Exploring risk through pervasive gaming
By Andrés Monroy-Hernández
MIT Media Laboratory
T540: Cognition and the Art and Science of Instruction
Harvard Graduate School of Education
Professor Tina Grotzer
This is project is not for the TE Specialization but uses technology and is not part
of another class.
Overall purpose
The goal of this project is to foster the understanding of risk by creating
and participating in pervasive games. Special focus will be directed towards risk
in the context of complex causality, such as the risk in virus-like phenomena.
As our society becomes more interconnected in the physical and digital
realm, complex phenomena have become common place. From viral marketing
to the avian flu virus, the understanding of risk within this context is important for
the full participation of individuals in society.
In the spirit of the Participatory Simulations (Colella, 1998) and the
Environmental Detectives (Klopfer, 2002) this project exposes learners to the
understanding objectives through games that take in the physical world. Inspired
by Constructionist philosophy (Papert, 1980), I try to go a step further by
encouraging participants to design their own games. This paper is a description
of the following:
• A two day workshop where middle-school students create and participate
in pervasive games that explore the topic of risk assessment. The
overarching goal of the activities is to promote the transfer of knowledge
through the exploration of structurally similar phenomena.
• The technical design and proof-of-concept prototype of the system where
the pervasive games will take place.
Understanding goals
Based on the research literature (Grotzer, 2005) there are core issues
involved in the perception of risk in complex phenomena. This project tries to
tackle these issues while promoting the use of iterative design.
1. Question: How can we find the hidden (or non-obvious) causes of an
infection1? Statement: Students will understand and appreciate:
• that risk can stem from unexpected causes
• that it is important to question assumptions
2. Question: Can an infection be caused by something far away or that
happened sometime in the past? Statement: Students will understand and
appreciate:
• that causes do not have to be in temporal and spatial proximity. For
example, there are theories that explain active hurricane seasons
as caused by an increase in water temperature in the arctic.
3. Question: Can an infection have multiple unrelated causes? Statement:
Students will understand and appreciate:
• that a centralized mindset is not always useful when analyzing
complex causality. For example, most people tend to think that
traffic jams are caused by a single event such as an accident or a
construction site, but oftentimes it is caused by the small
1
The word infection is used to refer to any type of phenomenon that exhibits virus-like
properties.
decentralized relationships between cars such as differences in
speed.
4. Question: Can we tell how risky something is? Statement: Students will
understand and appreciate:
• that when confronted with risk it is helpful to model it in a
quantitative and/or qualitative way, instead of viewing it as a
dichotomy (i.e. risky vs. not risky)
5. Question: What is the value of creating a model of a system and
redesigning it a few times? Statement: Students will understand and
appreciate:
• that model-building skills are helpful to convey and understand an
idea
• that iterative design is a useful skill
It is important to note that the goal of this intervention is not to present risk as
something to be avoided at all costs. After all we would not have learned how to
ride a bike if it was not for our ability to take risks.
Analysis of learning challenges
Most of the understanding goals described before were selected based on
what research has identified as being challenging features of understanding risk
in the context of complex causality. Therefore, the learning challenges and goals
in this project have an almost one to one relationship.
1. Tendency to attend to only obvious causes. When assessing risk there are
multiple factors that can come into play to determine how risky something
is. Research shows (Driver and Warrington, 1985) that often times
people’s explanations focus more on those “perceptually obvious” than
those less noticeable causes.
2. A centralized mindset is common in most children and even adults. The
idea of decentralized systems is not new, yet, most of us grow in a world
where the centralized mindset is prevalent. Assessing risk with a
decentralized mindset is powerful, however, it is challenging not only for
children but for all people. Leiser (1983) exemplifies it when writing about
children’s conception of Economics: “The child finds it easier to refer
unexplained phenomena to the deliberate actions of a clearly defined
entity, such as the government, than to impersonal ‘market forces’.”
3. Great difficulty is shown in understanding causes that involve probability.
Research shows that children can think in probabilistic terms, but they
greater difficulty in thinking about probabilistic causes (Kalish, 1998).
