Toward a Meta-design Method for Learning Games
Marne Bertrand
1 a
, Muratet Mathieu
2,3 b
and Sehaba Karim
4 c
1
ICAR UMR 5191, Universit
´
e Lumi
`
ere Lyon 2, France
2
Sorbonne Universit
´
e, CNRS, LIP6, F-75005 Paris, France
3
INS HEA, 92150 Suresnes, France
4
Universit
´
e de Lyon, CNRS. Universit
´
e Lyon 2, LIRIS, UMR5205, F-69676, France
Keywords:
Serious Games, Meta-design, Computational Thinking, Authoring Tools.
Abstract:
To support the appropriation of serious learning games by teachers, we are studying design methods that
focus on both design and use phases: meta-design methods. Our objective is to propose models and tools
for designing levels and scenarios for Blockly Maze and to provide the teacher with monitoring indicators
allowing him/her to appropriate and adapt the game according to the observed uses. In this article, we detail
the first contributions of this study based on qualitative data: analysis of the game Blockly Maze and interviews
with two teachers. First results draw three main needs: level design tool, scenario design tool and monitoring
tool. Beyond these needs we introduce underlying models of each of them and future works.
1 INTRODUCTION
Learning games offer significant benefits over tradi-
tional teaching tools (Freitas, 2006; Prensky, 2004).
Indeed, many authors consider learning games as
promising, especially to increase learner engagement
(Bouvier et al., 2014) and motivation (Garris et al.,
2002; Malone and Lepper, 2005), others are more
interested in modeling and evaluating player experi-
ence (Kiili, 2005; Sweetser and Wyeth, 2005) or in
using learning games to promote a more construc-
tivist learning (Bogost, 2007; Marne, 2019; Ryan
et al., 2012). However, their adoption, especially by
teachers, remains scarce (Li, 2018; Sardone, 2018).
To foster the adoption of learning games by teachers,
we work on the hypothesis that a participatory design
method such as meta-design might be suitable.
Meta-design is an advanced participatory design
method, in which the design process is centred on
the end users (“owners of problems”). However, end
users must continue to have the means to design dur-
ing the artefact use phase (Fischer and Herrmann,
2011). This is made possible, among other things,
by the underdesign that Fischer et al. (Fischer et al.,
2004) define as:
a
https://orcid.org/0000-0002-4953-9360
b
https://orcid.org/0000-0001-6101-5132
c
https://orcid.org/0000-0002-6541-1877
“[. . . ] underdesign aims to provide social and
technical instruments for the owners of prob-
lems to create the solutions [of their problems]
themselves at use time.
Our work questions this approach in the context
of the use of serious games by teachers as owners of
problems.
The work presented in this paper reports the be-
ginning of a project on this topic. We focused our
study on the teaching of computational thinking, and
the implementation of a meta-design approach for
learning games. This project aims to exploit an ex-
isting learning game and to study to what extent it
is possible for teachers to appropriate it (underde-
sign). We investigated how allowing them to mod-
ify its content according to their pedagogical needs
and observed or intended uses. We chose Blockly
Maze
1
(BM), among other things, because it is well-
tested and free software. Indeed, inspired by Scratch
(Resnick et al., 2009), BM has been reused in many
derivative works. For instance, the Hour of Code
2
,
Algorea
3
, and more than 400 derivatives on the forge
Github
4
. The fact that the source code is available,
1
Blockly Maze is a learning game developed by Google:
https://blockly.games/maze consulted on 2021/02/03
2
https://code.org/ consulted on 2021/02/03
3
https://algorea.org/ consulted on 2021/02/03
4
https://github.com/google/blockly-
370
Bertrand, M., Mathieu, M. and Karim, S.
Toward a Meta-design Method for Learning Games.
DOI: 10.5220/0010530203700376
In Proceedings of the 13th International Conference on Computer Supported Education (CSEDU 2021) - Volume 2, pages 370-376
ISBN: 978-989-758-502-9
Copyright
c
2021 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
freely reusable, and widely tested has particularly in-
terested us to design and evaluate methods and tools
allowing meta-design and underdesign for this learn-
ing game.
