Can cooperative video games encourage social and motivational inclusion of at-risk students?

Abstract

Can the challenges encountered in cooperative video games encourage classroom inclusion? And can this experience be translated into curriculum engagement? This study describes a 3 week intervention with game-based learning activities in eight lower secondary classrooms (N=190). The intervention combined the use of the coop action role-playing game Torchlight II and analogue gamification aimed at including 32 students challenged by social difficulties and lack of motivation. The video game was used to create more inclusive classrooms by increasing students' opportunities for participation through collaboration in teams. The students also participated in game-related Danish (L1) and Mathematics activities. Effects on social well-being, learning and motivational patterns were measured through teacher assessment combined with the Children's Perceived Locus of Causality-scales (c-PLOC). The results show multidimensional effects including positive impact on at-risk students' well-being and reduced experiences of external regulation to participate in Mathematics and Danish. The qualitative analysis confirms the positive findings, but also shows how the intervention created ambiguities surrounding the relationship between game activities and curriculum-related assignments. The findings indicate that the impact of game-based classrooms is not due to their fun element, but rather how they enable reframing of social participation and students' engagement with the curriculum.

Full text

1
Can cooperative video games encourage social and motivational inclusion of at-risk students?
Thorkild Hanghøj, Andreas Lieberoth and Morten Misfeldt
Thorkild Hanghøj is an Associate Professor in game-based learning at the Department of Psychology and
Communication at Aalborg University. His main research interests are games and literacy, game-based
teacher roles, and the use of games for furthering inclusive education. Andreas Lieberoth is a PlayTrack
fellow at the Aarhus University Interacting Minds Center, and an Assistant Professor in educational
psychology at the Danish School of Education. He studies play, games, motivation and technology though
the lenses of educational, social and cognitive psychology using experiments and mixed methods
research. Morten Misfeldt is a Professor at the Department of Learning and Philosophy at Aalborg
University. His research interests include the use of ICT in mathematics curriculum and mathematical
practices in various areas of society. Address for correspondence: Dr Thorkild Hanghøj, Aalborg
University, Department of Communication and Psychology, A. C. Meyers Vænge 15, DK-2450 Copenhagen
SV, Denmark. Email: thorkild@hum.aau.dk
Abstract
Can the challenges encountered in cooperative video games encourage classroom inclusion? And can this experience be
translated into curriculum engagement? This study describes a 3 week intervention with game-based learning activities
in eight lower secondary classrooms (N=190). The intervention combined the use of the co-op action role-playing game
Torchlight II and analogue gamification aimed at including 32 students challenged by social difficulties and lack of
motivation. The video game was used to create more inclusive classrooms by increasing students’ opportunities for
participation through collaboration in teams. The students also participated in game-related Danish (L1) and
Mathematics activities. Effects on social well-being, learning and motivational patterns were measured through teacher
assessment combined with the Children’s Perceived Locus of Causality-scales (c-PLOC). The results show
multidimensional effects including positive impact on at-risk students’ well-being and reduced experiences of external
regulation to participate in Mathematics and Danish. The qualitative analysis confirms the positive findings, but also
shows how the intervention created ambiguities surrounding the relationship between game activities and curriculum-
related assignments. The findings indicate that the impact of game-based classrooms is not due to their fun element, but
rather how they enable reframing of social participation and students’ engagement with the curriculum.
Keywords
Game-based learning, commercial video games, cooperative gaming, social play, inclusion, at-risk students,
motivation, framing
Since the dawn of online gaming, a growing literature has demonstrated how playing challenging
games together can encourage friendship, social inclusion, offline social support, social self-
efficacy, and general identity building around the shared pursuit (Trepte, Reinecke & Juechems,
2012; Domahidi, Festl & Quandt, 2014; Kaye, Kowert & Quinn, 2017). In this same period, digital
games have become a key medium in the education field (Persico et al., 2017), but the combined
social and curricular promises of gaming have hereto not been studied in the context of classroom
inclusion interventions.
The spread of gaming has inspired researchers and educators to adopt various forms of game-based
learning activities within educational contexts, including analysing games as texts (Beavis,
Dezuanni & O’Mara, 2017), teaching with games as tools (Squire, 2004), learning through game
design processes (Kafai & Resnick, 1996), or using game-like elements to “gamify” educational
practices (Sanchez, Young & Jouneau-Sion, 2016). Of particular interest to the present study, the
Level Up survey of teachers (Takeuchi & Vaala, 2014: 50) found that educators commonly use
games for improving inclusion of at-risk students, with 55% of respondents especially valuing
games for their ability to motivate low-performing and disabled students. In terms of measurable
impact, researchers investigating this broad spectrum of proposed benefits have found an equally

2
broad spectrum of impacts depending on purpose (Connolly et al., 2012; Young et al., 2012; Clark,
Tanner-Smith & Killingsworth, 2016).
Studies of online gaming outside school have suggested potential social benefits in terms of
creating a sense of closeness, friendship and belonging, especially when linking online and offline
relationships (Domahidi, Festl & Quandt, 2014; Kowert & Oldmeadow, 2014; Sundberg, 2018).
Existing literature on games and inclusion of at-risk students has, however, tended to focus on the
use of games made specifically for learning to address specific learning difficulties (Stewart et al.,
2013). Even though such “serious games” are predicated on the promise of “fun” and may offer
advantages in terms of specific students’ needs, and align with curricular aims, games designed for
instructional purposes sometimes suffer from poorly designed gameplay that can create limited
student motivation (Rigby & Ryan, 2011; Wouters et al., 2013). For this reason, teachers and
researchers have more recently turned to the use of commercial off-the-shelf (COTS) games such as
The Sims, World of Warcraft or Minecraft (Steinkuehler, 2011; Lacasa, Méndez & Martínez, 2008;
Callaghan, 2016). A new generation of game using teachers (Lieberoth & Hanghøj, 2017) thus
emphasize the practical need to both invent games of their own and create congruent links between
curricular aims and games, which were not initially designed for educational purposes (De Grove,
Bourgonjon & Van Looy, 2012).
Some of these COTS-game interventions have been targeted towards at-risk students. One
longitudinal study, for instance, investigated a two year experimental curriculum aimed at boys at-
risk of dropping out of upper secondary school (Westin & Lange, 2012). The study demonstrated
promising results in terms of socialising at-risk students around World of Warcraft-related learning
activities, which were looser organised than the fixed time schedules in formal education. Another
study demonstrated how struggling male readers, who played World of Warcraft on a regular basis,
were able to read complex instructional game texts up to six grades levels above their measured
ability to read similar school texts (Steinkuehler, 2011). A number of studies have also described
how the educational use of commercial games such as The Sims (Lacasa et al., 2008) or Minecraft
(Callaghan, 2016) afford students new trajectories for participation and collaboration, which may
create meaningful links to life experiences outside school. This aligns well with research on
inclusion of at-risk students, which shows that differentiated instruction that offers varied forms of
active participation, may be a valuable way of creating more inclusive classrooms (Gibson, 2013).
In this way, it may be argued that the many motivations to play (Rigby & Ryan, 2011) and varied
opportunities for social participation offered by complex commercial video games is a key reason
for using “real” games to include atrisk students.
In this study, we therefore investigate the “School at Play” approach for using commercial video
games to support inclusion of students at-risk, which we define as students, who are perceived by
teachers as being inactive participants in the classroom mainly due to social difficulties – e.g.
shyness, disruptive behaviour or social exclusion by classmates – but also due to lack of motivation
to learn and low academic performance (Dyssegaard & Larsen, 2013). Following a relational and
situated perspective, at-risk students are primarily viewed as being in social and motivational
difficulties, and to a lesser degree as having specific disabilities or special needs (Tetler &
Langager, 2009). Thus, the at-risk students’ social and motivational difficulties are primarily
understood through the students’ lacking possibilities for meaningful participation within the
learning community of the classroom. This broad categorization of at-risk students differs from
more specific approaches, which identify at-risk students in relation to pre-defined categories such
as age, gender, ethnicity, socioeconomic status, behaviour, learning disabilities and/or academic

