A personalized support tool for the training of mindful walking: The mobile “MindfulWalk” application [original]

Universität Ulm | 89069 Ulm | Germany Fakultät für

Ingenieurwissenschaften,

Informatik und

Psychologie

Institut für Datenbanken

und Informationssysteme

A personalized support tool for

the training of mindful walking:

The mobile “MindfulWalk” application

Masterarbeit an der Universität Ulm

Vorgelegt von:

Julian Frank

[email protected]

Gutachter:

Prof. Dr. Manfred Reichert

Dr. Rüdiger Pryss

Betreuer:

Dr. Rüdiger Pryss

2017

Fassung November 23, 2017

2017 Julian Frank

This work is licensed under the Creative Commons. Attribution-NonCommercial-ShareAlike 3.0

License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/de/

or send a letter to Creative Commons, 543 Howard Street, 5th Floor, San Francisco, California,

94105, USA.

Satz: PDF-L

TEX2ε

Abstract

Digital health prevention is a trend that becomes increasingly important in various do-

mains. Health insurers crave for effective methods that can be offered to their customers.

Moreover, smart mobile devices pose many advantages as they can be easily used in ev-

eryday life without being burdensome. Taking these advantages into account, completely

new applications become possible. This thesis presents an application that is intended to

support users to walk mindfully. It is a mobile personalized tool that senses the walking

speed and provides haptic feedback thereof. The procedure of mindful walking, the

technical prototype as well as preliminary study results are presented and discussed.

The reported user experience and the study result indicate promising perspectives for

a tool that supports a mindful walking behavior. Altogether, the use of modern smart

mobile device sensors paves the way for useful mobile application in the context of health

prevention in particular and health care in general.

iii

Contents

1 Introduction 1

1.1 Motivation.................................... 1

1.2 Relatedwork .................................. 3

1.2.1 Computer-supported mindfulness . . . . . . . . . . . . . . . . . . 3

1.2.2 Smartphone-supported mindfulness . . . . . . . . . . . . . . . . . 4

1.3 Approach .................................... 5

1.4 Structure .................................... 6

2 Implementation 7

2.1 Requirementsanalysis............................. 8

2.1.1 Functional requirements . . . . . . . . . . . . . . . . . . . . . . . . 8

2.1.2 Non-functional requirements . . . . . . . . . . . . . . . . . . . . . 10

2.2 Concept..................................... 11

2.3 Implementation................................. 14

2.3.1 iPhoneApp............................... 16

2.3.2 WatchApp ............................... 16

2.4 Validation and Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

2.4.1 Functional requirements . . . . . . . . . . . . . . . . . . . . . . . . 18

2.4.2 Non-functional requirements . . . . . . . . . . . . . . . . . . . . . 20

2.4.3 Limitations ............................... 22

3 Study 23

3.1 Method ..................................... 23

3.1.1 Participants and design . . . . . . . . . . . . . . . . . . . . . . . . 23

3.1.2 Procedure................................ 23

3.2 Results and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

4 Summary and Outlook 31

Introduction

Mindfulness has become the focus of interest in psychology and medicine more and more

[

Mar12

]. Chronic diseases are often treated with mindfulness-based therapies [

CBM+12

SES+12

] and the state of mindfulness is significantly related to several indicators of

psychological and physical health [

CBM+12

SES+12

KSR11

]. Mindfulness is a skill

that can be obtained by regular training while intentionally and non-judgmentally paying

attention to the present moment. Mindfulness can be exercised by anyone to reduce

stress (mindful-based stress reduction; MBSR), calm anxiety, sleep better, improve focus

and more. Due to its growing popularity, many mobile applications emerged to support

and track mindful exercises. The focus of a mindful exercise can be any experience

of the present moment like objects (e.g., chair, plant, etc.), environmental influences

(listening) or body-related processes (breathing, walking). Mindful walking as proposed

by Thich Nhat Hahn [

Han

] has the advantage of easy integration into our every-day life.

Everyone walks every day. Walking is usually just a way of getting from one place to

another. Therefore mindful walking has the potential of giving everyone daily access

to mindfulness practice without having to learn new techniques. A mobile application

can remind people on a regular basis to walk mindfully and to observe and track the

exercise, giving real-time interactive feedback using sensor measurement.

1.1 Motivation

We live in a very busy world, the pace of life is often frantic and we’re always doing

something. The mind doesn’t get any rest since it’s always occupied. But we rely on our

mind to be happy, content, emotionally stable as individuals and also kind thoughtful and

1 Introduction

considerate with others. We need our mind to stay focused, creative, spontaneous and

to perform at our best. But instead, we get stressed with all the thoughts and difficult

emotions going round that we don’t know how to deal with and the result of that is that

we are not longer present in the world we live in; we miss out on the things that are most

important to us.

The goal of mindfulness is the appreciation and experience of the present moment in

order to get less distracted and not lost in thought by stepping back and looking at

thoughts and emotions from a different, non-judgmental point of view.

Mindfulness can be described as the intentional and non-judgmental attention to ex-

periences of the present moment [

KZ90

]. Mindfulness is rooted in eastern meditation

traditions and has become highly relevant in basic research [

THP15

] as well as in health

[

KSRF15

], and clinical science [

KLF+13

]. The body is the focus of many mindfulness

exercises applied for example in Mindfulness-based Stress Reduction (MBSR) [

KZ90

]

and Mindfulness-based Cognitive Therapy (MBCT; [

SWT02a

]). In these exercises, the

participants are trained to intentionally and non-judgmentally attend to automated body-

related processes (e. g. breathing, walking). The aim of these exercises is to gain the

ability to flexibly switch from the “doing mode” into the “being mode of mind”, which is

characterized by the concrete and accepting perception of moment-by-moment experi-

ence and allows disengagement from dysfunctional cognitive processes [

Wil08

]. The

ability to disengage in turn is considered to reduce depressive symptoms. One of these

body-related mindfulness exercises is walking mindfully [

Han

]. The participants are

instructed to become conscious of the otherwise automated walking process by walking

slowly and taking small steps. Mindful walking has, for example, been shown to be benefi-

cial for patients with depression [

PSTS14

] as well as for patients with diabetes [

GHTS16

Mindful walking can be exercised anytime by anyone. Hanh even proposed to walk

mindfully all the time, anywhere we go [

Han

]. All one has to do is walk slowly and notice

each step. Explicitly attend to breathing and walking and nothing else. It takes a little

practice and some may find it hard to integrate such mindful exercises in today’s hectic

world. Also, participants often experience that their mind wanders away from being

1.2 Related work

mindful and they have difficulties to redirect their focus to the mindful exercise, especially

patients suffering from depression [

RADM14

]. Moreover, specifically novices are often

insecure on whether they practice mindfulness the way it is intended. Applications pro-

viding support in practicing mindfulness may help to stay mindful and in refocusing the

attention during mindful exercises, in turn, might be helpful to address these problems.

Almost everyone has a smartphone and it can help people to walk mindfully. A mobile

application could manage user-created schedules and notify the user to start a mindful

walk training. It could support the user during the training by ensuring the user isn’t

walking too fast so that he is actually walking mindfully. As a bonus, a mobile application

could measure and collect walking data during the exercise to let the user review past

exercises and monitor progress. Since mindful walking has to be trained, an app could

help inexperienced users to get into mindful walking, to stay focused during the exercise

and to train mindful walking on a regular basis. Since mindfulness isn’t only beneficial for

patients with acute illnesses, a mobile application could also attract younger individuals

which may be hard to get in contact with from a clinical setting [PPL+17b].

