Gamification, mHealth and user adherence

Full text

Gamification, mHealth and user adherence
Rui Carvalho
Faculdade de Ciências,
Universidade do Porto, Portugal
up201404194@fc.up.pt
Diogo Machado
Instituto de Telecomunicações
Faculdade de Ciências,
Universidade do Porto, Portugal
dmachado@dcc.fc.up.pt
Pedro Brandão
Instituto de Telecomunicações
Faculdade de Ciências,
Universidade do Porto, Portugal
pbrandao@dcc.fc.up.pt
Abstract—MyDiabetes is a mobile application targeted for the
management of type I diabetes. A considerable challenge faced by
diabetes’ management applications is user adherence and moti-
vation. In this article we will describe the different gamification
techniques implemented in the MyDiabetes mobile application
in order to tackle the mentioned challenges and increase the
number of daily user records. To evaluate the success of the
implemented gamification techniques, a survey was conducted
at the endocrinology service of S. João’s Hospital. Twenty three
participants, diabetics type I, participated in the survey. Different
types of users’ motivations resulted in contrasting opinions about
particular gamification elements. Nonetheless, the implemented
gamification was, in general, well accepted. The gamification
elements with the lowest classification were streaks and social
comparison of points. Even so, these elements obtained an accep-
tance rate of 83%. There was a common preference for elements
that promote better glycaemic management. Classic gamification
elements that promote points, levels and objective completion
were not disregarded by the inquired patients, but were given
a secondary roll in usage motivation. Although promising, it
was not possible to determine the long-term effectiveness of the
implemented gamification. Further trials are planned to verify
the impact of gamification on long-term user adherence.
Index Terms—gamification, diabetes, mobile health, motivation
I. INTRODUCTION
Diabetes is a chronic disease that occurs when the pancreas
does not produce enough insulin or when the body is not able
to use effectively the produced insulin. It is one of the leading
causes of death in the world [1]. However, when managed cor-
rectly patients can lead a normal life without complications [1].
To achieve a correct diabetes daily management is a challenge
for diabetics. It is a demanding, repetitive and tedious set of
daily tasks that are neglected by most people with this chronic
disease. These include registration of glycaemia, meals intake
(carbohydrates and if possible lipids and proteins), insulin
taken, exercise, diseases, weight and the analysis of the effects
those options had on the glycaemic control.
Presently, a series of devices and sensors are available to
assist in diabetes management. Although useful, these devices
do not offer autonomous diabetes management. User input
is still an essential requirement for mobile applications to
evaluate correctly the user’s health state. This need in many
cases becomes a drawback, as users do not feel motivated
to do this task. It should be noted, that this obstacle does
not originate necessarily from poor usability or design. In
an evermore technologically driven society, where everything
is connected and interacts, patients expect their devices to
function autonomously with minimal to no user input. In this
article we propose the use of gamification as a motivational
tool to overcome the lack of user adherence and disposition to
register events.
Gamification is defined as the use of game elements and
techniques in serious contexts [2]. It aims to improve the
interaction between users and the application in question,
for a purpose other than pure entertainment while using the
motivational principles that games provide. As a working basis,
we used the Android mobile application MyDiabetes
1
, for type
I diabetes management. The application stores diabetes related
information, calculate insulin doses and display the information
in the form of graphs. In addition, within the MyDiabetes
application we are developing an advice system to guide and
council the user in various diabetes related occurrences. This
feature relies on the recorded user’s information to ascertain
the user’s current state and advise accordingly. As such, there
is the need for the user to register information, which leads to
the need for motivation and adherence.
In order to motivate the user, the MyDiabetes application
had to undergo an initial gamification process and redesign. In
Parra, R. [3], some elements of gamification were implemented,
followed by a questionnaire to understand the user’s acceptance
of the gamification applied. The main elements introduced in
the system were: points, levels and badges complemented with
objectives with varied difficulties for the user to complete
by recording new data. The objectives serve as long term
objectives and as motivation to insert new records. By inserting
new records and unlocking achievements the user is rewarded
points, that serve as immediate incentive. After gathering a
certain amount of points the user levels up and unlocks more
advanced objectives. The described system and the interactions
between the implemented elements is represented in Figure 1.