The limits of the original Google version of
Blockly Maze are the same limits as those found in
many other learning games: (1) limited number of
levels, Blockly Maze offers 10 levels covering a few
programming skills; (2) non-modifiable scenario,
the course of these 10 levels is a linear sequence that
the teacher cannot modify; (3) the players’ activity
is not recorded, hence the teacher cannot retrieve any
trace of the learners’ performance. For these reasons,
it is difficult for the teacher to understand the practices
and difficulties of the learners and thus to appropriate
and adapt the learning game to the pedagogical con-
text.
Our goal is to provide tools and methods enabling
teachers, more or less comfortable with teaching com-
puter science, to use Blockly Maze, to build their
own sequences of levels corresponding to their needs,
to build levels that complement those available, and
to follow the learners’ activity to better understand
their use and thus adapt learning. Therefore, our ulti-
mate goal is to build a meta-design context based on
a generic model for the design and development of
levels.
Our first step is the deconstruction of Blockly
Maze, both on a conceptual perspective and on a soft-
ware one. Thus, we intend to develop explicit models
of the levels and their educational content. This paper
describes this first part of our work.
In the first part of the paper, we present Blockly
Maze and the scientific issues we are studying, as they
emerge from the different obstacles arising from the
literature and the practice of teaching computational
thinking. In the second part, we present our method-
ological context and our choices. Finally, before con-
cluding, we present our first results in the third part.
2 INTRODUCING BLOCKLY
MAZE AND OUR SCIENTIFIC
PROBLEMS
In order to work on meta-design, we chose an open
source learning game (source code available, freely
reusable, and modifiable). We chose Blockly Maze
because of the popularity of the Blockly engine,
but also because of the wide adoption of Blockly
Maze (BM) by many organizations such as the Hour
of Code or Algorea. The concept of this learning
games/network/members consulted on 2021/02/03
Figure 1: Screenshot of Blockly Maze showing block pro-
gramming to guide the avatar toward the exit.
game developed by Google is to teach programming
through the guidance of an automatic avatar in a maze
(see figure 1). The guidance is done by program-
ming with blocks of instructions, inspired by Scratch
(Resnick et al., 2009).
However, the original BM from Google only
offers 10 levels. They train to the use in-
structions sequences, conditions (if/else), loops
(repeat/until), including nested ones, and basic al-
gorithmic (left-hand method). This is rather limited
considering the various needs of teaching computa-
tional thinking (Brunet et al., 2020; Vandevelde and
Fluckiger, 2020). Thus, many derived works (forks)
and learning games similar to BM have emerged, of-
fering numerous other levels. We mentioned the Hour
of Code, which offers a large selection of additional
levels dealing with multiple aspects of computational
thinking distributed in several “courses”. The “Hour
of Code” has a wide choice of scenarios (combining
other types of learning games than BM), which means
that, for a teacher, an informed choice of the resource
or course to use requires a global vision of the avail-
able options, and therefore also requires a significant
amount of time to grip the whole platform. We also
mentioned Algor
´
ea (L
´
eonard, 2020), it offers learning
games that are close to BM (although, apparently on
a different code base). Like the Hour of Code, these
learning games are provided with a variety of other
educational resources, structured to prepare a French
programming competition: Algor
´
ea. The learning
games provided are puzzles that can be solved with
blocks like in Blockly (in addition to the Python code
and Scratch blocks). Nevertheless, only a few of them
use a maze as puzzle. Those maze games only focus
on a few aspects of computational thinking: instruc-
tion sequences, function calls, simple or nested loops,
conditional instructions. The learning games based
on block programming which address more complex
notions are not based on maze puzzles.
In any case, it seems difficult for a teacher, who
is familiar with these three resources, to build a cur-
riculum for his or her learners by picking out what is
Toward a Meta-design Method for Learning Games
371
Figure 2: Screenshot of a sample Blockly Maze scenario provided by APPLiq.
needed for them. Moreover, the resources are not de-
signed for this kind of flexibility, but rather to be used
in the form of indivisible sequences of several levels.
Given these limitations, our problem is therefore
to provide teachers with the proper tools and methods
to be able to appropriate BM. On the one hand, by
giving them the opportunity to understand how BM
works and how it is structured (instrumentation), and,
on the other hand, by enabling them to adapt, mod-
ify, remix BM and its levels (instrumentalisation and
meta-design) (Rabardel, 2003).
