Help write my assignment online), as in most studies
investigating the effect of sleep on associative inference, the sleep period consisted
of a daytime nap. Besides the fact that duration of sleep is likely to affect memory
consolidation process (Achermann & Borbély, 1999; Ambrosini & Giuditta, 2001;
Scullin & Gao, 2018: 2024 – Write My Essay For Me | Essay Writing Service For Your Papers Online), some researchers have also demonstrated that sleeping in a
laboratory setting is susceptible to affect sleep quality and dreams (Agnew Jr.,
Webb, & Williams, 1966; Baekeland & Hoy, 1971; Metersky & Castriotta, 1996;
Toussaint et al., 1997) An habituation phase is usually worthy (Le Bon et al., 2001).
For these reasons, in the present study we preferred a full night’s sleep occurring in
the usual home setting of participants. Although the assessment of overall sleep
duration and quality relies on subjective and behavioral assessments, which are less
precise than polysomnography, we suggest that a typical sleep context situation
could be informative and complementary to data obtained in a laboratory setting.
Dreaming is a prominent feature of sleep. Schredl (2017) has suggested how
dream content is related to memory performances. In search of similarities between
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
SLEEP IMPROVES ASSOCIATIVE AND RELATIONAL MEMORY 173
dream and events encountered during waking period, Vallat, Chatard, Blagrove,
and Ruby (2018: 2024 – Write My Essay For Me | Essay Writing Service For Your Papers Online) evaluated how voice-recorded dream reports were linked with
waking life events of participants. They demonstrated that memories of life events
(encountered during wake) included in dreams were susceptible to changes over
times. For instance, the emotional tones tend to be attenuated toward a more
neutral valence when they are reported within dream reports (Vallat et al., 2018: 2024 – Write My Essay For Me | Essay Writing Service For Your Papers Online).
Wamsley and Stickgold (2019: 2024 – Online Assignment Homework Writing Help Service By Expert Research Writers) replicated their own previous findings showing that
dreaming of a virtual maze navigation task can improve performance at this same
task. We propose in the present study to investigate the links between written
dream reports and associative and relational learning. For this purpose, besides the
memory task, we proposed to participants a computerized diary to collect written
dream reports and sleep information.
Problematic
The present study is conducted to investigate sleep benefit on associative and
relational memory performances when a usual sleep setting and schedule are
proposed. Our main expectation is to observe a quantitative benefit on memory
performance for participants who slept after learning compared with those who
remained awake. We also expected that participants indicating dreams about the
study would have better memory performances on both relational and associative
tasks. We also postulated that emotional salience would influence memory performance and that this influence would be different between participant who slept and
those who did not.
Method
Participants
The experimental sample consisted of 63 participants (age 21.03 3.47
years, gender 54 female, nine male). They were randomly assigned to the two
experimental groups: the sleep group (N 31) and the wake group (N 32).
Participants were recruited through online social media and advertisements posted
in the corridors of a French university. Students who participated in this study were
eligible for course credits. Inclusion criteria were absence of medically asserted
sleep disorder, of medically asserted psychiatric disorder, and of chronic medication
able to interfere with sleep and memory. The study was performed in accordance
with the ethical requirement of the French and European legislation. All participants were instructed about the procedure and completed an informed consent that
guaranteed confidentiality and anonymity of investigations. An identification
number corresponding to each participant was provided. Before experimentation,
for the purpose of pretesting, 42 other students were recruited among the graduate
students of French University (16 for the face data set pretest, 14 for the
photography data set pretest, and 12 for the full protocol testing).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
174 RIBEIRO, GOUNDEN, AND QUAGLINO
Materials
The memory task (material and procedure steps) is based on the work of Alger
and Payne (2016: 2024 – Do my homework – Help write my assignment online) who used an AIT. As a reminder, the rationales of AIT can be
described as such: “If A is associated with B and B is associated with C, then A is
related to C” (see Figure 1).
