IMAGINATION, COGNITION AND PERSONALITY, Vol.

29(1) 3-29, 2009-2010

THE MEMORY EXPERIENCES AND DREAMS

QUESTIONNAIRE (MED-Q): A VALIDATED
MEASURE OF DREAM REMEMBERING

CAROLINE L. HORTON, PH.D.

Leeds Metropolitan University, United Kingdom

MARTIN A. CONWAY, PH.D.

University of Leeds, United Kingdom

ABSTRACT

Dream recall has been investigated via various means, often relying upon
self-report judgments. Such methods, as well as individual differences and
cognitive correlates of dream recall, rarely acknowledge models, theory, or
empirical work concerning waking memory. Study 1 presents the develop-
ment and psychometric validation of the Memory Experiences and Dreaming
Questionnaire (MED-Q)—a 30-item measure of autobiographical and dream
memory sensations incorporating items on dreaming, sensory experiences,
and autobiographical remembering behaviors. It produces a single score and
can be broken down into its constituent factors: “awareness of dreaming,”
“daydreaming,” “dream sensations upon waking,” “déjà-states,” “compre-
hensibility of dream content,” and “senses.” Study 2 demonstrates the
validity of the MED-Q as compared with dream report indices of dream
detail. The MED-Q measures the extent to which a person engages with their
dream memories through both frequency and subjective characteristic ratings.
It is therefore novel in emphasizing the context of autobiographical memory
for dreaming.

INTRODUCTION
Any investigation of dreaming relies upon the validity of dream memories. A
reasonably extensive amount of work has been undertaken in order to investigate

Ó 2009, Baywood Publishing Co., Inc.

doi: 10.2190/IC.29.1.b
[Link]
4 / HORTON AND CONWAY

how dreams are recalled, indicating the characteristics of dream memories.

However, research and theories concerning dreams, especially with regard to
interpretation, often ignore the mediating role of memory; that is, that one’s dream
report may omit details or have been constructed after the dream had occurred. The
present article reviews how dreams are recalled, emphasizing the need to consider
processes and models of autobiographical memory; memory for one’s personal
experiences [1, 2]. Two studies are outlined, presenting the development of a
measure of dreaming in light of autobiographical memory sensations and experi-
ences. A new measure of dream memory is demonstrated to be psychometrically
valid (Study 1) and reliable (Study 2), while being novel in acknowledging
the overlap between dream and memory systems. The designed measure—the
Memory Experiences and Dreams Questionnaire (MED-Q)—measures the extent
to which a person engages with their dream memories through both frequency and
subjective characteristic ratings. This differs from previously adopted measures
attempting to quantify dream recall, often in an invalidated way.
Dream recall (DR) is notoriously poor. Most individuals accept that, while
they dream nightly, they rarely remember their dreams. Laboratory investigations
in which dreamers are systematically awakened in different stages of sleep have
provided information about the characteristics of dreams that would have been
forgotten under normal circumstances. Approximately 80% of dreams can be
recalled from rapid eye movement (REM) sleep, and approximately 50% of
dreams from non-REM sleep [3-5]. Three to six dreams can be gathered from
an individual who is awakened in every REM phase per night [6]. Out of the
laboratory there is much variance in the ability to remember dreams and in the
kinds of dreams experienced by different individuals.
The developmental nature of dreaming and DR has demonstrated that young
children are unable to recall dreams [e.g., 4, 7, 8]. It is not known whether or
not some kinds of dreams are experienced in young, pre-lingual children. DR
seems to decline slightly in older adults, although this may be a function of
general memory ability. Older adults also spend less time in REM sleep than
children and younger adults [9]. It is assumed by many theorists [e.g., 10-15]
that the proportion of REM sleep relates to learning and memory consolidation
of waking life experiences.

Measures of DR

A variety of measures of DR exist that largely aim to quantify how many

dreams are recalled in a given timeframe. Dream recall frequency (DRF) measures
include questionnaires, laboratory awakenings, or diary reports measures, which
count the proportion of days in which a dream report was produced. Alternative
measures focus upon dream detail or characteristic information.
Many investigations that rely upon laboratory awakenings in a tightly con-
trolled environment allow DR to be monitored throughout different stages of
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 5

sleep. The discovery of a strong correlation between dreaming and REM sleep
[16] provides a potential methodology for dream research. Previously, researchers
had to rely only on diaries in which people recorded their dreams at home, but
the accumulation of such data by this method can be slow, as people typically
report only two or three dreams each week. By contrast, when subjects are
brought into the laboratory and are awakened during REM sleep, several dream
reports can be collected each night. The laboratory method also has other advan-
tages: the conditions of recall can be controlled and manipulated so that the
effect of distractions and delay intervals can be systematically assessed and
pre-sleep stimulation can be presented so as to study its effect on dream content.
It has become possible, therefore, to apply the methods of cognitive psychology
to dream research and to search for commonalities between memory processes
in waking life and memory processes in dreams [e.g., 6, 17, 18]. However, dream
research is still entirely dependent on self-reports and there is no way in which
dream memories can be judged as accurate or not accurate. Studies are usually
restricted to relatively small numbers of participants and results have tended to
vary from one laboratory to another. Laboratory methods of investigating DR can
also be criticized for their lack of ecological validity, both in terms of memory
processing as well as dream content. In fact, Foulkes and Schmidt [19] demon-
strate that laboratory dream recall prevalence in young children may be reduced, in
comparison to home dream recall, due to the laboratory being a relatively bland
environment, lacking pre-sleep stimulation. Home-recalled dreams, therefore,
may be assessed as a more appropriate alternative methodology in terms of an
investigation of memory processes involved in dream recall. For whatever reason,
cognitive psychologists have so far shown rather little interest in dream research.
Differences have been found between laboratory- and home-recalled dreams,
in which spontaneously recalled morning reports are recalled [4]. REM-sleep
awakenings in the laboratory produce more mundane, realistic, coherent, and
well-formed dreams, which are distinct from the incoherent and strange dreams
in the home-dream reports [20]. Home-dream reports are more likely to have
occurred longer ago than laboratory dreams [4], which may account for the
characteristic and content differences between these two types of dream report.
Hall and Van de Castle [20] found that dreams spontaneously recalled in the
laboratory were significantly longer than spontaneous home dreams, suggesting
that greater selectivity takes place when at home—possibly due to individuals not
focusing upon their dreams when in their everyday routines. Home dreams
are more likely to be generally vivid, dramatic, and emotional than laboratory
dreams. While this could be accounted for by a lack of environmental input
affecting dream content when in a laboratory controlled environment, it may
well be that specific features of dreams, such as bizarreness or emotionality, make
them more memorable, as with waking events. Cipolli et al. [21] have noted a
clear memory advantage for bizarre, compared to non-bizarre, events within
dreams. It has been suggested that this effect represents the richer encoding of the
6 / HORTON AND CONWAY