To foster appropriation of BM by teachers through
mastery and adaptation, we relied mainly on the meta-
study of Dermeval et al. (Dermeval et al., 2018) to
explore available authoring tools. Our goal was to
find out which ones could be used to adapt the lev-
els or the succession of levels of BM. However, BM
is not intended to work with such tools. In our inves-
tigations, beyond this meta-study, we only found the
authoring tool APPLiq which can be used to adapt a
BM scenario, as it has already been tested with it by
its author (Marne and Labat, 2014; Marne, 2014).
APPLiq enables teachers/users to prepare and
provide learners/players with a succession of levels
(called activities) in a learning game, taking into ac-
count the pre-requisite and worked on pedagogical
objectives at each of these stages. This succession
of levels is not necessarily linear (see figure 2), and
may therefore depend on the actions and performance
of the learner/player (objectives worked on or not
worked on during the activities).
Nevertheless, APPLiq has several limitations. In
the context of our study, the most important limitation
is that this authoring tool only allows the adaptation
of the levels order. Therefore, while it enables teach-
ers/users to change the sequence and order of the lev-
els, they cannot change the BM levels themselves. In
APPLiq, one can add new levels in the description of
the model of a specific learning game. However, it
is the concern of the user who makes this modifica-
tion to make the necessary adaptations to the learning
game itself. Indeed, in MoPPLiq (Marne and Labat,
2014), APPLiq’s underlying scenario model, the ac-
tivities (levels) are modelled as black boxes.
In addition, the author of APPLiq provides a mod-
ified version of BM working with his authoring tool.
But the latter is derived from a version of BM released
by Google in 2012. Since then, BM has changed a lot,
and these very profound adaptations would have to be
made again.
In our particular situation, choosing APPLiq as a
tool enabling teachers to adapt BM implies, on the
one hand, providing a major update of BM and, on
the other hand, providing another authoring tool that
allows adaptation of levels.
For this work of learning games modelling and
development of authoring tools for teachers, our
approach is a design-based collaborative research
(Sanchez et al., 2017). We describe the three main
axes of this research in the next section.
3 APPROACH AND
METHODOLOGY
Our goal is to provide tools and methods inspired
by meta-design that allow teachers to master Blocky
Maze in order to design, use and adapt levels and sce-
narios that meet their pedagogical needs in the field
of computer science teaching.
To achieve this goal, our approach is to involve
teachers in order to discuss and conceptualize com-
putational thinking for pedagogical purposes in the
form of a concept map (see figure 3). To co-design
this map
5
, we chose to conduct qualitative interviews
with two teachers instead of quantitative survey based
on questionnaire. The first teacher is an experienced
mathematics teacher who teaches hospitalized sec-
ondary school children, and the other is the director
of children’s education at the same hospital.
The co-design of this map was, above all, the
main discursive tool we used to address many of the
main pedagogical issues of the teachers we worked
with. The map is therefore not intended to model
5
See the complete concept map (in French): https:
//mycore.core-cloud.net/index.php/s/NsntsseDxGUILM4/
download.
CSEDU 2021 - 13th International Conference on Computer Supported Education
372
Figure 3: Overview of our approach.
computational thinking in an exhaustive or consen-
sual way. Rather, it allows us to understand the ap-
proaches taken by the teachers in constructing lev-
els/scenarios and to identify possible difficulties they
might be experiencing. Thus, we were less interested
in the actual map itself than in discussing its content.
The summary of the interviews carried out with
each teacher regarding the conceptual map allowed us
to pinpoint:
The concepts taught and the teaching methodol-
ogy adopted by them.
The relevance of storytelling and games to foster
learning.
The need to adapt the BM levels to, among other
things, adjust statements’ wording, blocks used,
dialogues, and add features (variables, timers,
etc.).
The need for pedagogical monitoring indicators
enabling teachers to measure the progress of their
learners and identify their appropriation of the
concepts studied.
The need for meta-design indicators that make
it easier for teachers to adapt and update lev-
els/scenarios according to the uses observed.
The first study resulted in specifying three main
directions that bring together the main features of a
set of authoring tools dedicated to meta-design: (1)
Level design tool: this tool will allow the teacher to
implement new levels or to reuse existing levels fol-
lowing a simple and intuitive model and methodol-
ogy. Therefore, for each level, the aims are to be able
to (re)define the mazes themselves, the list of blocks
made available to the learner, the dialogues and the
conditions for triggering them, as well as other con-
straints: number of blocks available, time limits for
resolution, blocks already set, etc.; (2) Scenario de-
sign tool: this tool should enable the teacher to de-
fine the sequence of the levels and the conditions for
triggering them. It should allow the teacher to de-
fine a succession of specific levels for each learner
according to his or her performance; (3) Monitor-
ing tools: this tool should allow the teacher to col-
lect data related to interactions between the learner
and the levels/scenarios and to calculate indicators on
learning and the quality of teaching resources. These
are high-level indicators such as success rates, levels
completed or not, completion times, or more specifi-
cally the number of blocks used, the number of help
messages displayed, etc.
