ThesisPDF Available

Technology for Better Animal Care: Identifying the Dimensions for Increasing the Caretakers’ Awareness through Dog Activity Monitoring Systems

Authors:

Abstract

In the last ten years, wearable technologies for animals have become increasingly popular, and activity monitoring systems are one of the most commonly used types of technology. However, there is a limited amount of research on dog activity monitoring systems and their impact on the lives of caretakers and their awareness, despite the growing number of studies on wearable technologies for humans. This thesis aims to fill this gap by conducting a longitudinal study with 30 participants, exploring the dimensions of interaction with dog activity monitoring systems, caretaker personas related to the use of these systems, and their potential to contribute to the caregiving of dogs. The study involves participants using a specific dog activity monitoring device for six weeks, along with in-depth interviews, experience sampling method, and complementary questionnaires. The findings are used to develop the Dog Activity Monitoring Systems-mediated stage-based awareness model that explains how dog activity systems can mediate the human-dog relationship and support the caregiving of dogs.
TECHNOLOGY FOR BETTER ANIMAL CARE:
IDENTIFYING THE DIMENSIONS FOR INCREASING THE CARETAKERS’
AWARENESS THROUGH DOG ACTIVITY MONITORING SYSTEMS
A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY
BY
ASLIHAN TOKAT
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR
THE DEGREE OF DOCTOR OF PHILOSOPHY
IN
INDUSTRIAL DESIGN
MAY 2023
Approval of the thesis:
TECHNOLOGY FOR BETTER ANIMAL CARE: IDENTIFYING THE
DIMENSIONS FOR INCREASING THE CARETAKERS’ AWARENESS
THROUGH DOG ACTIVITY MONITORING SYSTEMS
submitted by ASLIHAN TOKAT in partial fulfillment of the requirements for the
degree of Doctor of Philosophy in Industrial Design, Middle East Technical
University by,
Prof. Dr. Halil Kalıpçılar
Dean, Graduate School of Natural and Applied Sciences
Prof. Dr. Gülay Hasdoğan
Head of the Department, Industrial Design
Assist. Prof. Dr. Gülşen Töre Yargın
Supervisor, Industrial Design, METU
Prof. Dr. Yasemin Salgırlı Demirbaş
Co-Supervisor, Veterinary Medicine, Ankara University
Examining Committee Members:
Prof. Dr. Bahar Şener Pedgley
Industrial Design, METU
Assist. Prof. Dr. Gülşen Töre Yargın
Industrial Design, METU
Assist. Prof. Dr. Güzin Şen
Industrial Design, METU
Assist. Prof. Dr. Aslı Günay
Media and Visual Arts, Koç University
Assist. Prof. Dr. Nazlı Cila
Industrial Design Engineering, TU Delft
Date: 25.05.2023
iv
I hereby declare that all information in this document has been obtained and
presented in accordance with academic rules and ethical conduct. I also declare
that, as required by these rules and conduct, I have fully cited and referenced
all material and results that are not original to this work.
Name, Last name : Aslıhan Tokat
Signature :
v
ABSTRACT
TECHNOLOGY FOR BETTER ANIMAL CARE:
IDENTIFYING THE DIMENSIONS FOR INCREASING THE
CARETAKERS’ AWARENESS THROUGH DOG ACTIVITY
MONITORING SYSTEMS
Tokat, Aslıhan
Doctor of Philosophy, Industrial Design
Supervisor: Assist. Prof. Dr. Gülşen Töre Yargın
Co-Supervisor: Prof. Dr. Yasemin Salgırlı Demirbaş
May 2023, 285 pages
In the last ten years, wearable technologies for animals have become increasingly
popular, and activity monitoring systems are one of the most commonly used types
of technology. However, there is a limited amount of research on dog activity
monitoring systems and their impact on the lives of caretakers and their awareness,
despite the growing number of studies on wearable technologies for humans. This
thesis aims to fill this gap by conducting a longitudinal study with 30 participants,
exploring the dimensions of interaction with dog activity monitoring systems,
caretaker personas related to the use of these systems, and their potential to
contribute to the caregiving of dogs. The study involves participants using a specific
dog activity monitoring device for six weeks, along with in-depth interviews,
experience sampling method, and complementary questionnaires. The findings are
used to develop the Dog Activity Monitoring Systems-mediated stage-based
awareness model that explains how dog activity systems can mediate the human-dog
relationship and support the caregiving of dogs.
Keywords: Animal-computer interaction, dog, animal welfare, dog activity
monitoring systems
vi
vii
ÖZ
HAYVANLARA DAHA İYİ BAKMAK İÇİN TEKNOLOJİ:
KÖPEK AKTİVİTE TAKİP SİSTEMLERİ İLE İNSAN FARKINDALIĞINI
ARTIRMA BOYUTLARININ BELİRLENMESİ
Tokat, Aslıhan
Doktora, Endüstri Ürünleri Tasarımı
Tez Yöneticisi: Dr. Öğr. Üyesi Gülşen Töre Yargın
Ortak Tez Yöneticisi: Prof. Dr. Yasemin Salgırlı Demirbaş
Mayıs 2023, 285 sayfa
Hayvanlar için giyilebilir teknolojiler giderek daha popüler hale gelmekte ve
sundukları akıllı deneyimler yoluyla hem günlük hayatı paylaştığımız evcil
hayvanların hem de insanların yaşamlarını iyileştirmeyi vaat etmektedirler. Son
yıllarda hem ürün hem de kullanıcı sayısı gittikçe artan ve son kullanıcıya yönelik
en yaygın giyilebilir teknolojilerden olan köpeklere yönelik aktivite takip
teknolojileri, evcil hayvan ürünleri endüstrisinde de yerini alarak yaygınlık
kazanmaya başlamıştır. Bununla birlikte, insanlar için giyilebilir teknolojiler üzerine
artan sayıda çalışma olmasına rağmen, köpek aktivite takip sistemleri ve bu
teknolojilerin hayvan bakım kalitesi, hayvan sahiplerinin yaşam biçimleri ve
farkındalıkları üzerindeki etkileri hakkında sınırlı sayıda araştırma bulunmaktadır.
Bu tez, köpek aktivite takip sistemleri ile etkileşimin boyutlarını, bu sistemlerin
kullanımıyla ilgili hayvan sahibi personaları ve bu teknolojilerin köpek bakım
kalitesine katkı sağlamak bakımından potansiyellerini, 30 katılımcının yer aldığı
uzun dönemli bir alan araştırması ile inceleyerek bu boşluğu doldurmayı
amaçlamaktadır. Çalışma, katılımcıların altı hafta boyunca belirli bir köpek aktivite
takip cihazını kullanmasının yanı sıra derinlemesine görüşmeler, deneyim örnekleme
yöntemi ve tamamlayıcı anketleri içermektedir. Bulgular, köpek aktivite takip
viii
sistemlerinin insan-köpek ilişkisine nasıl aracılık edebileceğini ve köpek bakımını
nasıl destekleyebileceğini açıklayarak, bu teknolojilerin tasarımında yol gösterici
teorik bir çerçeve oluşturmak üzere, Köpek Aktivite Takip Sistemleri-aracılı aşama
temelli farkındalık modelini geliştirmek için kullanılmıştır.
Anahtar Kelimeler: Hayvan-bilgisayar etkileşimi, köpek, hayvan iyi oluşu, köpek
aktivite takip sistemleri
ix
To my father
x
ACKNOWLEDGMENTS
It gives me great pleasure to acknowledge all those who have contributed to the
completion of my thesis. This journey has been long and challenging, but I appreciate
every moment of it, for it altered my life in ways I never imagined.
I would like to express my sincerest appreciation to my supervisor, Assistant Prof.
Dr. Gülşen Töre Yargın, and my co-supervisor, Associate Prof. Dr. Yasemin Salgırlı
Demirbaş, for their unwavering support, direction, constructive feedback, and
knowledge throughout the research process. Their guidance has been invaluable to
me in finishing this thesis, and I am deeply grateful for their mentorship.
I owe countless thanks to Gülşen Hocam, who has been far more than a supervisor
to me throughout the years, ever since my undergraduate studies. She has been a
wonderful mentor and role model, whom I can take as an example in every aspect of
life that I can think of. I feel honored to have her as a guide whenever I am doubtful,
and I thank the life plan that led me to meet my dear Hocam. She has helped me to
grow not only as a researcher but also as a person, and I will always cherish the
lessons I have learned from her.
I would also like to express my gratitude towards the members of the thesis
monitoring committee for their insightful comments and helpful suggestions. I am
thankful to Prof. Dr. Bahar Şener Pedgley and Assist. Prof. Dr. Aslı Günay for their
valuable contributions to my study, which have greatly enriched my work.
I would like to extend my sincere thanks to my dear friends Dilruba, Sıla, and Sarper,
for their endless support and companionship during this journey. Their
encouragement and unwavering support have kept me motivated and focused, even
during the most challenging times.
xi
I am grateful to Furkan Yaz for his technical assistance, which I greatly appreciated.
His invaluable support and technical expertise have made a significant contribution
to this study.
I would also like to thank my beloved cats, Tofu and Isot, for their companionship
throughout my Ph.D. journey. Their presence brought immeasurable joy during
challenging times, reminding me to take breaks, play, and find comfort in their
company. They provided a calming and soothing presence during late nights of
study, always by my side, offering their unconditional love and warmth. I am truly
grateful to have had Tofu and Isot as my furry co-pilots on this academic adventure.
Lastly, I would like to express my heartfelt gratitude to all the dog parents and their
beloved dog friends who participated in this study. Without your participation, this
study could not have been completed. I am truly humbled and grateful for your time,
effort, and trust in this study. I thank you all from the bottom of my heart for making
this journey possible.
Also, thanks to the METU Scientific Research Projects (BAP) for supporting this
study. 1
1 The field research within the context of this thesis is funded by the METU Scientific Research
Projects (BAP) Unit granted with the project number TEZ-D-203-2020-10260.