3
ability (McGrath & Van Bergen, 2015). Having a specific impairment such as ADHD, autism or
dyslexia may have a significant impact on social opportunities and motivations to learn. Diagnosis
may facilitate access to support, though should not be used as a gate keeping mechanism. However,
the individual perspective, where the child is labelled as the carrier of a specific diagnosis, also
entails a significant risk of stigmatising and locking students into marginalised positions (Kirby,
2017). Instead, this paper primarily classifies at-risk students according to how individual teachers
experience particular students’ lack of participation related to social difficulties and low academic
motivation, regardless of whether or not they have a diagnosis. This bottom-up or “local” approach
is informed by comprehensive research on inclusion, which shows that teacher-student relationships
are highly important in terms of including at-risk students, especially in relation to student
behaviour and low student performance (McGrath & Van Bergen, 2015).
More specifically, the aim of the paper is to measure the effect of an intervention designed to
leverage COTS game activities in combination with game-related assignments to reframe at-risk
students’ opportunities for classroom participation in the two school subjects Mathematics and
Danish. The intervention was based on the School at Play approach (www.skolenispil.dk), which
involves a combination of student collaboration through commercial multi-player video games (e.g.
Torchlight II and Minecraft) and teachers’ use of analogue gamification tools in order to promote
positive student behaviour as well as supporting students’ progression within game-related
curricular modules.
In summary, the current study represents a much needed contribution to the limited body of
empirical studies investigating how commercial video games can be used to include at-risk
students, which focuses on both social participation and motivation to learn. The mixed methods
approach provides valuable knowledge, which can both describe and explain the potential
challenges and possibilities inherent to game-based learning as a tool for combined social and
learning interventions. By combining a repeated measures approach with explanatory qualitative
observations, this study thus investigates the following overall research questions:
 How and to what extend did the School at Play intervention support changes in
participation, motivation and experience of disciplinary assignments for students?
 Will there be positive effects for at-risk students (broadly understood as students who
display difficulties related to social inclusion and classroom participation) from the
intervention, and is there a difference between the positive effects for at-risk students and
their peers?
The School at Play intervention
The School at Play approach was initially developed by a team of independent educational
consultants based on positive experiences with games in their jobs as teachers and special-needs
educators. The approach can be summed up as a combined pedagogical use of commercial games,
game-related assignments and analogue gamification of classroom behaviour in order to address
curricular and social aims concurrently.
The cornerstone of the School at Play-approach is using commercial video games to create
meaningful contexts for collaboration and learning. Students play and explore game worlds (e.g. the
co-op action role-playing game Torchlight II) in order to experience specific game mechanics and
tactics, but must also master disciplinary knowledge within Mathematics or Danish, that may grant

4
advantages when playing the game together. For instance, Torchlight II features a system of
wielding weapons which each take a set amount of time to swing and deal set amounts of damage,
in order to deplete opponents’ health points - see screenshot below.
Fig. 1. Screenshot from Torchlight II, which illustrates key game mechanics: character level,
weapon damage, and health points.
This process lends itself well to inferring efficient attack combinations with mathematics. One core
aim of the method is to establish an interplay between understanding how some game mechanics
relate to disciplinary knowledge in e.g. Danish and Mathematics, and, on the other hand, learning
how to use disciplinary knowledge in order to improve game play. In this way, the School at Play-
approach assumes that it is possible to create a two-way dynamic relationship when teaching with
commercial games, which may both increase students’ interest in using disciplinary knowledge to
become better players, and also use students’ interest in playing to increase their disciplinary
motivation to learn.
In addition to using COTS games for creating meaningful learning contexts, the School at Play
approach also involves a number of analogue gamification tools that facilitate game dynamics in the
classroom though “Portal Assignments”, the “Progress Bar” and the “Classroom Game”. The aim of
the “Portal Assignments” is to conduct classroom discussions, which may help students connect
disciplinary knowledge to in-game tasks. Hence the term “Portal”, this denotes a connection
between the game world and disciplinary activities. The example shown below links Torchlight II
and Mathematics by asking students to use their ability of division in order to address and improve
their understanding of in-game challenges.

5
A Health Potion gives 900 health points within 8 seconds. A Big
Health Potion gives 1.800 within the same time span.
How much health per second do you get from 1 Health Potion?
How much health pr. second do you get from 1 Big Health Potion?
Imagine that you are running away from a group of skeletons and
have 200 out of 500 health points. Which potion do you drink? Why?
Fig. 2: Portal assignment in Mathematics for Torchlight II
As students solve assignments, they may move their name on a Progress Bar, which is located on a
wall in the classroom. The Progression Bar ranges from 0 to 100% completion, with the additional
option of progressing up to 150%. Beforehand, the students are divided into three different levels of
expertise by the teacher, so low-performing students are given fairly easy possibilities for
progression.
In contrast to the other two tools, which link activities and curricular aims, the primary aim of the
Classroom Game is to promote positive student behaviour in the classroom. The game is based on
“Class Virtues” (such as arriving in class “On time” or showing “Respect” by not making
unnecessary noise.
The pedagogical values behind the School at Play approach stress the importance of giving positive
feedback by visibly acknowledging at-risk students’ achievements in-game and/or in-class, related
to both disciplinary knowledge and behaviour. In this way, the teacher plays a crucial role not only
in planning and staging the intervention, but also in facilitating dialogue, setting visible goals, and
providing relevant forms of feedback on students’ progression and behaviours. As such, the goal of
the School at Play intervention is to create new social spaces, give clear feedback on social
behaviour in a playful way, and facilitate added student identification with the goals of school
subjects though reframing in terms of how those subjects can also be relevant to being better game
players.
Improving student participation through games
Preceding the present study, qualitative observations of the School at Play consultants’ work
suggested that the combination of Torchlight II and analogue gamification elements opened new
possibilities for active participation through students’ knowledge sharing and collaboration, as well
as improved positive social relations (Hanghøj, 2015). According to frame theory (Goffman, 1974;
Waern, 2012; Lieberoth, 2015), game and play activities create interpretative frames, which
organize how players experience, communicate and negotiate meaning in social situations. This
paper therefore adopts the working assumption that curriculum-relevant elements found within
COTS games can be used by teachers to reframe students’ experience of school subjects, in order to
provide a new sense of shared meaning and individual motivation.
In order to understand how students experience motivational changes through COTS games, the
present study follows the perspective of Self-Determination Theory. Based on empirical studies,
this theory has been used to explain how games may motivate through an increased sense of
autonomy, relatedness, and competence (Ryan, Rigby & Przyblski, 2006). For the purpose of this
paper, we wish to focus on how students become motivated when making social and disciplinary