The hereby developed underlying mechanism enables mindful walking to be integrated

with a mobile crowdsensing service like TrackYourTinnitus [

PPS+17

PSLR17

]. This way

longitudinal data could be collected under real-life conditions using ecological momentary

assessment [

SPP+16

] to observe a variety of other factors which may affect test results

and are not covered in this work like e.g., time-of-day dependency [PPL+17a].

1.2 Related work

1.2.1 Computer-supported mindfulness

Most studies of computer-supported mindfulness involved mindfulness as a component

of a broader therapeutic intervention. Also, most of these studies only presented mind-

fulness techniques without providing interactive practices. The content was presented

using videoconferencing [

GNBBG08

] or web pages, sometimes enriched with audio or

video [

AK04

EABP08

LFV+10

TWO+10

KCKW12

]. Some researchers utilized web-

1 Introduction

pages on smartphones [

KFE+11

NDE+12

] or actual smartphone applications [

MKL+10

Glück and Marcker [

GM11

] investigated an interactive web-based mindfulness training

for thought distancing where participants imaginarily labeled clouds with distressing

thoughts or sensations on a screen with a blue sky to watch the clouds move across

the sky and out of sight. The application was designed to support affect labeling and

thought distancing but it basically just moved a cloud across the screen when hitting the

spacebar. The participants had to do all the work with labeling distressing thoughts and

associating the labels with the moving clouds. The experiment of [

BEC+12

] included a

mindfulness technique in a virtual environment for mood induction. Participants reported

low levels of difficulty to use and also high levels of satisfaction although the technique

didn’t provide an interactive component. Another study utilizing a virtual environment

was conducted by [

SGS07

] of the Meditation Chamber where participants had to focus

on their breath while an interactive feedback based on the user’s skin conductance,

respiration, and blood volume pulse was projected into the VE in form of an abstract

video.

1.2.2 Smartphone-supported mindfulness

There are countless mindfulness-based mobile applications (MBMAs) on the market.

Studies have shown that most of these apps are simply guided meditations, reminders,

and timers but most importantly the almost complete lack of evidence of the efficiency

of those applications [

MKHS15

PDHM+13

]. A sample search among the top-rated

MBMAs on the App Store during the course of this study has confirmed that these results

are still valid. All tested apps (Mindfulness [

Min

], Headspace [

lim

], ZMeditations [

For

SimpleHabit [

], 7Mind [

7Mi

] and Buddhify [

Eve

]) provide audio-guided meditations

and most of them offer reminders (Mindfulness, Headspace, SimpleHabit, 7Mind and

Buddhify). Some apps show mindfulness exercises based on goals such as sleep

better, reduce stress, calm anxiety or improve focus (Headspace, SimpleHabit, Buddhify)

but still, all exercises consist only of an audio track with no interactive feedback or

measurement. Only one tested app provides an optional Watch app to measure the

heart rate before and after an exercise (Mindfulness) and one application has a self-

1.3 Approach

assessment rating of mindfulness, concentration, and balance at the end of an exercise

(Buddifhy). Besides the limited feedback and interactivity, many apps only provide a

basic set of meditations. All other meditations or sets of meditations are chargeable per

exercise or subscription plan. Across all tested MBMAs in this and other studies, there

is almost a complete lack of interactive feedback and sensor measurement as well as

evidence of usefulness.

During the research for this study, only two interactive mindful walking apps have been

found. The Breathewalk-aware system [

YWLH12

] uses a complex net of sensors to

raise the user’s awareness of walking and breathing behaviors by providing multimedia

guidance on the smartphone.

The AmbientWalk App also takes breathing into account and generates a real-time

ambient sound based on the user’s breath and walking pace.

Both apps are research prototypes and not targeted at end users. Also, both apps

provide a quite intrusive audio-visual feedback to the user which can disturb the actual

mindful exercise.

1.3 Approach

The practice of mindfulness can be difficult for people with no or minimal experience

with meditation [

KZ05

SWT02b

]. A supportive tool must be easy to understand and use,

otherwise “naive meditators” will be discouraged to start practicing and will eventually

abandon practicing. Mindful walking is all about walking and intentionally attending to

the automated process of walking. This exercise should not be disturbed by attending

to a phone. To make a supportive tool for mindful walking as little intrusive as possible

the mindful walk app is split onto a smartphone and a smartwatch. All interactions can

take place on the phone benefiting from the larger screen while during the actual mindful

walk exercise nothing more than the watch has to be carried. The watch doesn’t need to

be attended to during the entire exercise so one can fully attend to walking.

Mindful walking can be best exercised when walking slowly so one can consciously

perceive every single motion of the walking process. The MindfulWalk app sets a target

speed for the mindful walking exercise which is slower than the “normal” walking speed

1 Introduction

by a selectable percentage. This way everyone can select an individual walking speed

for the exercise but it will always be

slower

than the normal walking speed. In order not

to exceed the target speed during the exercise the application must provide a gentle

non-intrusive feedback. The smartwatch can perform various haptic feedbacks from

a slight tap on the wrist up to vibrating so the user can keep attending to walking and

doesn’t have to look at a screen.

Since emotional states, as well as emotion dynamics, can mediate the effects of the

exercise [

PPLS16b

PPLS16a

], the experienced emotions on the dimensions

affective

valence, dominance

and

arousal

are also captured before and after the exercise using

the Self-Assessment Manikin [BL94].

1.4 Structure

The following chapters present how the prototype app was developed, how the pilot study

was designed and conducted and finally discuss the study results and give an outlook

how the app could be developed further. Chapter 2 focuses on the implementation of

the prototype, where first all requirements are analyzed and structured (cf, Sect. 2.1)

and then a concept is presented how to realize the latter (cf. Sect. 2.2). In the following

Section details of the implementation of both the iPhone App and the Apple Watch

Extension are presented. Finally the defined requirements are validated against the

finished prototype (cf. Sect. 2.4).

In Chapter 4 the details of the pilot study are described which was conducted to examine

the usefulness of the app. Section 3.1 explains the method used whereas in Section 3.2

the results are evaluated.

Finally, a summary and outlook is given.

Implementation

In order to conduct the study, a mobile application was developed. This application

consists of two parts. A smartwatch app and a smartphone app. This way the user can

comfortably interact with the questionnaires on the phone on a larger screen but doesn’t

need to carry the phone or look at it while walking since he gets feedback over the watch

on his wrist.

This chapter comprises the applications requirements, the concept to realize them

and a summary of the actual implementation followed by a concluding analysis of the

final app regarding the requirements but also limitations.

2 Implementation

2.1 Requirements analysis

2.1.1 Functional requirements

For a simple and easy-to-learn usage, the app will be divided into three sections, which

will be available in the main menu. These sections are:

1. General information

2. Schedules

3. Start

Each section must provide certain functionalities, which will be defined in the following.

For later on reference, the single requirements are labeled with (FR#1).

General information

FR#1.1 Display App information:

The user can view general information on the app

itself as well as instructions on how to use it. The instructions should give an

overview of the functionalities of the app and cover every use-case. The wording

should and be easily understandable for all user groups.

FR#1.2 Display General information:

This section serves general information on

mindful walking. What it is, why the user should train it and how to train it using

this app.

Schedules

FR#2.1 Create new schedule:

The app should be able to notify the user to train mind-

ful walking. The time and frequency of these notifications can be managed by the

users through schedules. When creating a new notification schedule the user can

set the day, time and interval (daily, weekly, every X hours) of the notification. The

user can create infinite schedules.

