Content uploaded by Jayne Gackenbach
Author content
All content in this area was uploaded by Jayne Gackenbach on Jan 04, 2016
Content may be subject to copyright.
International Journal of Dream Research Volume 4, No. 2 (2011) 63
D
IJoR
Commentary
The discussion between Hobson and Schredl (2011) regard-
ing the viability of the continuity versus discontinuity hypoth-
esis is considered herein from the perspective of the waking
experience of video game play and its impact on subsequent
dreams. While both dream researchers agree that there may
be both continuous and discontinuous dream elements, the
dream community, clinical and research, seem to be leaning
towards a continuity conclusion. This is justied if one con-
siders that while content analysis systems may fall short in
raw count of actual continuous information from waking to
sleep, when full access to a dreamers memory and experi-
ence base is available, i.e. as in clinical treatment or in rst
person dream diary accounts, then the apparently discon-
tinuous elements can often be traced through the dreamers
memory/semantic networks to include much more of the
dream. A wide range of evidence has supported the conti-
nuity hypothesis of dream content. Events, personality, and
pathology (Schredl & Hofmann, 2003) have all been dem-
onstrated to show a waking to dreaming inuence. Thus it
seems that an inherent shortcoming of most dream con-
tent analysis systems that researchers use is the lack of rich
connective associations that presumably only the dreamer
can provide. But this is only one of various shortcomings
with dream content analyses systems in addressing the
continuity-discontinuity issue.
A case in point of the limits of some dream content analy-
sis systems is in our program of inquiry into the dreams of
video game players. Historically, in the sleep and dream lit-
erature a related media experience, lms, has been used
as a presleep stimuli. They have been considered to be an
easily controlled and an impactful pre-sleep event which
allowed fairly easy evidence for dream incorporation and
thus the continuity hypothesis. Films have been used to
investigate stress, aggression (Foulkes & Rechtschaffen,
1964), dream intensity (Foulkes, Pivik, Steadman, Spear, &
Symonds, 1967), sound incorporation (deKoninck & Kou-
lack, 1975) and dream lag effects (Powell, Nielsen, Cheung,
& Cervenka, 1995; Nielsen, Kuiken, Alain, Stenstrom, &
Powell, 2004). The advantage of a lm is it allows pre-sleep
controlled manipulation in order to investigate incorporation
questions.
As our media landscape is changing, so too are our op-
portunities to use media while awake to investigate issues
of dream incorporation. The problem with lm, television
or radio is that they are all unidirectionally presented, or
‘pushed’ at the passive viewer. In real life we are not passive
viewers, but active participants. This active participatory el-
ement is captured in computer use and video game play. It
is increasingly being incorporated into previously push me-
dia in forms such as viewer call-ins or online voting. While
these early efforts at a push/pull media approach increase
audience engagement, they pale in contrast to video game
play, where the entire experience can be almost completely
interactive. Thus video game play offers an ideal pre-sleep
stimulus to further investigate the continuity hypothesis.
Stickgold, Hobson, Fosse and Fosse (2001) used the
method of pre-sleep video game play to investigate if epi-
sodic memories transitioned from waking to sleep. Isolated
The continuity vs. discontinuity hypotheses:
A consideration of issues for coding video game
incorporation
Commentary on “The continuity and discontinuity between waking and dreaming: A
Dialogue between Michael Schredl and Allan Hobson concerning the adequacy and
completeness of these notions”
Jayne Gackenbach, Tyler Sample, and Gabe Mandel
Department of Psychology, Grant MacEwan University, Edmonton, Canada
Corresponding address:
Prof. Jayne Gackenbach Department of Psychology, Grant
MacEwan University, 10700-104 Ave., Edmonton, Canada
T5J 4S2.
E-mail: gackenbachj@macewan.ca
Summary. In response to the discussion between Hobson and Schredl, the history of our program of research for cod-
ing dreams of video game players both after playing a game and without such consideration, was reviewed. While many
of our studies are about response style in dreams resulting from game play, we also have considered incorporation is-
sues. Some of our previous results seemed to favour the continuity hypothesis, while others favoured the discontinuity
perspective. Two approaches to coding gamers’ dreams were considered and critiqued. Some of these problems were
then taken up in a compilation of data from three previous research studies where games were played the day before a
dream and dream information was gathered. The 182 dreams were categorized into three groups, no game incorpora-
tion, partial game incorporation, and full game incorporation (i.e., the dream is the game). Individual difference and game
content variables were unrelated to incorporation into subsequent dreams. However, this classication of dreams did
result in various content differences.
Keywords: Video games; continuity; discontinuity; incorporation; dreams; content analysis
Commentary
International Journal of Dream Research Volume 4, No. 2 (2011)64
D
IJoR
elements of the video game Tetris were incorporated early
in the sleep cycle, but nothing appeared about the context
of playing the game (i.e., computer, keyboard). They con-
cluded that the lack of context cues in subsequent dreams
argues that the incorporation was not episodic. This study
demonstrated the usefulness of video game play to inves-
tigate dream incorporation factors. In a later study from
this same laboratory (Wamsley, Perry, Djonlagic, Babkes
Reaven, & Stickgold, 2010), they used an arcade type video
game in which the individual is downhill skiing, and exam-
ined its impact on sleep mentation. They found that 30% of
verbal reports after various sleep lengths were related to the
video game and concluded that, “the nature of this cogni-
tive ‘replay’ effect was altered with increasing durations of
sleep, becoming more abstracted from the original experi-
ence as time into sleep increased.” (p. 59).
However, it is important to keep in mind that video games
are not only a potential independent variable to be manipu-
lated pre-sleep, but that because they are increasingly per-
vasive in today’s youth and young adult culture, broader
questions of their impact on dreams need also to be con-
sidered. The video game playing history of all potential
research participants in any study where video gaming is
manipulated as a pre-sleep stimulus should be taken into
account. As Preston (1998, 2007) presaged, wide exposure
to virtual reality (VR) environments will allow individuals who
are not able to become deeply absorbed constitutionally to
have experiences of altered states of consciousness.
Dreams are but one of a series of “imaginal” realms, or
alerted states of consciousness, that have important im-
plications for consciousness and its development. Hobson
(2009) argues that REM sleep may be the fertile eld upon
which consciousness grows and that the phenomenal expe-
rience of REM sleep, dreams, thus are lled with all poten-
tial experiences (i.e., discontinuous from waking). Of course
everyone experiences REM sleep. But other imaginal ex-
periences may serve this function as well. For instance, Ma-
son and Orme-Johnson (2010) has shown that the practice
of meditation has profound effects on sleep EEG and on
dream content as predicted by Hunt (1991).
Another deeply absorbing state that is much more widely
practiced is the play of video games. This imaginal realm
and its compelling playability is also affecting dreams. To
be fair any compelling life experience informs subsequent
dreams (Schredl & Hofmann, 2003). The difference is the
wide spread use of games. Entertainment Software Ratings
Board (ESRB) estimates 65% of all American households
now have video game units as do virtually all cell phones.
The long term effects of video gaming have been the focus
of inquiry in a series of studies by Gackenbach and col-
leagues and are summarized in Gackenbach (in press) and
in Gackenbach, Kuruvilla, Dopko and Le (2010). Thus far
we have found that like meditators, video gamers report
more lucid and control dreams (Gackenbach, 2006; 2009a).
Such dreams are transition sleep states between REM and
waking according to recent research (Hobson, 2009). These
dream qualities allow for conict resolution in the dream. It
is possible that video games fulll the threat simulation role
of REM sleep which can be argued supports the discontinu-
ity function (Gackenbach & Kuruvilla, 2008a; Gackenbach,
Hill, & Ellerman, 2011). Our lab also reported increased bi-
zarreness in video gamers’ dreams which we argued was
not associated with day before the dream media exposure
(Gackenbach, Kuruvilla, & Dopko, 2009). Additionally, we
found an association between video gaming, dream bizarre-
ness and creativity (Gackenbach & Dopko, 2011) controlled
for amount of day before the dream game play. This bizarre-
ness nding was associated with history of game play and
supports the continuity hypothesis.
Our work implies that gaming may create a response
style which was learned in waking VR and has general-
ized to sleep. One example is the self reported lucidity of
gamers but especially and stronger is their self reported ac-
complishments of dream control. The adaptiveness of these
alternative game based response styles is also illustrated
in Gackenbach, Hill, and Ellerman’s (2011) study. While spe-
cic game content may or may not have been incorporated
into these soldiers dreams, the response of the high end
gaming group to war threat in their dreams was empow-
ered and adaptive relative to soldiers who rarely gamed.
The point is that while incorporation of games into dreams
is important when considering the continuity-discontinuity
hypotheses, there may still be broader based effects on
dreams which are important and potentially have real world
consequences.