Although promising, the implemented gamification mech-
anisms did not translate in more user adherence nor number
of records. To comprehend these results, an opinion survey
was conducted. This study was carried during one month (two
mornings per week) and had the participation of 14 diabetic
1See https://mydiabetes.dcc.fc.up.pt (in Portuguese).978-1-7281-8086-1/20/$31.00 ©2020 IEEE
Authorized licensed use limited to: CALIFORNIA INSTITUTE OF TECHNOLOGY. Downloaded on November 02,2020 at 00:50:36 UTC from IEEE Xplore. Restrictions apply.

Fig. 1. Gamification point, level and achievement interaction graph.
patients aged between 21 and 54, of the Endocrinology service
of the S. João’s hospital.
Questioned about the point and level system, 9 (64%) partic-
ipants considered both as relevant elements in the application.
When asked if they felt compelled to insert more records by
levels and points, the percentage of people that agreed dropped
to 8 (57%). It is possible to reason that people generally
consider points and levels an important part of a gamified
system, but these elements, by themselves, are not enough to
motivate and keep people engaged. Nonetheless, in some cases,
it is even possible to argue that some people expect gamified
applications to at least have these elements.
The achievement system together with the objectives/quests
had a similar response by the participants. When asked about
the relevance of this game mechanism, 10 participants (71%)
considered them relevant. This number decreased to 9 (64%)
when the participants were asked if they felt motivated to win
badges and complete all the achievements.
At this stage there were two elements thought, but not
implemented. One was based on competition elements with
a “classification table” and the other a community connection
using social network integration. The sensitivity of the topic and
data influenced our choice. We used the survey to perceive the
opinion of diabetic patients on these elements. The participants
had clear opinions on both concepts. Social network integration
was received with major discomfort, 12 (86%) participants
completely disagreed with the possibility of sharing any type
of game or health information. Contrasting, the introduction of
competition elements was well-received. 11 (77%) participants
stated they would appreciate and be motivated by in-application
competition, on the condition that all data would remain
anonymous. Some participants disapproved of in-application
competition. This originates from the fear of being demotivated
by the success of other participants.
Per the survey’s conclusions, we decided to further develop
the application’s gamification and redirect some of its focus to
the user’s health improvement, instead of points and levels.
II. BACKGROU ND
Games, specially video-games, are notorious for their ad-
dictive nature. On the other end, tasks concerning health and
education, although important, are considered tedious. The
application of gamification to other areas other than games is
seen as an opportunity to transform nonchalant exercises into
pleasurable activities. To achieve this, it is important to analyse
the components of games responsible for addictive behaviour
and apply it outside its gaming scope.
A. Game elements as gamification techniques
As stated by Souza-Júnior M. et al. gaming mechanics can
be used to engage and stimulate the user’s desire to address
health problems, when correctly implemented [4]. Werbach
and Hunter [5] proposed a framework, in which game elements
are presented in the form of a three-level pyramid. According
to the authors, the informed use of these elements will result in
a better gamified system. This pyramid consists of:
dynamics
,
mechanics and components.
The
dynamics
pyramid’s level is constituted by the most
conceptual elements of a gamified system, which structures the
game, but never enters the system directly. Its elements are:
the game’s limitations and rules (Constraints); sentiments, such
as accomplishment or victory (Emotions); the story-line that
drives the game (Narrative); the sense of progress (Progression);
and the ability to socially interact (Relationships).
The
mechanics
are composed by the basic processes that
drive the user to act, consequently leading to involvement.
Its elements are: Challenges; Chance; Competition; Feedback;
Cooperation; Resource Acquisition; Rewards; Transactions;
Turns and Win States.
Finally, the
components
are the specific implementations
of the pyramid’s upper levels. This layer is at the base of the
pyramid since it contains elements that are in direct contact
with the users. In our initial gamification implementation the
components implemented were the: Achievements, Badges,
Points, Levels and Content Unlocking.
B. Motivation
The feeling of motivation comes from the desire to achieve a
certain goal. Motivation can be categorized as:
extrinsic
, when
the objective is tied to external factors such as money or power;
and
intrinsic
, where the user consciously completes objectives
for the feeling of self-achievement and competence, without
expecting any external reward. Regarding gamification, both
intrinsic and extrinsic motivation types should be implemented.