In the next section, we present our work and its
results in each of these three directions.
4 FIRST WORK AND RESULTS
Following this meta-design approach, we have con-
ceptually deconstructed BM in order to develop three
models presented in the subsections.
4.1 Level Model
Interviews with teachers about the design of the con-
cept map showed us that the BM levels, as they cur-
rently exist, were not sufficient enough to meet the
needs (theirs, those of their colleagues) for teaching
computational thinking. We worked with them to dis-
cuss possible improvements and to establish a model
of BM levels, both on a conceptual standpoint and
within the source code itself.
Modelling BM levels was done in two steps: (1)
isolation of all the components of the BM source
code related to the description of levels, within a very
monolithic program (a single file) where these com-
ponents are spread over many functions and objects;
(2) Conceptualization of the level source code in the
form of a UML model (see figure 4).
The main features of this BM level model are, on
the one hand, the maze, the starting conditions (ori-
Toward a Meta-design Method for Learning Games
373
Figure 4: UML model for Blockly Maze levels.
entation, number of blocks available, etc.) which are
modelled as simple variables. On the other hand, the
message system (helpMessages) for players based on
their actions. This system is much more complex and
needs to be modelled with both a conditions system
and an event (trigger) system.
To contribute to verify this model, we have suc-
cessfully described again all the ten BM levels by in-
stantiating them in JSON files. Furthermore, we have
significantly reworked BM’s source code so it is now
able to read these JSON files. In addition, we are
preparing a pull request to be submitted to BM devel-
opers (Google), so that all potential users can easily
make new levels written with JSON.
Our first experiments with the teachers show that
we can work with them to write new levels with the
JSON model. The weakness of this approach is the
design of the help messages, which is still very com-
plex. Therefore, the next step in our work is to build
an authoring tool that generates these JSON files, and
helps to set up help messages for players.
The tiniest possible scale of the changes that
teachers might consider is the levels. We also intend
to allow the teachers to modify learning games on a
bigger scale: the sequence of levels (scenario).
4.2 Scenario Model
Our interviews confirm the teachers’ need to be able
to organize a progression through the different levels
according to their specific teaching context.
For this purpose, APPLiq seems relevant (Marne
and Labat, 2014). It is based on an XML model of
the scenario (MoPPLiq) in which the levels (activi-
ties) are considered as black boxes, meant to let the
players work on some specific sets of pedagogical and
Figure 5: MoPPLiq Entity-Relationship Diagram (Marne
and Labat, 2014).
playful objectives. Each activity can have several out-
put states depending on the set of objectives effec-
tively worked on. Each activity can also have several
input states restricting the possibilities of connecting
the activity to the rest of the scenario according to a
set of prerequisite objectives. The scenario is there-
fore described as a sequence of all the desired output
state/input state links (see figures 2 and 5).
When refactoring BM’s source code to be able to
load levels described with JSON, we also used the
API provided by APPliq’s author (Marne, 2014) to
embed the ability to load MoPPLiq models.
However, we are facing several issues: (1) The
new levels designed with the JSON model must also
be described (modelled) within APPLiq. Therefore,
we think that we should allow our level authoring
tool to communicate with APPLiq to prevent teach-
ers from having to describe their levels twice and ac-
cording to different approaches (BM JSON, on the
one hand, and MoPPLiq on the other hand); (2) AP-
PLiq enables teachers to build a scenario that adapts
to the learner’s performance according to the worked
on pedagogical objectives described with the output
states. However, in our BM level model, the complex
assessment of what is or is not worked on is not im-
plemented. We consider adapting our model and its
conditions and trigger system (currently related to the
help messages) to measure whether the objectives are
being worked on; (3) APPLiq provides a static sce-
nario. Therefore, adaptation to the performances of
learners is pre-designed. We are studying the possibil-
ity of modifying APPLiq to benefit from the results of
a monitoring system such as the one we have started
to model.