But to return to the continuity-discontinuity discussion,
our work seems to support both continuity (Gackenbach,
Sample, Mandel, & Tomashewsky, 2011; Gackenbach,
Rosie, Bown, & Sample, 2011; Gackenbach, Kuruvilla, &
Dopko, 2009) as well as for discontinuity (Gackenbach &
Kuruvilla, 2008a; Gackenbach, Hill, & Ellerman, 2011). Thus
it becomes important to consider how are game inuences
coded in dreams. There are two questions when consid-
ering the relevance of coding for video games. The rst is
what presleep gaming elements might be associated with
later dream incorporation and the second, how does one
code for game elements in dreams that may be different
than what normal dream coding systems allow?
Video Game Presleep Elements Potentially Pre-
dicting Incorporation
We have considered various types of pre-sleep gaming re-
lated information. This includes most often, individual differ-
ences in gaming history, such as frequency of play, length
of play, number of games played, and age begun play. How-
ever, in some studies we also considered susceptibility to
motion sickness, history of dream recall, gender, mediation/
prayer history among other individual difference variables
(see Gackenbach, in press for a review). We will next review
some of the game and dream content analysis issues that
can arise when considering game incorporation. These con-
siderations are not comprehensive but they are ones that
we bring new empirical evidence to bear upon later in this
paper.
The obvious starting point when examining more than one
game prior to sleep is type of game. Indeed, we are most
often asked what games produce specic dream effects.
Since there are literally thousands of games we have consid-
ered game genre in some of our work. Genre is a term used
to categorize things by a loose set of criteria. It is widely
used in the literary world and media studies but tends to
have no xed boundaries. Genre applied to video games
has a fundamental difference from other media applications.
Specically, video game genre’s have developed along the
lines of the nature of the interactions rather than visual or
auditory differences (Apperley, 2006). Unlike literary genre’s,
a video game genre is independent of its game play content
International Journal of Dream Research Volume 4, No. 2 (2011) 65
D
IJoR
Commentary
(Adams & Rollings, 2006). Thus a rst person shooter set
in the wild west is comparable to one set in the far future.
This categorization of game genres allows for comparisons
across genres. Game genre has been used to look at game
motivation (Tanis and Jansz, 2008), addiction (Huh, 2008)
and cognitive load (Gackenbach & Rosie, 2009). Addition-
ally, conceptual associations between genre and presence
have been made (McMahan, 2003; Gackenbach & Bown,
2011). In the gaming world (i.e. gamespot.com) genre’s
are increasingly being subdivided into more specic types.
Thus Halo 3 is not identied as only a “rst person shooter”
but as a “science ction rst person shooter”. These sub-
tler differentiations are becoming increasingly common, so
much so that there are now over 100 genre used in games-
pot.com. Indeed the entire issue of the relevance of genre
to understanding the video game experience is under seri-
ous consideration in the game studies literature (Clearwa-
ter, 2011). Despite these qualications, as in the meditation
literature, where it has become increasingly clear that you
cannot lump all types of meditation together, so too in the
gaming literature there is typically a lot of care placed on
what game is being considered or at least what genre. In
our laboratory we have increasingly collected information
on what genre of game is either preferred, most frequently
played or was last played (Gackenbach & Bown, 2011).
There are two additional generic concepts related to
video game content, which may inform game incorporation
into dreams. These are cognitive load and emotional inten-
sity. Easily available to researchers are the Entertainment
Software Rating Board (ESRB) rating categories for video
games. These can be adapted to dream content analysis for
games. The ESRB categorizes games along dimensions of
concern to parents regarding what their children would be
exposed to when playing a specic video game. But it also
allows consumers to be informed about the content of what
they are considering purchasing. There are seven ratings
categories: Early Childhood, Everyone, Everyone 10+, Teen,
Mature, Adults Only and Rating Pending. The categories as-
signed to each game by the ESRB are based on the degree
to which the game includes elements which are thought to
be problematic for children to be exposed to, and thus of-
ten emotionally evocative. Here are some examples of these
content descriptors:
▪Fantasy Violence: Violent actions of a fantasy nature,
involving human or non-human characters in situations
easily distinguishable from real life
▪Intense Violence: Graphic and realistic-looking depic-
tions of physical conict. May involve extreme and/or
realistic blood, gore, weapons and depictions of human
injury and death
▪Sexual Themes: References to sex or sexuality
▪Strong Sexual Content: Explicit and/or frequent depic-
tions of sexual behavior, possibly including nudity
▪Use of Drugs: The consumption or use of illegal drugs
Because games with these elements may be emotionally
evocative, they may be indicative of what games or game
elements are more likely to be incorporated into dreams.
Therefore the ESRB information on games may be useful in
determining which games or genre of games are more likely
to result in dream incorporation. The ESRB website has a
search tool to nd out how a game is rated. Thus Halo 3 is
rated Mature with blood, gore, mild language and violence.
Cognitive load is not easily obtained from the video game
industry as genre and ESRB ratings are, thus the cognitive
model of Das (2002) identifying planning, attention-arousal,
simultaneous and successive cognitive processing (PASS)
was adapted for our program of inquiry. Johnson (2008) has
shown a relationship between subscales on the PASS and
frequent internet use. She found “insignicant differences in
cognitive processing were most apparent between students
who reported frequent and infrequent recreational use of the
Internet (e.g., dating, downloading music and videos, play-
ing games)” (p. 2098-2099). However, her nding may be
in part due to the nature of her sample which was 72% fe-
male who reported (70 to 82%) never playing video games.
Indeed, Bowman and Boyan (2008) report cognitive skill
as related to game play, which was supported by Sherry,
Rosaen, Bowman, and Huh (2006). In any case the PASS
model was adapted to video games and the categories are
listed below (Gackenbach & Rosie, 2009):
1. Planning: An example would be the strategic pairing of
units in a real time strategy game, matching comple-
mentary classes in an MMORPG or RPG, or picking
plays that complement your team in a sports game.
2. Attention: How much attention did the game require in
order to be successful and to enjoy the game (i.e.: did
you have to pay attention to only one thing, multiple
things, successive things etc.)?
3. Physiological Arousal (i.e. increases in heart rate, breath
rate, blood pressure): An example would be heart rate
increasing during a ghting game, a close Player vs.
Player ght in an MMORPG, or a shootout in a hockey
game.
4. Simultaneous Processing: An example would be con-
trolling units on multiple battle fronts in a real time strat-
egy, changing lines during play in a hockey game, or
healing your party and attacking enemies in an MMOR-
PG or RPG (can involve multiple senses; sight, hearing
etc.).
5. Successive Processing: An example would be inputting
a complex sequence of buttons into the controller to
get your ghter to initiate a powerful combo in a ght-
ing game, waiting for an event to occur to allow you to
successfully initiate an action in an MMORPG or RPG,
or correctly manipulating 7 steps in sequence in order
to nish a puzzle.
The perception of the cognitive load of many games were
identied by over 1000 video game players who were asked
to evaluate three games each (Gackenbach & Rosie, 2009).
The reason we pursued this line of inquiry was because
such cognitive load information is not available about video
games, unlike genre where there seem to be clear denitions
emerging. Thus we have created an information base of the
cognitive load of specic games, which we can draw upon
in our work on dream content analysis of video game play-
ers dreams. In summary, game information (genre of game,
cognitive load of game and ESRB rating) can be culled and
matched with comparable dream information coded with a
content analysis system.
Coding for Video Games in Dreams
While the above considerations have been used in our
program of inquiry in order identify what presleep gaming
Commentary
International Journal of Dream Research Volume 4, No. 2 (2011)66
D
IJoR
information might predict incorporation, another problem
emerged as we moved down this path of dream content
analysis of video game players. We have used a variety of
dream content analysis systems in our laboratory. These
include examinations of, threat simulation (Gackenbach &
Kuruvilla, 2008a), bizarreness (Gackenbach, Kuruvilla, &
Dopko, 2009), lucidity/control (Gackenbach, 2006; 2009a),
war content (Gackenbach, Hill, & Ellerman, 2011) and broad
dream content analysis by judges using the Hall and Van-
deCastle system (Gackenbach & Kuruvilla, 2008b; Gacken-
bach, 2009b). The central problem with all these systems
is that they do not allow for virtual worlds to be coded in
dreams. Most were developed before the widespread use
of virtual type technologies so it’s not surprising that this
element is missing.
Here is an example with the content analysis of lucid/
control dreams (Gackenbach & Hunt, 2010). This is a dream
from a male hard core gamer, who had played from 4 to 7
hours the day before this dream. The games he had played
were rst person shooters including Half-Life 2 and Halo 3.
It should be kept in mind that what is interesting about the
rst person point of view (POV) games is that they do not
always allow a third person perspective. However, the real
self is actually in third person while playing a rst person
shooter and thus hours of being in that perspective may
have helped to mediate this dream.
I was in a desert. I looked bad, dusty. I saw my tiny silhou-
ette against a large sun, meaning I was watching myself,
in 3rd person. While I looked bad I didn’t feel bad. I was
indifferent to the “my” feelings. I came upon a carnival,
but it gets sketchy at that point. Eventually I’m driving a
car, again not at a real POV (point of view), but following
behind the car. It didn’t matter to me that I was crashing
into other cars or walls. My car caught re, I saw it melt
from within. I died not trying to escape. (Subject #27)
This gamer went on to report an interesting detachment from
the dream events when lling out the Metacognition, Affect,
Cognitive Experiences (MACE) questionnaire (Kahan, 2001)
as a follow-up to the subjects dream report:
As the car was burning I opened the door and leaned out
to leave but made the decision to stay inside instead be-
cause I was curious to see what I would look like burning
alive. While I felt the heat, smelt the smoke, I didn’t feel
any pain. I felt detached from the feelings, but recognized
that they were my own.