6
meaning out of their game experiences. According to Self-Determination Theory, meaning can be
understood as an experience, which arises when individuals pursue intrinsic purposes and are
reasonably successful in achieving those (Weinstein et al., 2013). Instead of simply testing for game
induced “fun”, we therefore use Children’s Perceived Locus of Causality (c-PLOC) scales
(Pannekoek, Piek & Hagger, 2014) as a measure of the balance between inner and outer motivations
present in Danish and Mathematics classes respectively. c-PLOC is based on Self-Determination
Theory (Ryan & Deci, 2000), which assumes that engagement stems from an interplay between
inner and outer drivers, leading an ongoing internalization of goals, demands and motives in
everyday life. According to this theory, the inner adaptations that occur in the course of
socialisation means that students gradually come to accept values advocated or modelled by peers,
teachers and parents, experiencing them as their own. Conversely, situations of insurmountable
demands or lacking needs fulfilment, e.g. in the context of particular school subjects, may represent
a disjuncture from inner motives, which according to SDT can be detrimental to motivation and
participation, as well as – in the long term – to human functioning more generally (Ng et al., 2012;
Ryan & Deci, 2000). The c-PLOC instrument’s five scales are intended to represent the continuum
between internal and external drivers for motivation (Ryan & Connell, 1989) including identified
regulation, denoting that individuals engage in behaviour because they understand the behaviour’s
relevance, and value the anticipated outcomes. In this way, the c-PLOC scale allows this study to
look beyond “fun”, and instead track changes in motivational dynamics including feelings of
external regulation and levels of integration between school demands and personal motives.
The first hypothesis (H1) heading into this effect study was thus that the School at Play experience
would afford new opportunities for social participation, and possibly for beneficial long term social
repositioning in relation to teachers and peers. Based on the goals explicit to the School at Play
method, we further test the hypothesis (H2) that the experience would create shared motivation for
participating in Danish and Mathematics, including an increased understanding of the
value/usefulness of those subjects when viewed through the lens of gaming.
Further as H1.1 and H2.1 respectively, we investigate how the students identified as being at-risk in
terms of social inclusion, would show the largest degrees of social and curriculum-motivational
improvement compared to their peers. For the purposes of this intervention, at-risk status was
assigned through a combination of teacher interviews, diagnoses (if any), and observations of the
eight classrooms before the interventions (see below). Specific needs and diagnoses were taken into
account only as a contributing factor in classroom inclusion/exclusion – not as a determining factor
of at-risk status, even though specific needs clearly affected different children’s participation
trajectories as discussed in the qualitative case analysis section below. This hypothesis was
informed by the broad tradition of research into inclusion of at-risk students mentioned earlier as
well as the fact that the School at Play methodology was first developed for special needs
classrooms, which was also reflected in the pilot where positive social outcomes were especially
observed for at-risk students (Hanghøj, 2015). We thus assumed that the game-related assignments
could not only create spaces for social participation for those positioned at the margin of typical
classroom activities, but also help these at-risk students with negative learning experiences to
identify with and develop positive identities as doers (Boaler & Greeno, 2000) of e.g. Mathematics
or Danish thus encouraging both intrinsic motivation and identified regulation as measured by the
c-PLOC instrument.

7
Method
The research was carried out as a parallel mixed methods intervention, which allowed a
combination of a repeated measures effect study with classroom observations and student
interviews. Here, we will first focus on statistical effects and then, in a more explanatory capacity
(as per Creswell & Plano Clark, 2013,) use three student cases illustrating convergent patterns of
social inclusion but divergent experiences of the game-related learning activities.
Participants
The intervention study involved eight classes from four schools, grades 3-6 (age 9-12), for a total of
190 students (age M = 11.48, SD = .84, 98 female). Out of this group, 32 students with 4 students
in each class (age M = 11.37, SD =.94, 12 female) had been identified as at-risk. In this way, there
was a lower percentage of female students among the at-risk students than in the overall student
group.
The participants were identified through teachers’ and researchers’ relational perspective on how
students were perceived as being in social and motivational difficulties within local classroom
settings. The identification and selection of at-risk students were identified through a combination
of teacher interviews conducted by the researchers prior to the intervention, which focused on how
the teachers perceived their students’ well-being and motivation to learn, existing knowledge of
student diagnoses (if any), and observations of the eight classrooms before the interventions. The
further process of identifying at-risk students involved structured observations in all eight
classrooms that focused on students who showed limited social and disciplinary participation. This
process involved collaboration among two researchers in order to avoid bias and ensure validity.
Based on these three different types of data, four “focus students” were identified by the researchers
for each of the eight classrooms and further validated in dialogue with the teachers. The students
were not aware that they were identified as at-risk students. Three to four weeks after the
intervention, interviews were conducted with all the focus students, as well as the teachers, which
focused on the impact of the intervention. In accordance with Danish research ethics standards first
teachers, and then parents of the students, gave informed consent to participate in the research. All
the four schools were located in or close to larger cities.
Although there were relatively more males in the at-risk group, the authors consider that the
difference in gender distribution did not have a significant impact on the results. Even though few
of the girls had experience with multiplayer action games, and were “forced” to play Torchlight II
at the highest level of difficulty, they eagerly accepted the game challenge and primarily reported
positively on their game experiences through interviews, surveys and classroom discussions. There
were no significant differences in age distributions between at-risk students and their peers.
Measures and instruments
The quantitative data involved two different kinds of measures: Teacher ratings were collected
before and after the intervention, and self-reported c-PLOC measures were obtained before, during
and after. The Children’s Perceived Locus of Causality Scale (c-PLOC) has previously been used to
assess motivation in school domains such as curriculum learning and physical education
(Pannekoek et al., 2014). The instrument contains five three-item subscales expressing different
situational motivators on a continuum of self-determination (as per Ryan & Connell, 1989):
intrinsic motivation (fun, enjoyment), identified regulation (wanting to learn, “seeing the point” of
activities), introjected regulation (internal impetus to participate or do well in own and others’
eyes), external regulation (regulation by demands, rewards or punishments) and amotivation
(participating with no conscious notion why). Heightened intrinsic motivation, identified regulation

8
and to a lesser extent introjected regulation are interpreted as improvements in regard to H2, as are
lowered scores on external regulation and amotivation. The c-PLOC scale was adapted to fit
Danish and Mathematics classes, back-translated from English, and pre-validated using students
outside of the project ( = .64 - .80). Teachers were also asked to rate students’ general
performance in Mathematics and Danish, as well as their social participation using a 6-point scale.
Procedure
The intervention was based on the School at Play approach, which is described in more detail
above. Teachers were prepared for the process through two whole day workshops, where they
played Torchlight II and learned to use the game tools of the intervention. The Mathematics
teachers were instructed how to identify basic game mechanics within Torchlight II such as attack
patterns, DPS (Damage Per Second), different types of healing potions, and how to use in-game
problems as meaningful contexts for mathematics assignments – e.g. how to use mathematics to
help decide which healing potion to use when under attack. Similarly, the Danish teachers were
introduced to the genre aspects of the action role-playing game and instructed on how to create
assignments, which allowed the students to write game guides and character analysis based on their
game experiences. Moreover, the teachers were also instructed on how to use the three analogue
gamification tools – Portal Assignments, Classroom Game, and Progress Bar – which aimed to
further classroom discussion and student reflection on the game-related assignments, classroom
behaviour, and curricular progression.
The classroom interventions consisted of three week teaching activities in all Danish and
Mathematics lessons combining Torchlight II, associated Mathematics and Danish assignments, and
classroom gamification. Each intervention started off with a 2-3 hour game session using a LAN
(Local Area Network) setup, where the students became immersed in the game world and learned
how to master the basic game play. The students played together in groups of 3-6 players against
the computer. All the groups were decided by the teachers and involved a mix of at-risk students
and regular students. The game was set to the highest level of difficulty (“Elite”) as to make the
game so challenging that the students were more or less forced to collaborate in order not to die and
progress together within the game.