4. Time delays and distance can prevent people from seeing a causal
relationship (Grotzer & Bell, 1999). Credit assignment is one of the pillars
of learning (Minsky, in print), but sometimes it is very challenging to
identify the causes of an event due to the difference in time or distance
between cause and effect. For example, is global warming related to
something that is happening in a different location?
Structure of Intervention
“Riskland”
If you wanted to really learn French, you cannot do better than living in
France for a while (Papert, 1980). With this idea in mind, the Logo programming
language was invented as way to give children the ability to live in 'Mathland' – a
computer environment where Mathematics must be learned in order to play,
explore and prosper.
Using this powerful notion of full immersion, this project tries to create a
microworld where in order to play, explore and prosper, one must learn about
risk. “Riskland” takes the form of a pervasive game where the learning goal is to
let participants be fully immersed in a virtual world where the understanding of
risk is necessary to succeed.
The game
The objective of the game is to earn points by collecting items scattered
around certain location (school, neighborhood or city). Participants are split into
teams and teammates are encouraged to help each other. The winning team is
the one that gets the highest score by the end of the session, which is not
necessarily the one that collects the most number of items. In order to emphasize
the collaborative nature of the game, only the team’s scores are published for
everyone to see. The personal scores are private by default.
Participants carry cell phones with them that, using cell phone tower IDs,
the system can determine the zone in which the participant is located. Given this
location, the cell phone presents the user with a list of possible items to be found
in that area with a description and picture of the specific location given as clues.
Fig 1. Cell phone application with location awareness.
The items to be collected are physical annotations embodied in
Semacode2 tags. Inspired by the use of Semacodes in the Semapedia project
(Rondeau & Wiechers, 2005), these paper tags are attached to physical spaces
such as walls, doors, etcetera. This how a Semacode looks like this:
2
Semacode is a trade name for machine-readable two-dimensional black and white
symbols that act as "barcode URLs." It is primarily aimed at being used with cellular phones with
cameras. ([Link]
Fig. 2. Example of Semacode printed on paper (Semapedia)
The tag itself only needs to have the Semacode, anything else printed on
the tag is for decoration purposes or to explain to people not participating in the
game what the tags are about.
When a participant finds an item he or she he uses a camera phone to
read the tag and record the finding into the system. Just for finding the item, the
participant’s team gets one additional point.
Fig 3. Participant finding Semacode in the street uses the Semacode reader (Semapedia, Semacode).
Once recording the finding and its corresponding point, the participant will
be presented with the choice of getting an additional reward that comes with the
risk of a penalty. The participants are presented with information that will help
them make the decision of whether or not the reward is worth the risk of the
negative consequence.
Rewards can be:
- Points ranging from 1 to 10.
- Ability to get double or triple points in the
future collections.
- Ability to get additional points if participant
engages in an interaction with another
teammate. Interactions consist of getting in
physical proximity with a teammate and
connect both cell phones via Bluetooth3.
Fig. 4. Users exchanging information
via Bluetooth (Corbis)
Penalties can be:
- Item collection disabled for certain period of time or in certain location.
- Item collection hindered by cutting in half or third future points collections.
3
Bluetooth provides a way to connect and exchange information between devices like
personal digital assistants (PDAs), mobile phones, laptops, PCs, printers and digital cameras via
a secure, low-cost, globally available short range radio frequency.
Awards and penalties are explicitly described as well as the percentage of
risk associated with it. For example, a screen could have a statement like this:
Fig. 5. Proof of concept application
Penalties can affect the individual that found it or they can spread to
others when interacting with them. The spreading of penalties depends on:
- The penalty. Some penalties cannot be spread
- The interaction with other players, mainly teammates.
- The interaction with the items. Some penalties can be transported from
one item to another.
The game changes as participants interact with items and with each other,
giving it dynamism and the ability to frame each activity to convey specific
understanding goals. All activities are recorded by the system.
The activities
During the workshop, participants first play a predefined game and later
customize their own game by placing Semacode tags in physical locations and
registering them into the system using their cell phones. When the registration
takes place, the participant has the ability to choose a reward and penality from a
menu.