2.1 Requirements analysis

FR#2.2 Edit schedules:

Existing schedules can be edited and deleted. In the edit-

mode, the user can tweak the same variables as in the create-mode.

Start

This section is where the actual mindful walking training starts. The training consists of

several steps and each step has to meet certain functionalities.

FR#3.1 Mood evaluation:

Using a self-assessment manikin[

BL94

] the user can rate

his/her mood.

FR#3.2 Normal walking speed measurement:

To measure the “normal” walking speed,

the user is instructed to walk normally until he hears a feedback sound to indicate

that the measurement is successful.

FR#3.3 Save data:

The collected data (mood and normal walking speed) should be

saved to a database locally as well as remotely along with a user ID to identify the

participant and a timestamp.

FR#3.4 Choose target speed:

Before the training starts, the user can select a target

walking speed based on the measured normal walking speed. The choice consists

of 10, 25 or 50% reduction of the normal walking speed.

FR#3.5 Save target speed:

The chosen target speed should be saved to the database

as well. Additional parameters like user ID and timestamp should be saved too, to

correlate the datasets later.

FR#3.6 General instruction:

Before the actual mindful walking training, general in-

structions on how to walk mindfully should be displayed.

FR#3.7 Mindful walking:

During the mindful walk training the app should constantly

measure the current walking speed and notify the user whenever the current

walking speed exceeds the target speed. The training should last at least 5 minutes

and the app should notify the user when 5 minutes have passed but the user can

cancel the training at any point, before or after the 5-minute recommendation.

2 Implementation

FR#3.8 Finish training:

When the user finishes the training the collected data should

be saved to the local and remote databases. The dataset should include: start

time, finish time, feedback frequency, feedback timestamps and average walking

speed.

2.1.2 Non-functional requirements

The non-functional requirements specify the quality and performance expected from the

app as well as the error handling. A comprehensive documentation of the implementation

of requirements and a good feasibility and flexibility.

NFR#1 - Quality

The Application should be

correct

. It may not show false information and all functionalities

should work correctly and as expected. It should be

reliable

and still work under

adverse conditions.

User-friendliness

is a major part of the quality including

consistency

regarding the design, language, and navigation. Well-know navigation- and design

concepts of mobile applications should be preferred. The user should easily reach

all functionalities and shouldn’t be disturbed or even restricted by the app. It should

rather be seen as a tool to conduct the mindful walking study. In terms of

security

the

app must collect all data anonymously. The privacy of the user must be preserved

and the participants should be informed about all data the app is saving. Data - user

matching only happens by a random and anonymous user ID. Last but not least the

application must be

robust

and should not crash after wrong or unexpected user input

or measurement data. This also affects the data upload to the remote database. If no

network connection is available the data is only saved locally and sent to the server as

soon as a connection is possible.

2.2 Concept

NFR#2 – Performance

Essential performance features are

response time behavior

computing expense

and

data throughput

. This means the app should quickly respond to user input and not freeze.

Computationally expensive operations should be avoided to keep down CPU usage and

thereby save battery life. The data throughput through network transactions should be

kept low by reducing the saved datasets to the essential parameters to save the user’s

data allowance. Also, weak network connections can lead to longer responding times if

the data packages are too big.

NFR#3 - Error handling

In case of an error or application failure, understandable and usable error messages

should be given to the user, for example in form of a dialog. Common errors in this

context are database and network errors but also measurement errors because of

hardware (e.g. gyroscope) disturbances.

NFR#4 - Documentation and Flexibility

Good software design includes appropriate flexibility, expandability, and feasibility. By a

modular design, the single components of the application can be delimited and handle

distinguished tasks (cohesion) while relying on each other as little as possible (coupling).

This supports easy editing of individual components.

Last but not least the compliance of the defined requirements has to be traceably

documented in the implementation.

2.2 Concept

Even though the main purpose of the app was to conduct the study, it is designed to be

easily extendable to make it usable for everyone in everyday use. In this case, it would

2 Implementation

be a personal tool to measure mindful walking exercises and to review his own data. For

the study though, all subjects will be using the same app on the same hardware. This

has an impact on the underlying data structure. Instead of one user with many exercises

there are many users with one exercise each. To make the app usable for the study but

also easily alterable with little effort the data model in Fig. 2.1 was developed.

Figure 2.1: Data Model

This way every subject has a distinct user for the study and the app can be altered for a

single user by omitting the user creation and creating a singleton user instead.

All training relevant data is stored in the training object. The single measurements during

the exercise are referenced in another table since there can be many of them for each

training.

The user itself is handled anonymously and is only identifiable by a randomly generated

string for the sake of the study. The user object can easily be extended with personal

information if needed. Not shown in the data model are the schedules, which are similar

to the training objects referenced by the user. They are left out in this figure since they

are not necessary for the study.

To make the application as easy to use for the subjects there is a big “Start”-Button on

the home screen of the app which directly creates a new user and starts a new training.

For the training itself, the user switches between phone and watch. The training starts

2.2 Concept

with the first self-assessment manikin on the phone after which the user is instructed to

walk normally for a while to determine his normal walking speed. To minimize false data

of the user standing, the user himself must start the measurement when he is ready to

walk by tapping a button on the watch. When the normal walking speed is determined

the user is notified on both the phone and the watch and the phone app automatically

switches to the next screen where the user can then select his target speed and duration

of the training.

After this initialization process, the user himself can start (and stop) the training by

tapping the button on the watch. The phone is not needed during training since the

measurement happens on the watch. During the training, the user can observe the

current duration and speed. The speed indicator doesn’t show the actual walking speed

of the user since this is not particularly of interest and could even be distractive. Instead,

it shows the current speed in relation to the target speed on a numberless speedometer

to give a hint on how much the user is walking too fast. When the target duration is

reached or the user is walking too fast he is notified by a slight vibration of the watch so

the user doesn’t even have to look at the watch once and can fully concentrate on the

mindful exercise.

When the user finishes the training by pressing the button on the watch, the mea-

surement data is transferred to the phone and the phone app automatically switches to

the next screen with the concluding self-assessment manikin.

The other features of the app are not targeted for the subjects. They comprise the whole

data management (creation and deletion of users, exercises and schedules) as well as

statistical insights into single exercises and static information about the app. The data is

represented in lists where the first page is a list of users, selection of a user shows the

users list of exercises and the selection of a training shows the details of this training.

All data can be transferred to a remote database for further investigation and statistical

calculations. For an end-user version of the app, this process needs to be automated to

ensure regular backups.

2 Implementation

2.3 Implementation

The app was developed on an iPhone SE (running iOS 11 beta) and an Apple Watch

Series 1 (running watchOS 3.2) using Xcode 9 beta and Swift 4. All data is created

and persisted using the CoreData framework which is based on the scheme shown in

figure 2.1. CoreData creates so-called managed objects based on this scheme which

are mutable objects of the defined entities with all their attributes and relationships.

The core of the application though is the walking speed measurement. The CoreMotion

framework provides access to the system-generated live walking data. The current pace

is calculated utilizing a step counter over time. This is as precise as it gets working with

the available sensors of the phone/watch. It has its limitations when it comes to very

slow walking speeds since the pedometer cannot distinguish single steps anymore and

returns a pace of zero.