He also reported that it was not a nightmare, he was not
scared, but that the dream was violent. Finally, he reported
that the dream was not lucid and that he had no control over
it. At rst glance this seemed odd. Our judges thought that
certainly he must have known or suspected it was a dream?
When asked “did you feel any emotions during the experi-
ences?” he commented:
Sort of. I knew what the person I saw as myself felt, but
didn’t share those feelings. Throughout the emotions of
disgust, loneliness, or excitement were all ones I thought
best t the “character” of myself based on the situation.
Then he was asked “Did you think about what you were do-
ing?” and he wrote:
I was constantly thinking about my every move, mak-
ing sure that whatever I did was in my best interest. If
anything was off-putting (the carnival owner, the desert) I
simply moved on.
Then he replied to this question “Did you think about what
was happening around you?” by saying:
I was constantly analyzing my surroundings... At the city
where I drove my car, I noticed the simplicity of the envi-
ronment, which seemed to be constructed out of simple
polygons.
With this nal comment it became clear that he thought he
was in a video game environment while in the dream. He
did not think it was real, thus his remarkable choices, but
nor did he recognize that it was a dream. This sort of game
within a dream presents a tricky question as just identifying
if a dream is lucid can be challenging for researchers, no
less when you add the dimension of virtual worlds repre-
sented in dreams.
Virtual Worlds in Dreams Coding with the Hall and
VandeCastle Scale
This section is an analysis, in part by one of our coders (i.e.,
Coder A) who was trained on the Hall and VandeCastle (JH-
VDC) dream content analysis scale (Hall and VandeCastle,
1966), as articulated by Domhoff (1996), regarding the prob-
lems he encountered when coding a video game players
dreams from a long series (Gackenbach, Sample, Mandel,
& Tomashewsky, 2011). Coder A is a lifelong serious gamer
thus he brings that expertise to bear on his observations of
virtual worlds in dreams.
Of the 131 Dreams that Coder A coded, a subset (n=23)
of dreams had the largest instance of virtual environments.
This group of dreams was compiled by the dreamer into a
set that had a major theme of video games present in the
dreams. We will now examine some of the subscales of the
HVDC coding system used and how the presence of a vir-
tual environment inside of a dream is problematic for proper
HVDC coding.
Characters
When coding for characters the judge needs to select from
a number of discriminators in order to properly label the ex-
istence of a character in the dream. While the system has
no problem stating the existence of a character, it does not
properly allow for virtual characters or a virtual relationship
to the dreamer. What this means is that the coding system
does not allow a dream character to ip in and out of a
game character. So in the burning to death dream there was
a dual perspective of self with one being in a game and the
other watching the game unfold.
Another problem is when controlling a virtual character in
virtual space, which is the essential game play experience.
Take dream 2-114 for example:
I’m playing a Metroid game and I enter a chamber lled
with red lava below. There was an area above that I
needed to get to but couldn’t. The lava below lowered
with time. I recall deliberately jumping in. In 2 replays of
this area, I went in the deep lava and sunk more than 2
screens in it. I “swam” in it by jumping repeatedly. I once
got close to my intended area, just 80 pixels short (I could
see the top with 32 px of extra space.).
While the dreamer does not say anything in particular about
actually using a character, it can easily be inferred by the
simple fact that he is using an imaginary character inside
International Journal of Dream Research Volume 4, No. 2 (2011) 67
D
IJoR
Commentary
one dream ego (game character or avatar) was engaged in
violence while the second dream ego (watcher) was not.
Since video games are normally based around the use of
violence against nonplayer characters in order to progress,
if video games were dreamt about, the dream would show
an unusual amount of aggression to and from the dreamer.
This has been the case in our analysis of gamer dreams
where dream ego’s virtual perspective was not accounted
for (Gackenbach, Matty, Kuruvilla, Samaha, Zederayko,
Olischefski, & Von Stackelberg, 2009) and is taken up in the
current inquiry.
In the dreams that Coder A coded there was no friendli-
ness recorded inside the dreams’ virtual space. There is the
potential for friendliness to appear in this type of dream.
This will most likely occur when a virtual character helps
the dreamer’s character to accomplish some goal, help de-
feat enemies or if the dreamer helps a virtual character in
some way. Likewise there was no sexuality in any of the
video game dreams coded in this one study. These may be
because the dreamer in the case that Coder A was coding
suffers from Obsessive Compulsive disorder and potentially
from several other problems.
On the speculative side, it is reasonable to assume that
there is the potential for such dreams to occur especially
since the sexual aspect in many games is becoming in-
creasingly apparent as graphics improve and clothing styles
evolve. Video games often portray a gender gap between
males and females, with males as strong and mighty and
females as weak and in need of saving. Even though, there
are more games coming out now then ve or ten years ago
which empower the female characters, and remove this
gender bias, women are still scantily clad. The difference in
appearance and actions could easily lead to the emergence
of sexual events in dreams about video games. Indeed in
several other studies we found greater sexuality in gamers
dreams relative to norms (Gackenbach & Kuruvilla, 2008b;
Gackenbach, et al., 2009; Gackenbach, 2009b). Thus the
actual coding of friendliness and sexuality in a context of
sensitivity to virtual elements would be useful for at least
some game genres.
Setting
The setting is where the dream takes place. Dreams rarely
occur in a void and as such there is normally something
within this category. When a dream contains a virtual envi-
ronment it becomes problematic to explain the setting. In
some dreams that Coder A coded, the dreamer started by
saying he was playing a game, this suggests that he was
indoors and probably in a familiar place. Unless the dreams
specied location, Coder A treated the dreams as though
they were played in the dreamer’s home, since this is where
the majority of people play video games. Although, with
hand held gaming devices and games on cell phones, it is
possible to play video games almost anywhere. Once the
dreamer began to describe the virtual aspects of the dream,
other problems emerged. Consider dream 2-97.
I’m playing some game - Sonic in a freely explorable city.
I nd an area with a very high slope resembling Ice Cap’s
mountain. I use the spin dash to climb it. I see the fa-
miliar mountain scene but in 16-bit color and true 3D is
used making it look so much better and much more real-
istic. The polygon count also seems very high, well into
the millions. I climb to the top in about a minute and a
half and look around. I can see one extremely large tree
about 400 feet tall and some human-built infrastructure 5
the Metroid game. The apparent problem is that while the
dreamer is using a character in the videogame there is no
way to properly code for the use of the virtual character. In
fact, the virtual character would not even be coded using
HVDC since the dreamer did not provide details about who
or what is being controlled in the video game.
The inability to separate the dreamer from the virtual world
may adversely affect the results of the data because the
dreamer is now being coded as if they are physically acting
in the video game. While one could argue that perceptually
the sense of self is transferred to the game avatar (Blas-
covich & Bailenson, 2011), there are also instances where
physical and virtual selves coexist in waking reality while
gaming, thus why not in subsequent dreams.
Activities
The activities section of coding classies what a character
does in a dream. As was discussed above, when a virtual
world, such as a video game, is present inside of a dream
the activities of the dreamer can be, he is playing a game,
or she is in a game. In the former case the dreamer is con-
trolling a virtual character in the game, which is participa-
tion in different activities. The dreamer often refers to this
virtual character in the rst person which skews the results
from an analysis of the dreamer to an analysis of a video
game character. Taking a piece from dream 2-114, “I went
in the deep lava” would be coded as though the dreamer
was moving into the lava. While the Hall and Van De Castle’s
system of dream coding does not code for the danger of the
objects present in the dream, moving into the lava could be
viewed as an attempt for the dreamer to harm himself and
coded under the social interaction - aggression category.
This apparent self harm was brought home in the letting the
“self” burn to death dream also discussed. However, in that
case there were two dream egos. The player of a game who
watched events unfold and the avatar in the game, who was
also clearly himself and who burned to death. If the dream
ego stays as one self-character then the problem could be
fairly easily rectied by describing in the activities category
when a character assumes control of a virtual character. If
this relationship is acknowledged in coding then where the
dream character is the character taking virtual control of the
virtual character then any following actions by the dreamer,
while they control the virtual character, are coded normally
through the virtual character. It is easy to think of this in
terms of a transfer of consciousness from one character to
another. In the case of two selves, each could be separately
coded.
