Ecological Momentary Assessment based Differences
between Android and iOS Users of the
TrackYourHearing mHealth Crowdsensing Platform
R¨
udiger Pryss1, Winfried Schlee2, Manfred Reichert1, Ira Kurthen3, Nathalie Giroud4,
Laura Jagoda5, Pia Neuschwander5, Martin Meyer5, Patrick Neff2, Johannes Schobel1,
Burkhard Hoppenstedt1, Myra Spiliopoulou6, Berthold Langguth2, and Thomas Probst7
Abstract— mHealth technologies are increasingly utilized in
various medical contexts. Mobile crowdsensing is such a tech-
nology, which is often used for data collection scenarios related
to questions on chronic disorders. One prominent reason for
the latter setting is based on the fact that powerful Ecological
Momentary Assessments (EMA) can be performed. Notably,
when mobile crowdsensing solutions are used to integrate EMA
measurements, many new challenges arise. For example, the
measurements must be provided in the same way on different
mobile operating systems. However, the newly given possibilities
can surpass the challenges. For example, if different mobile
operating systems must be technically provided, one direction
could be to investigate whether users of different mobile operat-
ing systems pose a different behaviour when performing EMA
measurements. In a previous work, we investigated differences
between iOS and Android users from the TrackYourTinnitus
mHealth crowdsensing platform, which has the goal to reveal in-
sights on the daily fluctuations of tinnitus patients. In this work,
we investigated differences between iOS and Android users
from the TrackYourHearing mHealth crowdsensing platform,
which aims at insights on the daily fluctuations of patients with
hearing loss. We analyzed 3767 EMA measurements based on
a daily applied questionnaire of 84 patients. Statistical analyses
have been conducted to see whether these 84 patients differ with
respect to the used mobile operating system and their given
answers to the EMA measurements. We present the obtained
results and compare them to the previous mentioned study.
Our insights show the differences in the two studies and that
the overall results are worth being investigated in a more in-
depth manner. Particularly, it must be investigated whether the
used mobile operating system constitutes a confounder when
gathering EMA-based data through a crowdsensing platform.
Index Terms— mHealth, crowdsensing, hearing loss, EMA
1Faculty of Computer Science, Engineering and Psychology, Ulm Uni-
versity, Germany
{ruediger.pryss,manfred.reichert,burkhard.hoppenstedt,
johannes.schobel
2University Hospital Regensburg, Germany
3Developmental Psychology: Infancy and Childhood, Department
of Psychology, University of Zurich, Zurich, Switzerland
4Cognition, Aging, and Psychophysiology Laboratory, Depart-
ment of Psychology, Concordia University, Montreal, Canada
5Division of Neuropsychology, Department of Psy-
chology, University of Zurich, Zurich, Switzerland
6University of Magdeburg, Germany [email protected]
7Department for Psychotherapy and Biopsychosocial Health, Danube
I. INTRODUCTION
Mobile crowdsensing is a technology that has proven its
worth for the medical domain. For example, in previous
works, we found interesting insights on data gathered with
the TrackYourTinnitus (TYT) mHealth crowdsensing plat-
form [1], [2]. Also other works have revealed new insights
when using mobile crowdsensing in the context of chronic
diseases [3]. In general, these findings are based on measure-
ments in daily life, i.e., ambulatory assessments / Ecological
Momentary Assessments (EMA). These measurements can
be realized by built-in sensors of modern smartphones and
the utilization of electronic questionnaires, which are filled
out by the users on the smartphone repeatedly over a longer
period of time [4]. Following this, new medical datasets
become possible [5]. However, the technical realization of a
platform that properly enables the aforementioned aspects is
a challenging endeavor. Based on experiences when running
TYT [6]–[8] for years, we were able to adjust the technical
platform to other healthcare questions. So far, the technology
was adjusted to medical questions on the loss of hearing,
the management of stress and diabetes as well as the sup-
port of pregnant women (see Table I). For data that was
gathered with TYT, we revealed investigation opportunities
that were not thought of when designing the technology,
e.g., to compare retrospective and prospective statements
of tinnitus patients [7]. Another observation that was not
initially intended constitutes the opportunity to compare the
behavior of TYT users with respect to their used mobile
operating system. Whether a user has given an answer with
an Android or iOS smartphone was by design only recorded
for testing purposes.