xii
TABLE OF CONTENTS
ABSTRACT .............................................................................................................. v
ÖZ ............................................................................................................................ vii
ACKNOWLEDGMENTS ......................................................................................... x
TABLE OF CONTENTS ........................................................................................ xii
LIST OF TABLES ............................................................................................... xviii
LIST OF FIGURES ................................................................................................ xix
LIST OF ABBREVIATIONS ............................................................................... xxii
1 INTRODUCTION ............................................................................................. 1
1.1 Problem Background ..................................................................................... 1
1.2 Aim of the Study and Research Questions .................................................. 14
1.3 Structure of the Thesis ................................................................................. 15
2 ANIMAL WELFARE AND ANIMAL-COMPUTER INTERACTION ........ 17
2.1 Animal Welfare ........................................................................................... 17
2.2 Animal-Computer Interaction (ACI) ........................................................... 23
2.2.1 Animal-Centered Design and Its Challenges .......................................... 26
2.2.2 Research Methods in ACI for the Identification of Animal Needs ......... 27
2.2.3 Theories, Models, and Frameworks within ACI ..................................... 33
2.2.4 The Review of the Existing Technological Applications for Tracking and
Monitoring of Dogs Concerning ACI ...................................................................... 42
2.3 Dog Activity Monitors as Interspecies Information Systems ...................... 47
2.4 Conclusions Regarding the Chapter ............................................................ 49
3 BEHAVIOR CHANGE AND SENSEMAKING ........................................... 51
xiii
3.1 Persuasive Role of Technology ................................................................... 51
3.2 Behavior Change Models, Theories, and Frameworks ................................ 57
3.3 Information Processing ................................................................................ 64
3.3.1 Reflecting on Self-Tracking Data ............................................................ 66
3.3.2 Data, Information, and Knowledge .......................................................... 76
3.3.3 Sensemaking Models and Theories ......................................................... 77
3.4 Conclusions Regarding the Chapter ............................................................. 82
4 METHODOLOGY .......................................................................................... 85
4.1 Methodology Selection ................................................................................ 85
4.1.1 Understanding Longitudinal User Experience ......................................... 85
4.1.2 User Diversity and Personas .................................................................... 92
4.1.3 Measuring Human-Dog Relationship ...................................................... 93
4.2 Device Selection .......................................................................................... 98
4.3 Participant Selection .................................................................................. 100
4.4 Research Materials ..................................................................................... 105
4.5 Procedure of the Study ............................................................................... 107
4.5.1 Pre-Usage Stage ..................................................................................... 108
4.5.2 Usage Stage ............................................................................................ 109
4.5.3 Post-Usage Stage ................................................................................... 110
4.6 Data Analysis ............................................................................................. 112
4.6.1 Data Preparation and the Theoretical Background for Qualitative Data
Analysis 112
4.6.2 Data Analysis to Identify Caretaker Personas ....................................... 115
4.6.3 Quantitative Analysis ............................................................................. 116
xiv
5 CARETAKER PERSONAS RELATED TO THE USE OF DAMS ............ 119
5.1 Exploring User Diversity ........................................................................... 119
5.2 Caretaker Personas .................................................................................... 121
5.2.1 Findings ................................................................................................. 125
5.2.2 Descriptive Statistics ............................................................................. 135
5.2.3 Hypotheses on Caretaker Personas ........................................................ 139
5.2.4 Statistical Analyses ................................................................................ 144
5.3 Discussion on Caretaker Personas ............................................................. 146
6 DIMENSIONS TO INCREASE HUMANS' AWARENESS VIA DAMS TO
IMPROVE THEIR CAREGIVING OF DOGS .................................................... 152
6.1 DAMS-Mediated Stage-Based Awareness Model .................................... 152
6.2 Descriptive Statistics ................................................................................. 154
6.2.1 ESM Survey Results .............................................................................. 154
6.2.2 MDORS T-Test Results ........................................................................ 158
6.3 Making Sense of the Tracking Data .......................................................... 161
6.3.1 Comprehending the Information ........................................................... 161
6.3.2 Contextualization of the Data ................................................................ 163
6.3.3 Sensemaking through Comparison with Other Dogs ............................ 165
6.3.4 Sensemaking through Comparison with Familiar Patterns ................... 168
6.3.5 Social / Collaborative Sensemaking ...................................................... 169
6.3.6 Sensemaking with the Assistance of Data Visualizations ..................... 171
6.4 Reflecting on the Tracking Data ................................................................ 174
6.4.1 Checking on Discrepancies ................................................................... 174
6.4.2 Seeking Guidance for Reflection ........................................................... 175
xv
6.4.3 Selective Focus ...................................................................................... 177
6.4.4 Data Handling ........................................................................................ 179
6.4.5 Tracking Trends in the Data .................................................................. 181
6.4.6 Self-Calibration ...................................................................................... 182
6.5 Behavior/Action Stage ............................................................................... 184
6.5.1 Effects on Lifestyle ................................................................................ 185
6.5.2 Change in the Caregiving Attitude/Behavior ......................................... 188
6.6 Implications of DAMS Use ....................................................................... 191
6.6.1 Increase in Caretakers’ Awareness ........................................................ 191
6.6.2 Perceived Effects on the Relationship ................................................... 193
6.6.3 Perceived Effects on Dog Welfare ......................................................... 193
6.7 Barriers to Long-Term Adoption of DAMS .............................................. 194
7 THE MODEL OF DAMS-MEDIATED STAGE-BASED AWARENESS:
DESIGN STRATEGIES ........................................................................................ 197
7.1 Design Strategies ....................................................................................... 198
7.1.1 Enhancing Comprehensibility of the Information ................................. 200
7.1.2 Providing Data Variety to Support Sensemaking .................................. 201
7.1.3 Ensuring Compatibility with Mental Models ........................................ 202
7.1.4 Enhancing Contextualization to Support Meaningful Reflection .......... 203
7.1.5 Providing a Basis for Meaningful Comparison ..................................... 204
7.1.6 Visualizing the Tracking Data to Enhance Understanding .................... 205
7.1.7 Showing Data History to Enable Users to Track Their Progress ........... 206
7.1.8 Supporting Social Sensemaking ............................................................ 206
7.1.9 Providing Guidance to Support Reflection ............................................ 208
xvi
7.1.10 Providing Improved Personalization for Meaningful Reflection .......... 209
7.1.11 Enabling Self-Calibration through Improved Guidance ........................ 210
7.1.12 Motivating to Support Action ................................................................ 210
7.2 Discussion .................................................................................................. 211
8 CONCLUSION ............................................................................................. 214
8.1 Revisiting the Research Questions ........................................................... 215
8.1.1 Q1: What are humans’ different concerns and behaviors that characterize
their caretaking fashion towards their dogs? ......................................................... 215
8.1.2 Q2: How do these concerns and behaviors vary among caretakers? What
are the implications of this user diversity on the design of DAMS in terms of
increasing human awareness? ................................................................................ 217
8.1.3 Q3: How do dog caretakers make sense of and reflect on the data collected
via DAMS? ............................................................................................................ 219
8.1.4 Q4: What are the dimensions to increase humans' awareness through
DAMS to improve their quality of caregiving (of their dogs)? ............................. 220
8.1.5 Q5: What are the design strategies to increase caretakers’ awareness of
their dogs via DAMS to support their caregiving? ................................................ 222
8.1.6 MQ: How can we improve humans' awareness of dogs to enhance their
quality of caregiving through the use of DAMS? ................................................. 223
8.2 Contributions of the Thesis ....................................................................... 224
8.2.1 Caretaker Personas ................................................................................ 224
8.2.2 DAMS-Mediated Stage-Based Awareness Model ................................ 224
8.3 Limitations of the Study and Future Directions ........................................ 225
REFERENCES ...................................................................................................... 229
A. Consent Form ................................................................................................ 253
xvii
B. Participant Information Survey Questions ..................................................... 254
C. ESM Survey Questions .................................................................................. 257
D. Study Procedure ............................................................................................. 259
E. Measurement Tools (Turkish Versions) ........................................................ 263
F. Measurement Tools (English Versions) ......................................................... 273
G. Study Cards .................................................................................................... 283
CURRICULUM VITAE ........................................................................................ 284
xviii
LIST OF TABLES
TABLES
Table 2.1. Five Domains model (Adapted from Mellor & Beausoleil, 2015). ........ 22
Table 4.1. Human-dog relationship measurement tools. ......................................... 95
Table 4.2. Caretaker sample distribution. .............................................................. 102
Table 4.3. Dog sample distribution. ...................................................................... 103
Table 4.4. Coding example from the second interview. ........................................ 114
Table 5.1. An example section from the task-based user segmentation matrix. ... 123
Table 5.2. Participant persona distribution based on the code repetition. ............. 124
Table 5.3. Persona characteristics and typical behaviors and participants showing the
characteristics of each persona. ............................................................................. 125
Table 5.4. MDORS scores of participants. ............................................................ 135
Table 5.5. C-BARQ scores of participants. ........................................................... 137
Table 5.6. Hypotheses for the relation between personas and C-BARQ scores. .. 143
Table 5.7. Information needs of personas. ............................................................ 151
Table 6.1. ESM results regarding the most used Fitbark app features. ................. 155
Table 6.2. ESM results showing the most useful Fitbark app features as perceived by
the participants. ...................................................................................................... 156
Table 6.3. ESM results of the participants’ use frequency of the Fitbark app. ..... 157
Table 6.4. Participants’ opinions about continuing to use/long-term adoption of the
product. .................................................................................................................. 158
Table 6.5. MDORS t-test results. .......................................................................... 159
Table 6.6. Codes and sub-codes related to sensemaking. ...................................... 160
Table 6.7. Codes and sub-codes related to reflection. ........................................... 173
Table 6.8. Codes and sub-codes related to behavior/action. ................................. 184
Table 7.1. Design strategies to support sensemaking, reflection, and action. ....... 199
Table 7.2. Design strategies matrix. ...................................................................... 213
xix
LIST OF FIGURES
FIGURES
Figure 1.1. Transtheoretical model of change (Adapted from Prochaska &
Diclemente, 1986). .................................................................................................. 10
Figure 1.2. Summary of the problem background. ................................................. 13
Figure 1.3. Structure of the thesis. .......................................................................... 16
Figure 2.1. Border Collie dog persona (Hirskyj-Douglas, Read, and Horton, 2017,
p.7). ......................................................................................................................... 31
Figure 2.2. MEAU stages and key aims (Ruge & Mancini, 2019, p.3). ................. 32
Figure 2.3. Canine-centered framework (Freil et al., 2017, p.105). ....................... 33
Figure 2.4. The AWAX iterative development model for the development of
interactive animal technology (Linden, Zamansky & Hadar, 2017, p.425). ........... 37
Figure 2.5. The welfare through competence animal objectives matrix (Webber,
Cobb & Coe, 2022, p.8). ......................................................................................... 38
Figure 2.7. Categories of existing technological applications in ACI studies. ....... 43
Figure 2.8. A dog wearing a pet activity tracker (on the left), the activity tracker
widget (on the right) (Alcaidinho et al., 2015, p.463). ........................................... 45
Figure 2.9. Data flow in an IIS consisting of stakeholders of different species (van
der Linden, 2021, p.5). ............................................................................................ 48
Figure 2.10. Key elements of interactions within an IIS (van der Linden, 2021, p.10).
................................................................................................................................. 49
Figure 3.1. The Functional Triad: Roles Computers Play (Fogg, 2003, p.25). ...... 52
Figure 3.2. Transtheoretical model of change (Transtheoretical model of change
(Adapted from Prochaska & Diclemente, 1986). .................................................... 59
Figure 3.3. The Behavior Change Wheel (Michie, van Stralen, and West, 2011, p.7).
................................................................................................................................. 62
Figure 3.4. General model of information processing (Heijs, 2006, p.45). ............ 65
Figure 3.5. The Stage-Based Model of Personal Informatics Systems (Li et al., 2010,
p.561). ..................................................................................................................... 70
xx
Figure 3.6. Lived Informatics Model (Epstein et al., 2015, p.5). ............................ 72
Figure 3.7. The Technology-Mediated Reflection Model (Bentvelzen, Niess &
Wozniak, 2021, p.6). ............................................................................................... 74
Figure 3.8. The DIKW hierarchy (Rowley, 2007, p.164). ...................................... 77
Figure 3.9. The Data/Frame Theory of Sensemaking (Klein et al., 2006, p.89). .... 79
Figure 3.10. A conceptual model of sensemaking in intelligence analysis (Pirolli
&Card, 2005, p.3). ................................................................................................... 81
Figure 3.11. Dervin’s Sense-Making Theory (Reinhard & Dervin, 2012, p.33). ... 82
Figure 4.1. List of available dog activity and behavior monitoring devices. .......... 99
Figure 4.2. Structure of the methodology. ............................................................. 104
Figure 4.3. Study cards. ......................................................................................... 106
Figure 4.4. Research kit. ........................................................................................ 107
Figure 4.5. Research outcomes. ............................................................................. 111
Figure 5.1. Persona scale - the spectrum of willingness for self-reflection. ......... 144
Figure 6.1. DAMS-mediated stage-based awareness model. ................................ 153
Figure 6.2. Fitbark home page (on the left), dog page with barkpoints data in the
circle (on the right). ............................................................................................... 162
Figure 6.3. Weekly view of data chart (on the left), top dog board (on the right). 166
Figure 6.4. Interactive data map of the daily rest levels of dogs registered in the
Fitbark database (retrieved from
https://public.tableau.com/app/profile/fitbark/viz/shared/KYMHPQ26B). .......... 167
Figure 6.5. Top dog board, discover friends, and pack request features on the app.