memory, hence possible greater memorability, as opposed to more bizarre events

occurring. However, as dreams reported from a home setting are likely to be
slightly older than the laboratory dreams in general (as it often takes longer for
a participant to report their dreams at home than in the laboratory), the home
dreams may have to be slightly more memorable in order to be recalled at all.
Thus, as the characteristics of dreams seem to alter their recallability, measuring
dream recall away from the laboratory may be advantageous when investi-
gating the memory processes involved. Indeed, understanding the influences of
different methodologies upon the characteristics of dream memories high-
lights the variability in types of detail of dream memories. Robert and Zadra
[22], for instance, describe retrospective versus prospective measures of
nightmare frequencies.
Montangero, Pasche, and Willequet [23] describe a more valid method of
gathering dream information by first waking an individual in a sleep laboratory or
at home, if EEG information can be collected appropriately (such as with the
Nightcap [24], a portable eyelid movement sensor), then reading that gathered
information back to the participant upon waking in the morning, and allowing
additional information to be reported. In defending this approach, the authors
note that nighttime awakenings cannot necessarily elicit lengthy dream reports
due to a fear that the participant involved will not be able to return to sleep
and that allowing a participant to wake fully would prevent subsequent valid
awakenings (and dream collections) later in the same night. This implies that
the many studies that have employed laboratory awakenings in order to yield
reports of sleeping mentation may not be comparable to DR when awake in a
valid way, if experimenters maintain a sleepy state at the time of recall. As a
result, in most cases dream memories should perhaps be collected from par-
ticipants while awake to improve reliability and comparisons across studies.
This point is especially poignant when considering the comparability of dream
memories with autobiographical memories for waking experiences.
Some researchers prefer a more ecologically valid method for collecting dream
reports whereby participants rely on subjective self-report measures in their
home environment. Typically this involves keeping a dream diary, from which
experimenters can assess either dream detail (the kinds of characteristics or
themes typical of dreams) or DRF (usually by counting how many reports were
elicited in the given timeframe). Questionnaires are also often used as a means
of determining DRF [25]. Self-report measures are easily administered and
cheaply produced, accessing large samples relatively quickly. The present article
describes the development and validation of a questionnaire measure of DR,
involving DRF and dream detail data, in relation to measures of autobiographical
memory sensations.
Self-report measures of dreaming and DR are few and far between. They have
been criticized for being subjective and therefore invalid measures [e.g., 26].
However, a self-report measure that would correlate highly with other measures,
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 7

such as characteristic ratings or diary measures, would therefore be a reliable and

usable measure of DR in a number of settings. For dreams to be recognized as
being memories for an experience, which itself cannot be objectively validated
or even (arguably) encoded during their occurrence, the characteristics of dream
memories need to be explored.
An alternative appropriate measure of DR therefore needs to focus upon
the characteristics of the memory, indicating the strength of the memory trace.
Such characteristics of dreams, in terms of measurable dimensions, were psycho-
metrically outlined by Kallmeyer and Chang [27], who identified four indepen-
dent factors within dreams: dream importance, dream vividness, dream useful-
ness, and dream recall. Thus, the use of these dimensions was encouraged within
dream research. Kallmeyer and Chang argued that the dimensions are indepen-
dent as a result of extensive factor analysis; however, the dream recall factor was
significantly correlated with dream importance (r = .24). This demonstrates the
need for dream and memory characteristics to be further investigated and their
structural relations understood.
Diary measures of DR provide rich data about dreams themselves, although
this is not always a convenient way of gathering data and ignores accurate DRF
measures. A more suitable measure of DR, or preferably dreams as memories,
would employ a questionnaire design. Investigations of DR have used a host of
measures. Questionnaire measures in the past have neither been standardized
nor validated and tend to include just two or three questions requiring participants
to estimate their dreaming habits. Schredl [25] reports that a single question
measure, a 7-point scale in response to the question “How often have you recalled
your dreams recently (in the past several months)?” is reliable over time (although
the time interval is not specified, it is implied to be over weeks). However, DR
can change with routine; thus, a questionnaire item such as this may not be entirely
valid for a particular time. The subjective experiences of dream memories have
not previously been explored. Waking autobiographical memories have been
characterized [e.g., 28], but not in relation to dream memories or experiences.
Dream memories tend not to be investigated by memory researchers.
Researchers who compare the dreaming and waking brain tend not to consider
the mediating role of memory in DR. We propose that in order to adequately
conceptualize dreams as memories, DR should be compared to sensations,
processes, and abilities of waking memory. Applying models of autobiographical
memory to DR should ascertain the extent to which the profile of dreams matches
that of waking autobiographical experiences.

Dreaming and Autobiographical Memory

Although much research has been conducted upon the characteristics of auto-
biographical remembering and false memories [e.g., 29, 30], relatively little
research has investigated the potential cognitive relationships between the
8 / HORTON AND CONWAY

characteristics of autobiographical memories and memories for dreams. Despite

this, evidence suggests that the characteristics facilitating waking event recall such
as bizarreness [31], emotionality [32], and personal salience also facilitate recall
in dreams when they are bizarre [21], emotional [33, 34], and personally salient
[33], for example. As dreaming is an autobiographical experience, it follows that
dreaming relies on autobiographical memory functioning and processes.
Measures of autobiographical memory can refer to sensory experiences or
characteristics as well as quantitative abilities or frequencies. Questionnaires
may not yield reliable measures of ability; however, they allow sensory experi-
ences to be investigated easily. Sensations of memory such as déjà-vu—an
experience whereby there is an overwhelming sense of familiarity juxtaposed
with knowledge that that familiarity is inappropriate [35]—can be insightful in
terms of memory functioning as well as dreaming. In experiential terms, déjà-vu
may feel extremely strange as if, for a fleeting moment, everything that is
perceived has already been perceived before. While the frequency of experi-
encing this sensation has been found to correlate with DR [36, 37], déjà-vu
experiences may also be associated with the confusion of dream content being
interpreted as having actually occurred [38]. The experience also typifies a
sensation of memory with great individual differences variation, just as with
DR. The questionnaire therefore consists of a number of items referring to
different déjà-states, as described by Neppe [38].
Sutin and Robins [28] emphasize the phenomenology of autobiographical
remembering. This can be conceived of in at least two ways: first in terms of
advocating a subjective method of measuring memory experiences; and second
in terms of providing insight into the cognitive sensations involved in memory
processes. Items included in the MED-Q share this focus. Sutin and Robins
reviewed literature and empirical studies of memory in order to find question-
naire items that may be appropriate investigations of the phenomenology of
memory. After a systematic psychometric process, the authors concluded
that 10 independent scales could adequately be used. These were: vividness,
coherence, accessibility, time perspective, sensory detail, emotional intensity,
visual perspective, sharing, distancing, and valence. These have not been investi-
gated thoroughly in relation to dream memories, or other internally-generated
experiences.
Of the few studies investigating the comparisons of autobiographical- and
dream-memory, much evidence has demonstrated the overlap between the
memory systems. Botman and Crovitz [18] found that retention function—a
mathematical depiction of the typical number of recallable memories from any
given period in the lifespan—was also reflected in the recall of dreams. Research
by Grenier et al. [39] also suggests that the similarities between dreaming and
autobiographical remembering extend beyond the characteristics of those
memories in terms of patterns of recall over the lifespan. This includes the three
main features of autobiographical remembering: childhood amnesia (the relative
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 9

inability to recall memories from early childhood); the reminiscence bump (the
disproportionately great recall of memories that occurred between the ages of
approximately 15 and 25); and a recency effect (a heightened propensity to
recall experiences that occurred most recently). This indicates that autobio-
graphical memories for dreams and the autobiographical system operate similarly
over the sleep-wake cycle. Once dreams are able to be recalled, they tend to be
very emotional in the early years [4, 40]. As earliest dreams are seemingly
memorable for some individuals [40, 41], they seemed to serve an appropriate
methodological validation for the questionnaire (see Study 2). That is, significant
correlations between diary measures for the characteristics of earliest as well
as recent dreams and the new questionnaire would provide evidence for the
questionnaire’s reliability and validity. Remembering early experiences have
also been found to correlate with DR [42, 43].
The development of the questionnaire described in the present article adopts a
unique perspective in that it requests information about sensations of autobio-
graphical memory for waking experiences as well as dreams. Autobiographical
memory and dreaming aimed to be explored via designing a questionnaire con-
sisting of items high in face validity and theoretical standing. The psychometric
validation of this measure of dreams as memories, through using items analysis,
indicated the types of characteristics, sensations and experiences that are indica-
tive of remembering dreams.