Several studies refer that the implementation of extrinsic
motivators in tasks where users do not feel intrinsically
motivated, successfully stimulates the user to still accomplish
them [6]. Given the importance of possessing both motivational
types, it is important to find a balance between intrinsic and
extrinsic motivation [6].
C. Player Types
Every person is unique. As persons, players have personal
tastes and different motivators. Bartle R., while studying
MMOGs’ (massive multiplayer online games) players, analysed
the different factors that motivate them [7]. In this study, he
created a framework that defines four types of players. The
Achievers
, are players who enjoy having an active in-game role
and are motivated by in-game accomplishments. The
Explorers
,
are players that enjoy interacting with the world and uncover
in-game aspects, history, features and mechanics.
Socializers’
main motivator is in-game player interaction. The
Killers
are
competitive players motivated by the feeling of being superior
to other players.
Authorized licensed use limited to: CALIFORNIA INSTITUTE OF TECHNOLOGY. Downloaded on November 02,2020 at 00:50:36 UTC from IEEE Xplore. Restrictions apply.

This player characterization describes clearly the major
MMOGs population. When considering the general population,
this method becomes inaccurate [6]. While flawed, this system
is still considered as an excellent tool to classify player types
and to create player oriented systems [5].
D. Behavioural persuasion
The Fogg’s Behaviour Model (FBM) is used in the study of
human behaviour. This psychological model defines the three
main factors that, when present, influence the performance
of a certain behaviour: motivation, the ability to execute
the proposed action and a trigger that prompts the user to
act [8]. Additionally, this model co-relates motivation and
ability. According to this model, the more complex an action
is (high ability) the more motivation is required for the user
to achieve the desired action.
Neither the FBM nor the Werbach and Hunter’s pyramid
give undisputed answers to what is a well gamified system.
Instead, they present useful mechanisms and the theory that
supports these mechanisms. Understanding both frameworks
allows developers to implement correctly a gamified system.
The MyDiabetes project, being a health related application,
requires special attention. Health management is innately
related to intrinsic motivation. Gamified health applications
should consider this relation as an opportunity to increase the
user’s motivation and consequently the user’s adherence.
III. REL ATED W OR K
The use of gamification in mHealth to improve adherence
and motivate the user is not novel. Although gamification is
presented in theory as the link between patients and mHealth
applications, it is important to assert its impact in a real life
scenario. Johnson et al. evaluated the impact of gamification
for health and well-being in their systematic review [2]. Results
were overall positive, as 59% of the analysed studies reported
that gamification had a favourable impact. The remaining
41% of the studies reported a neutral or mixed reception
of the gamified system. The authors further refer that these
results must be interpreted with caution given the small
study pool and the methodological limitations of many of
the considered studies. Other considered causes for the mixed
reception of some of these studies were: the manner in which
it was applied e.g. disproportionate feedback, and the possible
incompatibility between the gamification techniques applied
and the target users. In the review, most of the different
studies implemented more than one gamification technique,
but evaluated the gamified system as a whole. This process
does not allow the evaluation of the impact of each technique
individually. In order to verify individual contributions, the
authors established a connection between the most commonly
used techniques and the gamification’s success. The nineteen
studies evaluated, referred a total of 46 gaming elements
used. Reward elements (35%) were the most common element
type. The second and third most common elements were
avatars (13%) and leader-boards (13%). Five of the studies [9]–
[13], identified rewards as responsible for an increase in
user engagement. One of these studies in particular showed
that gamification incentives can contribute to an increase of
glycaemic records in adolescents with type 1 diabetes [11].
Two other studies [14], [15] associated an increase in user
engagement and motivation to a combination of reward to
leader-board elements. One study, targeting physical activity
motivation [16], verified that their reward approach was a
temporary success (8 weeks). On the long-term (20 weeks), the
reward system, was not enough to maintain the users engaged.
Although motivating, reward systems require other elements
to maintain the levels of motivation in users.
The MyDiabetes first gamification process supported this
element, as it was mostly focused on reward elements. While
upgrading the gamification system, we had the conscious
decision to diversify the gamification elements in order to
reach a long term success.
IV. DEV EL OP ME NT
People expect applications to be valuable, usable, and reliable.