............................................................................................................................... 171
Figure 6.6. Daily activity graph. ............................................................................ 172
Figure 6.7. Activity suggestions on the app (on the left), informative blog posts sent
via email (on the right). ......................................................................................... 176
Figure 6.8. Data visualizations. ............................................................................. 178
Figure 6.9. Journal feature. .................................................................................... 179
Figure 6.10. Weekly and monthly graph views. .................................................... 182
Figure 6.11. Daily activity goal settings and goal reminders. ............................... 186
xxi
Figure 6.12. Fitbark collar-mounted device (retrieved from
https://www.fitbark.com/). .................................................................................... 195
Figure 7.1. DAMS-mediated stage-based awareness model. ............................... 198
xxii
LIST OF ABBREVIATIONS
ABBREVIATIONS
ACI Animal-Computer Interaction
ANT Actor Network Theory
C-BARQ Canine Behavioral Assessment and Research Questionnaire
DRM Day Reconstruction Method
HCI Human-Computer Interaction
DAMS Dog Activity Monitoring System
ESM Experience Sampling Method
GPS Global Positioning System
IIS Interspecies Information System
iOS iPhone Operating System
MEAU Method for Evaluating Animal Usability
MDORS Monash Dog-Owner Relationship Scale
MTHD More Than Human Design
PTSD Post Traumatic Stress Disorder
SAR Search and Rescue
TTM Transtheoretical Model of Change
UI User Interface
UX User Experience
1
CHAPTER 1
1 INTRODUCTION
1.1 Problem Background
Interactive technologies have long been in the daily lives of both humans and
animals. They increasingly become more embedded in every aspect of life, changing
the way how human and non-human animals live. Today, animals come into contact
with technologized environments, systems, and products on a day-to-day basis.
While there have been interactive technologies for non-human animals like robotic
milking systems and biotelemetry devices for quite some time, they have typically
been designed without taking into consideration animal factors such as their
cognitive, physiological, and behavioral characteristics, as well as their needs and
preferences, as noted by Mancini (2011). Nevertheless, the lack of an animal-
centered approach during the design and development of such technologies is likely
to affect animals’ welfare adversely as their capabilities, needs, and experiences are
often not considered. While ubiquitous computing technologies continue to become
an integral part of human life increasingly, concerns over the underrepresentation of
animals and the prevalence of anthropocentric approaches in the design of animal
technologies have increased (Mancini, Lawson & Juhlin, 2017).
Along with the increasing concerns over this issue, Animal Computer Interaction
(ACI) has emerged as a research area expanding the boundaries of a relatively mature
field, Human-Computer Interaction (HCI), to include non-human animals as users
for the design and development of technology (Mancini, Juhlin, Cheock, van der
Linden & Lawson, 2014). The development of ACI studies is essential, interactive
technologies have the potential to ensure animals' welfare in an economically
2
sustainable way (Jukan, Bruin & Amla, 1994). These technologies can improve
animals' well-being by providing ways to fulfill their needs, support them in their
assigned functions, and promote the relationship between humans and animals by
enabling communication through various means (Mancini, 2011).
Within the ACI field, dogs hold a unique position as they are firmly ensconced in
human society as companions. Dogs are the oldest domesticated animal, and they
have been living with humans for approximately 30,000 years (Gompper, 2014).
Dogs have been part of humans' daily lives and our evolutionary path. They possess
a unique ability to comprehend human social and gestural cues, which sets them
apart from all other non-human mammals, likely due to their co-evolution with
humans (Hare & Tomasello, 2005). Moreover, due to both their social proximity to
humans and their unique capabilities, they are assigned a variety of roles in human
society, including search and rescue (SAR), bomb and drug detection, assistance,
hearing assistance, guide dogs, medical alert, PTSD/emotional support dogs, and
pets (companion animals) (Freil et al., 2017). Especially in homes, dogs find
themselves in increasingly technologized environments. As their co-evolution with
humans continues, it is for sure that they will be more engaged in interactive
technologies in the upcoming years. Considering that HCI has provided multiple
benefits for humans working with technology by increasing their efficiency,
effectiveness, and productivity; similarly, the development of ACI can offer similar
benefits to dogs interacting with technology (Freil et al., 2017). Besides, dogs'
special social skills make them suitable candidates for ACI studies, probably more
than any other species, because they are easier to work with.
Dogs are the most widely kept pet animal globally, with increasing adoption rates
and spending on related products and services (Grand View Research, 2019).
However, research suggests that the inadequate knowledge of dog owners about their
pets' health and behavior can have negative impacts on dog welfare. As an example,
according to a survey conducted by Rohlf and colleagues (2010), even dog owners
who are considered to be committed fail to follow responsible dog ownership
3
practices, including confinement, registration, microchipping, desexing,
participation in formal obedience training, and regular socialization practices. The
survey also found that certain aspects of dog welfare have worsened instead of
improved in recent years. As cited in Philpotts, Dillon, and Rooney (2019), the top
five welfare concerns related to caretaker practices are as follows;
pedigree or poor breeding practices (Rooney & Sargan, 2010; Packer,
Murphy and Farnworth, 2017),
obesity (Degeling, Kerridge and Rock, 2013; Luno et al., 2018),
dog behavior and training (Blackwell, Bradshaw, and Casey, 2013; Todd,
2018),
dog purchasing and relinquishing behaviors (PDSA, 2017; Packer, Murphy
and Farnworth, 2017; Summerton, 2015; Sandoe et al., 2017),
dog companionship or being left alone for extended periods (RSPCA, 2018;
PDSA, 2017; Norling & Keeling, 2010).
Although caretakers have access to a wealth of information through various
sources such as online resources, volunteer organizations, and veterinarians, it is
surprising that dog welfare continues to decline. The ways in which people care
for their dogs have changed over the years due to changes in human lifestyles.
McGreevy and Bennett (2010) explain that this shift is reflected in what humans
currently expect from their pets and our ability to meet their needs. For instance,
caretakers now spend a lot of money on grooming and dog clothing with the goal
of making their pets happier. Over the past few years, the prevalence of obesity
and obesity-related health issues in dogs has increased significantly, leading to a
decline in their quality of life (Degeling, Kerridge, and Rock, 2013; Luno et al.,
2018; Greenebaum, 2010). Moreover, research indicates that there is a lack of
understanding among dog caretakers about certain aspects of their dogs'
behavior, such as trainability (Mirko, Doka & Miklosi, 2013), play signals (Tami
& Gallagher, 2009), emotional arousal (Kerswell, Butler, Bennett & Hemsworth,
2010), and acute stress (Mariti et al., 2012). Furthermore, a thorough survey
4
conducted on dog owners has revealed that many of them overestimate their
dogs' cognitive abilities (Howell, Toukhsati, Conduit & Bennett, 2013).
Moreover, caretakers’ attribution of anthropomorphic behaviors to dogs, such as
associating certain dog behaviors with their feeling and expression of guilt,
without any sound scientific evidence, might lead to unrealistic expectations
from companion dogs, which in turn might bring about potential relationship
breakdowns (Horowitz, 2009). Additionally, they often fail to recognize severe
signs of common diseases in older dogs, indicating a lack of understanding of
critical issues related to dog health and behavior (Davies, 2011). While
caretakers may not intend to cause any harm or suffering to their dogs, this lack
of awareness can result in various problems, including dysfunctional human-dog
relationships, behavioral issues in dogs, and reduced quality of life for both
parties (Salgırlı et al., 2012).
Research indicates that human behavior and the quality of care provided to dogs
can impact their emotional and physical health. For example, human behaviors
such as positive reinforcement (Deldalle & Gaunetand, 2014), affiliation
(Horvath, Doka & Miklosi, 2008), human attention (Schwab & Huber, 2006),
and safety (Gacsi et al., 2013) are known to contribute to positive emotional
states in dogs, which are likely to produce positive behavioral outputs lead to
positive emotional states in dogs, which can result in positive behavioral
outcomes. Dogs also demonstrate attachment behaviors toward humans that
resemble the bond observed between infants and their caregivers (Serpell, 1996),
as defined in Bowlby’s attachment theory (1958). This similarity is highlighted
by their tendency to engage in proximity-seeking behaviors when the attachment
figure is absent, which serves as a coping mechanism for dealing with stress, as
evidenced by studies such as Schoeberl et al. (2012). Relatedly, Dogs who are
considered by their caretakers to be “meaningful companions” or “social
partners” tend to have lower levels of cortisol in their saliva, which is an indicator
of reduced stress (Schoeberl et al., 2012, p.199). Likewise, clinical studies
demonstrated that interacting with dogs provides several psychological health
5
benefits for humans (Barker & Wolen, 2008; Schneider et al., 2014). Based on
this information, it can be concluded that human attitude is an essential factor in
moderating the human-dog relationship. Therefore, promoting positive human
behavior can enhance the relationship between humans and dogs, resulting in
mutual advantages.
Despite current access to the vast amount of data sources online, finding the most
accurate information on animal health and well-being can be challenging for
caretakers. Especially on welfare-related critical issues, the data must be
appropriately presented to the caretakers to ensure proper guidance (Davies,
2011). It is also of particular significance that the provided information is
accurate, informative, and specific to the individual and species to avoid
unwanted consequences. In this sense, the technologies that utilize smart sensors
to monitor the behavior and health of animals appear to be a promising way for
informing dog owners about their companion animals' behavior and health.
As the Internet of Things (IoT) becomes more widespread, wearable
technologies for animals are also becoming more popular. Such wearables can
enhance the lives of both humans and animals by offering smart features and
experiences. Wearable technologies are one of the most commonly used types of
IoT devices available. These devices provide caretakers essential health-related
data about their companion animals, such as daily activity and sleep levels,
energy expenditure, and rest time. The pet wearables market estimated will grow
at a 14.3% CAGR until 2030 (Grand View Research, 2023). These products on
the market are primarily targeted toward dogs, most probably because the dog
segment accounted for the biggest share of the pet products market by %39 in
2021 and is expected to expand in the near future (Grand View Research Report,
2022). It is evident that the quantity of wearable gadgets designed for dogs and
their human users has increased in the past decade and is predicted to continue
to grow in the upcoming years.
6
However, although there has been a significant amount of research on wearable
technologies for humans, there is limited number of studies examining the effects
of pet activity monitoring devices on the lifestyles of caregivers and the well-
being of animals. As an example of the studies on wearables for animals,
Alcaidinho et al. (2015) examined whether using dog activity monitors can
reduce the return rate of newly adopted dogs from a shelter. Twelve participants
were provided information about their newly adopted dogs' daily activity and rest
levels via a commercially available dog activity tracker attached to dogs’ collars
and synched with the mobile application for eight weeks. The study involved
conducting surveys with adopters, at one-week and one-month intervals post-
adoption, to investigate their experience of the technology and its impact on their
relationship with their dogs. The study showed that providing health and activity-
related data to adopters through a mobile application has resulted in a decrease
in dogs’ re-relinquishment rates. The findings also showed a reported change in
both dogs’ and adopters’ habits regarding increased activity levels and time spent
together based on the information provided via the trackers. Thus, the results
supported the hypothesis that framing dog monitoring data leads to behavior
change in humans, similar to how framing personal tracking information affects
people’s behavior and health.