STUDY 1

Method
Item Selection

Items relating to sensations of dreaming, daydreaming, and autobiographical

remembering were devised based upon having high face validity. Nineteen
hypothesized constructs were devised that related to dream recall and sensa-
tions of autobiographical memory. Each construct was assessed with at least
three questionnaire items. The constructs were: daydreaming [44]; déjà-vu [37];
general sleeping patterns; emotions in dreams [33]; remembering dreams; detail
of dreams; routine of recall; post-dream experiences; senses in dreams; com-
prehensibility; attitudes toward dreaming [45, 46]; intensity of dreams [33];
individual differences in dreams; ruminating, continuity between dreaming and
waking [47]; the development of remembering dreams; perspectives (field or
observer) in dreams; returning/involuntary memories; and feelings of control
in dreams. Most items directly reflected dreaming behaviors, in referring to
frequency of experiencing particular behaviors, assessing depth of detail or vivid-
ness, or accounting for the continuity between dreaming and waking behaviors.
A number of items also referred to experiences of déjà-vu. All questions could be
answered on a 5-point rating scale. These used a variety of frequency measures,
10 / HORTON AND CONWAY

such as “always,” “for some of my dreams but not for others,” “sometimes,”
“for some parts of my dreams but not for others,” “never” as well as ratings
using particular timepoints, such as “daily,” “once/twice a week,” “a few times
a month,” “once a month,” “a few times a year/never.”
Demographic information was also collected, including a question on how
many hours the sample tended to sleep for each night. A mixture of both positively
and negatively scored items was used; that is, for the positive items a lower
score indicated a greater likelihood of recalling dreams (such as 1 for “daily”
or “always”). For the negative items, scores were reversed so that summing
the scores from each item would give a total score for likelihood of recalling
dreams. The result was that all items with a lower score reflected a heightened
propensity to recall dreams. The Memory Experiences and Dreams Questionnaire
(MED-Q) had thus been devised using a number of new measures in the field
of dream memory consisting of 90 questions requesting self-report judgments on
dream behaviors.

Participants

Two hundred thirty-three individuals, opportunistically sampled and mainly

comprising undergraduate BSc Psychology students (30 males, 203 females;
mean age = 24, median = 20 years) from the University of Leeds, participated in
this study by completing a paper questionnaire.

Procedure and Analysis

Questionnaires were distributed individually and in classes for course credit.

Participants were asked to complete the questionnaires in their own time and
instructed to answer all questions. Descriptive responses to all items were studied.
An iterative items (reliability) analysis was conducted in order to ascertain that
only psychometrically stable items were included. Items were removed based on
strict statistical criteria in order to reduce the MED-Q to a concise form. The
finalized MED-Q was factor analyzed using principal components analysis and
its factor structure theoretically confirmed.

Results
Trends

Trends concerning frequency of dream recall and the level of detail recalled
from dreams emerged from the questionnaire responses. Approximately half
(49.1%) of participants felt that they dreamed daily, and 37.5% answered
once/twice a week. The remaining participants felt they dreamed a few times a
month (7.8%), once a month (3.0%), or a few times a year/never (2.6%). Thus,
participants felt that they dreamed frequently. However, asking participants
how often they remembered their dreams gave rise to a different profile of results
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 11

with: just 18% responding daily; 42.1%, once/twice a week; 26.6% a few times
a month; 7.3% remembering their dreams once a month; and 6% a few times
a year or never. This shift toward more occasional DR was reflected in the
responses to the question, “How often do you forget your dreams?” although
the pattern did not mirror those of remembering dreams exactly, as there was
a tendency to answer with a high frequency response in both instances: 21.1%
felt they forgot their dreams daily; 28.8% once/twice a week; 27.6% a few
times a month; 6% once a month; and 6.5% a few times a year or never.
The mean response to the demographic question, “For how many hours
do you sleep each night?” was 2.74 (SD = 1.04), indicating that on average
participants slept for about 7 hours each night. Fourteen items were skewed to
the left (items with a mean below 2) indicating a high frequency of occurrence,
while only 1 (“How often do you experience a sense of revelation upon waking,
despite the dream being confusing?”) was skewed to the right (mean > 4),
indicating a consistently low frequency of occurrence. Those items skewed
to the left related to attitudes toward dreaming, indicating a positive attitude
overall. Participants used the full range of responses available to them in the
closed-ended questions, with the exception of question 58 (“Do you feel that there
are differences between people in the way that they dream?”), whereby no one
responded with “no—people dream in the same ways.” Responses to “Do you
have a generally accepting attitude toward dreams?” omitted “never.”
An open-ended question was included in the questionnaire asking partici-
pants why they felt that there would be differences between the ways in which
people dream. The question included personality, memory abilities, creative
abilities, and whether people wanted to remember their dreams as examples.
Only four responses were collected: they referred to levels of arousal while
asleep and waking; how well rested the individual was; individual differences
in defense mechanisms; and age in that younger people do not remember their
dreams, presumably referring to a cognitive developmental idea of remem-
bering dreams.
Normality of data was indicated from the comparable spread of responses
along items, and was explicitly tested by calculating z-scores and with the
Kolmogorov-Smirnov test. As a result, Pearson’s correlation coefficients were
calculated on this data. High correlations were found between items of the
questionnaire.

Reliability Analyses

In order to produce a psychometrically validated questionnaire measure of DR,

an items analysis (reliability analysis) was performed on the aforementioned
89 [closed-ended] items. The first reliability analysis on these 89 items produced
a Cronbach’s alpha of .869 (.873 when based upon standardized items), indi-
cating reasonable internal consistency of the questionnaire. Performing the items
12 / HORTON AND CONWAY

analysis on only the 80 items that were found to be normally distributed, slightly
reduced the alpha coefficient to .857. The inter-item correlations increased from
.067 to .070. The value was still exceedingly low, reflecting that certain items were
negatively correlated with other items, resulting in a mean inter-item correlation
close to 0. A number of items were removed in an iterative style and the items
analyses re-conducted. As a result of the reduction in Cronbach’s alpha1 the
non-normally distributed items were re-entered into the items analysis.
First, some items were removed in a block due to being relatively low in face
validity. They referred to developmental aspects of dreaming, such as dreaming
similar content when younger to the current time. While these developmental
aspects are worthy of study, these items only questionably loaded onto the overall
designated construct of DR. In addition, items concerning opinions on other
people’s dreaming behaviors were also removed due to being low in face validity
as, while they corresponded to attitudes toward dreams generally, they did not
correspond to attitudes to one’s own dreams. They also provided confusing output
in the correlation matrix; that is, some of the correlations were negative. In
addition, the item concerning opinions on whether differences existed between
people in the ways that they dreamed was also removed as the responses were
not distributed evenly (there were no “never” responses). In addition, some of the
reversed items still produced a substantial number of negative correlations with
other items, indicating an unclear relationship to the overall construct of recalling
dreams, so they were also removed from subsequent analyses.
The iterative items analysis thus began with the remaining 68 items. That is, an
items analysis was conducted and output scrutinized. One item would be selected
for removal as based upon the following criteria: being low in face validity;
correlations were either low or negative; item means being especially low or
high; an item’s deletion would have little effect on the overall scale mean, but a
great effect on the scale variance; the item had a low corrected item-total cor-
relation (a final value of .5 was aimed for); the Cronbach’s alpha would rise if
the item was deleted; and low squared multiple correlations were found for the
item, indicating that the item was poorly predicted by other items. After removing
the identified item, the analysis was re-conducted. This process was repeated
until a succinct number of items remained which all contributed statistically to
an overall construct of DR. The initial items analysis on 68 items produced a
Cronbach’s alpha which had increased from a previously low value to .903 (.905
when based on standardized items) and a mean inter-item correlation of .123
which, although was substantially improved, still required increasing further.
Table 1 conveys the summary statistics for the resulting items. After the
iterative analyses a 30-item questionnaire remained. Questionnaire items are