However, it emerged that this information can also be
used to compare differences between Android and iOS users
with respect to their demographic and health characteristics
Project Medical Aspect URL
TrackYourTinnitus Tinnitus http://www.trackyourtinnitus.org
TrackYourHearing Hearing Loss http://www.trackyourhearing.org
TrackYourStress Stress http://www.trackyourstress.org
Chrodis+ Diabetes http://chrodis.eu
MyKind Pregnant Women http://www.mykind.info
TABLE I: mHealth Crowdsensing Projects
on tinnitus [9]. In this work, in turn, we took a closer
look at the current dataset of the TrackYourHearing (TYH)
mHealth crowdsensing platform, which addresses the daily
fluctuations of users with a hearing loss. Note that the latter
constitutes one of the top causes of years lived with a
disability [10]. The investigation we conducted in this work
aims at two goals:
•Can we reveal differences between Android and iOS
users of the TrackYourHearing mHealth crowdsensing
platform based on their collected EMA data?
•Can we confirm the results of the TrackYourTinnitus
mHealth crowdsensing platform on differences between
Android and iOS users as shown in [9]?
The remainder of this paper is organized as follows. In
Section II, relevant related work will be reviewed. Section
III, in turn, provides background information on the TrackY-
ourHearing mHealth crowdsensing platform. In Section V,
the materials and methods used for the data analysis are
described. Section VI presents the obtained results, while
Section VI discusses them. Section VIII finally concludes
the paper with a summary and an outlook.
II. RELATED WORK
Three categories of related work are relevant in the context
of this work: (1) Approaches that deal with mobile crowd-
sensing and EMA in the healthcare domain, (2) approaches
on differences between Android and iOS in the context of
mobile healthcare applications, and (3) mobile applications
focusing on hearing loss. Regarding the first category, some
recent works exist that deal with generic crowdsensing
approaches to enable human-subject studies [11]. However,
the use of crowdsensing in the context of chronic diseases
is still rare. One example for chronic disorders is the Track-
YourTinnitus mHealth crowdsensing platform [1], [2], [12].
In turn, technical solutions that enable EMA measurements
without using mobile crowdsensing technology have been
presented with valuable healthcare results [13], [14].
Regarding the second category, approaches that compare
differences between Android and iOS in the context of
mobile healthcare applications are of particular interest. In
[15], [16], for example, security issues for Android and
iOS mHealth applications are discussed. The authors of
[17], in turn, discuss guidelines of Android and iOS to
create applications that are able to support personal health
records. In general, many works exist that aim to evaluate
the differences and quality of mHealth applications that
were developed for both mobile operating systems [18]–[20].
Approaches that directly compare Android and iOS users
exist beyond healthcare questions [21]. Android and iOS
users were directly compared only in a few studies [22].
These studies have found that iOS users are more likely
female, in the mid-30s, with a graduate degree, in a higher
income group, with more technology knowledge, and that
iOS users spend more time using applications than Android
users. Clinically relevant differences were reported for the
SmokeFree28 (SF28) application [22]. For example, iOS
users downloaded the app more likely for a serious attempt
to quit smoking. Android users, in turn, took stop-smoking
medication more often. Another study investigated whether
iOS and Android users differ in personality traits [23]. Only
a few and small differences were found. Altogether, research
on mobile crowdsensing based differences between Android
and iOS users is still in its infancy.