9
Fig. 3. Example of a classroom playing Torchlight II in a LAN set up.
After the initial game session, the students would take part in classroom discussions about their
game experiences, which would be linked to disciplinary content through Portal Assignments. For
the remaining part of the intervention, the students would be working with game-related
assignments and only play the game additional 2-3 units per week.
Two weeks before the intervention, students were asked to fill in pre-test surveys online in class.
This procedure was repeated during the intervention’s second week, and in a post-measure three
weeks after the intervention. Before the intervention and three weeks after, teachers were also asked
to rate each student’s general performance in Mathematics and Danish, as well as their general
social well-being understood as social participation in classroom activities. Specifically, the
procedure was put in place to explore the following hypotheses:
 H1: Students’ inclusion and well-being will improve, as quantified though teacher assessed
social participation before and after the School at Play intervention and reported through
student and teacher interviews.
o H1.1: At-risk students will display larger positive changes than their peers on the
factors described in H1.
 H2: Students engagement and performance in Danish and Mathematics will improve as
measured on c-PLOC scores as well as teacher assessed performance for each of the two
subjects and reported through student and teacher interviews.
o H2.1: At-risk students will display larger positive changes than their peers on the
factors described in H2.
Data analysis
The quantitative data was analysed using IBM SPSS 24. In order to mitigate the risk of type two
error from multiple comparisons, false discovery rate correction (Benjamini & Hochberg, 1995)

10
was used to calculate q-values within each block of repeated tests (
http://www.sdmproject.com/utilities/?show=FDR). FDR-corrected values are reported alongside the
p-values, and should be interpreted according to the same alpha-standards. The qualitative data
from the student interviews was transcribed, coded and subjected to thematic analysis in order to
identify key analytical themes, which could expand the quantitative findings (Braun & Clarke,
2006). Three representative cases exemplifying similarities and differences across the analytical
themes “collaboration within the game”, “social relations”, “motivation to play games at school”
(intrinsic motivation) and “motivation to learn through games” (identified regulation) were chosen
for presentation below.
Results
Statistical analyses showed a fairly complex pattern of changes to motivation and achievement in
the 190 students. A series of FDR-corrected repeated measures ANOVAs were used to first identify
effects on students in general. Table 1 and figures 4-6 summarize changes over time with asterisks
denoting significant drops and increases as revealed by Bonferroni post-hoc tests.
There were changes to most variables: Teacher assessed social participation rose significantly from
before to after the intervention. Further, amotivation (Danish), and experiences of being externally
pressured to participate (Mathematics) decreased significantly between all three time points.
On the more puzzling side, some variables like identified regulation (Danish), and intrinsic
motivation (Mathematics) decreased. In contrast to the simplistic “games make learning fun, fun
makes better learning”-notion found in writings on “serious games”, the pattern revealed here is
complex. As such, while hypothesis 1 was supported by significant rises in social participation
reported over time, there was only partial support for hypothesis 2: While performance increased
for Danish, this was not the case for mathematics. And while scores on factors related to external
pressure to participate decreased, which is a positive development in motivational terms, more
intrinsic factors did not rise accordingly. It should also be noted that, as seen in fig 4 & 5 some of
the significant effects displayed in table 1 are due to changes occurring after the intervention was
over, as students returned to regular classroom activities. The results show that motivation to
participate cannot easily be reduced to one factor, or be assumed to generalize between Danish and
Mathematics assignments. These findings are further discussed in the qualitative analysis below.
df
F
p
q

p2
Social participation
1
5.25
.001
.001**
.09
Danish
1
15.53
.001
.001**
.01
Intrinsic Motivation
2
4.05
.23
.23
.01
Identified regulation
2
22.63
.001
.001**
.14
Introjected regulation
2
10.87
.01
.01**
.07
External regulation
2
2.15
.09
.11
.02
Amotivation
2
7.84
.001
.001**
.05
Mathematics
1
3.31
.07
.09
.02
Intrinsic Motivation
2
13.31
.001
.001**
.09
Identified regulation
2
4.79
.01
.01**
.03

11
Introjected regulation
2
8.14
.001
.001**
.06
External regulation
2
14.63
.001
.001**
.10
Amotivation
2
1.97
.14
.15
.01
Table 1: Motivational changes over time. df = degrees of freedom, F = ANOVA test value, p and q =
nonadjusted and adjusted significance levels (* = significant at the .05 alpha level, ** = significant at the .01
alpha level),

p2 = ANOVA effect size. Social participation, Danish and Mathemathics (in bold) are teacher
assessed measures from before to after the intervention. The five motivation/regulation variables under
Danish and Mathematics (non-bold) show changes in c-PLOC scores related to each of those two subjects as
measured before, during and after the intervention. See also fig. 4 & 5 for a graphical representation and
statistics related to changes in c-PLOC scores between the three time points.

12
Motivation changes over time: Danish
Motivation changes over time: Mathematics

13
Fig. 4 & 5: c-PLOC motivations over time: ** p < .01 change between two timepoints, ** at the line
beginning denotes change from pretest to posttest.
Effects on at-risk students
In order to test hypotheses 1.1 and 2.1 that at-risk students would benefit significantly differently
from their peers, at-risk status was added as a between-subjects factor in the repeated measures
model described above. A significant interaction effect, f(1) = 13.81 , p < .001, q = .003

p2 = .07,
was detected for social participation. A split-file comparison showed that even though at-risk
students and their peers followed similar positive trajectories, the effect size for at-risk students f(1)
= 17.45 , p <. 001, q < .001

p2 = .36 (a medium effect according to Cohen, 1992) was appreciably
above that seen for other students f(1) = 6.26 , p = .01, q = .02

p2 = .04. No such interaction,
however, was found for subject-related performance or motivation. The primary statistical analysis
thus confirms H1.1 but not H2.1.
As can be seen in figure 6, however, at-risk students’ average experiences followed unique
trajectories for many variables. It is, however, worth noticing that all ten average scores ended up
converging roughly with those of their classmates’, perhaps revealing a process of shared social
expectations around Danish and Mathematics alike.

14
Intrinsic motivation – Danish and Mathematics

15
Identified regulation – Danish and Mathematics

16
Introjected regulation – Danish and Mathematics

17
External regulation – Danish and Mathematics

18
Amotivation – Danish and Mathematics
These nuances were explored by using split-file procedures to compare the change trajectories of
the 32 at-risk students with those of their 158 peers: While the peer group displayed a significant
increase f(1) = 11. 27, p = 0.001, q = .003

p2 = .07, in teacher-assessed Danish performance as a
result of the intervention, this development gain was only seen at trend-level in the at-risk group
after FDR-correction, f(1) = 4.25 , p = .048, q = .08

p2 = .12. This less robust learning gain is the
only parameter where the at-risk group did not show more promising developments than the peer
group, perhaps suggesting that while most students were ready to engage and learn, the at-risk
Fig. 6. Comparing c-PLOC motivation changes in at-risk students vs. peers.