1 The internal consistency may be more viewed as being more important in this questionnaire than
the mean inter-item correlation, due to the hypothesized non-unidimensionality of the questionnaire.
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 13

Table 1. Summary Item Statistics for the 30 Items

in the Final MED-Q

Mean Minimum Maximum Range Variance

Item means 2.67 1.52 4.02 2.51 0.49

Item variances 1.16 0.44 1.86 1.42 0.11

Inter-item

Covariances 0.25 –0.01 0.98 0.99 0.02

Inter-item

Correlations 0.22 –0.01 0.71 0.71 0.02

detailed in Table 2. Some remaining items had low means; however, they
remained in the questionnaire because this reason alone did not suffice in
removing them. Their variance was suitable and they tended to be especially
high on face validity. All item total correlations were over .3. While it would
be ideal for values to approach .5, only few items remained and those below .3
tended to have high squared multiple correlations. Again, removing them on
this basis alone did not suffice. Also, as the questionnaire was designed so as
not to be unidimensional, lower coefficients were expected. No items in the
remaining 30-item questionnaire were negatively framed; that is, all items
were able to be summed in order to produce a single value reflecting DR. A
lower score indicated a higher propensity to recall dreams.
Finally, the Cronbach’s alpha was found to be .891 (when based on stan-
dardized items it is similar at .893) and the inter-item correlation had increased
to .218.

Factor Analyses

A series of factor analyses were conducted in order to ascertain the structure

of the concise MED-Q. Initial unrotated solutions produced a first factor, with
an eigenvalue of 7.42 and 24.74% variance accounted for, with all 30 items
loading onto it with loading scores of at least .35. Most loading scores were
around .5. This indicated a degree of unidimensionality; all items were measuring
the same construct of remembering dreams. Seven other factors were produced
with eigenvalues greater than 1, although their distributions were confusing and
a scree plot seemed to better imply that just six factors were extracted altogether.
In order to clarify the factor structure, a number of analyses were conducted,
including unrotated analyses and Varimax rotations forcing three, four, five, and
14 / HORTON AND CONWAY

Table 2. Items Comprising the Final MED-Q

1. How often do you feel you dream?

2. How often do you experience the sensation of daydreaming?
3. How frequently do you find yourself ruminating?
4. How often do you remember your dreams?
5. How frequently do you find yourself ruminating and unable to return to a task?
6. How often are you aware that you have been dreaming?
7. How often do you experience the sensation of involuntary memories
interrupting you?
8. How often do you experience a sense of revelation upon waking?
9. How often are you distracted by daydreaming?
10. Do you feel you often understand why you dreamed particular material?
11. Do you dream emotionally positive dreams?
12. How frequently do you wake with a lingering sensation of bizarreness?
13. How frequently do you wake with a sensation of clarity of understanding?
14. How often do you experience a sense of revelation upon waking, despite
the dream having been confusing?
15. How frequently do you wake with insights into the meaning of the dream?
16. Do you feel you understand why you dream the material you dream?
17. How often does a memory of a dream involuntarily return to mind?
18. How frequently does a feeling of a dream, although not its memory,
involuntarily return to mind?
19. Do you have particularly emotional dreams?
20. Do you feel that you remember your dreams routinely?
21. Are you ever aware of being able to use the sense of smell in your dreams?
22. Do you dream especially intense or clear dreams?
23. Do you dream especially vivid dreams?
24. Are you ever aware of being able to use the sense of taste in your dreams?
25. Have you ever experienced the sensation of déjà-vu?
26. Have you ever experienced the sensation of déjà-entendu?
27. Have you ever experienced the sensation of déjà-fait?
28. Have you ever experienced the sensation of déjà-arrive?
29. How intense are your daydreams?
30. How intense are your dreams?
All items were scored on a 5-point scale with a lower score indicating a higher propensity
to recall dreams.
Items 1-18 responses: daily, once/twice a week, a few times a month, once a month, a
few times a year/never.
Item 19 responses: usually, for some of my dreams but not for others, sometimes, for some
parts of my dreams but not for others, never.
Item 20 responses: extremely consistent, pretty consistent, not sure, pretty inconsistent,
extremely inconsistent.
Items 21-28 responses: (yes) frequently, (yes) occasionally, I have been aware of this,
(no) but I probably could, (no)—I doubt I am able to use that sense.
Items 29 and 30 responses: extremely vivid and sensory-perceptual in nature, usually
vivid and sensory-perceptual in nature, not vivid but clear, occasionally vivid but usually
more vague, if I [day]dream at all they are vague.
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 15

six factors. There were few differences between the factor structures of the
Varimax rotated principal components analysis and the structure whereby six
factors were forced. Thus, the finalized factor structure is detailed in Table 3.
The initial (unforced) structure contained a separate factor for “intensity of
dreams” and “awareness of having dreamed,” which were combined into “aware-
ness of dreaming” in the final analyses, and also a separate factor (factor 8) for
“emotionality,” although this only contained the two emotion-related items as
well as the anomaly “do you feel that you remember your dreams routinely?”
This was amended in the forced structure so the anomalous item loaded onto
factor 1, alongside other similar items; “Do you dream emotionally positive
dreams?” was grouped with the other senses items; and “do you dream particularly
emotional dreams” was omitted from interpretations as its loading score was
< 0.4 (in actuality it loaded onto factor 1 with a low loading score of just 0.37).
The final factor structure of the Varimax rotated principal components analysis,
forcing six factors, was as follows: Factor 1: “awareness of dreaming” (eight
items); Factor 2: “daydreaming” (six items); Factor 3: “dream sensations upon
waking” (six items); Factor 4: déjà-states (four items); Factor 5: “comprehens-
ibility of dream content” (three items); and Factor 6: “senses” (three items).
Finalized “factors” can be achieved by simply summing the responses to the
items that make up that factor. Comprising item numbers are stated in brackets
in Table 3. For instance, to calculate a score for the “daydreaming” factor,
one would need to sum raw scores for items 2, 3, 5, 7, 9, and 29 for each indi-
vidual participant.
Values in Table 3 correspond to the degree to which each item loads onto the
factors described. Specifically, factor loadings represent the partial correlation
between the item and the rotated factor, ranging from 0 to 1. It can be seen that
in all factors values are high, with the possible exception of the final item
corresponding to dreams being emotionally positive. However, this item has
been included, nonetheless, due to it not meeting criteria for exclusion during
the items analysis.
Discussion
This study produced a clearly structured 30-item measure of DR based upon
items pertaining to the sensations of dreaming and autobiographical memory
more generally, as being high in face validity.
No differences in DR were found between males and females. However,
the sample was predominantly female, making clear comparisons difficult. This
gender bias was the result of recruitment difficulties. As the questionnaire was
lengthy, non-students had little incentive for its completion. (Students could be
rewarded with course credits, resulting in a largely student-based sample.) The
majority of Psychology undergraduates are female; however, the literature on
gender differences in DR does not provide reliable and convincing evidence to
suggest that females are more likely to remember their dreams than males.
 Table 3. Loading Plot Showing the 30 Items of the MED-Q and its Varimax
Rotated Component Matrix

Awareness Dream Compre-

of Day- sensations Déjà- hensibility
Item dreaming dreaming upon waking states of content Senses
16 / HORTON AND CONWAY