In another study, it was found that even a simple GPS-enabled collar can improve
human-dog relationships by opening up new forms of interaction (Weilenmann
& Juhlin, 2011). Also, Vaataja et al. (2018) investigated the caretakers’
motivation to use dog activity monitoring devices. In the study, researchers
conducted semi-structured interviews with seven Finnish dog activity tracker
users combined with an international online survey. The semi-structured
interviews aimed to identify how dog caretakers use dog activity monitors in
everyday life, their motivations, and goals to use such devices, their user
experiences, and the overall impact of device use on lifestyles. The findings and
insights gained through the interviews were confirmed and supported by an
international online survey. The study revealed that these devices were primarily
7
utilized to monitor dogs’ health, behavior, and learning-related issues and
balance daily activity levels and rest. The insights gained via the device served
as a motivational factor for behavior change in caretakers to better respond to
dogs’ needs. Zamansky et al. (2019), on the other hand, investigated users’
perceptions of dog activity tracker use and their experiences with these devices.
In the study, eighty-one users of a particular dog activity tracker were recruited
through social media and participated in a questionnaire. The study revealed that
the device use resulted in an improvement in the quality of caregiving and
increased awareness of caretakers’ responsibility for their dogs’ well-being.
Moreover, these devices interestingly led to an increase in caretakers’ own
activity levels and encouraged them to be more active together with their
companion dogs during the day.
As can be seen, there are few qualitative studies that examine dog activity
monitoring devices from a variety of perspectives. However, these studies have
limitations as they do not thoroughly examine how dog technologies affect
human behavior change. Although there are a few longitudinal studies
(Alcaidinho et al., 2015; Zamansky et al., 2019), and some use a large sample
(Zamansky et al., 2019), they do not examine how caretakers interpret dog
tracking data and do not provide an in-depth understanding of data use of
caretakers. Current research mainly concentrates on how dog monitoring
technologies influence the relationship between humans and their dogs. Thus,
how these technologies affect human behavior, how caretakers make sense of
dog tracking data, and their impact on the quality of care is not studied as
holistically as they are with humans.
In recent years, computer technologies have become more ubiquitous, and they
have had a significant impact on human behavior. The relationship between
technology and human behavior is symbiotic: technology affects human
behavior, while human behavior affects the usage of technology (Slob &
Verbeek, 2006). While the original purpose of computer technology was not to
promote behavior change, in recent years, researchers have become interested in
8
using these technologies to promote positive changes in behavior. This area of
study is known as persuasive technology, which refers to interactive computing
systems designed to modify people's attitudes, behaviors, or both (Fogg, 2003).
To better understand technology’s persuasive potential, it would be helpful to
mention the different roles that computer technologies play in human life.
On the functional triad framework, Fogg (2003) proposes that computing
technologies have three essential functions from the users’ perspective: tools,
mediums, and social actors. In their role as tools, computer technologies aim to
equip users with new capabilities, allowing them to do activities more easily and
effectively. Computer technologies as tools can influence people’s attitudes or
behaviors in specific ways, such as by making the predetermined goals easier to
achieve, guiding people through a process or experience, or performing
calculations or measurements that motivate them. Computer technologies also
function as mediums. These technologies have the ability to influence people's
attitudes and behaviors by simulating experiences and enabling them to explore
cause-and-effect relationships within those experiences. As social actors,
computer technologies can reward people with positive feedback, model a target
behavior or attitude, and provide social support to shape their behavior or
attitudes. According to Fogg (2003), the information and feedback provided via
interactive technologies are essential motivators for people to perform a
behavior.
One example of persuasive technologies is personal health informatics systems.
These systems allow individuals to modify their behavior by monitoring
themselves and analyzing data, all with the goal of reaching a specific target. By
collecting and examining data, these systems help users attain their objectives by
presenting the data clearly and providing feedback when necessary (Fogg, 2003).
Along with providing self-monitoring data, persuasive technologies utilize
different strategies to encourage behavior change in individuals. Fogg (2003)
describes seven types of behavior change techniques included in persuasive
technologies: tunneling, tailoring, suggestion, self-monitoring, surveillance, and
9
conditioning. Proper use of persuasive technologies has the potential to enhance
people's awareness and motivation to perform desired behaviors. Previous
studies have explored wearable fitness trackers for humans, which are a form of
health informatics system, and have found that providing personal health-related
insights by such devices can result in long-term behavior change in users (Choe,
Lee, Munson, Pratt & Kientz, 2013). Another study examining the effects of
using a fitness tracker on people’s activity levels revealed that the device use
resulted in a significant increase in participants’ activity levels (Cadmus-Bertram
et al., 2015). Moreover, further studies indicate that activity monitoring devices
help users gain a more comprehensive understanding of their behaviors and
activities within the context of the data provided by these devices (Fritz, Murphy
& Zimmermann, 2014).
Along with technology's persuasive role, it is also crucial to consider how
individuals change their behavior. The Transtheoretical Model (Figure 1.1),
developed by Prochaska and Velicer (1997), The Transtheoretical Model,
developed by Prochaska and Velicer in 1997, outlines five stages: pre-
contemplation, contemplation, preparation, action, and maintenance. These
stages are called the stages of change. According to Prochaska and Velicer
(1997), these stages are referred to as the stages of change, and the primary
strategy for promoting positive behavior change is to create awareness related to
the issues associated with current behavior to move from the pre-contemplation
to the contemplation stage. If people are not aware of their problematic behaviors
or the need for a behavior change, it is unlikely that any change will occur,
whether it is adapting current behaviors to become healthier or adopting new
desired behaviors, awareness is necessary for change. The model suggests that
awareness can be achieved through knowledge. This includes informing people
about their current problematic behaviors, the potential outcomes, and alternative
behavior patterns. (Prochaska & Velicer, 1997).
10
Figure 1.1. Transtheoretical model of change (Adapted from Prochaska &
Diclemente, 1986).
As discussed earlier, the lack of knowledge about dogs' health, behavior, and
responsible owner practices among caretakers has negative effects on dog welfare.
Although caretakers do not intentionally cause harm to their dogs, the reported
deterioration of dog welfare in homes is primarily related to this lack of awareness.
Therefore, activity and behavior monitoring systems designed for dogs can promote
positive behavior change in humans, similar to personal health informatics systems'
effects on human behavior. As these devices provide feedback to caretakers, they
can improve dog welfare by increasing their awareness of their dogs' needs. Dog
activity monitoring devices are similar to human fitness trackers, using
accelerometers to measure physiological parameters such as activity levels, walking
distance, energy expenditure, and sleep quality. These devices can also connect to
computing applications to track dogs' health and behavior over time, motivating
caretakers to keep track of their dogs' progress and adjust their behavior to improve
the quality of their care. Studies show that using self-monitoring techniques either
by technological interventions or by diary methods is found to be motivating for
people to change their behaviors to be more active and lose weight in their daily lives
11
(Munson & Consolvo, 2012; Wang et al., 2014; Fritz, Murphy & Zimmermann,
2014; Normand, 2008; Burke, Wang & Sewick, 2011). Moreover, data-driven
feedback and information provided by these devices can also encourage people to
change their behavior, as exemplified in studies with human activity trackers
(Collins, Cox, Birds & Harrison, 2014; Consolvo et al., 2008; Cuttone et al., 2013;
Fritz et al., 2014; Hori et al.; 2013; Kay et al., 2012; Li et al., 2011).
Using state-of-the-art technology to increase human awareness to support dog
welfare has a two-fold effect. Evidence suggests that supporting dog welfare through
improved caregiving can also benefit humans, as studies show that human-dog
interactions affect human psychological and physiological health (Beck & Katcher,
2003). Thus, pet activity monitors can make a significant contribution to the well-
being of both humans and dogs not only by improving physical activity but also by
increasing caretakers’ awareness of their dogs and enhancing the quality of their
caregiving. In this line, it is important to examine how systems for monitoring dog
activity can mediate the relationship between humans and dogs, as it may reveal
many intervention areas to support caretakers in reflecting on the tracking data and
guide their behavior to make well-informed/data-driven decisions regarding dog
care, and thus, indirectly support dog welfare.
To sum up, as can be implied from this chapter, caretakers lack a thorough
understanding of - or misinterpret - the health and behavior of their pet dogs, with
potential implications for dog welfare. Therefore, to contribute to dog welfare in
domestic settings, there appears to be a need to inform caretakers about their
companion dogs’ health and behavior. Today, with the increasing popularity of
animal technology (Grand View Research, 2020), various commercially available
devices for dogs are increasingly being used by consumers. These devices are
growingly using smart sensing technologies that collect different types of data and
enable different forms of human-animal communication that were not possible
before. While these technologies have the potential to assist human users in a variety
of ways, they also introduce extra complexity in two ways. First, as animals become
targets of such technologies, they are no longer passively exposed to the technology
12
but turn into stakeholders in the interaction processes (Westerlaken & Gualeni,
2016). Animals and humans alike are involved in complex interactions between
humans, animals, and technology. Second, humans have to deal with increasing
amounts of data on a daily basis due to their daily interactions with data-driven
technologies. However, how caretakers interpret the tracking data collected and
provided by these devices and how this data affects their caregiving practices
remains unknown. Moreover, little design knowledge has been formalized on how
to design dog activity monitoring systems to provide monitoring data in a meaningful
way to guide human behavior and improve the quality of dog care. Therefore, there
is a growing need for research on how such technology is actually used and what
effects it has on dogs and caretakers. Thus, the primary work of this thesis is
concerned with exploring how to improve the quality of human care of dogs by
increasing their awareness through dog activity monitors. The summary of problem
background can be seen on Figure 1.2.
13
Figure 1.2. Summary of the problem background.
14
1.2 Aim of the Study and Research Questions
This thesis aims to develop a theoretical model on how dog activity monitoring
systems (DAMS) for companion dogs can mediate the human-dog relationship to
improve humans' caregiving by examining the potential and possibilities of these
technologies. To achieve this aim, the major question the study targets to answer is:
MQ: How can we improve humans' awareness of dogs to enhance their quality of
caregiving through the use of DAMS?
To answer these question the sub- questions are as follows:
Research Questions:
Q1: What are humans’ different concerns and behaviors that characterize their
caretaking fashion towards their dogs?
Q2: How do these concerns and behaviors vary among caretakers? What are the
implications of this user diversity on the design of DAMS in terms of increasing
human awareness?
Q3: How do dog caretakers make sense of and reflect on the data collected via
DAMS?
Q4: What are the dimensions to increase humans' awareness through DAMS to
improve their quality of caregiving (of their dogs)?
Q5: What are the design strategies to increase caretakers’ awareness of their dogs
via DAMS to support their caregiving?
15
1.3 Structure of the Thesis
Figure 1.3 outlines the structure of the thesis. While literature review is covered in
Chapters 2 and 3 constituting the background of the study, Chapters 5, 6 and 7
answers the research questions. Finally, the Conclusion Chapter revisits the research
questions and discusses the contributions and the limitations of the study.
16
Figure 1.3. Structure of the thesis.
17
CHAPTER 2
2 ANIMAL WELFARE AND ANIMAL-COMPUTER INTERACTION
This chapter provides an overview of the existing literature in the field of animal-
computer interaction (ACI) to serve as a basis for the thesis, which aims to develop
a theoretical model of how dog activity monitoring systems for companion dogs can
improve humans' caregiving by examining the potential and possibilities of these
technologies.