Regarding the third category, approaches exist that deal
with mobile technology in the context of hearing loss [24],
[25]. However, these works do not directly measure param-
eters on the hearing loss, they rather measure parameters
that may negatively affect the onset of a hearing loss (e.g.,
through a loud environment). In the context of EMA and
mobile technology in general, other works exist that have
shown its usefulness for healthcare questions on hearing
loss [26]. Finally, systematic literature reviews show that
many mobile apps beside mobile crowdsensing exist that
address a hearing loss [27], [28]. However, to the best of
our knowledge, none of these works have compared Android
and iOS users as shown in the work at hand.
III. TRACKYOURHEARING PLATFORM
Mobile crowdsensing is characterized by the following
aspects. Different sensing paradigms are utilized to relate
crowd users to sensing tasks on one hand [29]. On the
other, a sophisticated crowdsensing platform must be de-
veloped to enable measurements by crowd users. In the
context of mHealth questions, three technical components
are particularly necessary to provide a proper crowdsensing
platform. First, a proper data model including a flexible API
to handle the data exchange must be developed [8]. Second,
an architecture must be defined that reflects the needs for the
collection procedure [11]. Finally, mobile apps must be real-
ized that enable a collection procedure that is welcome by the
users [30]. These aspects are considered by the TrackYour-
Hearing (TYH; https://www.trackyourhearing.org) mHealth
crowdsensing platform, which is built on four technical
components. First, it offers a website for user registration
and other user-related features (e.g., account management).
Second, it offers an Android and iOS application. Third,
a MySQL database is used for the central repository for
the data collected. Fourth, a RESTful API is provided that
enables the communication between the mobile applications,
the website, and the database [8].
In general, TYH was developed to collect EMA of in-
dividuals with a hearing loss. On the one hand, EMA is
based on a set of electronic questionnaires, which are repeat-
edly (registration, daily) administered to the users on their
smartphones. On the other hand, EMA is based on sensor-
based measurements of the environmental sound level. Yet,
the measurement of the environmental sound level is only
realized if users actively give consent for this measurement
when registering to TYH for the first time. In doing so, we
consider the privacy of the users. In general, TYH users
accomplish three fundamental phases. First, they have to
register through the website or the mobile apps. Second,
users have to fill in three so-called registration questionnaires
once. The latter capture the current hearing loss situation,
demographic data (e.g., birthday), and other hearing loss re-
lated parameters. The completion of these registration ques-
tionnaires is compelling for the users who want to use the
features of the continuous mobile crowdsensing procedure
(i.e., the daily measurements). Also, during the second phase,
users have to accept or adjust a notification schema. The
notification schema determines how often and in what way
(i.e., fixed or random points in time) the daily assessment
questionnaire is applied. The number of daily assessments,
in turn, is restricted to 12 times per day. Third, after the
registration questionnaires have been accomplished and the
notification schema is determined, users can start with the
daily assessments. For the application of the daily assessment
questionnaire, notification features for both Android and iOS,
as well as a notification algorithm, were realized. After a
notification appears, the user may click on it. In the latter
case, the mobile application is started (if not already running)
and the daily assessment questionnaire is directly displayed
to the user for completion. Note that users can also fill in
the questionnaires without a notification whenever they want.
While filling out this questionnaire, the environmental sound
level is measured in users who agreed to this. The result is
then either transferred through the API to the database or
locally cached if the device is offline after completing the
questionnaire. A more detailed technical description of the
presented features can be found in [7], [8], [31].
IV. MATERIAL AND METHODS
This section provides materials and methods that were
used for analyses. Note that TYH is currently only available
in German, while English and French versions are under
development. The mobile apps have been officially released
to the Google Playstore and the Apple App Store. So
far, users that have been registered to TYH mainly stem
from Switzerland 54%, Germany 22%, and the USA 6%.