Fig. 6. Participat pasting printed Semacode in a physical location (Semapedia)
During the game, participants are expected to confront the understanding
goals when trying to make sense of the game as players and designers.
Debriefing sessions will consist of discussion and reflection on the activities as
well as through experiments that will lead students to make analogies of the
abstract underlying structure of the game with similar scenarios from real life.
Facilitators will encourage learning by analogies and will use those analogies to
confront students with their naïve assumptions about their experiences in other
settings.
Workshop
The activity will be a two day workshop that will follow a structure like this:
• Day 1
o 9:00 AM. Give students an overview of the day’s activities and
overall goals. For example:
“Welcome everyone, today we are going to learn about risk.
We will learn new ways of thinking about risk and its causes.
We are going to do this by playing some games. How many
of you like games? What type of games? Video games?
Well, today we are going to play a game using your cell
phones, it is part video game and part real-world game.
We’ll be walking, talking and having fun. The first half of the
day we will… and then we will finish by doing…. Ok?”
o 9:15 AM. Introduction. Students will be asked to thinking about their
current understanding of risk. This will also help facilitators get an
idea of what are the preconceptions students bring with them.
Sample questions to direct the conversation:
Can you give me examples of situations where you have
encountered risk?
How do you decide whether or not to take risk?
How does your perception of risk is affected by how big or
small the consequences are? (Extent of consequence)
Do you think risk is related to luck? (Deterministic vs.
probabilistic)
Can you always tell when something involves risk? Can you
remember any situation when you or someone else did
something and later found unforeseen consequences? (Non-
obviousness)
o 10:00 AM. Describe purpose of the game and split students into
two groups. Let them choose a name for their team.
o 10:30 AM. First game takes place to get students accustomed to
the system and experience the concepts in real life. This section
appeals to those learners that enjoy playing games and
manipulating toys, like in the case of Montessori learning objects
(Zuckerman, in preparation)
o 11:30 AM. Everybody gets together to have a brief discussion of
the experience. Sample questions for the discussion:
What strategies did you use while playing the game?
What made you take the risk of going after a reward? How
did you come up with that conclusion?
Did you encounter a situation where you didn’t know why
you got less or more points than you thought? Why do you
think that happened?
o 12:30 PM. The winning team is announced.
o 12:40 PM Lunch break.
o 1:00 PM. Facilitators explain that now it is time for them to design
their own game by hiding items and assigning them rewards and
penalties. They are informed that the game they are designing will
be played by the other team. The learning objective for the design
of customized games is mainly to foster a deeper understanding of
risk concepts because creative design is one of the cognitive tasks
that promote development of analogies (Ganter et al, 2001)
Students should accomplish a higher level of mastery in the
understanding of the game and risk itself. Through the design of
their own games students will think about how to challenge the
other players by thinking about making clever combinations of
award-penalty pairs that use some of the learning challenges
themselves.
o 1:30 PM. Groups go and design their game. This resonates with
learners more inclined to learning by design and construction.
(Zuckerman, in preparation). Facilitators need to make sure that the
level of difficulty is maintained within reasonable standards. The
parameters to be observed are:
Location of items. Make sure they are not in dangerous or
inaccessible locations.
The points awarded should correspond to the risk and the
penalty. For example, a reward of 1 extra point with a 99%
risk of loosing 10 points might not be the best arrangement.
The facilitator should point out this, but the students
themselves should think about this and making the final
decisions.
o 2:30 PM. Everyone gets together again. Facilitators indicate that
new each group can play each others game.
o 3:30 PM. Time is up. Discussion of experiences guided by
questions from the facilitators. These questions should lead
students to:
Answer any questions they had about the structure of the
game.
Make analogies of their experience with real-world
phenomena. Leveraging the power of analogical reasoning
(Gillespie, 1999).
Think about their strategies and ways of thinking while they
were playing the game.
Sample questions:
• Let’s list all the types of risk/rewards you experienced
and the strategies that we can use to make a
decision.
• If you could name the strategies you used, how would
you name them?
• Can you think of examples in the real life where you
have encountered similar decision making situations?