Social Interaction
The HVDC dream coding system breaks social interaction
down into three categories: Aggression, Friendliness, and
Sexuality. As was discussed in the last section, if the dream-
er is controlling a virtual character and does something that
the dreamer would not normally do with his own body in a
dream, i.e., deciding to burn to death, the results can be
misinterpreted (the coders initial impression that he must
have known it was a dream). When playing a video game the
player is often combating other virtual non player characters
(the articial intelligence of the game engine). This combat
can be very aggressive and violent. When combat is dreamt
about, if the dreamer is not linked to the virtual character,
which is actually doing the combat, the aggressive actions
recorded to the dreamer may be signicantly higher than
how the dreamer might actually behave. This is illustrated
if we look back again at the burning to death dream. The
Commentary
International Journal of Dream Research Volume 4, No. 2 (2011)68
D
IJoR
times as distant and easily seen due to its gigantic size. I
charged back down as I urgently needed to do something
but don’t recall what. Faulty collision detection prevented
extreme speeds - I got the fastest speeds through falling
which took a lot of it.
This dream is packed with scenery descriptors but the
problem is that the dream is almost completely inside a vir-
tual environment. According to the HVDC system of dream
analysis the setting would be in the dreamer’s house where
he is playing a video game. This would cause the loss of
almost the entire dream setting since it cannot be coded.
The setting is not hearsay, as in Domhoff’s instructions,
rather it is a perceptual experience in a dream about be-
ing in a video game. The freely explorable city is not really
there and as such is left out of the setting criteria. Some of
the dream content will nd its way under the Objects head-
ing, like the mountains and trees, but not where the majority
of the dream is actually occurring. This might be xed by
adding a condition to the setting criteria which would cre-
ate a separate grouping of virtual setting. For example the
above dream, the city would be coded with the condition
modier of virtual, outdoors, and unfamiliar. The simple ap-
plication of this modier to the setting would keep the virtual
and physical dream worlds separated if needed for analysis
or could be combined to get an overall analysis of settings
encountered. However, as real world physics rarely applies
to virtual worlds, witness how many houses in Second Life
oat in the air, then assuming the lava is “outdoors” might
be a stretch.
Emotions, Success, Failure, Good fortune and Misfor-
tune
These categories are not signicantly affected by the pres-
ence of a virtual environment inside a dream. If the charac-
ters are properly selected, the use of any of these categories
is not reliant on the setting. If the characters are controlling a
virtual character and one of these categories is needed, the
selection of one of these characters should link back to the
character that is controlling them. For example dream 2-74
describes the effort needed for the dreamer to defeat the
main enemy from the dream:
The main enemy I recall is a long gray snake-like object
nearly 50 feet long and 18 inches in diameter having a
head nearly a foot wide. You had to stab it in 3 spots in
order. The rst was far from the head, about 1/3 of the
way to tail. The second was about 1/6 of the way to the
tail from the head and the third was the top of the head.
The rst was very easy and the second was quite easy as
well. The third was hard as you needed a lot of force. I
recall playing the game twice. The rst time was where I
couldn’t defeat the snake and the second time I did, with
a second to spare and it took me 5 tries to get the head.
The nal success of the dreamer with defeating the enemy
would be coded as though dreamer was actually there to
defeat the enemy since it was his effort that resulted in the
success. Likewise, emotions are felt through the character
playing the game and as such they are the ones that will
feel happiness, apprehension or any other emotions. Good
fortune and misfortune should also be coded towards the
character controlling the virtual character since a misfortune
in the game results in a problem for the dreamt character to
overcome and good fortune in a game saves the character
from having to use effort to pass an obstacle. Indeed we
have found in past research on gamer’s dreams, differences
from norms in good fortune and misfortune (Gackenbach, et
al., 2009; Gackenbach & Kuruvilla, 2008b). Finally, the cat-
egories of objects and modiers were the least problematic
when coding virtual environments in dreams.
While the HVDC system does account for most dream
events, it does not account for how to code virtual envi-
ronments inside a dream. The system was designed in the
1960s when computers were very basic and as such it’s
not surprising that they did not include a coding system for
virtual environments. This analysis illuminates some of the
problems with using a standardized act frequency method
of dream content analysis for virtual worlds within dreams.
Alternatively we have approached the content analysis of
dreams for gaming information from a grounded theory per-
spective.
Grounded Theory Approach to the Content Analy-
sis of One Game in Dreams
In one study (Gackenbach, Rosie, Bown, & Sample, 2011)
we needed to develop dream content analyses system for
one video game, Mirror’s Edge, as we were looking at game
incorporation as a function of delity and interactivity of ex-
posure to the game. We took a grounded theory approach
to the development of this content analysis system. That is,
judges would read the dreams and view the game to see if
there are specic elements that emerge that might inform
our content analysis. Five categories emerged:
1. Primary in-game elements - Game elements that are
essential to the mechanics of Mirror’s Edge and/or must
be interacted with while playing the game. List items
can be agged with ‘EG’ (extra-game) or ‘LE’ (labora-
tory elements) when the particular instance of dream
content does not occur in a Mirror’s Edge setting and/
or context.
2. Secondary in-game elements - Complementary or rein-
forcing to primary elements – can either be static parts
of setting, interacted with in non-necessary ways or
used infrequently. List items can be agged with ‘EG’
(extra-game) or ‘LE’ (laboratory elements) when the
particular instance of dream content does not occur in
a Mirror’s Edge setting and/or context.
3. Conceptual themes from Mirror’s Edge – Things like
mazes or roof tops.
4. Physiological and psychological responses – Fears of
heights would be one example that could be associated
with Mirror’s Edge.
5. Laboratory elements – Things like the video goggles
from the laboratory session would be included herein.
We expected the highest dream incorporation in the high
delity/high interactivity condition. Incorporation was as-
sessed by subject self-report and judges’ evaluations. The
independent variable of delity was especially strong both
in the manipulation and in the subsequent dream incorpora-
tion for self report while interactivity became the dominant
variable when viewed from the judges’ perspectives. Given
the conicting results as a function of dream coding source,
subject versus judge, it’s clear that even with a scale devel-
oped for the particular elements of a specic game there are
still problems in coding for games in dreams.
International Journal of Dream Research Volume 4, No. 2 (2011) 69
D
IJoR
Commentary
Present Study
Over the decade of this research program we have col-
lected dreams from the night before, where a video game
was played the day prior to the dream. Selective data from
three studies were collapsed as similar information was
gathered each time. This allowed a more comprehensive
analysis of game incorporation into dreams. The purpose
of this inquiry, is to address some of the concerns just out-
lined with coding games in dreams. These game-dream re-
ports were gathered from (Study 1: Gackenbach & Kurvilla,
2008a; Study 2: Gackenbach & Dopko, 2011; and Study 3:
Gackenbach, Rosie, Bown, & Sample, 2011; Study 1 was
conducted during the 2007-2008 academic year. Study 2
was conducted during the 2008-2009 academic year. Study
3 was conducted during the 2009-2010 academic year. All
subjects were drawn from students in introductory psychol-
ogy who participated for course credit at the same western
Canadian university). Data explored as potentially predictive
of dream incorporation of games included: presleep infor-
mation from individual differences to game characteristics,
game incorporation criteria, and some content analysis of
dreams. The questions asked in this compilation of data are:
1. Is there anything that distinguished the dreams that
showed some game incorporation from those who were
rated as no incorporation?
2. Is type of incorporation, dream is game versus gaming
is in dream, associated with other dream or game con-
tent information?
Method
Information (n=182) was gathered from previous research
studies where the dreamer reported having played a game
the day before the dream. This information was garnered
from three previous studies in our laboratory over a four
year period. Three types of information were available for
each game-dream occurrence either as previously coded
or coded for the purposes of this inquiry. This information is
summarized below:
1. Information about games included:
a. ESRB: emotionality of game content
b. Cognitive load: based on PASS model with non-
subject game players having previously rated each
game
c. Game Genre information
2. Information about Players:
a. Sex
b. Gaming history of subjects
c. Media use the day before dream
d. Length of playing video game on the day before the
dream
3. Information about Dreams
a. Game in dream
i. Type of incorporation
1. Dream is game world
2. Video game is being played in dream
3. Some references to games other than the pre-
vious two
ii. As a function of selected HDVC scales (i.e. char-
acter)
iii. As a function of selected ESRB content catego-
ries
b. Dream content scales independent of game in dream
considerations
i. Lucid/control/watching coding
ii. HVDC sum scores
Specically while we knew which video game the research
participant played, we entered its genre, taken primarily
from www.gamespot.com; cognitive load, taken from the
game data base of Gackenbach and Rosie (2009); and
ESRB information, obtained from www.esrb.org. As noted
earlier gamespot has over 100 genre, but there are levels of
classication such that one can use only the base level of
genre as was the case in the present study.
Information about the players was already available as
was some of the dream content scale information. However,
additional dream content coding had to be done specically
for this inquiry. It should be noted that the HVDC training of
all judges was interconnected. That is, the Study 1 coder
was the rst research assistant to be trained and she trained
the Study 2 and Study 3 coders. All achieved at least 80%
concurrence in HVDC coding with the Study 1 coder.