For the presented analysis, results of the daily assessment
questionnaires were mainly used, for which the questions
are shown in Table II. It is noteworthy that three types of
user interface elements were used to give answers: sliders to
enable visual analogue scales (VAS), yes/no questions as well
as self-assessment manikins (SAM; [32]). In the first data
preparation step, all test users were removed, resulting in 84
users with 3767 filled out daily assessment questionnaires,
with an inter-assessment interval of at least 15 minutes. In
addition, we analyzed one of the registration questionnaires,
which comprises five questions on gender, handedness, age,
and whether a patient wears a hearing aid, and whether a
patient has actually a hearing loss. This questionnaire was
filled out by all of the 84 analyzed users.
V. STATISTICS
All statistical analyses were performed with SPSS 25.
Means (M) and standard deviations (SD) were calculated as
descriptive statistics. Assessments with an inter-assessment
interval <15 minutes were deleted. Android and iOS users
were compared with Fishers Exact Tests (FET), t-tests for
independent samples, and multilevel models. To analyze the
Question Scale
1
Do you wear your hearing aid right now? Y/N
2
To what extent do you perceive your hearing loss right now? VAS
3
To what extent are you limited in your daily life by your hearing
loss right now?
VAS
4
Do you feel emotionally charged by your hearing loss right now? VAS
5
How is your mood right now? SAM
6
Do you feel stressed right now? VAS
7
Do you feel irritable right now? VAS
8
Do you feel exhausted right now? VAS
9
To what extent were conversations of the last hours burdensome? VAS
9b
Regarding the latter question, if you had no conversations, please
indicate this?
Y/N
10
Do you physically perceive ambient noises negatively right now? VAS
VAS=Visual Analogue Scale, Y/N=Yes/No-Question
SAM=Self-Assessment Manikins
TABLE II: TrackYourHearing Daily Assessment Questions
items of the registration questionnaire, which was provided
once to the users, FET and t-tests were used. T-tests were
also used to analyze the following variables: days between
first and last daily assessment questionnaire within a user,
amount of daily assessment questionnaires per user, and
hours between each of the daily assessment questionnaire
across users. Linear multilevel models with two levels were
used for the items of the daily assessment questionnaire
with a numeric rating scale (i.e., Questions 2-10, except
Question 9b). These repeated daily assessments are nested
within users, so that assessments were level-1 and users
level-2 in the multilevel models. The linear multilevel models
were performed with the full maximum likelihood estimation
and a random intercept. For the binary Questions 1 and
9b, which show again the mentioned nested data structure,
multilevel models for dichotomous outcomes, i.e., binary
logistic models within the Generalized Estimating Equations
(GEE) were performed. In all multilevel models, the main
effect of the mobile operating system (Android vs. iOS with
Android being the reference coded as 0 and iOS coded
as 1) was evaluated and the scores of the items of the
daily assessment questionnaires functioned as the dependent
variables. All statistical tests were performed two-tailed. The
significance value was set to p<.05.
VI. RESULTS
In total, N= 84 users participated in the analysis. All
users filled in the registration questionnaire. In total, the users
3767 daily assessment questionnaires. Of the N= 84 users,
n= 37 used iOS, while n= 47 used Android. The Android
users provided 2450 daily assessment questionnaires and the
iOS users 1317. Table III presents the results of the compar-
ison between Android and iOS users with t-tests and FET,
which were performed to compare the users in the registra-
tion questionnaire items. No significant differences between
Android and iOS emerged. Table IV shows the results of the
multilevel models comparing the Android and iOS users in