Regarding diabetes management, people expect to keep input
to a minimum, and receive helpful feedback and information.
Gamification can be used to motivate users to engage in certain
behaviours, otherwise it is seen as a burden or a task.
In order to improve the previous gamification system, we
considered the opinions stated on the survey mentioned above,
literature review and medical experts’ opinions. The following
subsections describe the new gamification elements that have
been implemented in the new version of MyDiabetes.
A. “Daily Information” section
The daily information window is a summary of the user’s
daily diabetes management results. It contains the glycaemic
values’ average, variability and information about the user’s
condition through colour feedback. As represented in figure 2
the average and variability values change colour according to
the user’s state: green represents good values; yellow represents
values requiring caution and red represents values that require
the user’s full attention. The average blood glucose values are
divided into five intervals: recommended,
70 −180 mg/dL
(green); low,
54 −69 mg/dL
(red); very low,
<54 mg/dL
(red); high,
181 −250 mg/dL
(orange); and very high,
>
250 mg/dL (red).
Variability values are divided into three ranges: optimal,
<33%
(green); acceptable,
33 −36%
(orange); and not
recommended,
>36%
(red). Both selected variables are
standardized CGM (Continuous Glucose Monitor) metrics for
clinical care [17]. Colour feedback was used because it is
considered a source of information with the ability to influence
a person’s mood and feelings, also playing an important role
in shaping attitudes [18].
These values’ precision is bound to the daily record of at
least 6 glycaemias. The need of frequent records influenced
the inclusion of a streak mechanic. Streaks are a habit-forming
technique that compensates users who consistently complete
the required objectives [19]. In the MyDiabetes application the
user receives a point multiplier for achieving a streak, which
Authorized licensed use limited to: CALIFORNIA INSTITUTE OF TECHNOLOGY. Downloaded on November 02,2020 at 00:50:36 UTC from IEEE Xplore. Restrictions apply.

Fig. 2. MyDiabetes health window (on the left) and its Daily Information
view (on the right).
will influence all future records while on a streak. This element
can motivate the user to fulfil glycaemic summary requirements.
When the user completes the six daily glycaemia records, a
specific achievement notification appears congratulating the
user. With the objective now complete, the daily information
text referring the amount of measurements missing for complet-
ing the objective, will switch to “Congratulations on reaching
today’s objective”. These details, praising the user, functions
as extra rewards that add to the user’s intrinsic motivation to
have a successful glycaemia control.
B. Social comparison system
The Community statistics window represented in figure 3
displays a summary of the user’s current best score and streak
together with an anonymous community ranking containing
in-game and health statistics.
Fig. 3. Community statistics open view.
The user’s data is uploaded daily and stored in the MyDia-
betes server. The server creates a file every Sunday containing
anonymous weekly obtained points, streak values, glycaemia
value quality and number of hypo/hyperglycaemias. There is a
scheduled task to calculate the glycaemic value quality using
the Time In Range (TIR). TIR is the number of glycaemia
values in the range considered to be acceptable for the general
population, i.e., between 70 and 180. For equal TIRs, users
are ranked by the variability of the values. All the MyDiabetes
definitions and features were developed with the support of a
group of endocrinologist experts from the S. João’s hospital.
The statistics are sent to and used by the mobile application to
compare its user’s values with the ones from the community.
The results are always presented as a percentage in order to not
demotivate the user. For example, we believe that users prefer
to know they have better health values than 40% of users, than
to know they are the 140
th
user with better glycaemic values.
This feature is not cherished by all users. Considering this, we
have implemented the social comparison system as an optional
feature. To have access to it, the user must explicitly select
and activate this feature.
C. Widget
A widget is a graphical element that can be on the
smartphone’s home screen. These allow an “instant” view
of the most relevant data and features of an app, allowing the
user to quickly access it without opening the app.
Our implementation of this feature, shown on figure 4,
contains a graph with the last 5 records in the last 24 hours,
the values and date of the last record made and a plus sign
that functions as quick-access to the records’ activity in the
MyDiabetes application.
Fig. 4. Widget displayed on a mobile phone (on the left) and the isolated
widget window (on the right).