Do you dream especially vivid dreams (23) 0.74

Do you dream especially intense or clear dreams? (22) 0.71
How often do you remember your dreams? (4) 0.70
How intense are your dreams? (30) 0.70
How often are you aware that you have been dreaming? (6) 0.65
How often do you feel you dream? (1) 0.64
Do you feel that you remember your dreams routinely? (20) 0.49
Do you have particularly emotional dreams? (19) 0.37

How often are you distracted by daydreaming? (9) 0.83

How frequently do you find yourself ruminating? (3) 0.76
How often do you experience the sensation of 0.75
daydreaming? (2)
How frequently do you find yourself ruminating and unable 0.70
to return to a task? (5)
How often do you experience the sensation of involuntary 0.64
memories interrupting you? (7)
How intense are your daydreams? (29) 0.59
How often do you experience a sense of revelation upon 0.79
waking, despite confusing dream? (14)
How frequently wake with sensation of clarity of under- 0.74
standing? (13)
How often do you experience a sense of revelation upon 0.70
waking? (8)
How frequently wake with lingering sensation of bizarreness? (12) 0.54
Frequency of feeling of a dream, although not memory, 0.50
involuntarily return to mind? (18)
How often does a memory of a dream involuntarily return 0.48
to mind? (17)

Have you ever experienced the déjà-fait? (27) 0.83

déjà-vu? (25) 0.76
déjà-arrive? (28) 0.73
déjà-entendu? (26) 0.67

Do you feel you understand why you dream the material 0.86
you dream? (16)
Do you feel you often understand why you dreamed 0.77
particular material? (10)
How frequently wake with insights into the meaning of the 0.76
dream? (15)

Are you ever aware of being able to use the sense of taste 0.74
in your dreams? (24)
MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE

Are you ever aware of being able to use the sense of smell 0.69
in your dreams? (21)
Do you dream emotionally positive dreams? (11) 0.45
/ 17
18 / HORTON AND CONWAY

Structure of MED-Q

Despite the sample being student based and disproportionately female, it was
large enough for adequate factor analyses and correlations to be conducted.
This was based on the fact that a previously conducted unpublished study [48]
in which a similar questionnaire was distributed to undergraduate Psychology
students at the University of Durham, produced an average correlation coefficient
between all variables of 0.28. According to Baggaley [49], who stipulated that
the required sample size should equal this value multiplied by the number of
items, a sample size of at least 185 was required (90 items × 2.06). This was well
exceeded, so the analyses can be considered statistically thorough. High cor-
relations were also found between all items. In addition, the mean correlation
coefficient between all 89 questionnaire items included in the initial reliability
analyses was .09 from this study. This increased to .22 for the 30 items on the
final MED-Q. As this is high and similar to that found in the previous study, a
degree of reliability over time for these measures has been demonstrated.
While it may be deemed that the MED-Q has a reasonably complicated struc-
ture with six factors, the analyses were rigorous and systematic, and the factor
analyses were confirmatory (in that they related to hypothesized constructs)
rather than being totally exploratory. Also, there were many similarities between
the factor structure of the questionnaire pre- and post-reliability analyses, indi-
cating that the items were removed based upon statistical grounds while the
theoretical structure was upheld.
Certain items on the questionnaire had a potentially discontinuous response
scale. Specifically, items 27-49 referred to frequencies of particular dreaming
behaviors; however, instead of a structured pre-defined frequency scale, a more
subjective scale was included (usually: for some dreams, sometimes; for some
parts of dreams, never). All these items, with the exception of item 43 (Do you
dream particularly emotional dreams?) were removed. This demonstrates the
advantages of clearly defining frequencies with objective timepoints in order to
specify its ordinal nature.

Findings from the MED-Q

While this study aimed to devise a measure of DR, it also provided some
initial data on the norms of dreaming in a student population.
The open-ended responses to question 84 (“Why do you think there are dif-
ferences between people in the way that they dream?”) produced only four
responses; however, these referred to widely documented theories of individual
differences in DR (such as cortical arousal, personality dimensions of creativity,
differences in memory abilities, and age) reflecting cognitive developmental
theories of dreaming. Participants were, therefore, aware of the processes sur-
rounding DR. This is reassuring considering the self-report nature of the MED-Q
as a measure of DR.
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 19

Overall there was less variance in the MED-Q scores than anticipated.
However, investigating this in terms of individual differences in other memory
abilities or personality traits would further elucidate the extent of the variance in
the MED-Q scores [see 50].

STUDY 2

Study 1 outlined the development and psychometric validation of a 30-item

questionnaire measure of DR. Study 2 aimed to extend the validity of the MED-Q
by correlating its scores with alternative measures of DR. Measurements were
taken from actual dream reports to see if more commonly used diary-style
paradigms could predict scores on the newly formed measure, which shed light
onto the reliability of these two distinct forms of measuring dream memories.
Diary measures are comparable to questionnaire measures in that they provide
self-reported data, gathered out of the laboratory. In addition they can provide a
richer account of the phenomenology of dream memory. The present study aimed
to extend the development a psychometrically sound questionnaire measure of
dream memories, based upon items relating to dreaming and autobiographical
memory, by classifying the behaviors associated with recalling recent and
earliest dreams. Such an approach may help to clarify whether a likelihood of
recalling dreams is a cognitive ability or one that is more associated with certain
personality traits.
Dream memories may be similar to childhood memories, as both lack suffi-
cient knowledge, cognitive processes, and frontal development for a referenced
memory to be encoded. Tulving [51] refers to such contextless memories as “free
radicals” as they are neither truly episodic (their source is forgotten) nor semantic,
yet they are both autobiographical and experiential. These memories are often
encoded without contextual information, unless, for instance, that is especially
pertinent to the memory itself. For dreams this is likely due to attenuated frontal
activation during sleep, which renders encoding of sleep mentation difficult.
For childhood memories, in line with Domhoff [7] and Howe and Courage’s [52]
ideas concerning the development of cognitive abilities,2 the lack of a sophisti-
cated cognitive structure into which newly acquired memories can be assimilated
results in autobiographical memories being difficult to encode sufficiently for
long-term storage and subsequent recall. Dreams and even waking thoughts
to an extent share this profile of free radicals, thus emphasizing the overlap
between dreaming and autobiographical memory.
Due to such a cognitive developmental consideration of remembering dreams
as autobiographical experiences, the present study requested information on older

2 For Domhoff [7], one can only possess the ability to remember dreams once sufficient cognitive
structures have developed. For Howe and Courage [52], young individuals may not recall autobio-
graphical experiences due to the lack of development of a cognitive self.
20 / HORTON AND CONWAY

and recent dreams. While the MED-Q was administered, participants were
asked to report three recent and three earliest dreams. These were counted for a
simple quantitative measure of DR and analyzed for detail and sensory richness,
in order that they could be correlated with the MED-Q scores. The present study
demonstrates that the MED-Q is both a reliable and valid measure of DR.

Method

Participants, Materials, and Design

The same 233 participants, described above, completed this second part of the
questionnaire. In addition to the 90 items described in Study 1, which were
designed to measure dream memories, the second part required participants to
recall three early and three recent dreams. The dreams were reported alongside the
estimated time of occurrence. Participants were instructed to describe their three
most recent, as well as their three earliest, dreams that they could remember.
They were reminded that they should not feel embarrassed by the reported
material as it was not going to be made available beyond the experiment nor
content analyzed, and it was going to be dealt with anonymously and confi-
dentially. Participants could report as much as they wished. In addition, they
were asked to report reasons why they felt they had dreamed the material and
reasons why they felt they could remember that particular dream. Reports were
scored according to detail (the extent to which events from the dream were
reported) and sensory richness (sensory-perceptual and emotional experiences)
on a 5-point scale of: 1 = none, 2 = very few details/experiences reported, 3 =
some details/experiences reported but it is clear that there are omissions, or the
sensation was not strong, 4 = lots of detail/a strong sensation reported, and 5 =
all details seem to be recalled; there are no gaps in the report. In addition, for
the sensory richness measure, a score of 5 indicated that the sensations were so
profound they may well have indicated that participants were woken up by
the experiences, or there was a continuity of the experience into waking life,
such as waking up crying or screaming. These ratings can be alikened to the
detail and episodic richness ratings of the Autobiographical Memory Interview
(AMI) [53] respectively, although on a 5- as opposed to a 4-point scale.