19
group was more heterogeneous in their challenges to participation and learning, even though their
motivations and participation changed in a number of ways.
Firstly, the at-risk students displayed a significant decline in feelings of external regulation
(Danish), f(2) = 4.69, p = .01, q = 03

p2 = .15 pre-to-during the intervention. This improvement
was not seen in the peer group. Further, both at-risk students f(2) = 8.67, p = .001, q = .007,

p2 =
.25, and their peers f(2) = 8.56, p < .001, q = .002

p2 = .07 experienced declines in external
regulation (Mathematics), but the effect size seen for at-risk students represented a three times
larger effect size. Together, these findings suggest that one psychological result of the intervention
was to significantly reduce at-risk students’ experience of being cajoled or coerced into
participating in learning activities.
Secondly, while the peer group displayed a somewhat inexplicable decline in intrinsic motivation
(Danish) f(2) = 3.80, p = .02, q =.03

p2 = .03 pre-to-post, and intrinsic motivation (Mathematics)
f(1.89) = 9.92 , p < .001, q < .001 p2 = .08 (Huynh-Feldt corrected) pre-to-post and intervention-
to-post, the at-risk students maintained a positive intrinsic motivation trend for Danish and neutral
trend for math. Similarly, while the peer group display a significant pre-to-post drop in identified
regulation (Mathematics) f(2) = 6.4, p = .002 , q = .004

p2 = .06, the at-risk group showed no such
decline.
As such, although the immediate analysis did not find any interaction effects for engagement and
learning in the subjects of Danish and Mathematics, suggesting that the gains for at-risk students
were the same as for other students, some support for hypotheses 2.1 was found with this second,
more nuanced analysis, of the at-risk group: At-risk students experienced unique improvements in
engagement, but with the negative twist that the group, on average, did not experience the same
clear learning gains as their peers. In conclusion, the study found the strongest effects of the GBL-
intervention on social participation, including selective stronger effects for at- risk students, while
results were mixed for learning and engagement in the specific subjects of Danish and Mathematics.
The at-risk group was, however, not a homogeneous category, so let us expand our understanding
through deeper qualitative acquaintance with three cases, illustrating typical instances of how
students from the at-risk group changed positions in the classroom.
Explanatory expansion: Students’ perception of game-related assignments
In accordance with the explanatory mixed methods design (Maxwell, Chmiel & Rogers, 2015), the
statistical analysis is now expanded through three representative student cases, which are used to
analytically unpack how the game-based intervention influenced the students’ experience of
collaboration and the school subjects. In this type of mixed methods analysis, qualitative methods
can be used to qualify (triangulate), challenge or expand on quantitative results depending on how
findings from the two lenses overlap (Greene, 2007). The examples are based on thematic analysis
interviews conducted with all at-risk students after the intervention and aim to show both patterns of
convergence and patterns of variation.
The thematic analysis clearly showed that the at-risk students enjoyed playing Torchlight II,
especially their collaboration within the game. The students described how the co-op game
experience required them to help each other, share in-game knowledge and develop tactics in order
to progress within the game. Several students described how their mutual dependencies in the game
also lead to the development of new positive relations, fewer conflicts and sometimes even new
friendships with their classmates. This finding corresponds well with the teachers’ assessment of

20
the at-risk students’ increased well-being. Moreover, the students’ experience of a more playful
learning environment may also explain the drop in their perceived external regulation in relation to
the school subjects.
Bellow follows three examples with three different students, who all experienced a positive change
of social relations during the game activities, but experienced a mix of negative and positive
changes in relation to the game-related assignments.
The first student example is the fifth grader Lisa (age 11), who was selected as an at-risk student
due to her socially marginalised position in the classroom. She had no diagnosis but often ended up
in arguments with other students when doing group work and preferred to work alone. Lisa clearly
experienced positive social relations in relation to her classmates during the Torchlight II game
sessions, where she both gave and received a lot of help. In Lisa’s own words, the game sessions
made her:
…want to collaborate more with others, because normally in the groups, I always get into
an argument and I have to go into another group or just work alone. But if we got into an
argument in the game, we managed to get it solved (…) It was like, ‘Argh, I’m in the grotto
and there are tons of monsters, and could you please just come and help me?!’ And it was
like: ‘Okay, we’ll come and help you’. And then they just came in and then we’ll just fight
and then it’s just ‘Yeah!’.
As the example shows, Lisa valued how students helped each other when fighting monsters in the
game. During game-related assignments Lisa also displayed high levels of personal engagement,
expressing both intrinsic motivation and identified regulation, e.g. when writing up scenes from the
game in Danish. But as she already performed well in Danish and Mathematics, she did not
experience the same major change as she did with her transformed social status in the class.
The second example is the sixth grader Martin (age 12), who was relatively new to his class and had
recently been transferred from a class with special needs students. Martin was diagnosed within the
autism spectrum and mostly preferred to work alone. Below Martin is asked how the game sessions
changed his collaboration with a girl in his game group:
K: I work much more together with Sandra
I: You didn’t do that before?
K: If I did it before, it was because I was forced, but…
I: Now it’s more voluntary?
K: Yes, now it’s like… (…) Now, we’ve actually become friends
As the example shows, Martin experienced a new mode of social connection when playing together
with Sandra, who he had previously felt “forced” to collaborate with in the class. This type of
observation was recurrent across students and classrooms, and shows how the game experience
opens up new social positions as collaborators and friends. At the same time, Martin was not
particularly motivated by the game-related assignments. As shown below:
I: What about the work sheets where you both had to do Mathematics and Danish? How
was it to work with that?

21
M: It’s was just the same, it was… I didn’t quite understand it. It was just normal math
assignments except that they said something with Torchlight, where it was put in like…
Okay, of course we didn’t have plus and minus questions, then it would be the same as 5 + 7
and then if this guy is in level 5, what would it be if advanced 2 levels, it was so…
I: So it didn’t mean anything to you that it was related to Torchlight II?
M: No, it’s just like if you play a game and then you just put a skin on… It’s completely the
same, it just looks different… It was like, I couldn’t really care
I: So you didn’t feel more like doing them when they were about Torchlight II
M: Nah
As the excerpt shows, Martin, who was a high performer in Mathematics, did not experience the
framing of the game-related assignments as more meaningful or motivating than other school
assignments. His identified regulation is already in place when it comes to Mathematics, and so he
struggles to align the new experience with his existing mastery.
The third and final example is a fifth grade girl, Nanna (age 11), who was often very quiet in her
class and rarely spoke up. She was diagnosed with dyslexia and performed poorly on both reading
and writing. Just like the other at-risk students, Nanna became very engaged in playing Torchlight
II with her classmates. In her own words: “It was quite fun because you were on a team and then
you helped each other. No one was left behind by our team”.
After a game session, Nanna became very engaged when writing a guide for playing Torchlight II
in Danish and her high level of engagement was emphasized as being unusual by her teacher. From
the student interview:
I: What was it that you liked about it [the assignment]?
N: I liked that you had to write it for someone, which I didn’t know, which I hadn’t heard of.
You did not know the person. I thought that was kind of fun
I: And do you think that was different from the other assignments, which you are given in
school?
N: Sort of
I: What do think that was different?
N: Usually, you just have to write to your teacher or someone you know. But here you had to
write for someone, which you have never heard [of] or met
This student liked the assignment, because it was challenging to write a guide for someone, who did
not know the game well already – a task aimed at someone other than her teacher. In this way, she
viewed the assignment as an opportunity to meaningfully communicate her game experience to
other potential players showing signs of both introjected social expectations and intrinsic
motivation.
Taken together, the three examples clearly illustrate how all the at-risk students valued the
experience of collaborating in Torchlight II. Supporting H1.1, the game sessions provided the at-
risk students new social positions within each game team, which offered them positive possibilities
for participation in the learning community of the classroom. Thus, the three examples showed
signs of developing new friendships, solving conflicts and helping each other.