Did you act the same way as in the game? If not,
why?
• How is the game similar to real life? How is it
different?
• What type of penalties did you use when designing
your game? Which ones do you think are the most
challenging for other players and why?
• How did creating the game help you see the different
things to look for when assessing risk? What
elements of the game do not apply to real life?
• In what scenarios do you think similar strategies to
the ones used in the game might or might not work in
real life? Why?
• How would you advise others to act on similar
situations?
• Do others team members affect your decision
making? Would you have done the same if you were
playing by yourself? Does the same happen to you in
real life?
o 4:30 PM. Facilitators indicate that students can stay to discuss with
their teammates ways to improve their game design and playing
strategies for the next day session. This initiates the first iteration of
the design of their game. Facilitators should help teams in the
design of their games based on the different strategies and
challenges that came up during the discussion, which should
resemble the concepts listed in the understanding goals.
• Day 2
o 9:00 AM. Welcoming to day two. Facilitators present the goals and
schedule for the day.
o 9:30 AM. Students continue in the improvement of their game and
creation of strategies.
o 10:30 AM. Groups play each others game. It is assumed that the
difficulty of the game is higher as well as their skills for playing the
game.
o 11:30 AM. Everyone gets together and a winning team is
announced. Teams start working on a poster that summarizes their
experiences and strategies on how to approach the game and the
applications of the skills in other risk assessment activities.
Facilitators should help students in defining the structural elements
of the posters which ideally should include:
Lists of social and natural phenomena that share
characteristics with the game, specifying the differences as
well.
Mind maps on all the concepts touched upon the activities.
Workflows on different strategies used during the game and
noting how they can be applied in real life.
o 12:00 PM. Lunch break.
o 1:00 PM. Poster creation continues.
o 2:00 PM. Poster presentation takes place. Facilitators lead the
discussion and wrap up by giving a didactic presentation that
encapsulates the understanding goals. Each team has 20 minutes
to present their poster and defend their ideas. The organizers
should highlight those concepts presented by the students that
match the understanding goals.
o 3:00 P M. Group assessment takes place. The questions should be
very similar to the assessment done the first day, but giving more
emphasis to what was experienced in the games
o 4:00 PM. Ends.
The discussion session is perhaps one of the most important moments of
the workshop because it is the main opportunity to foster high-road transfer. The
facilitator should encourage students to challenge what is presented and
question each other their understanding (Feltovich, 1993). Questions should lead
to the description of the general structure of the game and then the facilitator
should point out explicitly the general features of the activity (Perkins and
Salomon, 1988). Some of the questions should lead to the finding of similarities
between the models presented in the activity and other phenomena students
know, for example:
Diseases
Rumors
Sexual education
Smoking
Pollution
Justification for the Design
There is a long tradition of games in the style of what here is proposed.
Examples of this are scavenger and treasure hunts and more recently
Geocaching (Peters, 2004). Geocaching, is an outdoor treasure-hunting game in
which participants (called "geocachers") use a Global Positioning System
receiver or other navigational techniques to hide and seek containers (called
"geocaches" or "caches") anywhere in the world. A typical cache is a small
waterproof container containing a logbook and "treasure" (usually toys or trinkets
of little monetary value). Some variations of the game include a point system to
enhance game play (Wikipedia). At the same time, these types of hunting games
have been mentioned by educators (Lary, 2004) as a valuable tool in educational
environment. Similarly, successful implementations of pervasive games for
education can be exemplified by the game Savannah (Benford et al, 2004) where
kids play at being lions by navigating in augmented environments with mobile
handheld devices. Nevertheless, the idea of letting people not just participate, but
also create their own games has not been widely explored. Through the design
of their own games, learners will take the active roll of creators not just
consumers of ideas. The creative design process is a cognitive task where
analogy development is manifested (Gentner et al,2001). The hope is that the
creations of analogies will occur and if it does, research shows that it can have a
very positive impact on learning.
1. One of the goals of the activities in this project is to appeal to a wide range
of people, something that Resnick and Silverman (2005) call “wide walls”.