The game in dream coding, as dened herein, was done
for this combined data set by two research assistants. Spe-
cically, in consultation with the lead author they read all
dreams and determined if the game was in the dream along
these lines:
1. Dream is game world
2. Video game is being played in dream
3. Some references to games other than the previous two
It should be noted that in order to do this coding both cod-
ers had some video game playing experience. However, if
they were unsure of a coding for a game they researched
the game online. Initially they coded the dreams in terms of
non-game elements that were not already available (i.e. lucid
dreaming related variables).Then they coded if the dream
was the game or if gaming was in the dream. Finally, these
research assistants coded the dreams in terms of specic
gaming elements, i.e. dream characters as game characters
(see results section for a detailed list of what was coded).
They were also asked if the game in the dream matched the
game that was played and nally they made an overall incor-
poration rating along a 5-point likert type scale, from highly
incorporated to no dream incorporation. These two coders
each coded the same eight dreams along the dimensions
just listed with correspondences per dream ranging from
56% to 89% with an average of 75% concordance.
Results
Since the primary question was differentiating game incor-
poration from no game incorporation, three groups of game/
dream pairs were identied based upon the judges evalua-
tions. Three game in dream coding questions were used to
derive these dream groups. Judges were asked to deter-
mine if in the dream the dream ego is in the game world, that
is the dream is the game, or if in the dream playing a video
game was mentioned or, nally, if in the dream, games were
mentioned.
Of the 182 dreams collected from the three previous stud-
ies where a game was played the day before the dream, 115
dreams were classied as not having the game in the dream
(no game group), 26 dreams were identied as either playing
Commentary
International Journal of Dream Research Volume 4, No. 2 (2011)70
D
IJoR
a video game in the dream or some other mention of games
in the dream (partial game group). Finally, 41 dreams were
identied as the dream is the game world (game group).
While there was some overlap between the two dream cat-
egories for gaming it was minimal. Specically, only one of
the 41 game group dreams also had playing a video game
mentioned. The partial game group included eight dreams
where playing a video game was mentioned and 18 where
games were mentioned (i.e., my mom was shopping for the
new x-box controller or I was playing hockey). Any reference
to playing any type of game was included in this category
because there are so many games that are both real and
video like hockey or jigsaw puzzles. This differentiation be-
tween inclusion and non-inclusion of gaming, and between
two types of inclusion (game group and partial game group),
allows us to consider some of the virtual worlds in dreams
coding problems highlighted earlier, as well as the question
of continuity versus discontinuity.
Verication for this dream group classication was ob-
tained through two one way ANOVA’s. One was for dream
group on judges overall incorporation rating and the other
was for game matching dream rating. Both were signicant
(overall incorporation F(2,177)=117.00, p<.0001 and game
matching dream F(2,176)=105.446, p<.0001). In both cases
dreams from the game group were more likely to be rated as
evidencing more overall game incorporation and as having
the game elements in the dream match the game elements
of the game played prior to sleep. For overall incorporation
there was no difference between the no game group and the
partial game group with the most incorporation in the game
is dream group. The game is dream group was the strongest
match followed by the partial game group and then the no
game group. All groups differed in post hoc analyses
The results are broken down into three sections of de-
pendent variables: background of the subjects, game quali-
ties, and dream content. The last is subdivided into content
analysis examining game related content versus content
analysis irrespective of game elements. In all sections ei-
ther one way ANOVA’s or chi-squares, as appropriate, were
computed with dream group (i.e., no game, partial game,
and game) as the independent variable.
Background of the Subject
This information was available from each of the previous
studies and included which study they were drawn from,
sex of subject, gaming history classication, and media use
the day before the dream. Because these variables were
not always collected in the same way sometimes they had
to be reduced to present/absent or high/low. This included
media use and game history which were sometimes mea-
sured along a likert scale. While individually each backgroud
variable is signicant across dream type groups, sans the
media use one, there was no difference as to the distribution
of each variable across dream groups. Specically, the most
cases were drawn from Study 3 (χ2(2)=43.923, p<.0001),
more males were represented (χ2(1)=45.098, p<.0001), and
more high end gamers (χ2(1)=73.066, p<.0001). However, it’s
important to keep in mind that none of the chi-square cal-
culations for each of these variables with dream group were
signicant. That is, while there were more cases from Study
3, they were evenly distributed across all dream groups. So
too with gender and game history.
As for media use the day prior to the dream three types
of information were gathered in each study; audio media,
video media, and interactive media. These were classied
as present or absent and chi-squares were computed for
each media type as a function of dream groups. One analy-
sis approached traditional signicance levels (audio media:
χ2(2)=5.356, p<.069), while video media use nor interac-
tive media use were not signicantly different as a function
of dream group. It should be noted that interactive media
use does not include video game play as that was inquired
about in separate questions. As for the audio media use this
could include listening to music or talking on the phone. In
all three dream groups subjects reported more often hav-
ing done one of these audio media activities than not, but
slightly less so for those whose dreams were categorized as
game dreams (82% versus 94 and 95 percents for the other
two dream groups).
Video Game Played
The next question addressed in these analyses is, are there
qualities about the game that might predict incorporation or
the lack thereof into subsequent dreams? Four types of in-
formation were available to address this question: length of
video game play the day before the dream and three game
type variables: genre, cognitive load and ESRB rating. Keep
in mind that while most subjects dreams were not found to
incorporate a game they all reported having played a game
the day prior to the dream and most were high end gamers.
Thus, it’s important to see that there were no dream group
differences in the length of game play (F(2,179)=0.212, ns).
All groups reported playing between 1 to 2 hours ranging
from less than an hour to more than four hours.
This lack of dream group difference was replicated for
the other game information variables. Seven general game
genre were identied from gamespot (i.e., rst person
shooter, music, action/adventure, sport/racing, role play-
ing, casual, and strategy) but the chi-square as a function of
dream group was not signicant (χ2(12)=16.568, ns). How-
ever, across groups there was a difference in genre’s played
(χ2(6)=52.599, p<.0001) such that rst person shooters were
played the most often (31%) followed by action/adventure,
sports/racing, and role playing each with about 16%. Less
frequently played were the music and casual genre averag-
ing 9% each and nally strategy was reported as least often
played at 5%. Keep in mind that the genre classications
were based upon gamespot.com and not from the players
who only reported one game they had played.
A repeated measures ANOVA on the ve cognitive
load variables for the video games played by dream
group was computed. The interaction was not signicant
(F(2,134)=0.012, ns) nor was the main effect for dream group
(F(2,134)=1.962, ns). However the main effect for cognitive
load was signicant (F(1,134)=72.436, p<.0001). The high-
est reported cognitive load by gamers in a separate study
(Gackenbach & Rosie, 2009) but applied to these dreams
was successive processing (x=2.16, standard error=.016)
followed by planning (x=2.312, standard error=.068) and
simultaneous processing (x=2.26, standard error=.073).
At the low end was physiological arousal (x=2.171, stan-
dard error=.055) and nally attention (x=1.641, standard er-
ror=.068). When each cognitive load was examined sepa-
rately as a function of dream group, all were non-signicant:
planning (F(2,136)=1.124, ns), attention (F(2,136)=2.233, ns),
physiological arousal (F(2,136)=0.689, ns), simultaneous
processing (F(2,136)=1.719, ns), and successive processing
(F(2,136)=0.781, ns). In other words, the type and amount of
cognitive load required for the wide range of games played
did not differentially discriminate between incorporation or
International Journal of Dream Research Volume 4, No. 2 (2011) 71
D
IJoR
Commentary
not into subsequent dreams.
The nal game type variable examined was the ESRB rat-
ings. These were considered as overall ESRB rating and six
mean scores from the 31 specic content categories. None
of these dependent variables showed a difference as a
function of dream group (main effect F(2, 165)=.982,ns) and
interaction (F(2,165)=.522, ns). However, there was a main
effect for content categories (F(1,165)=73.745, p<.0001)
such that violence has the most descriptors (x=.137, stan-
dard error = .008) followed by language (x=.124, standard
error = .011). No games were rated as having gambling. The
rest fell in between these extremes.
In conclusion, there was nothing about the games played
per se that predicted incorporation or the lack thereof. How-
ever, as will shortly be seen this is not the case when dream
content was considered.
Dream Content Analysis
There were several ways that these dreams were all con-
tent analyzed, in the original study or additionally for this
combined effort. Some of the new analyses were focused
on game imagery in the dreams. But additionally all dreams
were examined for lucid/control types of imagery and those
dreams that had not been content analyzed using the HVDC
were so analyzed for this study by Coder A. Thus two con-
ceptual types of analyzes were available for all 182 dreams,
game relevant and non-game relevant. These will be sepa-
rately presented.
Game Relevant Dream Content Analysis
In addition to the three broad questions regarding game
in dream, which created the three dream groups (no game
group, partial game group, and game group), dreams were
also coded with adjusted HVDC scales with a focus on gam-
ing. Specically, judges were instructed to indicate present
or absent for several HVDC categories relative to the game
components of the dream. These analyzes are summarized
in Table 1.
Fifteen chi-squares were computed with all but four evi-
dencing signicant differences between dream groups.
For the 11 chi-square tests that were signicant the partial
game group was roughly the same as the no game group in
4 cases, fell between the other two groups in 4 cases, was
highest in one case and matched the game group in one
Table 1. Chi-squares and Percent Present of Selected HVDC Content Categories Coded for Game Type Content.