22
When asked about the game-related assignments, the students’ responses were rather mixed. As
also seen in the quantitative data, the student interviews indicated high degrees of intrinsically
motivated activity surrounding Torchlight II and the classroom game, suggesting identified and
introjected regulation toward e.g. the group’s shared needs. At the same time some students were
indifferent and even negative toward the attempt of linking commercial games with school
curricula.
Even though at-risk students clearly enjoyed playing Torchlight II together in teams, there was great
variation in the students’ experience of the analogue gamification tools and the game-related
assignments, which ranged from being seen as highly engaging, slightly more meaningful than the
usual assignments to being perceived as irrelevant. This ambiguity toward the curricular reframing
of the game intervention may explain why the intervention did not lead to an increase in intrinsic
motivation for the at-risk students, and even a decrease for the total group of students in Danish.
Moreover, these findings indicate that the success of using game interventions to increase student
motivation to a large part depend on designing game-related assignments, which can provide
relevant and meaningful matches between particular games and curricular aims.
As such, it is easy to see how social inclusion was accomplished, but also why there is not one clear
picture for the 32 at-risk students. On one hand, the students differed in their social transformations
and existing motivations. On the other hand, the students’ abilities to identify with the link between
well-known school subjects, and the various types of game activities occurring during the
intervention, represents a tangible challenge for the design and operationalisation of the game-based
learning intervention: This challenge requires teachers to develop a subtle ability to align game-
frames and school-frames though assignments, classroom management and sensitivity to each
student’s meaning-making processes.
Discussion and conclusions
The object of this study was, on one hand, to assess the effects of a multifaceted game-based
intervention on students’ achievement, social inclusion and motivations as measured by the
Children’s Locus of Causality scales, and on the other, to identify the selective impact for at-risk
students, understood as all students in need of support for classroom inclusion.
The study found a significant effect in terms of increased social participation for the at-risk students
in comparison with their peers, which was a confirmation of hypothesis H1.1. Moreover, we found
a major decrease in external regulation for the at-risk students, which means that the intervention
lessened experience of pressure for taking part in the subjects compared to the pre-measure. This is
a partial confirmation of hypothesis H2.1, which predicted that at-risk students would display larger
positive changes than their peers on the c-PLOC scores and teacher assessed performance in Danish
and Mathematics. Indeed, motivation measures changed to a degree, where at-risk students came to
rest en par with their peers. In this way, the game-based intervention appears to have increased
inclusion for the at-risk students. Based on these positive findings, we strongly recommend the use
of co-op games to further social inclusion of at-risk students. The co-op game format, where players
play together in teams against the computer, is a rather unfamiliar game format to many teachers
and students, and is relatively unexplored in educational research. Game research outside
educational contexts emphasize how co-op games require players to reach social agreement on
common goals in spite of conflicting interests (Linderoth, Björk & Olsson, 2014) as well as the
importance of coordinated eye contact, when co-op games are played through physical co-presence

23
(Mauer, Lanke & Tscheligi, 2018). Therefore, it is important to choose co-op games, which fit with
the demands of specific educational contexts – e.g. in terms of giving students the necessary time
for mastering the basics of the game, while ensuring that the gameplay is sufficiently engaging and
challenging.
For the majority of the students in the study, we are not able to confirm the two hypotheses on
social participation (H1) and student engagement as well as performance in Danish and
Mathematics (H2). However, it is encouraging to note that feelings of amotivation and external
regulation decreased. Drops in lack of motivation are interesting, because they indicate positive
changes that are not necessarily measured by the four other scales. Simply put, the reduction in lack
of motivation suggests that a student loses the sense of “just rolling with it”, instead seeing personal
meaning or clear extrinsic reasons to engage. However, the c-PLOC data do not represent one clear
finding. Some of the “positive” motivations, i.e. intrinsic motivation and identified regulation, were
reduced after the intervention. These drops are less explicit for the at-risk students, which suggest
that the overall effect was better for them than their peers.
The decline in several variables, which we would expect to rise, may result from a simple training
effect as students got used to answering the surveys, or reversion toward the mean, which are both
common problem with repeated survey questions (Ellis, 1999; Cohen, Cohen, West & West, 2003).
It is, however, also likely that some students compared the experience of “regular” Mathematics
and Danish classes, to the more interesting game experience, when they answered the post-test
questions. Conversely, some students may have felt at home and been intrinsically motivated by
their established social roles and personal identification with curricular aims. Having that balance
between inner and outer motifs disrupted may have resulted in negative experiences with the game
intervention, or at least some ambiguity in relation to how the school subjects were perceived to
relate to the game activities, as seen in the quotes in the previous section.
However, these methodological challenges do not fully explain the lack of impact of the
intervention on most students’ positive motivations. As the three interview examples show, the at-
risk students valued the social participation but responded differently to the game-related
assignments and the curricular reframing of their game experiences. Due to this variation it is
difficult to point out direct links between “game effects” and academic performance, as the students
had quite different experiences of the game-related assignments. Thus, some students experienced
that the game-related assignments were irrelevant, whereas other students experienced them as
more meaningful learning activities than everyday school assignments. This finding is further
elaborated in a qualitative study of the students’ experience of the Danish assignments for the
Torchlight II intervention, which shows that the aim of supporting meaningful curricular reframing
of game experiences to a large degree depend on the actual design of specific game-related
assignments such as writing guides, character analysis or fictional stories (Hanghøj, 2017). In
summary, qualitative analysis may help us understand that even though interventions with multi-
player COTS games may increase students’ abilities to collaborate and sense of well-being, such
positive effects do not automatically extend to students’ experience of subject-specific learning
activities. In this way, there is a need for future research on how students experience shifts of
interpretive framing between taking part in game activities and in curricular activities.
The qualitative findings only provide limited insights into potential links between an increase in the
at-risk students’ social participation (H1.1) and in academic improvement (H2.1). One exception is
the third empirical example with the dyslexic student, which indicates that the combination of social