The widget’s graph allows users to evaluate the flow of their
values and verify tendencies. Having the last record at the
bottom of the graph, allows for a more detailed description of
the last carbohydrates, insulin, and glycaemia. The information
displayed in the widget allows users to make practical and fast
reflections of their health condition.
This element functions as a facilitator, increasing the user’s
ability to interact with other gamification elements within
the application. Since the widget does not possess a trigger
response, it is not possible to say if it will induce adherence.
However, we can affirm that this element, as it increases the
application’s simplicity, considering B.J. Fogg’s work [8], will
decrease the user’s necessary motivation to insert new records.
Authorized licensed use limited to: CALIFORNIA INSTITUTE OF TECHNOLOGY. Downloaded on November 02,2020 at 00:50:36 UTC from IEEE Xplore. Restrictions apply.

D. Notification system redesign
An important mechanism to support motivation is feed-
back [5]. It should be immediate, frequent and noticeable.
In the previous gamified version, the notifications lacked
these characteristics. The new version contains 4 types of
achievement notifications that improve this feedback element.
We separated the achievements notifications into: health
badge notification, streak notification, daily objective noti-
fication and logbook entry notification. This division was
necessary as the user can win more than one badge at the
same time. This way, no new achievement is omitted. All
achievement notification types have different colours which
helps to distinguish and recognize the different notifications.
E. Badge system redesign
The previous MyDiabetes version was composed by a
series of record objectives distributed by three difficulty levels.
Records have a major importance for a correct diabetes
management, nonetheless, the user should strive to improve all
diabetes’ related metrics. In light of this, we have included, in
the new version, three new health related objectives present in
the three, already existing, difficulty levels. The objectives cover
three important aspects of diabetes management: maintaining a
good average value for glycaemia, maintaining a low glycaemic
value variability and to keep most glycaemic values under
the limits defined by the medical expert accompanying the
user. The easier difficulty states that the user must maintain
these objectives 7 days in a row. As the difficulty increases,
the number of days in a row also increases. The values
used for the objectives were defined with the cooperation
of endocrinology experts. The addition of these health badges
introduces intrinsic motivation to a system usually responsible
for extrinsic motivation. These badges can function as a double
reward. The user receives not only a badge but also the
realization that this action was a diabetes achievement, which
will translate in better health prospects. More than a trinket,
health badges can be an element that proves to the user the
importance of consistence.
V. RE SU LTS
In order to evaluate the success of the different implemented
gamification elements, we conducted a new survey. In this new
survey we gathered, during two months (two mornings per
week), 23 participants with ages ranging from 18 to 51 years,
with a mean of 30 years. The survey consisted of nine questions
about the different new features and updates. Each question
could be ranked from one to five; 1 being unimportant and 5
very important. Respondents were given a short introduction
to the application’s new features and asked to use it during the
interview. The obtained results are shown in Table. I.
The first question focused on the new widget element.
Participants, when asked if the new widget would have an
impact on their diabetes management, responded positively.
Immediate feedback was appreciated by all the participants.
The shortcut feature to record new values was also appreciated,
but considered secondary.
The second question evaluated the medical feedback (“Daily
Information”). Participants were asked to evaluate the overall
feature and its colour feedback properties. This feature received
a mainly positive, but weaker, response. The survey’s first two
participants were not introduced to the colour feedback, so
there is no answer for them.
Now focusing on the new mechanics, the participants were
asked to evaluate the record streaks. This mechanism divided
opinions. Two participants evaluated it with the lowest score,
affirming that this mechanic was demotivating. Other two
participants evaluated the streaks negatively, justifying that
they did not consider this mechanism useful. Seven users had
a neutral response to the streaks and twelve participants had a
positive response to this feature. Considering these results, this
mechanic, although important to motivate and maintain regular
use, can be demotivating for a considerable amount of users.
The next two questions addressed the Community statistics.
This new competition feature, in order to rank users, contem-
plates both in-game points and glycaemic value quality. In
the forth and fifth question, participants had to evaluate the
new feature’s ability to motivate users to both insert more
records and improve glycaemic values, in order to reach a
higher in-game rank position. In terms of motivation to record
more, four participants responded negatively. Two of these
participants disfavour the competition aspect of the feature.
The other two participants acknowledged that this feature
would not motivate them to increase their amount of records.