Associations material. There were 80 visual associations between color
photographs of faces (A and C) in a light gray background and pictures of objects
(like manufactured objects, scenes, animals, etc.) presented in two blocks of 40
associations. Using the labeling used before, color photographs of faces were
Elements A and C and pictures of objects were Elements B.
All the photographs of faces were different (A and C), whereas the pictures of
objects (B) were identical from one block to another. Across the various associations, the position of the photographs of faces (A or C) and the pictures of objects
(B) were randomized (left or right) during the learning phase. All pictures of
objects (B) were displayed with 350px width and 275px height. All photographs of
faces were displayed with 220px width and 275px height.
The faces (A and B) consisted of 80 different photographs (front-faced) from
the FEI database (Thomaz & Giraldi, 2010 – Essay Writing Service: Write My Essay by Top-Notch Writer). There were 40 female faces and 40
male faces. A pretest on 10 participants served to the selection of the faces with the
most neutral expression from this database (a slider ranging from negative, neutral,
and positive was proposed).
The pictures of objects (B) consisted of 40 images (14 negatives, 14 positives,
and 12 neutral) from the OASIS database (Kurdi, Lozano, & Banaji, 2017). The
OASIS database offers 822 images divided into several themes such as manufactured and animate objects and scenes. We proposed images with different
emotional salience. A pretest (N 14) was conducted to ensure that the content of
the image was well understood and that the emotional valence evoked by the
picture of the object was the most positive, neutral, or negative possible (Alger &
Payne, 2016: 2024 – Do my homework – Help write my assignment online). Emotional valence was evaluated using the Self-Assessment Manikin
Figure 1. The AIT was composed of a learning phase (AB and BC), an immediate associative
recognition phase (AB and BC), and two types of delayed recognition. AIT Associative Inference
Tasks.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
SLEEP IMPROVES ASSOCIATIVE AND RELATIONAL MEMORY 175
(Bradley & Lang, 1994). The Self-Assessment Manikin can be used to assess
emotional valence or arousal on a 9-point scale. On this scale are five highly stylized
drawings of a person (the manikin) expressing different dimensions of emotion. For
valence, emotion ranged from very positive to very negative with the center
corresponding to neutral emotion.
Finally, to avoid any confusion between the pictures of objects, we selected
only one image per theme proposed in the OASIS database. In addition, to avoid
confusion between photographs of faces and pictures of objects, we excluded
pictures of objects containing faces facing the camera.
Four different lists of associations were created and counterbalanced among
participants. Associations in Lists 1 and 2 were identical, but the Blocks 1 and 2 were
inverted. In other words, the first block of 40 associations in the List 1 was identical to
the second block in the List 2. In the same way, associations presented in Lists 3 and
4 were identical but Blocks 1 and 2 were inverted. Objects (B) that were paired with
a female face (A or C) in the Lists 1 and 2 were paired with a male face (A or C) in
the Lists 3 and 4. Presentation order within each block was fully randomized.
Tasks procedure. The experimental procedure consisted of a learning phase, an
immediate associative recognition phase, a delayed relational recognition phase, and a
delayed associative phase (see Figure 1 below). The whole AIT was computerized and
elaborated with HTML5, JavaScript, and PHP languages. The software was accessible
within any up-to-date web browser and compatible with smartphone devices.
During learning phases, the instructions indicated to the participant that he or
she will see pairs of photographs composed of a face and an image for 5 s and that
he or she is requested to memorize these pairs and recognize them later. For each
pair, they were encouraged to imagine a link between the two photographs as
proposed by Alger and Payne (2016: 2024 – Do my homework – Help write my assignment online). The first 40 associations (AB) were displayed
on the screen for 5 s each and were followed by a white fixation cross displayed on
a black background for 1 s. The 40 following associations (BC) were displayed on
the screen for 5 s each and were followed by a white fixation cross displayed on a
black background for 1 s.