24
participation in a game and meaningful orientation toward other possible readers may lead to
increased disciplinary motivation and academic improvement. This corresponds with Steinkuehler’s
(2011) findings regarding improved reading capabilities for male World of Warcraft gamers, when
reading game-related texts. In this way, it is important to conduct further research on the links
between commercial games and literacy practices in relation to at-risk students (Beavis, Dezuanni
& O’Mara, 2017).
The analysis of the quantitative data shows that the at-risk students benefitted from the intervention.
However, the qualitative data shows how social participation and academic performance do not
necessarily coincide. At the same time, the quantitative data showed that the intervention did not
significantly increase the regular students’ motivation or social participation. This raises a key
dilemma: Is it meaningful to use commercial games in education to support at-risk students’ social
participation and reduce external regulation, if the game interventions are unable to further motivate
regular students? Based on the overall positive findings documented through interviews and
measurements, it is clear that the co-op video game Torchlight II can be used to improve social
well-being and contribute to a more playful and perhaps even more caring classroom culture (Adler,
2002), which is of particular value for at-risk students. At the same time, it is not unlikely that some
students may have enjoyed the game intervention, but felt insufficiently engaged in the game-
related assignments.
Following this dilemma, it is important to discuss the meaning of social participation when using
commercial co-op video games within educational contexts. Social inclusion is always tied up with
choice and possibilities for exclusion (Tetler & Langager, 2009). By using a co-op game such as
Torchlight II, the students in this study experienced a sense of mutual engagement and a playful
feeling of being “forced” to side on teams together with their classmates against a common enemy,
i.e. the computer. This created a strong shared sense of belonging and common endeavour within
the “hybrid site” of the game, which involved shared virtual and physical co-presence in the
classroom (Bailey & Burnett, 2014). The intense collaborative gameplay may also explain why the
study did not document significant examples of social exclusion, which might have been expected
due to the students’ different game preferences and player habits outside school. In this way, we
will argue that the use of co-op games with mixed teams (at-risk and peers, boys and girls) may be
able to overcome some of the challenges that arise when using games, which are strongly preferred
or disliked by specific student groups.
Finally, it is important to mention the role of the teachers in the project as a key factor in being able
to facilitate the game sessions and help the students to identify with the subject-specific content and
aims of the game-related assignments. None of the teachers in the project had any previous
knowledge of Torchlight II before being part of the intervention and taking part in the teacher
training. Based on the teachers’ feedback, it was clear that they did not have sufficient game
literacy or knowledge of the game in order to frame and scaffold the game-related assignments for
the students. This may also explain why some students were unable to create visible links between
their enjoyments of the game sessions with their experience of the school subjects. In this way,
there is a need for more detailed research on how teachers make links between games and
curriculum and students’ game experience (De Grove, Bourgonjon & Van Looy, 2012; Sandford,
Ulicsak, Facer & Rudd, 2006)).
In conclusion, the present study showed that interventions with commercial video games and
analogue gamification can be used to include at-risk students, but that this requires careful

25
reframing of game experiences into subject-specific aims and assignments. Further, complex
interventions may strike different chords with different students, depending on existing
identifications with subjects and social dynamics, leading to complex shifts in motivational
dynamics. While intrinsic motivation, the typical measure of “fun”, did not turn out to be central,
the intervention studied here appears to have broken down feelings of external pressure, perhaps as
the first stop of more meaningful identification between individual motivations, social groups and
school subjects. Future studies should pay attention to these complexities when discussing the
motivational effects of games and game-like activities, and strive to dissociate the effects of
different intervention elements – ranging from teacher’s work to reframe game experiences in terms
of school subjects, over the achievement of new social positions afforded by transformation of
classroom dynamics, to the pure unbridled fun and freedom of playing co-op games together in
class. Particular attention should be paid to the choice of specific co-op games when trying to
balance social participation with learning opportunities in relation to selected game problems.
Turning a classroom into a LAN gaming set up is definitely not without practical difficulties.
However, based on the positive findings in this study, we strongly recommend the use of co-op
games as pedagogical tools for creating playful and socially inclusive classrooms, which can level
the playing field and position at-risk students as active participants in overcoming shared challenges
with their classmates.
Acknowledgements
The authors would like to thank all the teachers, students and research assistants involved in the
research project for making this study possible. The School at Play research project has been
supported by a grant from the Egmont Foundation. Also thanks to Runic Games for offering free
Torchlight II licenses to be used in the project. The authors confirm that there are no known
conflicts of interest associated with this publication.

26
References
Alder, N. (2002). Interpretations of the meaning of care: Creating caring relationships in urban middle school
classrooms. Urban education, 37(2), 241-266.
Bailey, C., & Burnett, C. (2014). Conceptualising collaboration in hybrid sites: playing Minecraft together
and apart in a primary classroom. In C. Burnett, J. Davies, G. Merchant & J. Rowsell (Ed.) New
literacies around the globe (pp. 70-91). London: Routledge.
Beavis, C., Dezuanni, M., & O'Mara, J. (Eds.). (2017). Literacy, Learning, and Digital Games: Serious Play.
London: Routledge.
Benjamini, Y., & Hochberg, Y. (1995). Controlling the False Discovery Rate: A Practical and Powerful
Approach to Multiple Testing. Journal of the Royal Statistical Society. Series B (Methodological),
57(1), 289–300. https://doi.org/10.2307/2346101
Bessiere, K., Pressman, S., Kiesler, S., & Kraut, R. (2010). Effects of Internet use on health and depression:
A longitudinal study. Journal of Medical Internet Research, 12(1), 6–7.
https://doi.org/10.2196/jmir.1149
Boaler, J. & Greeno, J. G. (2000). Identity, Agency and Knowing in Mathematical Worlds. In J. Boaler (Ed.)
Multiple perspectives on mathematics teaching and learning (pp. 171-200). Westport, CT: Ablex
Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative research in psychology,
3(2), 77-101.
Callaghan, N. (2016). Investigating the role of Minecraft in educational learning environments. Educational
Media International, 53(4), 244-260.
Clark, D. B., Tanner-Smith, E. E., & Killingsworth, S. S. (2016). Digital games, design, and learning: A
systematic review and meta-analysis. Review of educational research, 86(1), 79-122.
Cohen, J., Cohen, P., West, S. G., & West, L. S. (2003). Applied multiple regression/correlation analysis for
the behavioral sciences (3rd ed.). Mahwah, NJ: Erlbaum.
Connolly, T. M., Boyle, E. A., MacArthur, E., Hainey, T., & Boyle, J. M. (2012). A systematic literature
review of empirical evidence on computer games and serious games. Computers & Education, 59(2),
661-686.
Creswell, J., & Plano Clark, V. L. (2011). Designing and Conducting Mixed Methods Research (2nd ed.).
London: SAGE Publications.
De Grove, F., Bourgonjon, J., & Van Looy, J. (2012). Digital games in the classroom? A contextual
approach to teachers’ adoption intention of digital games in formal education. Computers in Human
Behavior, 28(6), 2023-2033.
Domahidi, E., Festl, R., & Quandt, T. (2014). To dwell among gamers: Investigating the relationship
between social online game use and gaming-related friendships. Computers in Human Behavior, 35,
107–115. https://doi.org/10.1016/j.chb.2014.02.023
Dyssegaard, C. B. & Larsen, M. (2013). Evidence on Inclusion. Danish Clearinghouse for Educational
Research. Aarhus University. Link:
http://edu.au.dk/fileadmin/edu/Udgivelser/Clearinghouse/Evidence_on_Inclusion.pdf
Ellis, M. V. (1999). Repeated Measures Designs. The Counseling Psychologist, 27(4), 552–578.
https://doi.org/10.1177/0011000099274004
Gibson, L. (2013). Differentiated Instruction and Students with Learning Disabilities. In J. P. Bakken, F. E.
Obiakor & A. F. Rotatori (Eds.) Learning Disabilities: Identification, Assessment, and Instruction of
Students with LD. Advances in Special Education, Vol. 24. Emerald Group Publishing Limited, pp.161-
183.
Goffman, E. (1974). Frame analysis: An essay on the organization of experience. Harvard University Press.
Greene, J. C. (2007). Mixed Methods in Social Inquiry. San Francisco: Jossey-Bass / Wiley.
Hanghøj, T. (2015). The School at Play: Repositioning students through the educational use of digital games
and game gynamics. In Proceedings of the 9th European Conference on Game-Based Learning.
ECGBL 2015 (pp. 227-236). Academic Conferences and Publishing International.
Hanghøj, T. (2017). Read This or Die!: Including at-risk students through game-related literacy practices. In
Proceedings of the 11th European Conference on Game-Based Learning. ECGBL 2017 (pp. 219-226).
Academic Conferences and Publishing International.