their scores of the daily assessment questionnaires. In three
questions, iOS users scored significantly higher (all p<.05)
than Android users: Questions 7 (irritable), 8 (exhausted),
and 10 (ambient noises negative). Scores on Question 1
Android iOS Statistics
Gender male 25 (53.2%) 18 (48.6%) FET: p=.826
female 22 (46.8%) 19 (51.4%)
Hearing Ability
no problem 16 (34.0%) 14 (37.8%)
FET: p=.875
problem in both ears 27 (57.4%) 21 (56.8%)
problem in right ear 1 (2.1%) 1 (2.7%)
problem in left ear 3 (6.4%) 1 (2.7%)
Hearing Aid no 37 (78.7%) 23 (62.2%) FET: p=.144
yes 10 (21.3%) 14 (37.8%)
Handedness
ambidextrous 0 (0%) 1 (2.7%)
FET: p=.207left 1 (2.1%) 3 (8.1%)
right 46 (97.9%) 33 (89.2%)
Days between first and last daily
assessment questionnaire within a user M (SD) 26.26 (38.68) 20.09 (20.23) T(82)=.88; p=.382
Age M (SD) 67.84 (12.55) 67.06 (9.32) T(81)=.31; p=.755
Amount of daily assessment questionnaires per user M (SD) 52.13 (67.64) 35.59 (31.00) T(82)=1.38; p=.173
Hours between each of the daily assessment questionnaires
across users M (SD) 11.83 (78.68) 13.46 (53.00) T(3681)=-.67; p=.506
TABLE III: Results of the Fishers Exact Tests and t-tests
Android
(Intercept)
iOS
(Alteration compared to Android)
Question 2 Estimate (SE) .11 (.02) .04 (.02)
Statistics T(77.56)=6.67; p<.001 T(79.10)=1.66; p=.101
Question 3 Estimate (SE) .11 (.02) .04 (.03)
Statistics T(76.96)=6.42; p<.001 T(78.09)=1.57; p=.120
Question 4 Estimate (SE) .10 (.02) .04 (.03)
Statistics T(75.50)=5.87; p<.001 T(77.40)=1.60; p=.114
Question 5 Estimate (SE) .79 (.02) -.05 (.03)
Statistics T(75.13)=39.26; p<.001 T(77.33)=-1.67; p=.100
Question 6 Estimate (SE) .11 (.02) .03 (.02)
Statistics T(77.90)=7.45; p<.001 T(80.13)=1.11; p=.269
Question 7 Estimate (SE) .13 (.02) .61 (.03)
Statistics T(76.33)=5.52; p<.001 T(79.23)=17.72; p<.001
Question 8 Estimate (SE) .15 (.02) .08 (.03)
Statistics T(70.18)=6.96; p<.001 T(72.08)=2.42; p=.018
Question 9a Estimate (SE) .15 (.02) .04 (.03)
Statistics T(78.76)=7.43; p<.001 T(80.41)=1.20; p=.235
Question 10 Estimate (SE) .11 (.02) .07 (.03)
Statistics T(77.27)=6.59; p<.001 T(79.75)=2.61; p=.011
TABLE IV: Results of the Linear Multilevel Models
were not significantly different for Android and iOS users
(p=.489), as were scores for Question 9b (p=.779).
VII. DISCUSSION
Compared to our analysis on TYT and differences between
Android and iOS users [9], this work has revealed different
results. First of all, the registration questionnaire that was
used for TYH showed no difference for Android and iOS
users. For the variable age, significant differences were
obtained for TYT, but not for TYH. As we have much less
daily assessments gathered by TYH than TYT users, we
firstly had a look at the daily assessment questionnaire of
TYH and whether there exist differences before analyzing
the other two registration questionnaires. Table IV shows
that there are significant differences for three questions of
the daily questionnaire between Android and iOS users.
Therefore, it is worth to investigate the other registration
questionnaires of TYH. However, it must be considered that
the amount of assessments for TYH is sparse compared to
TYT and a larger sample must be reconsidered for TYH.
For TYT, in turn, differences between Android and iOS
users on the daily assessment questionnaire have not been
investigated so far [9]. Another observation that was new
compared to the study on TYT is to have a look at the
amount of assessments users provided per mobile operating
system on average. As can be obtained from Table 1,
Android users provide apparently more daily assessments
(i.e., daily assessment) than iOS users do, although there
is no significant difference ascertainable. Referring to the
two questions raised in the introduction, we can conclude
as follows: Concerning Question 1, we found significant
differences of Android and iOS users for TYH based on their
given EMA-data. Concerning Question 2, this study found
differences between Android and iOS users, but not the same
differences as the previous study on TYT [9].