CSEDU 2021 - 13th International Conference on Computer Supported Education
374
4.3 Monitoring Model
Based on discussions with the teachers, we also iden-
tified the need for a tracking and monitoring sys-
tem. Its purpose is to identify whether the levels and
scenarios designed meet teachers’ needs after one or
more play sessions by the learners.
We decided to use a version of xAPI adapted to se-
rious games (Serrano-Laguna et al., 2017) as a basis
for our tracking system. xAPI allows defining indi-
cators (statements specified with triplets actor verb
object) which are stored in a Learning Record Store
(LRS).
We identified two different levels of monitor-
ing: monitoring within the level, and monitoring be-
tween the levels. With the teachers help, we de-
signed statements they felt relevant: the time taken
to complete the levels (Actor initialized level,
Actor exited level); the completion of the lev-
els (Actor completed level, Actor unlocked
level); the use of external support (Actor
unfocused game Windows, Actor focused game
Windows); the number of blocks used and the num-
ber of tests of their program ran by the learner (Actor
interacted blocks, Actor executed program).
We are currently implementing this xAPI monitor-
ing system in BM’s source code.
To conclude, thanks to the work carried out with
the teachers who accompanied us, we were able to
model three significant aspects of BM: the levels, the
scenario and the monitoring.
5 CONCLUSIONS AND FUTURE
WORK
To foster adoption and appropriation of learning
games we planned to rely on the meta-design ap-
proach (Fischer and Herrmann, 2011), and we fo-
cused on learning games for computational thinking
such as Blockly Maze (BM). Meta-design goal is to
enable teachers to act as designers of learning games
both at the initial design stage and in the use stage.
In this paper, we have presented a description of
our work in progress. The work is focused on the
deconstruction of BM, both from a conceptual and a
software standpoint. Our objective is to obtain, with
the help of participating teachers, explicit models of
the levels and the educational contents. This work
was required to provide an underdesigned (Fischer
et al., 2004) version of BM. This means a usable,
“turnkey” BM, but above all a learning game that of-
fers the main features required to provide an effec-
tive and facilitated instrumental genesis for teachers
(Rabardel, 2003).
In collaboration with the teachers involved as con-
tributors, we experimented a research method based
on the design of a concept map related to a teach-
ing field (computational thinking). The map’s design
brought to light many issues related to their teaching
as well as the needs around BM.
Based on these needs, and always in collaboration
with the teachers, we developed three models describ-
ing BM. Then, we implemented them in a new ver-
sion of BM, which we will submit to its authors (pull
request). The Level Model, implemented in JSON,
proved to be able to describe all the current BM lev-
els, as well as to be used to design new ones with
teachers. For the Scenario Model, we choose MoP-
PLiq and APPLiq and plan to evolve them to answer
our research questions (Marne and Labat, 2014). We
have started work on a Monitoring Model based on
xAPI.
The research reported in this paper was severely
limited by the COVID-19 pandemic, which prevented
us from working with as many teachers as we would
have liked. Therefore, our current and future work
consists of collaborating with more teachers in (1) de-
veloping levels to enrich and test our model of BM
levels; (2) improving and experimenting with the sce-
nario design and adaptation; and (3) implementing the
monitoring of learners to allow teachers to validate
their BM adaptations.
ACKNOWLEDGEMENTS
The authors would like to acknowledge The Univer-
sity Lumi
`
ere Lyon 2 for the APPI 2020 Grant.
REFERENCES
Bogost, I. (2007). Persuasive Games: The Expressive
Power of Videogames. MIT Press, Cambridge, UK.
Bouvier, P., Sehaba, K., and Elise, L. (2014). A trace-based
approach to identifying users’ engagement and quali-
fying their engaged-behaviours in interactive systems.
Application to a social game. User Modeling and
User-Adapted Interaction (UMUAI’14), 45:413–451.
Brunet, O., Yessad, A., Muratet, M., and Carron, T. (2020).
Vers un mod
`
ele de sc
´
enarisation pour l’enseignement
de la pens
´
ee informatique
`
a l’
´
ecole primaire. In Di-
dapro 8 – DidaSTIC, Lille, France.
Dermeval, D., Paiva, R., Bittencourt, I. I., Vassileva, J., and
Borges, D. (2018). Authoring tools for designing in-
telligent tutoring systems: A systematic review of the
literature. International Journal of Artificial Intelli-
gence in Education, 28(3):336–384.