Categories Characteristic Chi-Square % of dream groups yes
Characters
Self is game character χ2(2)=114.928, p<.0001
No game - 1.8%
Partial game – 0%
Game – 75.6%
Other is game character χ2(2)=29.098, p<.0001
No game – 4.5%
Partial game – 8%
Game – 36.6%
Self changes into game character χ2(2)= 10.196, p<.006
No game - 0%
Partial game – 0%
Game – 7.3%
Self controls game character χ2(2)= 6.155, p<.046
No game - 0%
Partial game – 4%
Game – 0%
other χ2(2)= 3.36, ns
Activities (Physical, Movement, Location Change,
Verbal, Visual, Auditory, Thinking) χ2(2)= 76.873, p<.0001
No game – 12.5%
Partial game – 20%
Game – 85.4%
Emotions
Positive emotions χ2(2)= 0.154, ns
Negative emotions χ2(2)= 16.505, p<.0001
No game – 1.8%
Partial game – 4%
Game – 19.5%
Neutral emotions χ2(2)= 17.190, p<.0001
No game – 0%
Partial game – 0%
Game – 12.2%
No emotions χ2(2)= 1.798, ns
Objects
(Architecture, Household, Food, Travel,
Streets, Regions, Nature, Body Parts,
Clothing, Commun-ication [including
technological], Money, Miscellaneous)
χ2(2)= 33.398, p<.0001
No game – 7.1%
Partial game – 12%
Game – 46.3%
Implements
weapons χ2(2)= 33.765, p<.0001
No game – 1.8%
Partial game – 0%
Game – 29.3%
recreation χ2(2)= 5.25, p<.072
No game – 0%
Partial game – 4%
Game – 4.9%
Descriptive Ele-
ments
(Colour, Size, Age, Thermal, Velocity, Linear-
ity, Intensity, Evaluation, Temporal Scale) χ2(2)= 4.503, ns
Commentary
International Journal of Dream Research Volume 4, No. 2 (2011)72
D
IJoR
case. This substantiates the concept of that group as falling
between the other two.
The category of characters was further broken down into
virtual world dimensions (i.e., game character versus per-
sonal self). The game group had more of each type of self
than the other groups except for the category of self con-
trols game character. In that case it was the partial game
group that was highest which makes sense if you’re playing
or talking about a video game in a dream. The most direct of
the ways to view self in dreams relative to gaming, was the
self is the game character which constituted 75.6% of the
41 dreams that were determined to be the game world.
Two other HVDC type scales were subdivided, emotions
and implements. Negative and neutral emotions were high-
est in the game dream group. There were no dream group
differences in coding for positive emotions or coding for the
lack of emotions. Finally, the game group had highest activi-
ties, settings and objects that were game related.
The other content analyses focused on the game in the
dream were for social interactions. While friendliness coding
followed HVDC, the negative side of social interactions were
modeled after ESRB content descriptors because they were
derived specically from video games, coded by ESRB per-
sonnel for each game and included a wider range of topics
than most dream coding systems allow. These 31 descrip-
tors were clustered into six categories; violence, sex, drugs
and alcohol, language, humor and gambling. Coders were
instructed to count the number of incidents of each ESRB
type content category in each dream relative to gaming in
the dream. The results of these analyses are portrayed in
Table 2.
Not surprisingly the sum of all ESRB codes favoured
the game group which was primarily accounted for by the
violence coding. While the means for the non-signicant
categories are not presented in Table 4, they collectively
favoured the gaming group. Interestingly, gambling was
highest in the partial game group.
As with the ESRB game in dream content coding, friend-
liness was a count of the number of incidents of game
relevant friendliness. This time the category was dened
by HVDC. The one way ANOVA approached signicance,
F(2,176)=2.788, p<.06. The game group fell between the
partial game group and the no game group in terms of total
number of game related friendly interactions in the dream.
No Game Relevance Dream Content Analysis
Two types of nongame related dream content analyses
were done on these dreams. Traditional HVDC was done
but only subscale sums were compared across dream
groups. Also coded were various scales examining the lu-
cid/control dimension. Of the 13 HVDC subscale sums only
3 evidenced signicant or near signicant dream group dif-
ferences: aggression, F(2,110)=4.358, p<.01; friendliness,
F(2,110)=2.667, p<.07; and good fortune, F(2,110)=2.749,
p<.068. It should be noted that dreams were not coded that
were less than 50 words on the HVDC. Aggression was sig-
nicantly higher in the game group than in the partial game
group but the no game group was not different in aggres-
sion from either of the other two groups. On the ip side
friendliness was signicantly lower for the game group than
for the partial game group with the no game group falling in
between. Finally, the good fortune sum was highest for the
two gaming type groups relative to the no game group.
Finally, the lucid/control dimension was coded along sev-
eral dimensions which are portrayed in Table 3.
While not lucid, various prelucid or lucid related variables
showed dream group differences. These included two of the
control variables and four items from the MACE. Thus one
gets a picture of dreams that are games, as ones where the
Table 2. Descriptive Statistics and F-values for ESRB Dream Coding.
ESRB based Dream Coding
*(N/mean/SD)
No
game in
dream
game
mentioned
in dream
dream
is game F-values
ESRB violence 115 26 41 F(2,181) = 9.145, p < .0001
0.38 a0.42 a1.51 b
0.894 1.172 2.58
ESRB Sex F(2,181) = 0.524, ns
ESRB Drugs & Alcohol F(2,181) = 0.147, ns
ESRB language F(2,181) = 1.733, ns
ESRB Humor None coded
ESRB Gambling 115 26 41 F(2,181) = 3.069, p < .05
0 a0.04 b0 a
0 0.196 0
Sum of ESRB category 115 26 41 F(2,181) = 8.322, p < .0001
0.4261a0.5 a1.561 b
0.93716 1.27279 2.69304
Note. ab Duncan post hoc test results are indicated in the superscripts of a,b. Those that are the same did not differ while those letters that are different
indicate group differences.
*Of the 11 ESRB content categories related to violence only one was signicantly correlated to count of violence in dreams, animated blood
[r(170)=0.376,p<.0001]. None of the other specic content categories from the ESRB website for the games played correlated signicantly with any
of the ESRB type dream content coding categories.
International Journal of Dream Research Volume 4, No. 2 (2011) 73
D
IJoR
Commentary
dream ego is in control of themselves and other charac-
ters while commenting to self, being focused on task and
reecting on thoughts/feelings and not having difculty in
accomplishing tasks.
Finally, several pearson product moment correlations
were computed between very similar dream content scales
coded with gaming in mind and coded by separate sets of
judges as part of a general coding of all dreams from the
previous studies. Ten such tests were possible and most
were signicant. These are portrayed in Table 4.
It might be concluded from these results that coding sep-
arately for game elements makes no difference from general
HVDC coding. However, there are some notable exceptions.
The virtual nature of characters was articulated the most
in this new coding scheme and when that was done there
was no correlation to the original HVDC coding. Emotions
and objects also failed to correlate, although the subset of
objects called weapons might correlate but the data output
from the Domhoff method did not provide that information.
Additionally, two of these 10 correlations were with ESRB
based coding schemes which were broader based than the
classic HVDC system. Interestingly, in one case this correla-
tion was the highest, aggression, and in the other case the
lowest (marginally signicant), sex. In a sense it’s surprising
that there were even these magnitudes of correlations as
only about 1/3 of the dreams coded had game elements.
But then while signicant, the correlations were low to mod-
erate in magnitude and the traditional HVDC coding result-
ed in few signicant dream group differences.
Discussion
The history of our program of research for coding dreams
of video game players both after playing a game and with-
out such consideration was reviewed. While many of our
studies are about response style in dreams resulting from
game play, we also have considered incorporation issues.
Some of our previous results seemed to favour the conti-
Table 3. Descriptive Statistics and Chi-Squares or ANOVA’s for Lucid/control Type Variables.
Lucid/control Type Variables*
No
game
Partial
game Game Test
Lucid Dream F(2,178) = 1.087, ns
MACE Sum Score (sum of judges yes responses to each
question) F(2,178) = 1.593, ns
Did the dreamer choose between two or more options χ2(2) = 0.043, ns
Did the dreamer comment to him or herself about any per-
son or event 33.9% 26.9% 58.5% χ2(2) = 9.408, p = .009
Did something or someone suddenly capture the dream-
ers attention χ2(2) = 1.561, ns
Did the dreamer focus for a period of time on accomplish-
ing a particular task 33.9% 42.3% 61.0% χ2(2) = 9.066, p = .011
Did the dreamer experience any unusual difculty in ac-
complishing anything he/she was trying to do 13.4% 26.9% 7.3% χ2(2) = 5.172, p = .075
Was the dreamer concerned about the impression he/she
made, how the dreamer looked or how the dreamer ap-
peared to others
χ2(2) = 0.137, ns
Did the dreamer feel any emotions during the experiences χ2(2) = 0.750, ns
Did the dreamer think about their own thoughts or feelings 19.6% 7.7% 31.7% χ2(2) = 5.788, p = .055
Did the dreamer think about what he/she was doing χ2(2) = 2.198, ns
Did the dreamer think about what was happening around
the dreamer χ2(2) = 2.718, ns
Control of dream self (high=more) 104 25 40 F(2,178) = 2.922, p = .057
2.65 ab 2.44 a3.13b
1.245 1.417 1.09
Control of dream characters (high=more) 64 19 20 F(2,178) = 4.314, p = .016
0.81 a0.58 a1.2 b
0.5 0.507 1.152
Control of dream environment F(2,178)=0.767, ns
Dream ego stance F(2,178)=0.379, ns
Dream ego as watcher F(2,178)=1.018, ns
Note. * Continuous variables when signicant are expressed in number of cases, mean and standard deviation.