VIII. SUMMARY AND OUTLOOK
Although this work on TYH has shown different results to
the previous study on TYT, an investigation with the goal to
reveal medically relevant differences between Android and
iOS users based on EMA-data gathered by a crowdsensing
platform should be generally taken into account. In particular,
it must be further worked on the question whether the used
mobile operating system can reveal new insights and whether
it should be considered as a potential confounder in future
studies. However, although TYH has a smaller database
compared to TYT, again differences between Android and
iOS users have been obtained. Therefore, we currently focus
our research on various different directions. One direction
constitutes the application of machine learning techniques to
see whether we are able to predict the used mobile operating
system based on daily EMA-data. Another direction is an in-
depth investigation on how the user interfaces of the different
mobile operating systems may bias the user when filling out
EMA questionnaires. Altogether, differences of Android and
iOS users in the context of mHealth and chronic disorders
can be considered as a promising target in general.
REFERENCES
[1] W. Schlee et al., “Measuring the moment-to-moment variability of
tinnitus: the trackyourtinnitus smart phone app,” Frontiers in Aging
Neuroscience, vol. 8, 2016.
[2] T. Probst et al., “Emotional States as Mediators between Tinnitus
Loudness and Tinnitus Distress in Daily Life: Results from the
“TrackYourTinnitus” application,” Scientific Reports, vol. 6, 2016.
[3] A. Vasilateanu, I. Radu, and A. Buga, “Environment crowd-sensing
for asthma management,” in E-Health and Bioengineering Conference.
IEEE, 2015, pp. 1–4.
[4] T. Trull and U. Ebner-Priemer, “Ambulatory assessment,” Annual
review of clinical psychology, vol. 9, pp. 151–176, 2013.
[5] S. Engel et al., “Ecological momentary assessment in eating disorder
and obesity research: a review of the recent literature,” Current
psychiatry reports, vol. 18, no. 4, pp. 1–9, 2016.
[6] R. Pryss et al., “Mobile Crowd Sensing Services for Tinnitus As-
sessment, Therapy and Research,” in IEEE 4th Int’l Conf on Mobile
Services. IEEE Computer Society Press, 2015.
[7] ——, “Prospective crowdsensing versus retrospective ratings of tin-
nitus variability and tinnitus–stress associations based on the tracky-
ourtinnitus mobile platform,” International Journal of Data Science
and Analytics, pp. 1–12, 2018.
[8] ——, “Requirements for a flexible and generic api enabling mobile
crowdsensing mhealth applications,” in 4th Int’l WS on Requirements
Engineering for Self-Adaptive, Collaborative, and Cyber Physical
Systems. IEEE, 2018, pp. 24–31.
[9] ——, “Differences between Android and iOS Users of the TrackY-
ourTinnitus Mobile Crowdsensing mHealth Platform,” in Proc. 31th
IEEE International Symposium on Computer-Based Medical Systems.
IEEE Computer Society Press, 2018.
[10] T. Vos et al., “Global, regional, and national incidence, prevalence,
and years lived with disability for 328 diseases and injuries for 195
countries, 1990–2016: a systematic analysis for the global burden of
disease study 2016,” The Lancet, vol. 390, no. 10100, pp. 1211–1259,
2017.
[11] H. Xiong et al., “Sensus: a cross-platform, general-purpose system
for mobile crowdsensing in human-subject studies,” in Proc of the
Int’l Conf on Pervasive and Ubiquitous Computing. ACM, 2016, pp.
415–426.
[12] T. Probst et al., “Emotion dynamics and tinnitus: Daily life data from
the “trackyourtinnitus” application,” Scientific reports, vol. 6, 2016.
[13] U. Ebner-Priemer and T. Kubiak, “Psychological and psychophysio-
logical ambulatory monitoring,” European Journal of Psychological
Assessment, vol. 23, no. 4, pp. 214–226, 2007.