For these participants, competition is not motivational. This
apprehension to competition, in some cases, is connected to
the user’s diabetes management self-consciousness.
In terms of motivation to improve glycaemic control, three
people maintained their previous negative vote, but the feedback
was more positive. This question allows us to verify that people
accept more naturally a feature, even if it is disliked, when it
is targeted to their well-being.
The following two questions compared the old gamification
badge implementation against the new health badges. Once
more, the survey’s participants chose health as a main motivator
as the positive levels of acceptance increased from 56.5% for
record-based badges to 91.3% for health-based badges.
Lastly, considering the importance of feedback, we displayed
the new achievement notifications and asked the participants to
evaluate how perceptible the notification was. The notification
was considered to be perceptible by most participants, but there
were two users who had issues. It should be noted that one of
these users, who gave a score of 1, had cataracts and, for this
reason, was not able to read the notification. The other user,
who voted with a 3, saw the notification, but could not recall
the won badge type. This user feedback was important in order
to reconsider colour schemes, contrast and the animation of
the achievement notifications.
Feedback seems the most unanimous feature, as every
participant agreed on the importance of feedback.
The competition side of the application as a social aspect is
one of the most complex gamification elements to implement
successfully, as Werbach and Hunter [5] refer in their work. The
Authorized licensed use limited to: CALIFORNIA INSTITUTE OF TECHNOLOGY. Downloaded on November 02,2020 at 00:50:36 UTC from IEEE Xplore. Restrictions apply.

implementation of the Community statistics and its inherent
competition served as a motivator for players who fit to some
degree to the killer player type. Considering the survey’s results,
we believe the competition aspect of the application is a positive
addition, able to motivate a majority of users.
For the participants, gamification elements with a medical
aspect were unanimously important. We believe that applying
gamification in a health related application should always have
an innate health connection.
1 (dislike) 2 3 (neutral) 4 5 (likes)
0 0 0 12 11
0 0 2 6 13
2 2 7 6 6
Motivation
to record
more
(number of
answers)
3 1 7 8 4
Motivation
to improve
glycaemia
control
(number of
answers)
2 1 4 10 6
Record
(number of
answers)
1 2 7 10 3
Health
(number of
answers)
0 1 1 8 13
1 0 1 9 12
User's Feedback evaluation
(number of answers)
User's Widget evaluation
User's 'Daily Information'
User's Streaks evaluation
(number of answers)
User's
'Community
Statistics'
evaluation
User's
Badges
evaluation
TABLE I
QUESTIONNAIRE RESU LTS’TAB LE .
VI. CONCLUSION
In this work we upgraded the MyDiabetes mobile applica-
tion’s gamification system. The component updates and new
implementations were developed having into account the theory
referred in the background section and the previous survey.
The feedback from the conducted survey was overall positive.
Although, some implemented gamification elements were not
unanimously well-received, the majority of participants received
the new implemented gamification elements with clear signs of
motivation. Competition elements while demotivating for some,
are a necessity for killer type players. We believe that compe-
tition elements can be beneficial to improve both adherence
and motivation, if implemented with care. Considering other
types of mHealth applications, we concluded that patients tend
to better accept gamification when it is targeted to improve
their well-being, and not necessarily just for game’s sake.
However, we believe that no amount of gamification can
compensate for the lack of features that are already part of
the users’ daily life. We have anecdotal evidence of this, as
patients contacted were keen on the requirement for support
of importing data from glycaemic continuous monitors. This
enables automatic transfer of glucose values from the sensors to
the mobile device. So the need to manually add every measured
glycaemic value, insulin dose administrated, carbohydrates
consumed among other important diabetic related facts to these
tasks is seen as an overwhelming effort, which may undermine
the adherence. We will investigate this further in the future
survey, where long-term usage will be tested.
REFERENCES
[1]
World Health Organization, “10 facts on diabetes,” accessed:
08/May/2020. [Online]. Available: https://www.who.int/features/factfiles/
diabetes/en/
[2]
D. Johnson, S. Deterding, K.-A. Kuhn, A. Staneva, S. Stoyanov, and
L. Hides, “Gamification for health and wellbeing: A systematic review
of the literature,” Internet interventions, vol. 6, pp. 89–106, 2016.