In this chapter, firstly, different views regarding the definition of animal welfare and
assessment of animal welfare are discussed. Then, key terms and concepts related to
animal-computer interaction (ACI), including user-centered design and animal-
centered design, are defined. Also, the history and the current state of the ACI field
are discussed. Then, existing methodological approaches, theories, frameworks, and
applications in the field of ACI are reviewed. Following this, a brief overview of the
current ethical procedures in animal research is presented, followed by a review of
the recent research and practice regarding dog tracking and monitoring technologies
in ACI.
2.1 Animal Welfare
The design of interactive technology with an animal-centered perspective requires a
clear understanding of animal welfare. The idea of animal welfare can be compared
to concepts like quality of life and well-being (Webber, Cobb & Coe, 2022). The
state of an animal's welfare can vary from very poor to very good, and this depends
on various factors that impact the animal's life (Broom, 1996). According to the OIE
World Organization for Animal Health (2013), animal welfare refers to an animal’s
18
physical and mental state concerning the environment where it lives and works. The
reasoning for animal welfare is based on the idea that animals are sentient beings
that have the capacity to feel both positive and negative emotions and have a desire
for positive experiences (Boissy et al., 2007). Turner (2006, p.6) states that an animal
is sentient if “it is capable of being aware of its surroundings, its relationships with
other animals and humans, and of sensations in its own body, including pain, hunger,
heat or cold.”. Therefore, the well-being of animals is essential not just because it is
of instrumental value that humans confer on the animal as a means to achieve a
particular goal, but it is intrinsically valuable as an end in itself and worthy of moral
consideration (Rollin, 1992). That is, animals have value in their own right, and
because of that, it is the moral obligation of humans to ensure their quality of life.
However, this understanding that animals are sentient beings, and it is our moral
responsibility to provide a good life for them requires us to identify their needs first.
It is especially crucial for animals under human care (whether in domestic settings
as pets, captive animals in zoos, or test animals) where their environmental, social,
and behavioral options are often restricted within their living contexts (Coleman,
2018; Perdue, Sherwen & Maple, 2020). Yet, although animal welfare science and
animal ethics have a shared moral foundation, they should not be confused with one
another. Animal welfare science does not deal with how humans need to treat
animals; instead, it acts as a connecting concept between scientific research and
ethical considerations (Fraser et al., 1997).
Deciding on the state of an animal’s welfare is not an easy task because it is an ever-
changing state, depending on various internal and external factors. The Five
Freedoms identified by the Farm Animal Council (1979) outline the minimum
requirements for animal welfare as follows:
Freedom from hunger or thirst and malnutrition: by giving access to fresh
water and a balanced diet to maintain good health,
Freedom from discomfort: by providing a suitable environment that offers
shelter and rest,
19
Freedom from pain, injury, or disease: by providing preventative healthcare
or prompt treatment,
Freedom to express natural behavior: by offering enough space, proper
resources, and social interaction with other animals of their species,
Freedom from fear and distress: by ensuring that the animal's conditions and
treatment do not cause mental suffering.
The Five Freedoms principles were used as a guideline to determine the baseline of
an acceptable level of welfare that should be taken into account for the management
of settings intended for animals. These principles focus on minimizing suffering and
freedom from negative conditions with little or no consideration for the promotion
of positive welfare states. However, there are different views on these criteria as
contemporary approaches to animal welfare science underline the advancement of
positive states (Fraser, 2008). Besides, Dawkins (1990) states that when assessing
animal welfare, not just the risks to an animal’s survival but how an animal perceives
the situation from its point of view should be considered. This part, the animal’s
viewpoint, is integral to animal welfare science as understanding the subjective
experience of animals is the primary concern of the studies in this domain (Dawkins,
1990). How an animal perceives a situation is an entirely subjective experience
affected by how the environment it inhabits impacts its affective states (Broom,
1996; Mellor et al., 2020). Therefore, this subjective experience can only be assessed
and not measured (Rault, Webber & Carter, 2015).
There are different concepts regarding the assessment of animal welfare. The first
view is a functioning-based approach, considering the level of reproduction, physical
health, growth, and injury as indicative of animal welfare (McGlone, 1993). This
view suggests that “an animal is in a poor state of welfare only when physiological
systems are disturbed to the point that survival or reproduction are impaired.”
(McGlone, 1993). The second approach concentrates on the affective states of
animals, emphasizing that their feelings directly impact their welfare without
requiring that they necessarily affect their physical health (Dawkins, 1990). Thus,
20
evaluating animal welfare based on physical health or fitness alone is inadequate as
the concept of welfare also depends on the animal's emotional and mental state. For
instance, animals may experience psychological stress or anxiety that can negatively
impact their welfare, even if there are no obvious physical symptoms (Rault, Webber
& Carter, 2015). Therefore, considering the animal's subjective experience is
important when evaluating their welfare. On the other hand, the natural living
approach advocates that the extent to which animals can behave naturally is a
determining factor for animal welfare because “it is necessary over a period of time
for the animal to perform all the behaviors in its repertoire because it is all functional;
otherwise, it would not be there.” (Kiley-Worthington, 1989, p. 333). It is proposed
that animals have a nature or ‘telos’ that is made up of genetically encoded needs,
desires, and behaviors, and acting according to their telos is integral to good welfare
(Rollin, 1993). Dawkins (2021, p.1) provides a contemporary and animal-centered
perspective on animal welfare by defining positive welfare as "a combination of
good health and having access to what the animals themselves want".
Needing to address these different views regarding the assessment of animal welfare
and to extend the scope of conceptual frameworks identifying only negative welfare
states to include positive states as well (Farm Animal Welfare Council, 2009;
Webster, 2011; Edgar et al., 2013), the Five Domains of Animal Welfare Model was
devised to assess the welfare states of “sentient animals used in research, teaching,
and testing (RTT)” (Mellor & Reid, 1994, p.241). It provides a structured approach
to evaluate signs of both internal and external physical and functional conditions and
environmental factors, which then have an impact on the psychological experiences
of animals. The model comprises five domains, including four related to functional
variables: nutrition, environment, health, and behavior, and mental state (Table 2.1).
The first three domains mainly concentrate on the presence or absence of internal
physiological and survival-related factors such as nutrition, environment, and health-
related problems. The factors grouped under these three domains are crucial for the
functioning of animals’ genetically encoded biological mechanisms (Fraser &
Duncan, 1998; Panksepp, 2005; Denton et al., 2009). On the other hand, the fourth
21
domain includes situation-related factors linked with environmental conditions that
may restrict animals from performing their natural behaviors to the extent that would
potentially pose a challenge to their survival (Mellor et al., 2009). Once all the
internal and external factors in the first four domains of the model are systematically
evaluated, the emotional states resulting from these factors are accumulated in the
fifth "mental state" domain. The emotional experience of the animal is assessed in
this domain, which would determine the animal's overall welfare status (Mellor et
al., 2009).
The Five Domains Model, in contrast to the Five Freedoms, considers both positive
and negative mental states of animals. As a result, it is an effective method of
assessing animal welfare. The negative aspects stated in the Five Domains model
include “breathlessness, thirst, pain, hunger, nausea, dizziness, debility, weakness
and sickness, which are mainly associated with sensory inputs generated internally,
and anxiety, fear, frustration, anger, helplessness, loneliness and boredom, which are
associated mainly with the animal’s cognitive assessment of its external
circumstances.” (Mellor & Beausoleil, 2015, p.242). In addition to all aspects
concerning animal welfare discussed in this chapter so far, play is also found to be
related to animal welfare from four points onwards. First, it is seen as a possible
indicator of welfare, as it suggests an absence of threats to the animal's fitness (Fraser
& Duncan, 1998). Second, it is also associated with positive emotions in animals
(Fraser & Duncan, 1998). Moreover, it is also regarded as a method to improve
welfare because it provides long-term and short-term physiological and
psychological health benefits that may enhance animal welfare (Held & Spinka,
2011). Lastly, it holds the potential to contribute to well-being in animal groups as it
is socially contagious (Held & Spinka, 2011).
22
Table 2.1. Five Domains model (Adapted from Mellor & Beausoleil, 2015).
23
2.2 Animal-Computer Interaction (ACI)
Although animal welfare, animal ethology, and physiology have long been studied,
design for animals has traditionally been driven by economic interests and human
preferences rather than by an understanding of their evolutionary nature and their
welfare (Webber, Cobb & Coe, 2022). In line with this conventional view, up until
the turn of the century, animals’ requirements were often disregarded during the
design and development of animal technology as they were seen more as the subject
rather than system users (Hirskyj-Douglas & Read, 2014). However, with the
increasing integration of technology into human lives, it has been realized that
humans are not the only species that come into contact with interactive technologies.
Thus, it has become of interest how these systems affect animal behavior and the
human-animal relationship.
In line with the growing interest in this area, Animal Computer Interaction (ACI) has
emerged as a considerably new research field that was coined with the ACI manifesto
in 2011 (Mancini, 2011). It mainly studies “the interaction between animals and
computing technology within the contexts in which animals habitually live, are
active, and socialize with members of the same or other species, including humans”
(Mancini, 2011, p.1). It is a vast area of research as this ‘interaction’ will vary
substantially based on the context, environment, species, the category into which the
animal fits, including wild, domestic, working, farm, or laboratory animals, as well
as their individual differences (Mancini, 2011). Strongly influenced by the well-
established field of Human-Computer Interaction (HCI) in terms of methodological
approaches (Mancini et al., 2014; Resner, 2001; Westerlaken & Gualeni, 2014), ACI
focuses on the usability of technology intended for animals’ use and the user
experience of animals (Lee et al., 2006). Today, it has been seen that technology can
benefit both humans and animals in various ways, such as enabling human-animal
communication, monitoring animal health and behavior, supporting service and
24
working animals, and also for environmental monitoring and control in places where
animals live (Jukan, Masip-Bruin, & Amla,1994).
Along with the development of computerized technology, the advancement of the
ACI field could provide further advantages. As Mancini (2011) stated in the ACI
manifesto, the advancement of the field can;
Improve the human-animal relationship by enhancing interspecies
communication, which would lead to an increased understanding between
them.
Help to comprehend animals’ cognitive processes better through animal
behavior and usability studies with the help of animal technologies.
Increase the efficiency of animal conservation studies by guiding the design
of tracking and monitoring technologies to minimize their impact on animals
and maximize the reliability of the gathered data.
Contribute to the economic and ethical sustainability of the farming industry
and food production by giving animals greater control over their environment
or providing them with environmental enrichment to reduce their stress levels
and susceptibility to illness.
Be beneficial to specific human user groups as well, by exploring new ways
for eliciting requirements from non-verbal users or users with limited
cognitive abilities, by expanding the boundaries of HCI research.
Moreover, as ACI is naturally aligned with animal welfare (Rault, Webber, & Carter,
2015), the well-being of animals is one of the primary concerns for the studies in this
area.
It is essential to clarify how an animal is defined to identify the scope of ACI better
and differentiate it from HCI. The Oxford Dictionary (2019) offers two definitions.
In the first one, an animal refers to “a living organism that feeds on organic matter,
typically having specialized sense organs and nervous system and able to respond
rapidly to stimuli.” Based on this definition, humans are also included in the category
of animals. On the other hand, according to a second definition, which reflects a
25
more ordinary usage and the standard anthropocentric view, it means “an animal as
opposed to a human being” (The Oxford Dictionary, 2019). Following these
definitions, it is possible to look at ACI from two perspectives 1) as a subfield of
HCI focusing on non-human animals or 2) as an inclusive term covering HCI, and
Child Computer Interaction (CCI), considering humans as animals (Hirskyj-Douglas
et al., 2018).