[14] D. De Beurs et al., “The role of mobile phone technology in under-
standing and preventing suicidal behavior,” 2015.
[15] T. Dehling et al., “Exploring the Far Side of Mobile Health: Informa-
tion Security and Privacy of Mobile Health Apps on iOS and Android,”
JMIR mHealth and uHealth, vol. 3, no. 1, 2015.
[16] K. Huckvale et al., “Unaddressed Privacy Risks in Accredited Health
and Wellness Apps: A Cross-Sectional Systematic Assessment,” BMC
medicine, vol. 13, no. 1, p. 214, 2015.
[17] B. Zapata et al., “Mobile PHRs Compliance with Android and iOS
Usability Guidelines,” Journal of Medical Systems, vol. 38, no. 8,
p. 81, 2014.
[18] A. Azfar et al., “Forensic Taxonomy of Popular Android mHealth
Apps,” arXiv preprint arXiv:1505.02905, 2015.
[19] H. Seabrook et al., “Medical Applications: a Database and Character-
ization of Apps in Apple iOS and Android Platforms,” BMC research
notes, vol. 7, no. 1, p. 573, 2014.
[20] B. Zapata et al., “Empirical Studies on Usability of mHealth Apps: A
Systematic Literature Review,” Journal of medical systems, vol. 39,
no. 2, p. 1, 2015.
[21] Z. Benenson et al., “Android and iOS Users’ Differences Concerning
Security and Privacy,” in Extended Abstracts on Human Factors in
Computing Systems. ACM, 2013, pp. 817–822.
[22] H. Ubhi et al., “A comparison of the characteristics of ios and android
users of a smoking cessation app,” Translational Behavioral Medicine,
vol. 7, no. 2, pp. 166–171, 2017.
[23] F. G¨
otz et al., “Users of the main smartphone operating systems (iOS,
Android) differ only little in personality,” PloS one, vol. 12, no. 5, pp.
1–18, 2017.
[24] M. Zappatore et al., “Mobile crowd sensing-based noise monitoring as
a way to improve learning quality on acoustics,” in 2015 International
Conference on Interactive Mobile Communication Technologies and
Learning (IMCL). IEEE, 2015, pp. 96–100.
[25] Z. Qin and Y. Zhu, “Noisesense: A crowd sensing system for urban
noise mapping service,” in 22nd International Conference on Parallel
and Distributed Systems. IEEE, 2016, pp. 80–87.
[26] G. Galvez et al., “Feasibility of ecological momentary assessment
of hearing difficulties encountered by hearing aid users,” Ear and
hearing, vol. 33, no. 4, p. 497, 2012.
[27] B. Mart´
ınez-P´
erez, “Mobile health applications for the most prevalent
conditions by the world health organization: review and analysis,”
Journal of medical Internet research, vol. 15, no. 6, 2013.
[28] T. Bright and D. Pallawela, “Validated smartphone-based apps for ear
and hearing assessments: a review,” JMIR Rehabilitation and Assistive
Technologies, vol. 3, no. 2, 2016.
[29] H. Ma et al., “Opportunities in mobile crowd sensing,” IEEE Com-
munications Magazine, vol. 52, no. 8, pp. 29–35, 2014.
[30] K. Agrawal et al., “Towards incentive management mechanisms in
the context of crowdsensing technologies based on trackyourtinnitus
insights,” Procedia computer science, vol. 134, pp. 145–152, 2018.
[31] R. Pryss et al., “Mobile Crowdsensing Services for Tinnitus Assess-
ment and Patient Feedback,” in IEEE Int’l Conf on AI & Mobile
Services. IEEE, 2017, pp. 22–29.
[32] M. Bradley and P. Lang, “Measuring Emotion: The Self-Assessment
Manikin and the Semantic Differential,” Journal of Behavior Therapy
and Experimental Psychiatry, vol. 25, no. 1, pp. 49–59, 1994.