This also reflects the multiple intelligence theory set forth by Gardner
(1980). Given that framework we could map several intelligence types to
aspects of the game:
• Logical/mathematical. Coming up with a strategy for finding of items
and assessing risk in a quantitatively way
• Verbal/linguistic. Describing activities during the discussion and
poster presentation.
• Visual/spatial. Creating a mental image of the physical locations
where the game takes place.
• Bodily/kinesthetic. The activities involve walking and moving around
in a physical location looking for the hidden items.
• Interpersonal. Socialization is part of the team organization. Also,
having mental models of other players help make the game more
challenging during the design phase.
• Intrapersonal. Individual exploration and planning occurs during the
game.
2. Learners have the opportunity to confront these challenges while playing
two roles: designer and participant. While playing the game, the
understanding challenges are actually part of the game challenges,
fostering the understanding of the core ideas. Similarly to learning
analytical skills while playing chess. When the learners play the role of
designers, they have the opportunity to customize the game by placing
challenging obstacles that relate to the learning challenges as well. As
opposed to other game design tools that give a lot of freedom, in this case
the tool will give less freedom as to what options are available, giving
more emphasis to the understanding of the core concepts rather than the
game building itself. One can think of this design tool as a MS Word
Template where we can only change the content but the framework is
created for us.
3. The abstract nature of the game would allow for the use of multiple
analogies rather than being stuck to a specific example. All the
discussions should always link back to specific examples. This feature is
aimed at fostering transfer. The activities are abstract on purpose to
avoiding low-road transfer by attempting to emphasize deep structures
rather than surface features (Feltovitch et al, 1993). Ideally this will lead to
a high-road transfer (Perkins D.N. &Salomon, G, 1988) to a diverse
number of fields presented during the discussion session.
4. Facilitators end the sessions with a group discussion and reflection of the
events of the game that try to encourage the use of the metacognitive
layers through Socratic-like questioning. For example, one of the
metacognitive layers relates to the evaluation of the type of thinking
learners do, this is approached through suggested questions such as “In
what scenarios do you think similar strategies to the ones used in the
game might or might not work in real life? Why?”. Learners have the
opportunity to confront these challenges while playing two roles: designer
and participant. While playing the game, the understanding challenges are
actually part of the game challenges, fostering the understanding of the
core ideas. Similarly to learning analytical skills while playing chess. When
the learners play the role of designers, they have the opportunity to
customize the game by placing challenging obstacles that relate to the
learning challenges as well.
5. In the spirit of Constructivism, one of the goals of this project is to engage
students in the creative design of a game, not only to foster the model-
building skills but because “research has shown that the process of
creating models (as opposed to simply using models built by someone
else) helps develop a greater understanding of the concepts embedded in
the models” (Colella 2001).
6. Collaborative learning. Given the structure of activities it is expected to
foster collaboration.
7. Engagement. The activities should be fun for everyone. Participants in
previous similar activities have proven to be very engaging. This is what
an educator found in a Geocaching activity: “The kids were excited (as
were the parent volunteers who, in one case, had to be restrained!), and
two of them even mailed me thank you notes after the camp was over.
How often do students actually thank you for teaching them something?”
(Leary, 2005)
Critique of the design
Perhaps this project is too ambitious at trying to accomplish the
understanding of risk while also showing how to create a game. There were two
reasons why I decided to go this way. The first is that in my personal experience
the deepest learning experiences occur when I am building something rather
than being just an observer. The second is that the creative design functionality
has not been widely explored in the realm of pervasive gaming. I think running a
mini version of this workshop with other graduate students could give me a better
idea of how successful this could be and what changes can be made, perhaps
the creation of games should only be introduced in a longer workshop.
Given my limited experience in teaching it is hard to predict how the
sessions would be carried out, this has made the facilitator’s instructions lack
more detailed definition. This, again, can be improved by running a mini
workshop where I would get better sense of how the script of a session would
look like. The same critique applies to the description of the posters.
In terms of the technology, one of the biggest challenges is the User
Interface given the space limitations of the cell phone. The goal will be to opt for
simpler interfaces even at the cost of attractiveness.