Procedure

Questionnaires were distributed individually and in classes for course credit.

Participants were asked to complete the questionnaire, taking as much time as
was required, and instructed to answer all questions. If participants could not
remember dreams, they were asked to indicate so. Participants were also asked
to indicate if they were not reporting a dream due to its content that they
preferred not to reveal, so this missing data could be labeled appropriately.
Completion of the questionnaire took approximately 45 minutes. In order to
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 21

ascertain methodological reliability, the finalized 30-item MED-Q as resulting

from Study 1 were also correlated and regressed with the length, detail, sensory
richness, and other measures of the diary style reports of both the earliest
and recent dreams.

Results

For each of the three early and three recent dreams that were reported, the
time since occurrence (in months for early dreams, in days for recent dreams),
the age of the respondent when they had had the dream, length of the report
(word count), and the detail and sensory richness scores of these were recorded.
Table 4 shows the descriptive trends for these variables. A higher score for both
detail and sensory richness indicated more detail or sensory richness, respec-
tively, in the dreams. The N statistic depicts the number of participants who
managed to recall all three dreams at either the early or the recent stage. Of the
233 participants altogether, approximately just over half managed to recall all
the dreams that had been requested.
The lengths of the dream reports differed significantly between the early and
recent dreams (T (95) = –4.07, p < 0.01) with the early reports being on average
55 words longer than the recent reports. This is contrasted with a significant
difference in the detail scores of the reports (T (95) = –5.14, p < 0.001) whereby
the recent dreams were more detailed than the early dreams, with an approxi-
mate average mean difference of 0.4. However, the sensory richness scores,

Table 4. Descriptive Statistics for the Mean Early and

Recent Dream Reports

Std.
N Range Mean deviation

Time early (months) 123 716 175.77 118.97

Time recent (days) 119 3242 117.10 313.39
Age early 124 58 10.07 7.11
Age recent 119 70 23.02 9.20
Length early 124 250 43.20 35.76
Length recent 119 267 55.42 46.20
Detail early 124 5 3.06 0.89
Detail recent 119 5 3.50 1.03
Sensory richness early 124 5 3.88 0.82
Sensory richness recent 119 5 3.56 0.85
22 / HORTON AND CONWAY

which also differed significantly (T (95) = 3.59, p < 0.001), were in line with
the differences in length: the earlier reports scored higher on sensory richness
by approximately .25 on the 5-point rating scale.

Relationship to Gender

No significant differences between males and females were found on any of

the measurements (time, age, length, detail, and sensory richness) for either the
early or the recent reports.

Validity of the MED-Q

In order to assess the validity of the MED-Q total score as it relates to the
dream report measures, various correlations were computed and are summarized
in Table 5. Specifically, the detail and sensory richness ratings of both the early
and the recent dreams correlated significantly with the MED-Q scores. Although
these coefficients are not impressively high, there is enough shared variance to
indicate that dream reports and self-report questionnaire items are generally
measuring the same underlying phenomenon.
In addition, significant correlations indicated that the more time between the
occurrence of the earliest recorded dream, the less great the DR, and the younger
the age at the time of the recent dreams’ occurrence, the greater the DR. This led
to the investigation of associations with age generally. Indeed, total score was
significantly correlated with age in months (r = .243, p < 0.001) implying that
dream recallability decreases as one ages (note that although this relationship
is not negative, a lower score for dream recallability indicates heightened recall).
Many relationships between the lengths of the reports and other scores were
found: early lengths correlated with early sensory richness. Both early and recent
detail scores significantly correlated with early length. There was also a signifi-
cant correlation between the early and recent lengths, indicating a possible effect
of individual differences. Recent length scores correlated with both early and
recent detail scores and early sensory richness scores. Recent sensory richness
scores did not correlate with any length measurements. Early detail scores cor-
related significantly with both early and recent sensory richness scores, and
also with recent detail scores. Recent age scores correlated negatively with the
recent detail scores, indicating that the older the participant at the time of
dreaming, the more detailed the report. However, this was not the case in terms
of time since dream occurrence. Other significant correlations for the recent
detail scores were with both early and recent sensory richness scores. Both early
and recent sensory richness scores correlated with one another.
As demonstrated in Table 5, specific factor scores correlated significantly with
some of the dream report measures (apart from the senses factor). Awareness
of dreaming correlated with length of recent dreams (r = –.23, p < 0.01), detail of
early (r = –.22, p < 0.01), and recent dreams (r = –.29, p < 0.01), and sensory
 Table 5. Correlation Matrix Showing the Relationships between Dream Report Measures and
the MED-Q Total Score
Factor Factor Factor Factor Factor Factor Time Time Age Age Length Length Detail Detail SR SR
1 2 3 4 5 6 early recent early recent early recent early recent early recent

DMQ Total 0.770 0.658 0.750 0.514 0.582 0.645 0.228 0.045 0.066 0.234 –0.026 –0.003 –0.273 –0.212 –0.260 –0.326
Factor 1 0.320 0.471 0.258 0.408 0.409 0.130 –0.053 –0.051 0.011 –0.143 –0.234 –0.221 –0.286 –0.196 –0.293
Factor 2 0.325 0.217 0.222 0.276 0.208 0.149 0.310 0.295 –0.037 0.060 –0.191 –0.113 –0.096 –0.128
Factor 3 0.306 0.347 0.430 0.318 0.011 –0.107 0.253 0.003 –0.070 –0.288 –0.204 –0.224 –0.256
Factor 4 0.186 0.367 0.112 –0.055 –0.011 0.049 0.072 0.163 –0.158 –0.013 –0.197 –0.190
Factor 5 0.385 0.042 0.054 –0.049 0.037 –0.008 0.060 –0.089 –0.097 –0.191 –0.153
Factor 6 0.134 0.053 0.025 0.122 0.027 0.100 –0.101 –0.052 –0.034 –0.045
Time early 0.092 –0.174 0.877 –0.023 –0.053 –0.159 –0.107 0.022 0.104
Time recent 0.217 0.186 –0.058 –0.019 –0.063 –0.046 –0.109 0.030
Age early 0.168 0.082 0.157 0.067 0.111 0.162 0.189
Age recent 0.054 0.088 –0.141 –0.115 0.039 0.069
Length early 0.727 0.430 0.284 0.300 0.146
Length recent 0.219 0.459 0.249 0.158
Detail early 0.696 0.644 0.628
Detail recent 0.592 0.488
SR early 0.720

Values in bold type reflect correlations significant at the p < 0.05 alpha level.
MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE

Values in bold and italics indicate correlations significant at the p < 0.01 alpha level.
SR = sensory richness
/ 23
24 / HORTON AND CONWAY

richness of early (r = –.20, p < 0.05) and recent (r = –.29, p < 0.001) dreams.
Daydreaming related to time since the early dreams (r = .21, p < 0.05), age at the
early dreams (r = .31, p < 0.01), age at the recent dreams (r = .30, p < 0.01), and
detail of the early dreams (r = –.19, p < 0.05). Dream sensations upon waking
related to time since the early dreams (r = .32, p < 0.01), age at the recent dreams
(r = .25, p < 0.01), detail of recent dreams (r = –.29, p < 0.01), and sensory richness
of early (r = –.20, p < 0.05) and recent (r = –.22, p < 0.01) dreams. Déjà-states
related to detail of early (r = –.16, p < 0.05) and recent (r = –.16, p < 0.05) dreams,
and sensory richness of recent dreams (r = –.20, p < 0.05). Finally, compre-
hensibility of dream content related to sensory richness scores of recent dreams
(r = –.19, p < 0.05).
In summary, the daydreaming factor related to the time since the early dreams
had occurred. Other factors (namely, awareness of dreaming, dream sensations
upon waking, déjà states, and comprehensibility of content) related to measures
of the strength of the memory trace—that is, detail and sensory richness scores.
These measures are also appropriate measures of dream characteristics. In addi-
tion, the awareness of dreaming factor correlated with length as well as the detail and
sensory richness ratings, and is therefore an especially good measure of actual DR.
A host of regression analyses were also conducted, with dream report measures
found to significantly correlate with each factor entered as predictors. No models
suitably fitted the data.

Discussion

As the regression equations predicting the MED-Q with aspects of the dream
report measures were not significant, it appears as if the dream report aspect of
the questionnaire cannot accurately predict the total scores from the validated
items from the 30-item MED-Q. Thus, the different aspects of remembering
dreams may not be similarly composed of cognitive (i.e., memory based) aspects
as well as the detail, length, and sensory richness ratings of dream reports
themselves. It is worth remembering that these measures are different and thus it
is not surprising that the two aspects possibly do not tap into the same underlying
construct. However, they were significantly correlated and so there was some
degree of agreement between the measurements. In addition, there were some
significant findings when the factors of the MED-Q were regressed separately.
This does question the validity and comparability of the findings of studies
which have used questionnaire and diary-based measures of DR. This is especially
the case when considering how many items requesting participants to reveal
how frequently they thought they dreamed were removed from the question-
naire during the reliability analyses. In addition, the items reflecting opinions
about dreaming were, on the whole, the first to be removed. This may imply
that standardized measures such as this newly validated questionnaire should
be encouraged when attempting to assess DR, as opposed to relying upon
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 25

participants’ judgments, which have demonstrated a vast array of individual

difference effects. Incidentally, there is much disagreement over the theoretical
bases for such effects.
A clear profile emerged in Study 2 that demonstrated the aspects of auto-
biographical remembering that each factor of the MED-Q relates to. In particular,
factor 1 (“awareness of dreaming”) was associated with longer and more detailed
dream reports, while the “dream sensations upon waking” factor reflected the
sensory richness of the dream. The other factors were associated with some of the
measures from each of those aspects. The dream length and detail aspects may be
seen to be a measure of dream recall frequency and strength of the dream memory
trace, while the more sensory aspects may relate to the episodic nature of the
memory in a more experiential manner. The MED-Q thus seems to tap in to both
of these aspects of remembering.
The measures from the dream report part of the questionnaire that best cor-
related with the factors of the MED-Q were detail (early) and sensory richness
(early and recent). These ratings are very similar to the detail and episodic richness
(respectively for detail and sensory richness) ratings used in the Autobiographical
Memory Interview [53]. Thus, similar methodologies for rating memories and
dreams can be adopted. While statistical agreement between the MED-Q and the
dream report measures were found, this reflects how the MED-Q is a reliable
measure of DR. However measuring DR in numerous ways should also be
encouraged, as diary report measures or laboratory awakenings may still produce
different indices of DR, and they do not all seem highly reliable [4, 54].

GENERAL DISCUSSION

Taken altogether the MED-Q, composed of six factors, appears to be a well-

rounded measure of DR. Both detail ratings (from diary style measures) and
self-report questionnaires are significantly correlated, which has implications
for the methodological reliability of the construct of DR. In addition, investi-
gations of DR using simplistic measures of just one or two self-report questions
or detail measures of reports, for instance, are reminded that DR is composed of
a number of different factors. Individuals interested in specific aspects of DR
are encouraged to use items of this questionnaire that tap into those specific
constructs, such as detail, sensory richness, or more sensation-based experiences.
While this study has demonstrated that DR relates to aspects of autobio-
graphical remembering such as rumination or experiencing the sensations of
déjà-vu, there is much evidence to suggest that DR is at least as much a product
of individual differences in personality, as individual differences in memory
styles [26, 55]. Thus, the MED-Q needs to be correlated with a number of trait
measurements in order to further incorporate the MED-Q as a measure of DR
into the literature on dreaming, as well as to empirically investigate the contentious
relationships between DR and personality.
26 / HORTON AND CONWAY

A concise measurement of DR was developed, which was subsequently

found to relate to a number of alternative measurements of dream recallability,
such as time since the earliest remembered dream and detail and sensory richness
(like episodic richness) ratings of earliest and recent dream reports. Its relation-
ship with measurements of earliest dreams indicates two things. First, memory
for early experiences may require a particular autobiographical memory
profile that is conducive to recalling dreams. Second, early dreams may be
particularly detailed or vivid and therefore predictive of current dream patterns.
Such dreams that are characteristically detailed or episodically rich early in life
may continue to be so throughout life. This implies that individuals who are
likely to dream particular kinds of dreams may have a personality profile that is
sensitive toward DR.

REFERENCES

1. Conway, M. A. (1990). Autobiographical memory: An introduction. Milton Keynes:

Open University Press.
2. Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobio-
graphical memories in the self-memory system. Psychological Review, 107, 261-288.
3. Foulkes, D. (1962). Dream reports from different stages of sleep. Journal of Abnormal
and Social Psychology, 65(14), 14-25.
4. Foulkes, D. (1979). Home and laboratory dreams: Four empirical studies and a
conceptual reevaluation. Sleep, 2(2), 233-251.
5. Nielsen, T. A. (2000). A review of mentation in REM and NREM sleep: “Covert”
REM sleep as possible reconciliation of two opposing models. Behavioral and
Brain Sciences, 23, 851-866.
6. Meier, C. A., Ruef, H., Ziegler, A., & Hall, C. S. (1968). Forgetting of dreams in the
laboratory. Perceptual and Motor Skills, 26, 551-557.
7. Domhoff, G. W. (2001). A new neurocognitive theory of dreams. Dreaming, 11(1),
13-33.
8. Domhoff, G. W. (2002) The scientific study of dreams: Neural networks, cognitive
development and content analysis. APA Press.
9. Foulkes, D. (1982). A cognitive-psychological model of REM dream production.
Sleep, 5(2), 169-187.
10. Cavallero, C., Foulkes, D., Hollifield, M., & Terry, R. (1990). Memory sources of
REM and NREM dreams. Sleep, 13(5), 449-455.
11. Cipolli, C., Faglioni, I., Mazzetti, M., & Tuozzi, G. (2005). Consolidation effect of
the processing of declarative knowledge during human sleep: Evidence from long-
term retention of interrelated contents of mental sleep experiences. Brain Research
Bulletin, 65, 97-104.
12. Plihal, W., & Born, J. (1997). Effects of early and late nocturnal sleep on declarative
and procedural memory. Journal of Cognitive Neuroscience, 9, 534-547.
13. Rauchs, G., Bertran, F., Guillery-Gurard, B., Desgranges, B., Kerrouche, N., Denise,
P., et al. (2004). Consolidation of strictly episodic memories mainly requires Rapid
Eye Movement sleep. Sleep, 27(3), 395-401.
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 27