During the immediate associative recognition phase (AB/BC), each view
consisted of a face presented with four objects. The participant had to identify the
object (B) presented with the face (A or C) before. The distractors were already
seen objects that were paired with same-gender photographs of a face. Recognition
order was fully randomized inside each block. After each recognition, we proposed
an arousal question using the self-assessment manikin regarding the mental
association they elaborated. The wording for the 9-point scale used to rate arousal
was as follows (translated from French):
Irrespective of whether it is negative, positive or neutral, an emotion can provoke reactions
with different intensities. It can be calming or exciting. The left side of the scale below is used
to indicate an emotion that calms, soothes, relaxes. . . . The right side for an emotion that
activates or excites. . . . The center for an emotion that neither soothe or excite.
Questions were presented in a white font on a black background. The procedure
was identical for both the first block (AB) and the second block (BC) of 40
associations each.
The delayed relational recognition phase was proposed after 12-hr delay. For
each of the 40 relational recognition, a main photograph of a face (AC) was
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
176 RIBEIRO, GOUNDEN, AND QUAGLINO
presented alongside four other previously seen photographs of a face. One was
related to the main photograph and the other three distractors were related to other
faces across the learning phases. A question asked if the content of their dream was
linked to their answer.
For the delayed associative recognition phase (AB/BC), the task was identical
to the immediate associative recognition phase. In addition, participants were also
asked if the content of their dream was linked to their answer.
Sleep diary. The instruction of the diary indicated that as soon as they wake
up, the participants should use the diary on their smartphones to describe, as fully
as possible, all their dreams, impressions, and thoughts during the past night. They
were advised to put a text reminder on their alarm clock. All day long, if they
recalled a dream or a night, though, they were requested to add or update their
diary entry and indicate the time of the day.
The smartphone diary was developed to be ready to launch in seconds after
participants awakened. It was developed in dark shade to limit emission of light
(especially blue light), and all questions were displayed on the same page. A
textbook permitted the reporting of dream contents. List questions asked the
participant to indicate: the delay since awakening (in minutes), the sleep type
(night’s sleep or nap), subjective sleep quality ranging from 0 (good) to 4 (not good
at all), the sleep duration ranging from 10 min to 12 hr or more, and the links
between dream mentation and the current research. Each participant also selected
from a list, the response that corresponded to his or her situation: whether he or she
thought that the research did or not influenced the content of his or her dream. The
software automatically generated a timestamp when the diary was open.
The diary was identical for both groups. It was therefore necessary to provide
a task to which both groups could refer to answer questions about the study’s
influence on the content of their dream. For the sleep group, the instruction for this
task was to repeat internally for 5 min before going to bed this sentence “in my
dreams, I will see the things that I saw in this study.” For the wake group, as
participants had not seen the associations yet, they were instructed to repeat
internally for 5 min before going to bed the following sentence “in my dreams, I will
see the things that I ate today.”
To identify which items were incorporated in the dream report, four independent readers, blind to the objectives and assumptions of the present study, were
asked to identify which specific items the dream reports were related. The threshold
for considering that an item was incorporated into the dream content was set at the
majority of the readers (three or more designations out of four).
Questionnaire. We also used two questionnaires in this study. In the initial
questionnaire, we gathered demographic data (gender and age), dream recall
frequency (number of mornings that dreams were recalled during the last 21 days)
of participants, the average duration needed to fall asleep, and their usual sleep
duration. In the final questionnaire using a yes–no response, participants were
asked whether they were able or not to influence their dream according to the
instruction they received.
Dream evaluation. To summarize, dreams were assessed at three different
points in this study. In the Sleep Diary, a drop-down list labeled “Relationship to
current study” had two proposals. The first was “I don’t think the current study
influenced the content of this dream report” and the second was “I think the
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
SLEEP IMPROVES ASSOCIATIVE AND RELATIONAL MEMORY 177
current study influenced this dream report.” During the associative recognition,
there was a “Yes, I don’t know, or No,” question with the following wording: “In
your opinion, was the content of your dreams related to (or influenced) this
answer.” In the final questionnaire, the wording of the yes or no question was:
“Were you able to influence your dreams in the way indicated by the instructions?”