During a mindful walking training data has to be transferred back and forth between the

phone and the watch. Since live communication between phone and watch requires both

applications to be active and in the foreground there had to be a fallback solution since

this case cannot always be ensured. As soon as the user lowers his wrist the watch

screen turns off and the application switches to background mode. During training phone

and watch can get out of reach if the user walks without carrying the phone. The phone

screen turns off, switching the app to background mode if the user doesn’t interact with

it for some time. When the phone (or the watch) wants to send data to its counterpart

application it has to be checked for reachability first, meaning the counterpart app is

active and in the foreground. If this is not the case, data will be sent to a background

queue and delivered as soon as the counterpart app becomes active again (for instance

when the user raises his wrist to look at the watch) and can be shown immediately.

2.3 Implementation

1public func send(messages: [String : Any]) {

2if !session.isReachable {

3sendInBackground(messages)

4return

6// send message in foreground (live)

7session.sendMessage(messages, replyHandler: nil) { error in

8// handle errors

9print("error sending message: \(error)")

10 }

11 }

13 private func sendInBackground(_ message: [String : Any]) {

14 // WKSession must be activated

15 if session.activationState != .activated {

16 return

17 }

18 // cancel outstanding transfers

19 // only the current one is relevant

20 if session.outstandingUserInfoTransfers.count > 0 {

21 session.outstandingUserInfoTransfers.forEach {

22 $0.cancel()

23 }

24 }

25 // queues message on the other device

26 session.transferUserInfo(message)

27 }

Listing 2.1: Sending data to the counterpart

2 Implementation

2.3.1 iPhone App

The iPhone app is storyboard-based, meaning all screens are predefined in a storyboard

with a distinct ViewController for each screen. The screens are connected through

segues (transitions from one screen to another). This way the whole infrastructure of

the application could be built quickly, enabling the user to navigate through all available

views.

The goal was to make the single steps of the training as self-explanatory and simple as

possible and to avoid wrong input. The user can navigate linearly through the steps of a

training session and go back at any point to correct his or her input. Every screen shows

instructions on what to do.

The self-assessment manikin uses well-known radio buttons and can only be finished

after the user selected a value of each of the three dimensions.

After the self assessment-manikin, the normal walking speed is measured. The user is

instructed to switch to the watch for measurement and walk normally until the measure-

ment is complete which is signaled with a haptic and an acoustic feedback on the watch

while the phone automatically switches to the next screen.

On the following two views, the user can select both reduction of speed as well as the

duration of the training in a similar way on separate screens to maintain readability. The

last screen of the actual training shows to instructions to mindful walking. At the same

time, the user input is sent to the watch. The actual training then takes place on the

watch so the user doesn’t have to carry the phone.

When the phone receives the training results it automatically switches to the final self-

assessment manikin (i.e., which works exactly like the first one) and the results are

presented to the user while the training data is saved to the database. If the training is

interrupted at any point and doesn’t reach this final step, data will not be saved ensuring

only complete training objects.

2.3.2 Watch App

The watch app consists of two views: the normal speed measurement and the training

view. The user only interacts with the what to start and stop the measurement as soon

2.3 Implementation

as he is ready. All other interactions take place on the phone because of the bigger

screen. The first view only shows a disabled button to start the measurement for the

normal walking speed. It is enabled after the user finished the first questionnaire. Once

tapped it disappears and the message “Start walking” is shown. The user also receives

a feedback and a message when this measurement is completed. During measurement,

the pedometer measures ten speed values greater than zero and then calculates the

average which is considered the

normal walking speed

. This way some variations in the

subjects walking speed are taken into account but also the measurement doesn’t take

too long (about half a minute).

The calculated result is sent to the phone app and after receiving the user input the

watch automatically switches to the training view where the user can start and stop the

training while observing the current training time and speed. The interface is kept quite

simple to not distract the user from the mindful walking exercise. A speed indicator

shows when the user is walking too fast but stays at the same position when the subject

is walking too slow since this is not considered a mistake during the exercise. The user

will receive a feedback when his selected duration passed but the training will continue

until the user taps the stop button which he can do at any time even if the target duration

is not yet reached. At this point, all measurement data is sent to the phone and the

training is complete.

HealthKit

WatchOS apps are considered Foreground apps; they run only while the user interacts

with one of their interfaces. This is also due to the limited battery duration of the watch.

One problem when measuring live walking data on the Apple Watch is that the application

switches to background mode as soon as the user lowers his wrist (i.e., which everyone

does during walking). When in background mode, all processes of the application are

paused and tracking APIs like the pedometer won’t receive updates anymore. The only

exception tasks allowed to run in the background are URL sessions for networking tasks,

audio player to play audio in the background and the HealthKit WorkoutSession which

specifically targets workout tasks like walking. Using the WorkoutSession requires the

2 Implementation

user to explicitly grant the application access the HealthKit framework. The HealthKit

framework tracks and saves all kinds of health information and the user can specifically

decide which information the application can read or write.

For the purpose of the mindful walk exercise the app doesn’t need to read or write any

health data, it just needs access to the HealthKit framework itself in order to instantiate

HKWorkoutSession

enabling the application to keep tracking in the background. The

HKWorkoutSession

also fine tunes the Watch’s sensors for a specific activity, in this

case walking to generate higher-frequency data samples.

2.4 Validation and Limitations

2.4.1 Functional requirements

General information

FR#1.1 Display App information: Not yet implemented.

FR#1.2 Display General information: Not yet implemented.

Schedules

FR#2.1 Create new schedule:

The user can create new schedules by tapping the “+”-

Button in the schedules menu. Start date and time, as well as the repeat interval,

can be configured using iOS-PickerViews.

FR#2.2 Edit schedule:

All created schedules are shown in a list view. If the user

selects an existing schedule from the list it can be edited similarly to the creation.

Start

FR#3.1 Mood evaluation:

When starting a new training a self-assessment manikin is

shown which the user has to complete using radio buttons in order to get to the

next screen.

2.4 Validation and Limitations

FR#3.2 Normal walking speed measurement:

The phone app shows instructions on

what to do (switch to the watch, press start and start walking normally) while the

watch app measures the walking speed. When the measurement is complete a

feedback sound along with a vibration is played on both the watch and the phone

to notify the user.

FR#3.3 Save data:

The measured walking speed is transferred from the watch to the

phone and saved to the training object. It is not yet saved to the database since

the training object is saved as a whole at the end to guarantee consistent data.

FR#3.4 Choose target speed:

The user can select the target speed for the mindful

walking training relative to the measured normal walking speed. The choice

reaches from 1% to 50% in 1-percent-steps and can be selected using a picker

view. In the same manner, the duration of the mindful walking training can be

selected in 1-minute-steps starting from 5 minutes.

FR#3.5 Save target speed:

The target speed is saved to the training object. It is not

yet saved to the database since the training object is saved as a whole at the end

to guarantee consistent data.

FR#3.6 General instruction:

After the target duration is selected an instruction screen

appears, explaining the mindful walk training, how to walk and what to pay attention

to. No next button is displayed, instead, the user is instructed to switch to the watch

to start the measurement.

FR#3.7 Mindful walking:

During the mindful walking exercise, the watch constantly

measures the user’s current walking speed. Whenever this exceeds the target

speed plus a small threshold the user is notified by a vibration of the watch. The

user can also observe the current duration of the training and gets a feedback

from the watch when the selected target duration (during the study: 15 minutes)

is reached. However the training can be stopped at any time before or after the

target duration has been reached by pressing the “Stop”-Button on the watch.

FR#3.8 Finish training:

After completing the post-training self-assessment manikin

the training is saved to the device storage. The data set includes start and finish

time, both SAM-results, normal, target, and average walking speed, target duration

2 Implementation

as well all speed measurements during the training along with timestamps. The

data can be transferred to a remote database by pressing the button in the statistics

menu.