Yes/no variables are expressed in the percent of yes’s.
Commentary
International Journal of Dream Research Volume 4, No. 2 (2011)74
D
IJoR
nuity hypothesis others favoured the discontinuity perspec-
tive. Various approaches to coding gamers’ dreams were
considered and critiqued. Some of the problems were then
taken up in a compilation of data from three previous re-
search studies where games played the day before a dream
and dream information was gathered. All of these dreams
had been coded using the HVDC or were coded specically
for this compiled analysis using the HVDC. Additionally, an
initial effort to create a system of coding sensitive to gam-
ing in dreams was constructed and used to code all dreams
in this compilation. A second new dream content analysis
focused on the lucid/control dimension of dreaming.
Three groups of dreams were identied using the new
content analysis scheme. These were dreams where there
was no incorporation (no game group), dreams where there
was some mention of gaming (partial game group), and
dreams that were the game (game group). The classication
of these dreams was veried by judges overall incorporation
rating and matching game to dream rating.
The rst point to be noted was that these groups did
not differ in a variety of important variables including gam-
ing history and length of play of the specic game men-
tioned. Thus, the nding of 63% of the dreams not mention-
ing games points to a lack of incorporation or continuity.
This would seem to support the discontinuity hypothesis.
However, it’s important to keep in mind that this group of
subjects overall were male, high end gamers who were re-
cruited for one specic study. So the nding that most did
not evidence game incorporation may be due to their long
history of play, their choice of which specic game they
played or reported playing, or other individual difference or
situational considerations.
A similar distinction between partial incorporation and
dream as game was used by Murzyn (in press) in an analy-
sis of World of Warcraft (WoW) players dreams. Of the 233
WoW dreams she collected slightly less than half could be
identied as having some game element. While we collect-
ed dreams after a game was played and got a small number
of overall incorporations, she collected dreams from gam-
ers who claimed that the dream was about WoW, thus the
slightly higher incorporation rate.
In terms of the continuity hypothesis, of particular interest
were the dreams that did evidence some form of incorpora-
tion. These two dream groups (partial game and game) were
compared to the no game group dreams along three sets
of dimensions: subject differences, game content measures
and dream content judgements. As just noted subject vari-
ables showed no notable differences as a function of dream
group type nor did game information provide any specic
information to distinguish either between incorporation or
no incorporation or between partial and full incorporation.
That is, not to say that there are not predisposing subject
variables or game content that might be valuable to con-
sider, but they did not show up here. It may be, regarding
the game variables, that once you go beyond one specic
game there is too much variance to detect incorporation
and these generic measures did not capture the relevant
information.
So while we can’t yet say what individual difference or
situational information predisposes incorporation of games
into dreams, we can say that this manner of distinguishing
between dream groups is related to dream content differ-
ences. This goes to our second purpose for pursuing this
research. That is the problems inherent in coding dreams for
virtual worlds. As noted earlier, this will become increasingly
a problem as members of at least industrialized societies
are “living” in part in their own technologically created virtu-
al worlds. Thus coming to grips with these coding problems
is important for dream content analysis research to remain
current. Video game play is the ideal test case because it
represents the most immersive VR experience that is widely
available.
While a series of problems associated with coding gaming
dreams were discussed, this compilation of data attempted
to address only a few. First the issue of game characters in
dreams was addressed by creating relevant coding catego-
ries (see Table 1) which resulted in appropriate differentiation
between dream groups. The partial game dream group had
more instances of self controlling game characters, while
game group had more of the other three types of game
Table 4. Correlations between HVDC sum scores and game in dream coding sums.
HVDC Variables/Game Coding Variable Correlation Statistic
Activity sum of subscales/activity count in game elements r=.252, n=109, p<.008
Settings sum of subscales/settings count in game elements r=.273, n=109, p<.004
Objects sum of subscales/objects count in game elements r=.209, n=109, p<.029
Modifers sum of subscales/descriptive elements count in game elements r=.374, n=109, p<.0001
Aggression sum of subscales/ ESRB Violence subscale sum r=.468, n=111, p<.0001
Sexuality sum of subscales/ ESRB Sex subscale sum r=.166, n=111, p<.081
Character sum of subscales/sum of characters in game subscales r=.090, n=109, ns
Emotions sum of subscales/sum of positive and negative emotions count r=.124, n=109, ns
Friendly sum of subscales/sum of friendly game elements r=.372, n=109, p<.0001
Objects sum of subscales/sum of implements in game elements r=.140, n=109, ns
International Journal of Dream Research Volume 4, No. 2 (2011) 75
D
IJoR
Commentary
characters in dreams. Furthermore, this conceptualization,
unlike many of the others, did not correlate with standard
HVDC characters in dreams coding sum scores. For all the
rest of the HVDC categories, which were adjusted to code
counts of things relative to gaming in the dream, there were
dream group differences. There were also generally positive
correlations with standard (nongame sensitive) HVDC cod-
ing. This could mean that virtual considerations are not im-
portant or more likely in our opinion, that this approach was
too simplistic. Additionally, while signicant, the correlations
were moderate in magnitude.
A second alteration in coding game in dream content was
the adaptation of the ESRB to dream coding. Although it
also measured act frequency, there were more potential acts
that could be coded in any one type of content, i.e., blood
for the violence type content. This alternative scale (violence
or sex) was the highest correlation with standard HVDC ag-
gression sum scores and marginally correlated with HVDC
sexuality sum scores. Additionally, ESRB violence ratings of
dreams showed signicant dream group differences, while
sex did not. No dream group differences were evident for
the other ESRB content scales other than gambling. So
when the ESRB coding of gaming dreams was more violent,
these dreams were also coded as more aggressive using
the standard HVDC. Given that these were largely classic
male hard core gamers this is not surprising as the games
they tend to prefer involve a lot of violence.
This is not as dark as it appears on the surface as we
have found and pointed out in previous work. There was
no dream group difference in positive emotions while more
negative emotions were evident in the game dream group
but also more neutral emotions. This neutrality to violence
was nicely illustrated in the burning to death dream men-
tioned earlier. Indeed in some of our studies violence in
dreams was viewed as fun (Gackenbach, Hall, & Ellerman,
2011)! Consistent with the violence nding is that game
dreams were higher in activity and weapons. This activity
and weapons effects, we also found as differentiating mili-
tary dreams of high versus low end gamers (Gackenbach,
Hill, & Ellerman, 2011).
The cognitive elements of this higher activity were ushed
out in the lucid/control type coding and analyses. While the
game dreams were not judged to be lucid, they were judged
as highest in self and character control. Also in terms of
using the MACE with questions about the dream answered
by judges, game dreams were characterized as having
signicantly more self comments, thinking about thoughts
and feelings and focusing on accomplishing a task without
being thwarted in their intention. Games dreams were less
likely to be characterized by difculty in accomplishing a
task. The internal commentary of gamers and a lack of in-
tention being thwarted were found in Swanston and Gack-
enbach (2011) as was the higher dream control but no lucid
dreaming. Their sample is a different gamer group several
years later from the same university. This combination of
cognitive dream skills can be viewed as adaptive in the sol-
dier gamers who were judged as less susceptible to threat
in their dreams during military service (Gackenbach, Hill, &
Ellerman, 2001).
Conclusion
The issue of virtual world immersion during the day and its
implications for subsequent night time dreams was taken up
in two respects, incorporation and coding. Evidence from
previous work and the current inquiry support both continu-
ity and discontinuity. Some suggestions for adjusting dream
coding for the presence of virtual worlds were considered
and tested.
References
Adams, E. & Rollings, A. (2006). Fundamentals of Game Design.
Prentice Hall.
Apperley, T.H. (2006). Genre and game studies: Toward a criti-
cal approach to video game genres. Simulation & Gam-
ing, 37(1), 6-23.
Blascovich, J., & Bailenson, J. N. (2011). Innite reality: Ava-
tars, eternal life, new worlds, and the dawn of the virtual
revolution. NY: Harper.
Bowman, N. D. and Boyan, A. C. (2008, May). Cognitive Skill as
a Predictor of Flow and Presence In Naturally Mapped
Video Games. Paper presented at the annual meeting
of the International Communication Association, Mon-
treal, Quebec, Canada. DOI= http://www.icahdq.org/
conferences/2008/confprogram.asp
Clearwater, D. (2011). What denes video game genre? think-
ing about genre study after the great divide. Loading,
5(8), 4-28.