There is certainly a lot of work needed to be done before this can be
implemented in a school setting, but I think this paper help me shed a light on
those issues to be on the look out for. In the next weeks the focus will be on the
creation of the simplest version of both: the activities and the technology, leaving
aside the ability to design customized games. This will give me a product that I
can test in a lab environment and from which I can get more feedback from
experts. I expect to have multiple iterations of the same process until I have a
more refined version of this project.
Prototype
The initial idea is to implement the items as printed Semacode tags
(similar to those used by [Link]). Using cell phones with location based
services (Java API LBS JSR 1.79) the phone will tell the user where to look for
items and some clues. Once the user finds the item, it will use the camera phone
to retrieve the information from the Semacode.
The prototype is composed of a J2ME Semacode reader application that
utilizes Location Based Services and/or cell tower ID. The application assumes
the availability of camera features on the phone and data access to connect to
the server where the index of Semacodes and their reward/penalty information is
stored.
This project capitalizes on existing technology developed at the MIT and in
other places that would make it feasible to scale this to a larger audience. A
sample of the applications to be used as core components of this project are
here: [Link] along with instructions on
how to install them on a Nokia cell phone.
References
1. V Colella, R Borovoy, M Resnick (1989). Participatory Simulations: Using
Computational Objects to Learn about Dynamic Systems. Proceedings of
CHI ‘98, Summary.
2. E Klopfer, K Squire, H Jenkins (2002). Environmental Detectives: PDAs as
a Window into a Virtual Simulated World. Proceedings of the IEEE
International Workshop on Wireless and Mobile Technologies in
Education.
3. T Grotzer (2005). Learning to Think About Complex Causality and Risk -
Project Description. NSF Proposal.
4. O Zuckerman (in preparation). Historical Overview and Classification of
Traditional and Digital Learning Objects.
5. J Peters (2004). Complete Idiot’s Guide to Geocaching. Alpel Publishing.
6. L.M. Lary (2004) Hide and Seek: GPS and Geocaching in the Classroom.
Learning and Leading with Technology, March 2004 Vol. 31 No. 6.
7. S Papert (1980). Mindstorms: children, computers and powerful ideas.
MIT Press.
8. M Resnick (1994). Turtles, termites, and traffic jams: explorations in
massively parallel microworld. MIT Press.
9. S Benford, C Magerkurth, P Ljungstrand (2005) Bridging the physical and
digital in pervasive gaming. Communications of the ACM, 2005.
10. S Benford, D Rowland, M Flintham, R Hull, J Reid, J Morrison, K Facer, B
Clayton (2004) “Savannah”: Designing a Location-Based Game
Simulating Lion Behaviour. International Conference on Advances in
Computer Entertainment Technology.
11. Feltovich, P.J., Spiro, R.J., & Coulson, R.L. (1993). Learning, teaching,
and testing for complex conceptual understanding.
12. Wiske, M.S. (1998). What is Teaching for Understanding? In M.S. Wiske
(Ed.) Teaching for Understanding: Linking Research with Practice.
13. .Perkins, D. N., & Salomon, G. (1988). Teaching for transfer. Educational
Leadership, 46, 22-32.
14. Kalish, C.W. (1998). Young children’s predictions of illness: Failure to
recognize probabilistic causation. Developmental Psychology, 34(5).
1046-1058.
15. R Driver, L Warrington (1985) Students' use of the principle of energy
conservation in problem situations - Physics Education.
16. Leiser, D. (1983). “Children’s Conceptions of Economics- The Constitution
of a Cognitive Domain.” Journal of Economic Psychology.
17. Rondeau, A. & Wiechers, S. (2005) Semapedia – The Physical Wikipedia.
18. Resnick, M and Silverman B. (2005). Some Reflections on Designing
Construction Kits for Kids.
19. Wikipedia. [Link]
20. Gentner, D., Holyoak, K. J., & Kokinov, B. N. (Eds.) (2001). The analogical
mind: Perspectives from cognitive science. Cambridge, MA: MIT Press.
21. Semapedia. [Link]
22. Corbis. [Link]
23. Semacode. [Link]