2.4.2 Non-functional requirements

The non-functional requirements were constantly taken into account during implementa-

tion yet some ratings may be subjective.

NFR#1 - Quality

All displayed information is

correct

in the sense of instructions and explanations. As of

measured training data, false information is avoided by not saving incomplete (aborted)

training objects. A mindful walking training has to be completed as a whole and cannot

be paused and resumed later since this would not only corrupt data but also tear the

subject from its mindful state.

Since mindful walking can be exercised anywhere anytime the application must

reliably

work under diverse conditions. Speed tracking with the pedometer is independent of any

GPS-signal and also works while walking on staircases or similar. No internet connection

is required, only the Bluetooth connection between watch and phone and even that is

required only at the start end at the end of the training. Since the application only is a

support tool and the main focus of the user should be on the mindful walking exercise,

user-friendliness

is critical. A mindful walking training can be started by simply tapping

the big “Start”-Button on the home screen of the app. Navigation within the app follows

a well known horizontal page-to-page concept, allowing the user to go back and forth

at any time. The navigation bar at the bottom of the screen is available throughout the

app to

consistently

guide the user. No personal data is collected and the users are

referenced only by a random ID to avoid

security

issues. The app is

robust

since false

user input is prevented by strict input constraints. Data transmission between phone

and watch is assured by the background queue fallback. All data is saved locally on the

2.4 Validation and Limitations

device, no internet connection is needed. The data can be transferred to an external

database if a network connection is available.

NFR#2 - Performance

The app provides a quick

response time behavior

to user interactions during navigation

within the app. Small delays can occur during the data transfer from the watch to

the phone after the training when all measurement data is sent to the phone. Also

during walking, feedback to sudden speed changes can be delayed due to the tracking

mechanism with the pedometer which retroactively calculates the speed every few steps.

The

computing expense

is kept low, the application doesn’t perform any expensive

computations. This is especially critical on AppleWatch since apps using too much CPU

will be suspended by the operating system [

Inc

] due to the limited battery capacity of the

device.

Data throughput

to external services so far only happens manually by pressing

a button in the app. A network connection has to be assured beforehand, otherwise, no

data will be transferred. Besides that data transfers only happen over Bluetooth between

watch and phone. Since the measurement takes place on the watch, all measurement

data has to be transferred to the phone at some point which may lead to a small delay

depending on the duration of the training (up to 2 seconds for a 15-minute training).

NFR#3 - Error handling

Has yet to be implemented since it was not necessary for the study which was conducted

under supervision.

NFR#4 - Documentation and flexibility

The development of the application followed the model-view-controller pattern for ef-

ficient code reuse. This pattern is also well supported by the Xcode IDE. While the

CoreData framework handles the model layer and the views visually are created in

Interface Builder, one can concentrate on writing the controllers for the single views. This

2 Implementation

way the application can easily be maintained and expanded.

The single controllers rely on each other as little as possible (coupling) while every

component handles a distinguished task (cohesion). In addition to well-documented

source code flexibility is guaranteed.

2.4.3 Limitations

Speed measurement using a pedometer has some limitations going along with it. First,

sudden changes in speed can’t be noticed instantly since the current speed has to be

calculated retrospectively based on the number of steps taken in the last seconds so user

notifications about speed changes are always delayed. Second, speed measurements

lack precision since the step-length of the user isn’t determined and also won’t be

constant. This could be corrected to a certain degree by providing an input field for the

average step-length during mindful walking. Third, a certain minimum speed is required

for the speed measurement to work properly. If the walking speed during the exercise is

too slow (e.g., the subject already has a slow normal walking speed and chose a fairly

high reduction) most speed measurements will return zero (which is again ignored by

the app).

Further, the phone app, as well as the watch app, is required since they form a unit. One

app won’t work without its counterpart.

Study

To evaluate the usefulness of the MindfulWalk app a practical study was conducted

(mostly at Ulm University) from September to October 2017.

3.1 Method

3.1.1 Participants and design

The sample consisted of 30 participants (3 female) which were recruited via Email and

direct messages. Participation was voluntary and without any compensation. Most

participants were graduates and postgraduates. No personal information was raised.

This study used a one-group pretest-posttest design even though not only the change in

mood but also the participant´

s behavior during the exercise was of interest.

3.1.2 Procedure

First, participants were told what mindful walking was about and how to walk mindfully.

Also, a brief overview of the procedure was given. Next, the devices (iPhone and Apple

Watch) were handed out with the apps running. Participants then had to complete

the procedure shown in Fig. 3.1. They started the exercise themselves by tapping

the “Start”-Button



on the home screen which led them to the pre-exercise 5-point

Self-Assessment Manikin 2

[BL94] (cf. Fig. 3.2).

3 Study

iPhone iWatch

Start

User Input:

SAM

Questionnaire

Normal Speed

Instructions

Info Screen Start

Measurement

results

User Input:

Target Speed

& Duration

Monitoring Screen

Input Screen

Mindfuld

Walking

Instructions

Info Screen Start

{target speed, duration}

Measurement

Check for

Violation

Check for

Duration

Feedback

results

User Input:

SAM

walking period

reached

Questionnaire

store

data

General information 1

store

data 4

Figure 3.1: Walking Procedure

3.1 Method

Figure 3.2: Self-assessment manikin selected and unselected

3 Study

After completing the questionnaire an instruction to walk normally was displayed



and

the participants had to start the measurement with a button on the watch



(which is

enabled after the SAM has been completed) (cf. Fig. 3.3).

Figure 3.3:

Normal walking instruction on the phone (left) and start screen on the watch

with disabled (top) and enabled (bottom) button to start the measurement

The participants then walked normally for about a minute until they received feedback

in form of a sound a vibration, also a message on the watch was shown to return to

the phone since the phone was not carried during walking (cf. Fig. 3.4). The resulting

normal walking speed was then transferred to the phone via Bluetooth 4



Before the actual mindful walking exercise, the target speed and duration had to be

selected



. This was done for the participants to avoid too slow target speeds since the

speed selection depended on the normal speed and could therefore lead to very slow

3.1 Method

Figure 3.4:

Instruction to start walking after tapping the Measure-button (left) and feed-

back message after the measurement (right)

target speeds if too much reduction was selected. The selected target speed ranged

from 91-95% of the normal walking speed. The target duration was also selected for the

participants since all subjects were asked to walk for at least 15 minutes (cf. Fig. 3.5).

Next, the participants were again instructed how to walk mindfully and not to exceed their

target speed which will be indicated by a vibration of the watch. The measurement was

again started by the participant on the watch who could then walk anywhere for at least

15 minutes. Most participants decided to walk outside. During the exercise, subjects

could monitor their current speed and time on the watch (cf. Fig. 3.6). Whenever

the current walking speed exceeded the target speed the participant received a haptic

feedback on the watch 6

.

Finally, after tapping the “Stop”-Button on the watch, the participants had to complete

the post-exercise self-assessment manikin 7

.