Das, J. P. (2002). A Better Look at Intelligence. Current Direc-
tions in Psychological Science, 11, 28–32.
de Koninck, J. M., & Koulack, D. (1975). Dream content and
adaptation to a stressful situation [Abstract]. Journal of
Abnormal Psychology, 84(3), 250-260.
Domhoff, G. William. (1996). Finding Meaning in Dreams: A
Quantitative Approach. New York City: Plenum Press.
Foulkes, D., & Rechtschaffen, A. (1964). Presleep determina-
tion of dream content: Effects of two lms [Abstract].
Perceptual and Motor Skills, 19(3), 983-1005.
Foulkes, D., Pivik, T., Steadman, H. S., Spear, P. S., & Symonds,
J. D. (1967). Dreams of the male child: An EEG study
[Abstract]. Journal of Abnormal Psychology, 72(6), 457-
467.
Gackenbach, J.I. (2006). Video game play and lucid dreams:
Implications for the development of consciousness.
Dreaming: Journal of the Association for the Study of
Dreams, 16(2), 96-110.
Gackenbach, J. (2009a). Electronic media and lucid-control
dreams: Morning after reports. Dreaming, 19(1), 1-6.
doi:10.1037/a0013781
Gackenbach, J.I. (2009b). Video Game Play and Consciousness
Development: A Replication and Extension. Interna-
tional Journal of Dream Research, 2(1), 3-11. Retrieved
June 11, 2009 from http://archiv.ub.uni-heidelberg.de/
ojs/index.php/IJoDR/issue/view/71/showToc.
Gackenbach, J. I., & Bown, J. (2011). Video game presence as
a function of genre. Loading, 5(8)
Gackenbach, J. I., & Dopko, R. (2011). The relationship between
video game play, dream bizarreness, and creativity. Un-
published manuscript.
Gackenbach, J. I., & Kuruvilla, B. (2008a). The relationship be-
tween video game play and threat simulation dreams.
Dreaming, 18(4), 236-256.
Gackenbach, J. I., & Kuruvilla, B. (2008b). Video game play ef-
fects on dreams: Self-evaluation and content analysis
. Eludamos. Journal for Computer Game Culture, 2(2),
169-186. Retrieved from http://www.eludamos.org/in-
dex.php/eludamos/article/view/53/87
Gackenbach, J. I., & Rosie, M. (2009). Cognitive evaluation of
video games: Players perceptions. Paper presented at
the Future Play @ GDC Canada, Vancouver, British Co-
lumbia. Retrieved from http://portal.acm.org/citation.cf
Commentary
International Journal of Dream Research Volume 4, No. 2 (2011)76
D
IJoR
m?id=1639601.1639615&coll=ACM&dl=ACM&CFID=7
3975022&CFTOKEN=29423495
Gackenbach, J. I., Kuruvilla, B., & Dopko, R. (2009). Video game
play and dream bizarreness. Dreaming, 19(4), 218-231.
Gackenbach, J. I., Kuruvilla, B., Dopko, R., & Le, H. (2010).
Chapter 5: Dreams and video game play. In A. Soria,
& J. Maldonado (Eds.), Computer games: Learning ob-
jectives, cognitive performance and effects on devel-
opment (pp. 127-136). Hauppage, NY: Nova Science
Publishers.
Gackenbach, J. I., Matty, I., Kuruvilla, B., Samaha, A. N., Zed-
erayko, A., Olischefski, J., & Von Stackelberg, H. (2009).
Video game play: Waking and dreaming consciousness.
In S. Krippner (Ed.), Perchance to dream (pp. 239-253).
Hauppauge, NY: Nova Science Publishers.
Gackenbach, J. I., Rosie, M., Bown, J., & Sample, T. (2011).
Dream incorporation of video game play as a function
of interactivity and delity. Dreaming, 21(1).
Gackenbach, J. I., Sample, T., & Mandel, G. T., M. (2011). Dream
and blog content analysis of an OCD video gamers long
term diary. Dreaming, 21(2), 124-147.
Gackenbach, J.I. & Rosie, M. (2009). Cognitive Evaluation of
Video Games: Players Perceptions. Proceedings of the
2009 Conference on Future Play on @ GDC Canada,
Vancouver, British Columbia, Canada. Retreived Jan.
20, 2010 from http://portal.acm.org/citation.cfm?id=16
39601.1639615&coll=ACM&dl=ACM&CFID=73975022
&CFTOKEN=29423495.
Hall, C., & Van de Castle, R. (1966). The content analysis of
dreams. East Norwalk: Appleton.
Hobson, A., & Schredl, M. (2011). The continuity and discon-
tinuity between waking and dreaming: A dialogue be-
tween Michael Schredl and Allan Hobson concerning
the adequacy and completeness of these notions. In-
ternational Journal of Dream Research, 4(1), 3-7.
Hobson, H. (2009). The neurobiology of consciousness: Lucid
dreaming wakes up. International Journal of Dream Re-
search, 2, 41-44.
Hobson, J. A. (2009). REM sleep and dreaming: towards a the-
ory of protoconsciousness. Nature Reviews Neurosci-
ence, 10(11), 803-813.
Huh, S. (2008). The Inuence of Genre, Perception, and Person-
ality on Dependent Video Game Use. Paper presented
at the annual meeting of the International Communica-
tion Association.
Hunt, H. (1991). The multiplicity of dreams: Memory, imagina-
tion, and consciousness. Hartford, CN: Yale University
Press.
Johnson, G.M. (2008). Cognitive Processing Differences Be-
tween Frequent and Infrequent Internet Users. Comput-
ers in Human Behavior, 24, 2094–2106.
Kahan, T. L. (2001). Consciousness in dreaming: A metacogni-
tive approach. In K. Bulkeley (Ed.), Dreams: A reader on
the religious, cultural, and psychological (pp. 333-360).
New York: Palgrave.
Mason, L. I., & & Orme-Johnson, D. W. (2010). Transcendental
consciousness wakes up in dreaming and deep sleep
commentary on “the neurobiology of consciousness:
Lucid dreaming wakes up” by J. allan hobson. Interna-
tional Journal of Dream Research, 3(1), 28-32. Retrieved
from http://archiv.ub.uni-heidelberg.de/ojs/index.php/
IJoDR/article/view/595
McMahan, A. (2003). Chapter 3: Immersion, engagement and
presence. In M. Wolf and B. Perron (Eds.) The video
game theory reader, NY: Routledge (p. 67-83).
Murzyn, E. (in press). Incorporation of game elements in dreams:
Exploratory research into World of Warcraft dreaming.
In J.I. Gackenbach (Ed.), Video Game Play and Con-
sciousness, NY: Nova Science Publishers.
Nielsen, T., Kuiken, D., Alain, G., Stenstrom, P., & Powell, R.
(2004). Immediate and delayed incorporations of events
into dreams: Further replication and implications for
dream function. Journal of Sleep Research, 13(4), 327-
336. Retrieved June 1, 2009, from PsycINFO database.
Powell, R. A., Nielsen, T. A., Cheung, J. S., & Cervenka, T. M.,
(1995). Temporal delays in incorporation of events into
dreams. Perceptual and Motor Skills, 81, 95-104.
Preston, J. (1998). From mediated environments to the devel-
opment of consciousness. In J. I. Gackenbach (Ed.),
Psychology and the internet (pp. 255-291). San Diego:
Academic Press.
Preston, J. (2007). From mediated environments to the devel-
opment of consciousness II. In J. I. Gackenbach (Ed.),
Psychology and the internet (2nd ed., ). San Diego: Aca-
demic Press.
Schredl, M., & Hofmann, F. (2003). Continuity between wak-
ing activities and dream activities. Consciousness and
Cognition, 12, 298-308.
Sherry, J., Rosaen, S., Bowman, N. & Huh, S. (2006). Cognitive
Skill Predicts Video Game Ability. Paper presented at
the annual meeting of the International Communication
Association, Dresden International Congress Centre,
Dresden, Germany. DOI= http://www.icahdq.org/con-
ferences/program.asp
Stickgold, R., Hobson, J. A., Fosse, R., & Fosse, M. (2001).
Sleep, learning, and dreams: Off-line memory repro-
cessing. Science, 294(5544), 1052-1057. doi:10.1126/
science.1063530
Swanson, D., & Gackenbach, J. I. (2011). Morning after dreams
of video game play versus Meditation/Prayer. Paper
presented at the Paper Presented at the Annual Meeting
of the International Assocation for the Study of Dreams,
Netherlands.
Tanis, M. & Jansz, J. (2008). Gaming for Different Reasons:
What Motivates People to Play a Specic Video Game?
Paper presented at the annual meeting of the Interna-
tional Communication Association.
Wamsley, E. J., Perry, K., Djonlagic, I., Reaven, L. B., & Stick-
gold, R. (2010). Cognitive replay of visuomotor learning
at sleep onset: Temporal dynamics and relationship to
task performance. Sleep, 33(1), 59-68.