14. Rauchs, G., Desgranges, P., Foret, J., & Eustache, F. (2005). The relationship between
memory systems and sleep stages. Journal of Sleep Research, 14, 123-140.
15. Stickgold, R. (2006). A memory boost while you sleep. Nature, 444, 559-560.
16. Aserinsky, E., & Kleitman, N. (1953). Regular periods of eye motility and concurrent
phenomena during sleep. Science, 118, 273-274.
17. Badia, P. (1990). Memories in sleep: Old and new. In R. R. Bootzin, J. F. Kihlstrom, &
D. L. Schachter (Eds.), Sleep and cognition. Washington, DC: American Psycho-
logical Association.
18. Botman, H. I., & Crovitz, H. F. (1989-1990). Dream reports and autobiographical
memory. Perceptual and Motor Skills, 9(3), 213-224.
19. Foulkes, D., & Schmidt, M. (1983). Temporal sequences and unit composition in
dream reports from different stages of sleep. Sleep, 6(3), 265-280.
20. Hall, C. S., & Van de Castle, R. L. (1966). The content analysis of dreams. New York:
Appleton-Century-Crofts.
21. Cipolli, C., Bolzani, R., Cornoldi, C., De Beni, R., & Fagioli, I. (1993). Bizarreness
effect in dream recall. Sleep, 16(2), 163-170.
22. Robert, G., & Zadra, A. (2008) Measuring nightmare and bad dream frequency:
Impact of retrospective and prospective instruments. Journal of Sleep Research, 17,
132-139.
23. Montangero, J., Pasche, P., & Willequet, P. (1996). Remembering and communicating
the dream experience: What does a complementary morning report add to the night
report? Dreaming, 6(2), 131-144.
24. Mamelak, A. N., & Hobson, J. A. (1989). Nightcap: A home-based sleep monitoring
system. Sleep, 12, 157-166.
25. Schredl, M. (2004). Reliability and stability of a dream recall frequency scale.
Perceptual and Motor Skills, 98, 1422-1426.
26. Schredl, M., & Montasser, A. (1996-7). Dream recall: State or trait variable? Part I:
Model, theories, methodology and trait factors. Imagination, Cognition and
Personality, 16(2), 181-210.
27. Kallmeyer, R. J., & Chang, E. C. (1997) The multidimensional dream inventory:
Preliminary evidence for validity and reliability. Perceptual and Motor Skills, 85,
803-808.
28. Sutin, A. R., & Robins, R. W. (2007). Phenomenology of autobiographical memories:
The Memory Experiences Questionnaire. Memory, 15(4), 390-411.
29. Heaps, C. M., & Nash, M. (2001) Comparing recollective experience in true and
false autobiographical memories. Journal of Experimental Psychology: Learning,
Memory and Cognition, 27(4), 920-930.
30. Johnson, M. K., Suengas, A. G., Foley, M. A., & Raye, C. L. (1988). Phenomenal
characteristics of memories for perceived and imagined events. Journal of Experi-
mental Psychology—General, 117(4), 371-376.
31. McDaniel, M. A., & Einstein, G. O. (1986). Bizarre imagery as an effective memory
aid: The importance of distinctiveness. Journal of Experimental Psychology: Learn-
ing, Memory and Cognition, 12(1), 54-65.
32. Loftus, E. F., & Burns, T. E. (1982) Mental shock can produce retrograde amnesia.
Memory and Cognition, 10, 318-323.
33. Cohen, D. B., & MacNeilage, P. F. (1974). A test of the salience hypothesis of dream
recall. Journal of Consulting and Clinical Psychology, 42(5), 699-703.
28 / HORTON AND CONWAY

34. Hicks, R. A., Bautista, J., & Hicks, G. J. (1999) Boundaries and level of experience
with six types of dreams, Perception and Motor Skills, 89(3), 760-762.
35. Brown, A. S. (2004). The deja-vu experience. East Sussex, England: Psychology
Press.
36. Palmer, J., & Dennis, M. (1979). Cited in Neppe, V. M. (1983). The psychology
of deja-vu: Have I been here before? Johannesburg, South Africa: Wittwatersrand
University Press.
37. Zuger, B. (1966). The time of dreaming and deja-vu. Comprehensive Psychiatry, 7,
191-196.
38. Neppe, V. M. (1983). The psychology of deja-vu: Have I been here before? Johannes-
burg, South Africa: Wittwatersrand University Press.
39. Grenier, J., Cappeliez, P., St-onge, M., Vachon, J, Vinette, S., Roussy, F., et al.
(2005). Temporal references in dreams and autobiographical memory. Memory and
Cognition, 33(2), 280-288.
40. Bulkeley, K., Broughton, B., Sanchez, A., & Stiller, J. (2005). Earliest remembered
dreams. Dreaming, 15(3), 205-222.
41. Fiske, K. E., & Pillemer, D. B. (2006). Adult recollections of earliest childhood
dreams: A cross-cultural study. Memory, 14(1), 57-67.
42. Robbins, P. R., & Tanck, R. H. (1978). Early memories and dream recall. Journal of
Clinical Psychology, 34(3) 729-731.
43. Schredl, M., Morlock, M., & Bozzer, A. (1996). Childhood memories and adult
dreaming. Zeitschrift fur Psychosomatische Medizin und Pcyhoanalyse, 42(1), 25-33.
44. Singer, J. L. (2003). Daydreaming, consciousness, and self-representations: Empirical
approaches to theories of William James and Sigmund Freud. Journal of Applied
Psychoanalytic Studies, 5(4), 461-483.
45. Beaulieu-Prevost, D., & Zadra, A. (2005). Dream recall frequency and attitude towards
dreams: A reinterpretation of the relation. Personality and Individual Differences,
38, 919-927.
46. Beaulieu-Prevost, D., & Zadra, A. (2007). Absorption, psychological boundaries
and attitudes towards dreams as correlates of dream recall: Two decades of research
seen through a meta-analysis. Journal of Sleep Research, 16, 51-99.
47. Schredl, M., & Hofmann, F. (2003). Continuity between waking activities and
dream activities. Consciousness and Cognition, 12(2), 298-308.
48. Horton, C. L. (2004). Sensory-perceptual episodic memories for dreams: Individual
differences and characteristics. Masters dissertation. University of Durham.
49. Baggaley, A. R. (1982). Deciding on the ratio of the number of subjects to the
number of variables in factor analysis. Multivariate Experimental Clinical Research,
6, 81-85.
50. Horton, C. L. (Under review). Trait correlates of the memory experiences and dreams
questionnaire: A link between dreaming, memory and personality.
51. Tulving, E. (1983). Elements of episodic memory. Oxford: Clarendon Press.
52. Howe, M. L., & Courage, M. L. (1993). On resolving the enigma of infantile amnesia.
Psychological Bulletin, 113(2), 305-326.
53. Kopelman, M. D., Wilson, B. A., & Baddeley, A. D. (1989). The autobiographical
memory interview: A new assessment of autobiographical and personal semantic
memory in amnesic patients. Journal of Clinical Experimental Neuropsychology,
11(5), 724-744.
 MEMORY EXPERIENCES AND DREAMS QUESTIONNAIRE / 29

54. Schredl, M., Funkhouser, A. T., & Cornu, C. M. (2001). Reliability in dream research:
A methodological note. Consciousness and Cognition, 10(4), 496-502.
55. Schredl, M., & Montasser, A. (1996-7). Dream recall: State or trait variable? Part II:
State factors, Investigations and Final Conclusions. Imagination, Cognition and
Personality, 16(2), 239-261.

Direct reprint requests to:

Dr. Caroline Horton
Psychology, Faculty of Health
Leeds Metropolitan University
Calverley Street
Leeds
England
LS1 3HE
e-mail: [Link]@[Link]