3 Study

Figure 3.5: Target speed (left) and duration (right) selection

Figure 3.6: Speed and time monitor during the exercise

3.2 Results and Evaluation

In the pilot study, the mindful walking application was tested in

N= 20

participants. The

participants used the mindful walking application once;

n= 19

participants set

15min

training interval and

n= 1

participant set

10min

. The participants usual walking speed as

measured by the Apple Watch application was on average

M= 3.37km/h (SD = 0.35)

The average target walking speed the participants set for the mindful walking exercise

was

M= 3.16km/h (SD = 0.33)

, and the average walking speed during the mindful

walking exercise as measured with the Apple Watch application was

M= 3.13km/h

(SD = 0.28)

. A repeated measure analysis of variance (rANOVA) was conducted in

SPSS 24

to test whether the usual walking speed, the target walking speed, and the

walking speed during the mindful walking exercise are significantly different. The rANOVA

(Greenhouse-Geisser corrected) produced a statistically significant result indicating

relevant difference between the three walking speeds (usual walking speed, target

walking speed, and walking speed during the mindfulness exercise):

F(1.09; 20.67) =

29.13; p < 0.01

. Moreover, simple contrasts (with the walking speed during the mindful

walking exercise as reference) were performed within the rANOVA to evaluate the

hypotheses whether the usual walking speed is faster than the walking speed during the

mindful walking exercise as well as whether the walking speed during the mindful walking

exercise is not statistically different from the target walking speed the participants set for

the mindful walking exercise. The results were in correspondence with the hypotheses:

The walking speed during the mindful walking exercise was significantly slower than

the usual walking speed

(F(1; 19) = 30.48; p < 0.01)

and the walking speed during the

mindful walking exercise was not significantly different from the walking speed the users

set as target for the mindful walking exercise (F(1; 19) = 0.36; p= 0.55).

The results of the single participants are shown in Table 3.1.

3 Study

VB1AB1DB1normalSpeed targetSpeed targetDuration average VA2AA2DA2duration SD3speedDifference reduction

1 1 1 3,3326 3,1327 00:15:00 3,1308 -1 -1 -1 00:15:48 0.2399 0.06% 6%

1 -1 1 3,6609 3,4413 00:15:00 3,2416 -1 1 -2 00:15:26 5.069 0.11% 6%

1 0 1 3,3972 3,1594 00:15:00 2,8710 -1 1 0 00:17:17 0.4132 0.15% 7%

1 0 1 2,8813 2,6796 00:15:00 2,8782 2 1 1 00:15:19 1.2153 0.00% 7%

2 1 1 3,1246 2,9683 00:15:00 2,8598 -1 -1 0 00:15:50 0.3560 0.08% 5%

1 2 1 3,8577 3,5105 00:15:00 3,5536 2 -1 -1 00:16:27 0.5202 0.08% 9%

0 -1 1 3,3984 3,1945 00:15:00 3,3311 -1 1 0 00:27:27 0.4630 0.02% 6%

1 -1 1 3,3849 3,0803 00:15:00 2,9131 -2 1 1 00:00:00 0.8324 0.14% 9%

0 -2 1 2,9659 2,8177 00:15:00 3,0046 -1 -2 1 00:19:00 0.3764 0.01% 5%

1 -1 0 3,5547 3,3059 00:15:00 3,3760 -2 1 2 00:15:00 0.1846 0.05% 7%

1 -2 0 3,1330 2,9450 00:15:00 3,0311 -1 1 2 00:15:45 0.3795 0.03% 6%

2 -1 0 3,1966 3,0367 00:15:00 3,0019 1 -1 -2 00:15:33 0.5744 0.06% 5%

0 1 0 3,8879 3,6546 00:15:00 3,5313 -2 1 2 00:15:01 0.6937 0.09% 6%

-1 -2 0 4,0832 3,8790 00:15:00 3,4158 -1 1 -2 00:15:27 0.3818 0.16% 5%

1 -2 0 3,0755 2,8602 00:15:00 2,7641 -1 -1 2 00:15:11 0.4409 0.10% 7%

-1 -1 0 2,7749 2,5807 00:15:00 2,5684 -1 2 1 00:16:17 0.3930 0.07% 7%

1 0 1 3,7325 3,5086 00:15:00 3,4896 -2 1 2 00:16:16 0.5000 0.07% 6%

-1 0 -1 3,1076 2,9211 00:10:00 3,2245 -1 -1 -2 00:10:09 0.9117 0.04% 6%

1 -1 0 3,5788 3,3998 00:15:00 3,4546 -2 1 2 00:16:49 0.2514 0.03% 5%

1 0 0 3,3208 3,0884 00:15:00 3,0451 -2 1 0 00:15:05 0.3229 0.08% 7%

Table 3.1: Study Results

1VB/AB/DB = valence/arousal/dominance before exercise

2VA/AA/DA = valence/arousal/dominance after exercise

3MSD = standart deviation of single speed measurements

Summary and Outlook

In this work, an application developed to support individuals in walking mindfully is

presented. The application integrates a sensor to measure the walking speed in real

life situations and during mindful walking exercises. Moreover, a tool is implemented

to provide immediate haptic feedback on the sensed walking speed. This feedback

was integrated as it might prevent individuals from walking faster than intended and

from “mind wandering” during the mindfulness exercise. Haptic feedback might be more

suited in daily life than visual or auditory feedback as it is not always possible to look

at the smartphone screen to see visual feedback or to be in silent surroundings / to

wear headphones to hear auditory feedback. A pilot study was performed which showed

promising results. The data of the walking speed sensor showed that the participants

usual walking speed is faster than the walking speed during the mindful walking exercise.

In line with this result, [

YWLH12

] reported that their system to support mindful walk

also slowed down the walking speed. Moreover, the sensed walking speed during the

mindful walking exercise was not different from the walking speed the participants set as

target speed for the mindful exercise. The result that the participants could achieve their

intended “slow speed” might suggest that the haptic feedback function could be applied

successfully.

Yet, more evaluations are necessary and future experiments should include a control

condition. The latter without the haptic feedback function of the application, in turn,

would allow for more causal conclusions on whether the haptic feedback function of

the application is indeed helpful to walk mindfully as slowly as intended and to stay

mindfully during the exercise. It could also be possible that the feedback distracts

the participants from being mindful. Moreover, feedback of any kind could create a

4 Summary and Outlook

judgmental stance towards the walking, which is against the mindfulness principles. An

option in the application to turn the feedback on and off would not only be useful for a

follow-up study with a control condition but also give end-users the possibility to walk

mindfully without any distractions and only use the app as a silent tracking tool, if they

chose to do so. Also the stored data should also be synced with a cloud-based database

on regular basis to ensure backups. Further studies using psychometric questionnaires

(e.g.,

Five Facet Mindfulness Questionnaire

) are needed to explore how the feedback

function affects different aspects of mindfulness (e.g., nonjudging of inner experience

or nonreactivity to inner experience). Note that only a small sample was investigated

limiting generalizability of the results.

Larger trials could investigate whether using the mindful walking application can improve

mindfulness or also well-being in general. Additionally, further studies are needed to

investigate the health-related effects as well as the acceptance of the mindful walking

application in occupational settings (e.g., occupational health management). Further

comparisons between a condition using the mindful walking application and a condition

exercising mindful walking traditionally (e.g., guided by an expert, instructions delivered

by audio, or video) would reveal if the haptic feedback (and the mindful walking application

in general) are beneficial for walking mindfully. In summary, the newly developed mindful

walking application has been successfully applied to sense the walking speed and to

provide immediate feedback on the current walking speed. The results of the pilot study

are promising and the mindful walking application offers several opportunities for digital

health prevention.