Nevertheless, with an emphasis on the differences between human and non-human
animals, ACI is generally focused on the study of non-human animals regarding the
lack of research in this area. Focusing on non-human animals as their primary users,
ACI also seeks to adopt a user-centric approach to designing animal technologies.
User-centered design is a broad term meaning that the design process is shaped
around its intended users to meet their needs and preferences (Abras et al., 2004).
Thus, the key principle within user-centered design is the involvement of end users
in the design process to influence the design. It is thus essential to prioritize animal-
centeredness in ACI to ensure that design decisions are informed by the needs of
animals as its end-users, with the ultimate aim to provide technology that truly
benefits them.
In ACI, both terms ‘interaction’ and ‘user’ are utilized in a broad sense, including
whether the user interacts with the system actively and intentionally (Robinson et al.,
2014), actively and unintentionally (Mancini et al., 2015), passively and deliberately
(Cheok et al., 2011) or passively and unintentionally (Mancini et al., 2012). In
interaction design, it is given high priority that the needs and preferences of users
should be considered during the development of technology to enable the creation of
more usable systems and better user experience (Preece et al., 2015). To achieve this,
it is essential first to identify the requirements of prospective users to guide the
design and development of interactive technology. However, one of the most crucial
challenges within ACI research is eliciting requirements from non-human animals
who are non-verbal users (Hirskyj-Douglas et al., 2016). The biological differences
between humans and animals and the established anthropocentric approaches within
26
interaction design and HCI make it hard to understand what animals actually need or
prefer and make the right design decisions accordingly.
2.2.1 Animal-Centered Design and Its Challenges
One of the main challenges faced in ACI is to achieve animal-centeredness in the
design of animal technology, that is, identifying and prioritizing animal needs and
preferences at the center of the design. Adopting an animal-centered approach to the
design of animal technologies should be the main focus to ensure that technological
interventions result in long-term mental and physical benefits for animals (Webber,
Cobb & Coe, 2022). The field of Animal-Computer Interaction (ACI) suggests that
the interaction design methods employed in human-centered design projects can be
modified and applied to identify new possibilities for technology to enhance the
welfare of animals (Mancini, 2011; French, Mancini & Sharp, 2017). A challenge
exists, however, in determining what animals ‘need’ or ‘want’ (North & Mancini,
2016).
ACI, being a nascent field, shares close ties with HCI in terms of theoretical models
and research approaches. However, eliciting requirements from/identifying the
needs of animals is a significant challenge due to interspecies differences and
communication barriers (Zamansky et al., 2017), as most methods employed in HCI
are based on written or verbal communication. To overcome this challenge and
establish animal-centric approaches, researchers have investigated how various
methodologies from fields such as human-computer interaction (HCI) and child-
computer interaction (CCI) can be modified for use in ACI. In this section, the
methods used in ACI research have been reviewed by providing examples from the
literature.
27
2.2.2 Research Methods in ACI for the Identification of Animal Needs
Observation or ethnography is a widely used tool in animal science that has been
employed for a long time to understand animal behavior, particularly in natural
settings (Vicedo-Castello, 2017). As animals communicate with other animals
through behaviors, including gestures, postures, and sounds (Broom & Fraser, 2015),
these behaviors convey lots of information for researchers. Therefore, it is one of the
widely adopted requirement elicitation methods in ACI research. Ethnography is a
qualitative research method based on the observation of people to gain insights into
how they interact with the things in their natural environment (Hammersley, 2007).
Observation is also one of the most widespread ethnographic methods in HCI, in
which a researcher observes the actual behavior of users without directly interfering
with them. What makes observational techniques so useful for animal studies is that
they allow researchers to collect data directly from animals through the observation
of exhibited behaviors in their natural environment (Vicedo-Castello, 2017).
Therefore, analysis of animal behavior through observation is key to ACI research
to understand animals’ perceptions of a proposed design solution. Methods for
eliciting ethnographic data from animals have been previously applied in ACI
studies, as exemplified by Mancini et al. (2014).
Meyer, Forkman, and Paul (2014) have noted that animal behavior assessment has
been traditionally ethogram-based (a description of typical behaviors performed by
a species), as outlined by Martin and Bateson (1993). Ethograms have been used in
some ACI studies. For example, Baskin and Zamansky (2015) used ethograms in
their study to investigate dog user experience with interactive technology. The study
explored dogs' interactions with two digital games presented on a tablet. In this study,
the authors evaluated the behaviors of their participants with reference to a dog
ethogram. Moreover, observation of animal behavior is often combined with
physiological measurements for further interpretation and improved reliability
(North & Mancini, 2016). However, qualitative evaluation methods recently have
28
become more commonplace in animal research (Wemelsfelder, 2007; Uher &
Asendorpf, 2008; Meagher, 2009; Walker et al., 2010).
Observational methods have also been utilized for usability testing studies in ACI,
during which the researchers observed non-human users as they interacted with the
proposed systems (Ritvo & Allison, 2014; Westerlaken & Gualeni, 2014). However,
with concerns over the human exceptionalism inherent in ethnographic research
(Kirksey & Helmreich, 2010), the emergence of multispecies ethnography has
underlined that ethnographic studies should not be confined to humans as human
lives are entangled with the lives of other species. In the context of ACI, Mancini,
van der Linden, Bryan, and Stuart (2012) and Mancini, Harris, Aengenheister, and
Guest (2015) used multispecies ethnography in which observations of animal
behavior were combined with expert advice and involved caretakers as mediators to
investigate technology-mediated human-dog relationships. Similarly, North (2016)
suggests mitigating human supremacy in ethnography by combining it with
quantitative ethology-based approaches to analyze animal interactions and
behaviors, proposing a new method with the term ethographology.’ However,
although there is extensive literature on how to observe dogs' behaviors in laboratory
settings (Hasen, 2003; Quinn et al., 2007) using technology (Zeagler et al., 2016;
Gergely et al., 2014), there is currently no widely accepted approach in ACI for
studying dogs' behaviors within their natural domestic environments using
observational methods.
In ACI studies, it is often required to gather observational data from caretakers as
they are familiar with their dogs’ routines and behavior patterns. Studies on the
questionnaires used for dogs’ psychometric evaluation show that caretakers’
subjective assessments of their dogs’ behavior might lead to faulty results (Dodman,
Brown & Serpell, 2018). Thus, considering this issue, Hirskyj-Douglas (2017)
presented the dog information sheet (DISH) to inform caretakers/observers regarding
typical behaviors that dogs exhibit when interacting with technology. The tool is
developed based on the RSPCA (2015) dog behavior guidelines and a veterinary
29
consultant elaborating on this information. It is aimed with DISH to improve the
accuracy of human observers’ evaluation of dog behavior (Hirskyj-Douglas, 2017).
Another methodological approach in ACI is employing conversational/interview
techniques applied to collect data from experts or caretakers to identify animal
requirements. For example, in their study, Mancini et al. (2014) employed semi-
structured interviews with human caretakers by asking them questions about the
well-being and behavior of their companion dogs, their daily routines, and the
perceived benefits of technology for both humans and dogs. On the other hand,
Zeagler et al. (2016) conducted semi-structured interviews with experts to develop a
wearable interface for search and rescue (SAR) dogs to allow remote communication
with their handlers via a mobile application.
In addition to the methods in animal research mentioned so far, there are also a few
design methods adapted from the HCI field to understand the needs of animals. One
of the most promising techniques used to investigate animals’ design preferences in
ACI research is physical prototyping. Providing animals with prototypes of a
proposed system is found to be an effective way to allow them to express their
preferences directly and to gather feedback on possible design solutions. Physical
prototyping for requirement elicitation from non-human users through adopting a
research-through-design approach has been exemplified in several studies with
diabetes alert dogs (Robinson et al., 2014), with cancer detection dogs (Mancini,
Harris, Aengenheister & Guest, 2015), and with captive elephants (French, Mancini
& Sharp, 2015). Moreover, physical prototyping may allow the execution of
participatory design methods, such as co-design, by involving animal stakeholders
in the design processes. Taking its roots in user-centered design and participatory
design, co-design refers to the collaborative participation of both trained designers
and non-designers in the design process. (Sanders & Stappers, 2008). In their study
on two design projects, Westerlaken and Gualeni (2016) suggest involving animals
and humans in the design process as actors through multiple prototype iterations
(Westerlaken & Gualeni, 2016).
30
Another method that is transferred from the HCI methodology to ACI is personas. A
persona is a representation of a hypothetical user created based on either data or
assumptions considering the characteristics of the target user (Nielsen, 2017) used to
represent actual or potential users’ behaviors, goals, motives, and informational
needs (Blomkvist, 2002). The goal of using personas is to better inform the design
process about potential users (Pruitt & Adlin, 2006) to create more usable products
or systems. It is particularly a useful tool for the design and development of
interactive systems targeted towards animals, as their physiological and
psychological characteristics and requirements may be overlooked due to
interspecies differences, which could result in them being unable to perform the
desired task.
In ACI studies, Robinson et al. (2014) explored the use of dog personas to aid in
designing an emergency alarm system for diabetic assistance dogs to call for help in
case of an emergency. The dog personas in the study were generated based on the
researchers’ observations of the system’s potential users, a group of mobility service
dogs, and medical detection dogs. The personas included aspects related to the
system’s design, such as dogs’ size, age, breed, attitude, and play preferences and
behaviors. Additionally, concerning the dog personas, researchers also created
caretaker personas to present the human-dog relationship and the specific domestic
context. Building on this study, Hirskyj-Douglas, Read, and Horton (2017)
developed a set of dog personas to be used as a tool to represent dog requirements
for the design of screen systems. The personas created in the study are based on the
data gathered from dog caretakers through questionnaires. It is intended to present
different dog personalities with the related aspects that could guide the design of
screen systems for dogs, such as their general temperament, preferences, and
attention to technology and demographic information (Figure 2.1).
31
Figure 2.1. Border Collie dog persona (Hirskyj-Douglas, Read, and Horton, 2017,
p.7).
User involvement is an essential part of Human-Computer Interaction (HCI)
practices, as it provides essential guidance for designers and developers in the
creation of computer interfaces and interactions. However, since existing practices
are inadequate in obtaining guidance from non-human users, as they are mostly
based on verbal methods, Farrell, McCarthy, and Chua (2018) propose ways for
adapting expert techniques and processes from HCI to the field of Animal-Computer
Interaction (ACI). These methods include “controlled testing, direct observation,
heuristic evaluations, user profiling, interviews, focus groups, PICTIVE prototyping,
and cognitive walkthroughs”, particularly for the design and development of dog
training technologies (Farrell, McCarthy, and Chua, 2018, p.6).
In addition to the above-mentioned methods and approaches in ACI practices, the
assessment of usability in animal technology is another challenge that needs further
consideration. As usability is a key measure of user experience, usability assessment
should be an indispensable part of the development of animal technology. Usability
evaluation with dog users has been exemplified in many studies so far in ACI
(Mancini et al., 2016; 2015; Zeagler et al., 2014; Jackson et al., 2015; Bryne et al.,
32
2017). In their work, Ruge and Mancini (2019) highlight two primary difficulties
that arise when assessing usability for animals. The first challenge concerns the
variations in cognitive, physical, and sensory abilities between human evaluators and
animal users. The second challenge relates to the focus of most usability evaluation
techniques, which are primarily designed for human use and are therefore human-
centered. To address these issues, they propose the Method for Evaluating Animal
Usability (MEAU), in their study applied to evaluating the usability of different
access controls for Mobility Assistance Dogs (MADs) as users. MEAU aims to
create a framework to assess the usability of interactive technology for animal users,
considering their unique characteristics. It also aims to reinterpret established
interaction design principles to cater to animal-centric needs and requirements.