[3]
R. Parra, “Aumentar a interação com o utilizador –
MyDiabetes App.” Master’s thesis, University of Porto, Sep.
2017. [Online]. Available: https://mydiabetes.dcc.fc.up.pt/results/
teseFinal-RicardoParra-InteraccaoUtilizador.pdf
[4]
M. Souza-Júnior, L. Queiroz, J. Correia-Neto, and G. Vilar, “Evaluating
the use of gamification in m-health lifestyle-related applications,” in New
Advances in Information Systems and Technologies. Springer, 2016.
[5]
K. Werbach and D. Hunter, For the win: How game thinking can
revolutionize your business. Wharton Digital Press, 01 2012.
[6]
G. Zichermann and C. Cunningham, Gamification by design: Implement-
ing game mechanics in web and mobile apps. O’Reilly, 2011.
[7]
R. Bartle, “Hearts, clubs, diamonds, spades: Players who suit MUDs,”
Journal of MUD research, vol. 1, no. 1, p. 19, 1996.
[8]
B. J. Fogg, “A behavior model for persuasive design,” in Proceedings of
the 4th Int. Conf. on Persuasive Technology. ACM, 2009, p. 40.
[9]
J. Hamari and J. Koivisto, “’Working out for likes’: An empirical study on
social influence in exercise gamification,” Computers in Human Behavior,
vol. 50, pp. 333–347, 2015.
[10]
A. Allam, Z. Kostova, K. Nakamoto, and P. J. Schulz, “The effect of
social support features and gamification on a web-based intervention for
rheumatoid arthritis patients: randomized controlled trial,” Journal of
medical Internet research, vol. 17, no. 1, 2015.
[11]
J. A. Cafazzo, M. Casselman, N. Hamming, D. K. Katzman, and
M. R. Palmert, “Design of an mHealth app for the self-management of
adolescent type 1 diabetes: a pilot study,” Journal of medical Internet
research, vol. 14, no. 3, 2012.
[12]
S. Riva, A.-L. Camerini, A. Allam, and P. J. Schulz, “Interactive sections
of an internet-based intervention increase empowerment of chronic back
pain patients: randomized controlled trial,” Journal of medical Internet
research, vol. 16, no. 8, 2014.
[13]
T. A. Dennis and L. J. O’Toole, “Mental health on the go: Effects of a
gamified attention-bias modification mobile application in trait-anxious
adults,” Clinical Psychological Science, vol. 2, no. 5, pp. 576–590, 2014.
[14]
K. Thorsteinsen, J. Vittersø, and G. B. Svendsen, “Increasing physical
activity efficiently: an experimental pilot study of a website and
mobile phone intervention,” International journal of telemedicine and
applications, vol. 2014, p. 8, 2014.
[15]
Y. Chen and P. Pu, “Healthytogether: exploring social incentives for
mobile fitness applications,” in Proceedings of the second international
symposium of chinese chi. ACM, 2014, pp. 25–34.
[16]
C. Maher, M. Ferguson, C. Vandelanotte, R. Plotnikoff, I. De Bourdeaud-
huij, S. Thomas, K. Nelson-Field, and T. Olds, “A web-based, social
networking physical activity intervention for insufficiently active adults
delivered via facebook app: randomized controlled trial,” Journal of
medical Internet research, vol. 17, no. 7, 2015.
[17]
T. Battelino, T. Danne, R. M. Bergenstal, S. A. Amiel, R. Beck,
T. Biester, E. Bosi, B. A. Buckingham, W. T. Cefalu, K. L. Close et al.,
“Clinical targets for continuous glucose monitoring data interpretation:
recommendations from the international consensus on time in range,”
Diabetes care, vol. 42, no. 8, pp. 1593–1603, 2019.
[18]
S. Singh, “Impact of color on marketing,” Management decision, vol. 44,
no. 6, pp. 783–789, 2006.
[19]
D. Lovell, “What are streaks and why do they work?” Mar 2019, accessed:
08/May/2020. [Online]. Available: https://www.growthengineering.co.uk/
what-are- streaks/
Authorized licensed use limited to: CALIFORNIA INSTITUTE OF TECHNOLOGY. Downloaded on November 02,2020 at 00:50:36 UTC from IEEE Xplore. Restrictions apply.