List of Figures

2.1 DataModel................................... 12

3.1 WalkingProcedure............................... 24

3.2 Self-assessment manikin selected and unselected . . . . . . . . . . . . . 25

3.3

Normal walking instruction on the phone (left) and start screen on the

watch with disabled (top) and enabled (bottom) button to start the mea-

surement .................................... 26

3.4

Instruction to start walking after tapping the Measure-button (left) and

feedback message after the measurement (right) . . . . . . . . . . . . . . 27

3.5 Target speed (left) and duration (right) selection . . . . . . . . . . . . . . . 28

3.6 Speed and time monitor during the exercise . . . . . . . . . . . . . . . . . 28

List of Tables

3.1 StudyResults.................................. 30

Bibliography

[7Mi] 7MIND:7Mind.https://www.7mind.de. – last visited: 23.11.2017

[AK04]

ANDERSSON, Gerhard ; KALDO, Viktor: Internet-based cognitive behavioral

therapy for tinnitus. In:

Journal of Clinical Psychology

60 (2004), Nr. 2, S.

171–178. – ISSN 00219762

[BEC+12]

BAÑOS, R.M. ; ETCHEMENDY, E. ; CASTILLA, D. ; GARCÍA-PALACIOS, A. ;

QUERO, S. ; BOTELLA, C.: Positive mood induction procedures for virtual

environments designed for elderly people. In:

Interacting with Computers

24 (2012), Nr. 3, S. 131–138

[BL94]

BRADLEY, Margaret M. ; LANG, Peter J.: Measuring Emotion: The Self-

Assessment Manikin and the Semantic Differential. In:

Journal of Behavior

Therapy and Experimental Psychiatry 25 (1994), Nr. 1, S. 49–59

[CBM+12]

CHEN, Kevin W. ; BERGER, Christine C. ; MANHEIMER, Eric ; FORDE,

Darlene ; MAGIDSON, Jessica ; DACHMAN, Laya ; LEJUEZ, C. W.: Meditative

therapies for reducing anxiety: a systematic review and meta-analysis of

randomized controlled trials. In: Depress Anxiety 29 (2012), S. 545–562

[EABP08]

EISEN, Katherine P. ; ALLEN, George J. ; BOLLASH, Mary ; PESCATELLO,

Linda S.: Stress management in the workplace: A comparison of a

computer-based and an in-person stress-management intervention. In:

Computers in Human Behavior

24 (2008), Nr. 2, S. 486–496. – ISSN

0747–5632. – Part Special Issue: Cognition and Exploratory Learning in

Digital Age

[Eve]

EVERYWHERE, Mindfulness:

Buddhify

http://www.buddhify.com

. –

last visited: 23.11.2017

[For]

FORNIES, Hector R.:

ZMediations

http://www.zmeditations.com

– last visited: 23.11.2017

Bibliography

[GHTS16]

GAINEY, A. ; HIMATHONGKAM, T. ; TANAKA, H. ; SUKSOM, D.: Effects of

Buddhist walking meditation on glycemic control and vascular function in

patients with type 2 diabetes. In:

Complement Ther Med

26 (2016), S.

92–97

[GM11]

GLÜCK, Tobias M. ; MAERCKER, Andreas: A randomized controlled pilot

study of a brief web-based mindfulness training. In:

BMC Psychiatry

(2011), Nr. 1, S. 175. – ISSN 1471–244X

[GNBBG08]

GARDNER-NIX, Jacqueline ; BACKMAN, Stéphanie ; BARBATI, Julianna

; GRUMMITT, Jessica: Evaluating distance education of a mindfulness-

based meditation programme for chronic pain management. In:

Journal of

Telemedicine and Telecare 14 (2008), Nr. 2, S. 88–92

[Han]

HANH, Thich N.:

Thich Nhat Hanh on Walking

Meditation

https://www.lionsroar.com/

how-to-meditate-thich-nhat-hanh-on-walking-meditation/

– last visited: 23.11.2017

[Inc]

INC., Apple:

Leveraging iOS Technologies

https://developer.

apple.com/library/content/documentation/General/

Conceptual/WatchKitProgrammingGuide/iOSSupport.html

– last visited: 23.11.2017

[KCKW12]

KRUSCHE, Adele ; CYHLAROVA, Eva ; KING, Scott ; WILLIAMS, J Mark G.:

Mindfulness online: a preliminary evaluation of the feasibility of a web-

based mindfulness course and the impact on stress. In:

BMJ Open

(2012), Nr. 3. – ISSN 2044–6055

[KFE+11]

KRISTJÁNSDÓTTIR, Ólöf Birna ; FORS, Egil A. ; EIDE, Erlend ; FINSET,

Arnstein ; DULMEN, Sandra van ; WIGERS, Sigrid H. ; EIDE, Hilde: Written

online situational feedback via mobile phone to support self-management of

chronic widespread pain: a usability study of a Web-based intervention. In:

BMC Musculoskeletal Disorders

12 (2011), Nr. 1, S. 51. – ISSN 1471–2474

Bibliography

[KLF+13]

KHOURY, B. ; LECOMTE, T. ; FORTIN, G. ; MASSE, M. ; THERIEN, P.

; BOUCHARD, V. ; CHAPLEAU, M. A. ; PAQUIN, K. ; HOFMANN, S. G.:

Mindfulness-based therapy: a comprehensive meta-analysis. In:

Clin

Psychol Rev 33 (2013), S. 763–771

[KSR11]

KENG, Shian-Ling ; SMOSKI, Moria J. ; ROBINS, Clive J.: Effects of

mindfulness on psychological health: A review of empirical studies. In:

Clinical Psychology Review 31 (2011), S. 1041–1056

[KSRF15]

KHOURY, B. ; SHARMA, M. ; RUSH, S. E. ; FOURNIER, C.: Mindfulness-

based stress reduction for healthy individuals: A meta-analysis. In:

Psychosom Res 78 (2015), S. 519–528

[KZ90]

KABAT-ZINN, J.:

Full catastrophe living: The program of the Stress

Reduction Clinic at the University of Massachusetts Medical Center

. New

York: Delta, 1990

[KZ05]

KABAT-ZINN, Jon:

Coming to Our Senses: Healing Ourselves and the

World Through Mindfulness. Hyperion, New York, 2005

[LFV+10]

LJÓTSSON, Brjánn ; FALK, Lisa ; VESTERLUND, Amanda W. ; HEDMAN, Erik

; LINDFORS, Perjohan ; RÜCK, Christian ; HURSTI, Timo ; ANDRÉEWITCH,

Sergej ; JANSSON, Liselotte ; LINDEFORS, Nils ; ANDERSSON, Gerhard:

Internet-delivered exposure and mindfulness based therapy for irritable

bowel syndrome – A randomized controlled trial. In:

Behaviour Research

and Therapy 48 (2010), Nr. 6, S. 531–539. – ISSN 0005–7967

[lim]

LIMITED, Headspace meditation:

Headspace

https://www.

headspace.com. – last visited: 23.11.2017

[Mar12]

MARCHAND, William R.: Mindfulness-Based Stress Reduction, Mindfulness-

Based Cognitive Therapy, and Zen Meditation for Depression, Anxiety, Pain,

and Psychological Distress. In:

Journal of Psychiatric Practice

18 (2012),

S. 233–252

[Min]

MINDAPPS:

Mindfulness

http://www.mindapps.se/

themindfulnessapp. – last visited: 23.11.2017

Bibliography

[MKHS15]

MANI, Madhavan ; KAVANAGH, David J. ; HIDES, Leanne ; STOYANOV,

Stoyan R.: Review and Evaluation of Mindfulness-Based iPhone Apps. In:

JMIR Mhealth Uhealth 3 (2015)

[MKL+10]

MORRIS, E. M. ; KATHAWALA, Qusai ; LEEN, K. T. ; GORENSTEIN, E. E. ;

GUILAK, Farzin ; LABHARD, Michael ; DELEEUW, William: Mobile Therapy:

Case Study Evaluations of a Cell Phone Application for Emotional Self-

Awareness. In: J Med Internet Res 12 (2010), Nr. 2, S. e10

[NDE+12]

NES, Andréa A. ; DULMEN, Sandra van ; EIDE, Erlend ; FINSET, Arnstein

; BIRNA KRISTJÁNSDÓTTIR Ólöf ; STEEN, Ida S. ; EIDE, Hilde: The

development and feasibility of a web-based intervention with diaries and

situational feedback via smartphone to support self-management in patients

with diabetes type 2. In:

Diabetes Research and Clinical Practice

97 (2012),

Nr. 3, S. 385–393. – ISSN 0168–8227

[PDHM+13]

PLAZA, Inmaculada ; DEMARZO, Marcelo Marcos P. ; HERRERA-MERCADAL,

Paola ; GARCÍA-CAMPAYO, Javier ; EYSENBACH, Gunther: Mindfulness-

Based Mobile Applications: Literature Review and Analysis of Current

Features. In: JMIR mHealth and uHealth 1 (2013)

[PPL+17a]

PROBST, Thomas ; PRYSS, Rüdiger ; LANGGUTH, Berthold ;

RAUSCHECKER, Josef ; SCHOBEL, Johannes ; REICHERT, Manfred ;

SPILIOPOULOU, Myra ; SCHLEE, Winfried ; ZIMMERMANN, Johannes: Does

tinnitus depend on time-of-day? An ecological momentary

assessment

study with the “

TrackYourTinnitus

” application. In:

Frontiers in Aging

Neuroscience 9 (2017), S. 253–253

[PPL+17b]

PROBST, Thomas ; PRYSS, Rüdiger ; LANGGUTH, Berthold ;

SPILIOPOULOU, Myra ; LANDGREBE, Michael ; VESALA, Markku ;

HARRISON

, Stephen ; SCHOBEL, Johannes ; REICHERT, Manfred ; STACH,

Michael ; SCHLEE, Winfried: Outpatient Tinnitus Clinic, Self-Help Web

Platform, or Mobile Application to Recruit Tinnitus Study Samples? In:

Frontiers in Aging Neuroscience 9 (2017), April, S. 113–113

Bibliography

[PPLS16a]

PROBST, Thomas ; PRYSS, Rüdiger ; LANGGUTH, Berthold ; SCHLEE,

Winfried: Emotion dynamics and tinnitus: Daily life data from the

“TrackYourTinnitus” application. In: Scientific Reports 6 (2016)

[PPLS16b]

PROBST, Thomas ; PRYSS, Rüdiger ; LANGGUTH, Berthold ; SCHLEE

Winfried

: Emotional states as mediators between tinnitus loudness

and

tinnitus

distress in daily life: Results from the “

TrackYourTinnitus

”

application. In: Scientific Reports 6 (2016), February

[PPS+17]

PRYSS, Rüdiger ; PROBST, Thomas ; SCHLEE, Winfried ; SCHOBEL

Johannes

; LANGGUTH, Berthold ; NEFF, Patrick ; SPILIOPOULOU, Myra

; REICHERT, Manfred: Mobile Crowdsensing for the Juxtaposition of

Realtime

Assessments and Retrospective Reporting for Neuropsychiatric

Symptoms. In:

30th IEEE International Symposium on Computer-Based

Medical Systems (CBMS 2017)

, IEEE Computer Society Press, June 2017

[PSLR17]

PRYSS, Rüdiger ; SCHLEE, Winfried ; LANGGUTH, Berthold ; REICHERT,

Manfred: Mobile Crowdsensing Services for Tinnitus Assessment and

Patient Feedback. In:

6th IEEE International Conference on AI & Mobile

Services (IEEE AIMS 2017), IEEE Computer Society Press, June 2017

[PSTS14]

PRAKHINKIT, S. ; SUPPAPITIPORN, S. ; TANAKA, H. ; SUKSOM, D.: Effects

of Buddhism walking meditation on depression, functional fitness, and

endothelium- dependent vasodilation in depressed elderly. In:

J Altern

Complement Med 20 (2014), S. 411–416

[RADM14]

ROHDE, Katharina ; ADOLPH, Dirk ; DIETRICH, Detlef E. ; MICHALAK,

Johannes: Mindful attention regulation and non-judgmental orientation

in depression: A multi-method approach. In:

Biological Psychology

101

(2014), S. 36–43. – ISSN 0301–0511

[SES+12]

SEDLMEIER, Peter ; EBERTH, Juliane ; SCHWARZ, Marcus ; ZIMMER-

MANN, Doreen ; HAARIG, Frederik ; JAEGER, Sonia ; KUNZE, Sonja: The

psychological

effects of meditation: A meta-analysis. In:

Psychological

Bulletin 138 (2012), S. 1139–1171

Bibliography

[SGS07]

SHAW, Chris D. ; GROMALA, Diane ; SEAY, A. F.: The Meditation Chamber:

Enacting Autonomic Senses. In:

Proceedings of the 4th International

Conference on Enactive Interfaces 2007

, 2007 (Enactive ’15), S. 405–408

[SH]

SIMPLE HABIT, Inc.:

Simple Habit

https://www.simplehabit.com

. –

last visited: 23.11.2017

[SPP+16]

SCHLEE, Winfried ; PRYSS, Rüdiger ; PROBST, Thomas ; SCHOBEL,

Johannes ; BACHMEIER, Alexander ; REICHERT, Manfred ; LANGGUTH,

Berthold: Measuring the Moment-to-Moment Variability of Tinnitus: The

TrackYourTinnitus Smart Phone App. In:

Frontiers in Aging Neuroscience

8 (2016), December, S. 294–294

[SWT02a]

SEGAL, Z. V. ; WILLIAMS, J. M. G. ; TEASDALE, J. D.:

Mindfulness-based

cognitive therapy for depression – A new approach to preventing relapse

New York: Guilford Press, 2002

[SWT02b]

SEGAL, Zindel V. ; WILLIAMS, J. Mark G. ; TEASDALE, John D.:

Mindfulness-

Based Cognitive Therapy for Depression: A New Approach to Preventing

Relapse. The Guilford Press, 2002

[THP15]

TANG, Y. Y. ; HOELZEL, B. K. ; POSNER, M. I.: The neuroscience of

mindfulness meditation. In: Nat Rev Neurosci 16 (2015), S. 213–225

[TWO+10]

THOMPSON, Nancy J. ; WALKER, Elizabeth R. ; OBOLENSKY, Natasha ;

WINNING, Ashley ; BARMON, Christina ; DIIORIO, Colleen ; COMPTON,

Michael T.: Distance delivery of mindfulness-based cognitive therapy for

depression: Project UPLIFT. In:

Epilepsy & Behavior

19 (2010), Nr. 3, S.

247 – 254. – ISSN 1525–5050

[Wil08]

WILLIAMS, J. M. G.: Mindfulness, depression and modes of mind. In:

Cognit Ther Res 32 (2008), S. 721–733

[YWLH12]

YU, Meng-Chieh ; WU, Huan ; LEE, Ming-Sui ; HUNG, Yi-Ping: Multimedia-

Assisted Breathwalk-Aware System. In:

IEEE Transactions on Biomedical

Engineering 59 (2012), S. 3276–3282

Name: Julian Frank Matrikelnummer: 711148

Erklärung

Ich erkläre, dass ich die Arbeit selbstständig verfasst und keine anderen als die angegebe-

nen Quellen und Hilfsmittel verwendet habe.

Ulm,den .............................................................................

Julian Frank