Additionally, MEAU seeks to establish a process for evaluating animal usability that
recognizes the disparity between human evaluators and animal users (Ruge &
Mancini, 2019). The model involves seven distinct stages, as shown in Figure 2.2,
and includes creating use cases for the interaction to be evaluated.
Figure 2.2. MEAU stages and key aims (Ruge & Mancini, 2019, p.3).
Additionally, Freil et al. (2017) propose a dog-specific framework for analyzing
technological systems based on Don Norman’s seven-stage model (Norman, 2013),
a well-known and largely applied model to evaluate computer interfaces in HCI. By
adapting the framework for dogs’ interactions, researchers aimed to provide a tool
33
for the design and development of animal-centered technology (Figure 2.3). In the
model, interaction is separated into two phases: execution and evaluation. Execution
refers to the stage that the user decides on which action to perform on an interactive
system. Any failure here leads to the ‘gulf of execution’, which is the gap between
the user’s goal and the means to accomplish it. The user begins by setting a goal and
planning a sequence of actions to achieve it. Then, in the evaluation phase, the user
assesses the outcome of each action. Failure to understand the result of an action can
lead to the gulf of evaluation. When a user completes an action, they assess the
current state of the system, interpret the results, and compare them to their intended
objective. The framework is flexible enough to apply to different contexts, whether
a user is a dog or human, based on the assumption that every user shifts between
execution and evaluation phases during their interactions with a computerized
system (Freil et al., 2017).
Figure 2.3. Canine-centered framework (Freil et al., 2017, p.105).
2.2.3 Theories, Models, and Frameworks within ACI
This section briefly reviews the existing theoretical frameworks and models that
illustrate how animal-centeredness in technology design and development can be
achieved by placing animal welfare at the heart of these processes.
34
2.2.3.1 Actor-Network Theory
Actor-Network Theory (ANT) is a theoretical and methodological approach that
views both human and non-human actors as equal stakeholders in a constantly
shifting network of relationships. In ANT, the term ‘network’ refers to a system
composed of objects, actors, and relationships between human and non-human
agents that mediate one another, shaping the resulting actions and experiences
(Latour, 2007). Based on the theory of Latour, Verbeek (2011) suggests that humans
are not passively exposed to technology. Still, both technological artifacts and their
users could mutually shape their role in a technologically mediated interaction. It is
argued that this is also relevant in the case of animals that are involved in such
interactions. Depending on the context or network, an artifact is first interpreted by
a human or animal, and then it is utilized in one way or another (Verbeek, 2011). In
other words, by acknowledging both human and non-human stakeholders as
individuals and actors, Latour proposes a different view from human-centeredness
and argues that actors’ actions are not simply the result of their intentions. Rather
they are mediated by other interrelated factors, such as sociocultural and material
environments (Latour & Venn, 2002). Similarly, Haraway (2008) takes a
multispecies perspective and argues that humans and animals are interconnected by
the mere fact of existing together in the same world. Emphasizing the
interconnectedness of the living world, she asserts that it is wrong to regard humans
as separate from it. She states that; “If we appreciate the foolishness of human
exceptionalism, then we know that becoming is always becoming with, in a contact
zone where the outcome, where who is in the world, is at stake.” (Haraway, 2008,
p. 244).
Building on the perspective that ANT provides, design space has shifted its focus
from anthropocentric perspectives over the past decade placing humans at the center
and expanded to include approaches and methodologies of more-than-human design
(MTHD) (Coşkun et al., 2022). Products equipped with modern sensing and
processing capabilities have the ability to affect not just how other products react but
35
also how humans interact with them and with one another (Cila et al., 2017). This
shift has necessitated that designers and researchers to broaden their attention from
the traditional connection between users and products to encompass a variety of
products, services, and agents that have unique functions and interconnections with
one another. Furthermore, it has raised issues regarding the effectiveness of human-
centered design within this changing perspective (Coulton & Lindley, 2019;
Frauenberger, 2019; Giaccardi & Redström, 2020; Kuijer & Giaccardi, 2018).
Both ANT and MTHD are essential to mention in this context as they offer
alternative perspectives to the common anthropocentric approaches in HCI and
design research. In particular, ANT serves as a theoretical foundation for this study
by focusing on understanding the complex interactions and relationships between
human and non-human actors within a network. It emphasizes the idea that both
human and non-human actors have agency and can shape social interactions and
relationships.
In the study within this thesis, various actors, such as caretakers, dogs, and the dog
monitoring systems, are involved in a network. Aligning with the study's objective,
it is important to explore the ways in which these actors interact, influence each
other, and shape the caregiving practices of humans towards their companion dogs.
These systems, which monitor and track a dog's activity and behavior, have the
potential to mediate and influence the human-dog relationship. Having an ANT lens
can help explore how the introduction and use of dog activity monitoring systems
mediate human-dog relationships, influence human behavior, and shape the overall
caregiving dynamics. This perspective also aids in exploring the complex
interactions and influences between human and non-human actors, shedding light on
how these technologies can potentially improve humans' caregiving practices and
enhance the overall relationship with their companion dogs.
36
2.2.3.2 AWAX Model
Linden, Zamansky, and Hadar (2017) emphasize the importance of creating non-
verbal methods for understanding the needs of non-human users. To this end, they
propose “the Agility, Welfare as value and Animal eXpert involvement model
(AWAX)”, which integrates iterative prototyping, prioritizing animal welfare, and
direct involvement of animal experts in the development process (Linden, Zamansky
& Hadar, 2017, p.424). The model (Figure 2.4) entails collaboration among animal
experts, designers, and developers throughout an agile development process. In this
process, animal experts guide the design process and take an active role throughout
all stages of design, testing, and review, acting as a “surrogate stakeholder” for the
animal to ensure their needs are represented (Linden, Zamansky & Hadar, 2017,
p.53).
By incorporating animal experts into the agile development of animal technologies,
welfare concerns are addressed early in the process. The AWAX model illustrates
how the inclusion of animal experts guarantees the representation of animal needs
and the maintenance of animal welfare throughout iterative stages. Current
approaches to working with animals in technology development rely on physical
prototyping and obtaining feedback from the animals to iterate on the designs.
However, the absence of explicit models for eliciting requirements from animals
during technology development can jeopardize animal welfare by potentially causing
harm or inducing stress. Therefore, developers can utilize this model as a guiding
framework in the development of animal technology, ensuring the welfare of animals
is upheld (Linden, Zamansky & Hadar, 2017).
37
Figure 2.4. The AWAX iterative development model for the development of
interactive animal technology (Linden, Zamansky & Hadar, 2017, p.425).
2.2.3.3 Welfare Through Competence Framework
To prioritize animals as key stakeholders in technology design, Webber, Cobb, and
Coe (2022) propose the Welfare through Competence framework. This framework
integrates the "Five Domains of Animal Welfare" model with the "Coe Individual
Competence" model (Figure 2.5), offering a structured approach to defining
objectives that center on animals' needs. Its purpose is to guide interdisciplinary
38
teams in placing animals' interests at the center of animal technology design and
development.
The Coe Individual Competence Model highlights the importance of providing
animals with opportunities for choice, control, and variety, which contribute to their
development of competence (Coe, 2017). This approach is grounded in creating an
enabling environment that supports animals in attaining the necessary levels of
competence and agency. The Welfare through Competence framework provides a
systematic approach for assessing and identifying opportunities to enhance animals'
quality of life. Its application is particularly relevant in managed environments such
as farms and zoos, where promoting positive animal welfare is of utmost importance.
Designers can systematically explore design possibilities through an animal-centric
lens by analyzing how each competence principle from the Coe Individual
Competence model, as represented in the matrix, can positively impact the Five
Domains of animal welfare (Webber, Cobb & Coe, 2022).
Figure 2.5. The welfare through competence animal objectives matrix (Webber,
Cobb & Coe, 2022, p.8).
39
2.2.3.4 Animal Ethics
When applying theoretical and methodological frameworks in research with animals,
researchers have an ethical obligation to prioritize animal well-being and treat them
as sentient beings. Therefore, it is crucial to mention the current ethical frameworks
in animal research.
Currently, animals' involvement in research projects focused on developing animal
technology is regulated by existing ethical frameworks that abide by international
laws because there is not a formally established ethical protocol that focuses on the
animals as end-users in ACI research. The ethical concerns about animals started
originally with their use in laboratory experiments in the 1950s (Russel, Burch &
Hume, 1959). The ‘3Rs’ (Replacement, Reduction, and Refinement) approach, a set
of guidelines for animals’ use in testing processes presented by Russel and Burch
(1959), has become an internationally established principle. One of the most
extensive animal ethics legislation to date is the European Directive 2010/63/EU on
the protection of animals used for scientific purposes (EC, 2010). It applies to
scientific activities which include “any use of invasive or non-invasive of an animal
for experimental or other scientific purposes, with a known or unknown outcome, or
educational purposes, which may cause the animal a level of pain, suffering, distress
or lasting harm equivalent to, or higher than, that caused by the introduction of a
needle in accordance with good veterinary practice” (Article 3). The Directive also
recognizes animal welfare as “a value union enshrined in Article 13 of the Treaty on
the Functioning of the European Union (TEFU)” (Part 2). Moreover, the legislation
gives special attention to animals that are more closely related to humans from an
evolutionary standpoint, like non-human primates (especially great apes), or animals
that have a social connection with humans, such as companion animals like cats and
dogs (Parts 18, 21, 33). In the ACI manifesto, Mancini (2011, p.2) defined the
following ethical principles that researchers should be responsible for:
Recognize and appreciate the characteristics of all species involved in the
study without any discrimination.
40
Treat both humans and non-human participants with respect, consideration,
and care, based on their individual needs.
Conduct research with a specific species only if it aims to develop knowledge
or technology that benefits that species.
Protect both human and non-human participants from any physical or mental
harm by using non-invasive, non-oppressive, and non-depriving research
methods.
Allow both human and non-human participants to withdraw from the
interaction at any time, either temporarily or permanently.
Obtain informed consent from participants or their legally responsible
guardians before their involvement in the research.
Building on this initial consideration of animal ethics in ACI, researchers have
presented different ethical approaches for conducting animal studies. Vaataja and
Pesonen (2013) proposed design guidelines derived from the existing
frameworks in the literature by taking the 3Rs approach as their defining criteria.
Mancini (2016), on the other hand, suggested a welfare-centric ethics framework
recognizing consent as a vital requirement of participation. In this framework,
animals’ consent for engaging in research procedures is considered in two ways;
mediated and contingent consent. Mediated consent means obtaining consent for
animal participation in research from individuals who can understand the
potential impact of the research on the animal's well-being and have the legal
authority to give consent on their behalf. On the other hand, contingent consent
is based on the following criteria: 1) allowing the animal to adequately evaluate
the circumstance by providing them with ample opportunities to explore the
environment and research equipment before proceeding with the procedure, 2)
giving the animal the chance to make appropriate choices between different types
of interaction, such as choosing between reward systems based on food or play,
and 3) providing the animal with the opportunity to withdraw or withhold
engagement, such as having multiple escape routes and comfortable resting areas
(Mancini, 2016). However, the current ethical frameworks for animal use in
41
research mainly focus on the minimization of any negative impact of the research
on the individual animals’ welfare, involved generally through the
implementation of the 3Rs principles (Mancini, 2016).