Toward a Meta-design Method for Learning Games
375
Fischer, G., Giaccardi, E., Ye, Y., Sutcliffe, A. G., and
Mehandjiev, N. (2004). Meta-design: A manifesto for
end-user development. Communications of the ACM,
47(9):33–37.
Fischer, G. and Herrmann, T. (2011). Socio-Technical Sys-
tems: A Meta-Design Perspective. International Jour-
nal of Sociotechnology and Knowledge Development,
3(1):33.
Freitas, S. (2006). Learning in immersive worlds: A review
of game-based learning.
Garris, R., Ahlers, R., and Driskell, J. E. (2002). Games,
motivation, and learning: A research and practice
model. Simulation & gaming, 33(4):441–467.
Kiili, K. (2005). Digital game-based learning: Towards an
experiential gaming model. The Internet and Higher
Education, 8(1):13–24.
L
´
eonard, M. (2020). Score et chronom
`
etre sur l’interface:
Quels effets sur les r
´
esultats et la perception d’un con-
cours en ligne? In 8e Rencontres Jeunes Chercheurs
En EIAH, 2020, Poitiers.
Li, Q. (2018). Enactivism and teacher instructional game
building: An inquiry of theory adoption and design
consideration. Educational Technology Research and
Development, 66(6):1339–1358.
Malone, T. and Lepper, M. (2005). Making learning fun: A
taxonomy of intrinsic motivations for learning. Mak-
ing Learning Fun: A Taxonomy of Intrinsic Motiva-
tions for Learning, 3.
Marne, B. (2014). Mod
`
eles et outils pour la conception de
jeux s
´
erieux : une approche meta-design. Th
`
ese de
Doctorat en Informatique, Universit
´
e Pierre et Marie
Curie (UPMC), Paris.
Marne, B. (2019).
´
El
´
ements de r
´
eflexion autour de la con-
ception du jouet pour un jeu s
´
erieux. In Actes de
la 9
`
eme Conf
´
erence sur les Environnements Infor-
matiques pour l’Apprentissage Humain, pages 1–12,
Paris, France.
Marne, B. and Labat, J.-M. (2014). Model and Authoring
Tool to Help Teachers Adapt Serious Games to their
Educational Contexts. International Journal of Learn-
ing Technology, 9(2):161–180.
Prensky, M. (2004). Digital Game-Based Learning.
McGraw-Hill.
Rabardel, P. (2003). From artefact to instrument. Interact-
ing with Computers, 15(5):641–645.
Resnick, M., Maloney, J., Monroy-Hern
´
andez, A., Rusk,
N., Eastmond, E., Brennan, K., Millner, A., Rosen-
baum, E., Silver, J., Silverman, B., and Kafai, Y.
(2009). Scratch: Programming for All. Communi-
cations of the ACM, 52(11):60–67.
Ryan, M., Costello, B., and Stapleton, A. (2012). Deep
Learning Games through the Lens of the Toy. In
Meaningful Play 2012, pages 1–29, East Lansing,
USA.
Sanchez, E., Monod-Ansaldi, R., Vincent, C., and Safadi-
Katouzian, S. (2017). A praxeological perspective
for the design and implementation of a digital role-
play game. Education and Information technologies,
22(6):2805–2824.
Sardone, N. B. (2018). Attitudes Toward Game Adoption:
Preservice Teachers Consider Game-Based Teaching
and Learning. International Journal of Game-Based
Learning (IJGBL), 8(3):1–14.
Serrano-Laguna,
´
A., Mart
´
ınez-Ortiz, I., Haag, J., Regan,
D., Johnson, A., and Fern
´
andez-Manj
´
on, B. (2017).
Applying standards to systematize learning analytics
in serious games. Computer Standards & Interfaces,
50:116–123.
Sweetser, P. and Wyeth, P. (2005). Gameflow: A model
for evaluating player enjoyment in games. Comput.
Entertain., 3(3):3.
Vandevelde, I. and Fluckiger, C. (2020). L’informatique
prescrite
`
a l’
´
ecole primaire. Analyse de pro-
grammes, ouvrages d’enseignement et discours insti-
tutionnels. L’informatique, objets d’enseignements
enjeux
´
epist
´
emologiques, didactiques et de formation,
page 23.
CSEDU 2021 - 13th International Conference on Computer Supported Education
376