Study Design
Participants were invited to visit a web page that randomly assigned them to
one of the two groups (sleep or wake). The study lasted 3 days (Thursday, Friday,
and Saturday). A detailed description of the procedure is available in Figure 2.
The first day, all participants had to connect to the web page to acknowledge
a step-by-step procedure and fill-in the initial questionnaire. They were also called
for a telephone interview that made it possible for them to fully understand the
protocol. During this interview, we responded to any question they had about the
study and informed them of the opportunity to contact the researcher if they had
any further questions. They were also guided to install the diary on their
smartphone. For the first morning after the learning phase, participants were asked
to fill in their dream diary as soon as they woke up. They also had the opportunity
to complete it at will during the day. If participants reported dreams during the
night, the reports were automatically added to their morning reports. As in Alger
and Payne (2016: 2024 – Do my homework – Help write my assignment online), participants were only aware of the immediate and delayed
associative testing.
To start the learning phase, it was necessary to provide a valid identification
number, and as a protection against uncontrolled practice during the protocol, the
software recorded every launch and sent the data automatically to a secured server.
After a 12-hr delay during which they slept or stayed awake (depending on their
group), participants had to perform the recognition phases (see Figure 1).
The last day, they had to complete the final questionnaire. After completing
the questionnaire, they were automatically notified that the study was finished.
Figure 2. Procedure during the 3-day protocol for the sleep and wake groups.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
178 RIBEIRO, GOUNDEN, AND QUAGLINO
Results
All statistics in the following sections were performed using the R software
(Ihaka & Gentleman, 1996). For the initial questionnaire, comparisons were
performed on sleep quality and demographic data. No significant difference was
found between the two groups. The significance level was set at p .05 for all
comparisons.
Memory Performances
The memory performance scores were based on the number of correct
recognition for the immediate associative recognition phase (on 80), the delayed
associative recognition phase (on 80), and the delayed relational recognition phase
(on 40; Table 1).
For associative recognition, a repeated measure analysis of variance
(ANOVA) was performed (see Figure 3). Condition (sleep or wake) served as
between-subjects factor and type of recognition (immediate/delayed) served as
within factors. There was a significant effect on the type of recognition, F(2, 61)
98.99, p .001, and an interaction effect with testing periods (immediate or
delayed) of recognition and group, F(2, 61) 6.26, p .015. To be specific, the
sleep group score (M 65.4, SD 8.79) was not significantly different to the wake
group score (M 63.7, SD 10.1) during the immediate recall. However, the sleep
group score (M 58.1, SD 9.99) was significantly greater than the wake group
score (M 51.5, SD 12.7) during the delayed recall.
To take account of individual variations in immediate performances, a delta
score was calculated following the strategy used by Alger and Payne (2016: 2024 – Do my homework – Help write my assignment online). The
wake group (M 12.19, SD 9.18) was compared with the sleep group (M
7.29, SD 5.64). Result of the t test confirmed the hypothesis of a less important
deterioration of performance for the sleep group (p .007, Cohen’s d .63).
For each immediate and delayed recognition of a given association, four
different profiles are possible along the learning process: Incorrect–Incorrect,
Incorrect–Correct, Correct–Incorrect, Correct–Correct. We examine whether these
four profiles are represented differently in the two groups, as they refer to distinct
memory fates. For example, the correct–incorrect profile refers to associations that
are known but forgotten during the learning process and the correct– correct profile
refers to associations that are correctly learned and sufficiently consolidated during
the learning process. In other words, each fate can reasonably be considered as
Table 1
Mean (M) and Standard Deviation (SD) for the Recognition Score on the Three Phases of AIT
Phases
Sleep group
M (SD)
Wake group
M (SD)