27
Kafai, Y. B., & Resnick, M. (Eds.). (1996). Constructionism in practice: Designing, thinking, and learning
in a digital world. London: Routledge.
Kaye, L. K., Kowert, R., & Quinn, S. (2017). The role of social identity and online social capital on
psychosocial outcomes in MMO players. Computers in Human Behavior, 74, 215–223.
https://doi.org/10.1016/j.chb.2017.04.030
Kirby, M. (2017). Implicit assumptions in special education policy: Promoting full inclusion for students
with learning disabilities. Child & Youth Care Forum, 46(2): 175-191.
Kowert, R., & Oldmeadow, J. A. (2014). Playing for social comfort: Online video game play as a social
accommodator for the insecurely attached. Computers in Human Behavior, 53, 556–566.
https://doi.org/10.1016/j.chb.2014.05.004
Lacasa, P., Méndez, L., & Martínez, R. (2008). Bringing commercial games into the classroom. Computers
and Composition, 25(3), 341-358.
Lieberoth, A. (2015). Shallow gamification: Testing psychological effects of framing an activity as a
game. Games and Culture, 10(3), 229-248.
Lieberoth, A., & Hanghøj, T. (2017). Developing professional “game teacher” repertoires : Describing
participants and measuring effects in a Danish college course on game based learning. In Proceedings
of the 11th European Conference on Game-Based Learning. ECGBL 2017 (pp. 377–386). Academic
Conferences and Publishing International.
Linderoth, J., Björk, S., & Olsson, C. (2014). Should I stay or should I go? A Study of Pickup Groups in Left
4 Dead 2. Transactions of the Digital Games Research Association, 1(2), 117-145.
Maurer, B., Lankes, M., & Tscheligi, M. (2018). Where the Eyes Meet: Lessons Learned from Shared Gaze-
based Interactions in Cooperative and Competitive Online Games. Entertainment Computing, 27, 47-
59.
McGrath, K. F., & Van Bergen, P. (2015). Who, when, why and to what end? Students at risk of negative
student–teacher relationships and their outcomes. Educational Research Review, 14, 1-17.
Maxwell, J., Chmiel, M., & Rogers, S. E. (2015). Designing integration in Multimethod and Mixed Methods
Research. In S. Hesse-Bieber & R. Burke Johnson (Eds.), The Oxford Handbook of Multimethod and
Mixed Methods Research Inquiry (pp. 223–239). Oxford: Oxford University Press.
Ng, J. Y. Y., Ntoumanis, N., Thogersen-Ntoumani, C., Deci, E. L., Ryan, R. M., Duda, J. L., & Williams, G.
C. (2012). Self-Determination Theory Applied to Health Contexts: A Meta-Analysis. Perspectives on
Psychological Science, 7(4), 325–340. https://doi.org/10.1177/1745691612447309
Pannekoek, L., Piek, J. P., & Hagger, M. S. (2014). The Children’s Perceived Locus of Causality Scale for
Physical Education. Journal of Teaching in Physical Education, 33(2), 162–185.
https://doi.org/10.1123/jtpe.2013-0095
Persico, D., Bailey, B., Buijtenweg, T., Dagnino, F., Earp, J., Haggis, M., Manganello, F., Passarelli, M.,
Perrotta, C. & Pozzi, F. (2017). Gaming Horizons. Systematic Review and Methodological Framework.
Retrieved from http://www.gaminghorizons.eu/wp-content/uploads/sites/18/2017/05/D2.1-State-of-the-
Art-Literature-review.pdf
Rigby, C. S., & Ryan, R. M. (2011). Glued to games - How Video Games Draw Us In and Hold Us
Spellbound. London: Praeger.
Ryan, R. M., Rigby, C. S., & Przybylski, A. (2006). The motivational pull of video games: A self-
determination theory approach. Motivation and emotion, 30(4), 344-360.
Ryan, R., & Deci, E. (2000). Self-determination theory and the facilitation of intrinsic motivation, social
development, and well-being. The American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-
066X.55.1.68
Sanchez, E., Young, S., & Jouneau-Sion, C. (2017). Classcraft: from gamification to ludicization of
classroom management. Education and Information Technologies, 22(2), 497-513.
Squire, K. D. (2004). Replaying history: Learning world history through playing Civilization III.
Bloomington, IN: Indiana University.
Steinkuehler, C. (2011). The mismeasure of boys: Reading and online videogames. WCER Working Paper
No. 2011-3. University of Wisconsin.
Stewart, J., Bleumers, L., Van Looy, J., Mariën, I., All, A., Schurmans, D., ... & Misuraca, G. (2013). The
potential of digital games for empowerment and social inclusion of groups at risk of social and

28
economic exclusion: evidence and opportunity for policy. Joint Research Centre, European
Commission.
Sandford, R., Ulicsak, M., Facer, K., & Rudd, T. (2006). Teaching with games: using commercial off-the-
shelf computer games in formal education. Research report. NESTA Futurelab.
https://www.nfer.ac.uk/publications/FUTL49
Sundberg, M. (2018). Online gaming, loneliness and friendships among adolescents and adults with
ASD. Computers in Human Behavior, 79, 105-110.
Takeuchi, L. M. & Vaala, S. (2014). Level up learning: A national survey on teaching with digital games.
New York: The Joan Gantz Cooney Center at Sesame Workshop.
Tetler, S., & Langager, S. (Eds.) (2009). Specialpædagogik i skolen: en grundbog. København: Gyldendal.
Trepte, S., Reinecke, L., & Juechems, K. (2012). The social side of gaming: How playing online computer
games creates online and offline social support. Computers in Human Behavior, 28(3), 832–839.
https://doi.org/10.1016/j.chb.2011.12.003
Waern, A. (2012). Framing games. Proceedings of 2012 International DiGRA Nordic Conference, June 6–8,
2012, Tampere, Finland.
Weinstein, N., Ryan, R. & Deci, E. L. (2013). Motivation, Meaning, and Wellness: A Self-Determination
Perspective on the Creation and Internalization of Personal Meanings and Life Goals. In Wong, P.
(ed.). The Human Quest for Meaning: Theories, Research, and Applications. New York: Routledge, pp.
81-106.
Westin, T., & Lange, G. (2012). Gamers versus the Index. Creative Education, 3(8B), 25-30.
Wouters, P., van Nimwegen, C., van Oostendorp, H., & van der Spek, E. D. (2013). A Meta-Analysis of the
Cognitive and Motivational Effects of Serious Games. Journal of Educational Psychology, doi:
10.1037/a0031311
Young, M. F., Slota, S., Cutter, A. B., Jalette, G., Mullin, G., Lai, B., ... & Yukhymenko, M. (2012). Our
princess is in another castle: A review of trends in serious gaming for education. Review of educational
research, 82(1), 61-89.