Mancini and Nannoni (2022) suggest that while the 3Rs principles aim to protect
animals, they are rooted in a process-oriented ethical perspective that views
animals as tools in scientific processes. Therefore, they proposed an animal-
centered ethical approach that recognizes animals as independent and important
participants in the research process, with their own interests and the capacity to
give or refuse consent. They suggest four ethical principles, namely relevance,
impartiality, welfare, and consent, and a scoring system to evaluate the degree of
alignment between a research procedure and these principles. The aim is to assist
researchers and relevant authorities in evaluating how well a research procedure
adheres to these principles. This system is suggested to be used as a complement
to the 3Rs, assisting researchers in determining the circumstances in which
animal research is in the best interest of the animals involved, identifying ways
in which experimental procedures can be modified to improve their ethical
standards, and to recognize situations where non-animal methods should be
prioritized (Mancini & Nannoni, 2022). Moreover, in another study, Ruge and
Mancini (2022) developed an ethics toolkit to help researchers make ethically
sound decisions when working with animals and supporting animal-centered
research and design. The toolkit is made up of three templates, and each template
contains a series of questions to determine the ethical perspectives of the research
team and their project. Its use in animal research is designed to provide
researchers with a structured approach to defining the project's values and
understanding the ethical viewpoint that guides the team's interactions with
participant animals, handlers, and other stakeholders involved in the study (Ruge
& Mancini, 2022).
In this section, the existing ethical principles and guidelines for animal studies
have been reviewed. Consideration of these principles is essential in the design
and development of animal technologies to ensure that the studies conducted are
42
ethically appropriate. In addition, existing ethical frameworks for animal
research should be reviewed and adapted for the development of animal
technologies to be consistent with the advancement of animal technology. The
next section provides an overview of existing technological applications for
tracking and monitoring dogs.
2.2.4 The Review of the Existing Technological Applications for Tracking
and Monitoring of Dogs Concerning ACI
Numerous studies in ACI can be grouped under five categories: haptic interfaces,
screen interfaces, tracking and monitoring technologies, direct interaction sensors,
and auditory interfaces (Figure 2.7). However, as the study within this thesis focuses
on monitoring technologies, only studies in this domain are reviewed in this section.
In ACI, monitoring technologies were explored in many ways, including motion and
posture detection and activity and behavior monitoring studies. For example, Mealin
et al. (2016) used three-dimensional sensing hardware, Microsoft Kinect, for posture
and behavior detection and classification in dogs. The system was able to identify
the static postures of dogs, including standing, sitting, and lying, which can also be
used to observe stress behaviors. Pons et al. (2015) also used Microsoft Kinect for
cats to detect their location, body postures, and field of view. Microsoft Kinect was
also used in environmental enrichment studies, including captive animals such as
orangutans at the zoo (Scheel, 2018). Besides, Majikes et al. (2016) developed a
harness system composed of wearable sensors and devices to detect postures such as
sitting, standing, and eating. The study concluded that combining a computer-
assisted training system based on algorithmic interpretation with professional
training by humans would overcome problems related to ineffective timing during
dog training, thus increasing the success rate in training.
43
Figure 2.6. Categories of existing technological applications in ACI studies.
44
Also, in an attempt to create a lower cost and less subjective training method,
Brugarolas et al. (2013) developed a system that uses machine learning algorithms
for behavior recognition in dogs by using the data collected via an accelerometer and
gyroscope deployed on a vest. Extending on the behavior recognition research,
Valentin et al. (2015) created a collar system equipped with motion sensors to detect
the head gestures of working dogs. Each detected gesture by the system was paired
with a pre-recorded message that was delivered to humans via a smartphone. They
emphasized that working dogs have limited options for communicating large
amounts of stimuli that they perceive to humans, which results in a large information
gap between dogs and humans. Moreover, Ladha et al. (2013) developed a collar-
based system to record and analyze a set of behavior traits relevant to a dog’s well-
being, such as eating and sleeping patterns in their natural environments. Tracking
technologies seem to be a promising area for further research as they provide means
for understanding and measuring the behavior of non-spoken animals, which is
fundamental for the development of ACI methodologies.
In addition to the tracking technologies for animals mentioned above, fitness and
health trackers, which are one of the most popular devices for humans in the
wearables market, have taken their place in the pet industry as well. In recent years,
pet wearables have grown in popularity among pet owners. These gadgets are created
to keep an eye on the pet's location, monitor their fitness and activity levels, and give
insights into their health. Similar to human wearables, these devices consist of
hardware equipped with sensors that the pet wears and accompanying software that
the caretaker can access via a mobile app. For location tracking, pet wearables
usually use GPS or RF-based solutions, whereas activity trackers use accelerometers
and Bluetooth or Wi-Fi to send data. For instance, FitBark is an activity tracker worn
by dogs that measures their activity levels in a way similar to human fitness trackers.
The device captures raw accelerometer data, which is then converted into an
understandable format for human users, providing suggestions for interactions, such
as taking the dog out for more walks. These devices mainly provide health-related
data to pet owners about their pets’ daily activity levels, calories burnt, and rest and
45
sleep patterns, along with related suggestions. In dog monitoring technologies, the
primary user is humans. Eason (1988) outlined three different user types: the primary
user, the secondary user, and the tertiary user. The primary user is the individual that
will actively engage with the system, while the secondary and tertiary users are the
ones who may use the system occasionally or are impacted by its implementation
(Eason, 1988). Today, there are several health and activity trackers for pet dogs on
the market, such as FitBark, Whistle, Garmin, and PetPace, in addition to various
other commercial pet products, including pet cameras, automatic feeders, and
interactive toys.
In ACI research, a few studies have been conducted on pet wearables so far. As an
example of the studies on wearables for animals, Alcaidinho et al. (2015) examined
whether using pet activity trackers can reduce the return rate of newly adopted dogs
from a shelter. The study showed that providing health and activity-related data to
adopters through a mobile application resulted in a decrease in dogs’ re-
relinquishment rates (Figure 2.8). Also, the participants stated that the information
provided by the application helped them bond with their newly adopted dogs.
Figure 2.7. A dog wearing a pet activity tracker (on the left), the activity tracker
widget (on the right) (Alcaidinho et al., 2015, p.463).
Nelson and Shih (2017) studied how technology, data collection, and visualization
influence the way dog owners perceive and interact with their animals. They
presented the CompanionViz system, a prototype that provides caretakers with
46
details on their dogs' calorie intake, and activity. Twelve participants were surveyed
to assess their initial interest in the system, and then three users were chosen to test
it out in a field study. The feedback from these users suggested that the system led
to higher awareness, motivation, and curiosity about their pet's needs. Another study
found that even a simple collar equipped with GPS can improve human-dog
relationships by opening up new forms of interaction (Weilenmann & Juhlin, 2011).
Also, Vaataja et al. (2018) investigated the caretakers’ motivation to use these dog
activity trackers through interviews. The study revealed that caretakers use these
devices primarily to monitor health, behavior, and learning-related issues, balance
daily activity levels, and rest in dogs. However, the insights gained during device
use served as a motivational factor for behavior change in caretakers to spend more
time with their dogs. In addition to the use of commercial wearable devices for health
and activity tracking in dogs, Kumpulainen et al. (2018) aimed to classify seven
activities of dogs by using a three-dimensional movement sensor placed on a collar.
They argued that recognizing dog behavior would provide more information to pet
owners about their pets than just monitoring their vital signs via health and activity
trackers.
Additionally, Zamansky et al. (2019) conducted an empirical study to explore pet
owners' perceptions of a commercial dog activity tracker. Their research focused on
how and why commercial dog activity trackers were used by dog owners, the
influence of their use on pet and owner lifestyle, and the features of the trackers
perceived as significant by the pet owners. The findings revealed that the activity
trackers were perceived as factor increasing caretakers’ motivation to engage in
physical activity with their dogs, strengthening the human-animal bond, and
heightening caregivers' awareness of their pets' needs and resulting in a perceived
improvement in the quality of care. A number of participants reported an
improvement in their quality of caregiving and a greater understanding of their
animals' physical activity needs and overall well-being. Studies are conducted to
explore what motivates consumers to purchase companion animal technology as well
as any obstacles that may prevent them from doing so (Ramokapane, van der Linden
47
& Zamansky, 2019). The results of the study indicated that the primary barriers to
the adoption of pet wearables were their cost and durability, alongside users’
concerns related to the animal's welfare, perceived lack of usefulness, and accuracy.
This section has provided an overview of current technologies and research studies
related to tracking and monitoring dogs. The following section focuses specifically
on the concept of dog monitoring technologies as a type of interspecies information
system, which involves exploring the ways in which these technologies can facilitate
communication and exchange of information between humans and dogs.
2.3 Dog Activity Monitors as Interspecies Information Systems
Animals have traditionally been seen as either unintentional stakeholders or
resources in information systems. However, the development of new technology,
such as pet wearables, is allowing people to understand animals better and open up
new forms of communication between species that were otherwise left implicit or
misunderstood (Tami & Gallagher, 2009). Van der Linden (2021) proposes that this
creates an interspecies information system (IIS) where humans and animals are both
actors and stakeholders. The flow of data between participants of different species
in an IIS is demonstrated in Figure 2.9, with technology capturing data from one
species and using it to inform another species (van der Linden, 2021).
According to Van der Linden (2021), an IIS enables the exchange of data between
humans and animals, allowing humans to gain insights into the physical or
behavioral condition of animals. This knowledge can be used to intervene and affect
animals in positive or negative ways. However, some information systems, such as
pet wearables, exhibit a one-way flow of information. For example, in these systems,
the dog is monitored, and the software advises the owner on how to interact with the
animal. Meanwhile, the dog is not aware that it is part of this information system
(van der Linden, 2021).
48
Figure 2.8. Data flow in an IIS consisting of stakeholders of different species (van
der Linden, 2021, p.5).
Interventions from one species to another are informed by the information flow
within an IIS. Van der Linden (2021) suggests that in order to perform interspecies
interventions, it is important to understand the relationships between the various
components in an IIS, such as the actors of different species and the technology
involved. However, the impact of these interventions can be complex, affecting both
human and animal actors, as well as their surrounding social and organizational
environments. It is essential to consider the potential impact of these interventions
on each other (van der Linden, 2021).
The model in Figure 2.10 illustrates the flow of data and interactions between the
components of an IIS, enabling interspecies interventions that can affect processes
outside the IIS. For example, dogs can provide input to monitoring technologies like
activity trackers and vital sign sensors, which are then processed by information
technology, often in the form of software on a human's smartphone or computer. It
is important to be aware of the complexities of interspecies relationships and their
potential outcomes when considering the impact of interventions. The results of this
processing suggest interspecies interventions, which a human actor may enact, or
which may inform policy decisions outside the IIS. These interventions impact both
49
external processes, such as pet caregiving, and the human and animal actors
involved.
Figure 2.9. Key elements of interactions within an IIS (van der Linden, 2021, p.10).
2.4 Conclusions Regarding the Chapter
In conclusion, this chapter has presented a comprehensive overview of the key topics
and theoretical foundations that underpin this doctoral study. By inquiring into the
domains of animal welfare, animal computer interaction, and animal-centered
design, a solid understanding of the importance of considering animals as central
stakeholders in technology development has been established.
The discussion on research methods for identifying animal needs has shed light on
existing methodologies used to investigate the experiences and requirements of
companion dogs. By incorporating various theories, models, and frameworks such
as the AWAX model, the Welfare through Competence framework, and Actor-
Network Theory, a theoretical foundation has been revealed to achieve animal-
centeredness in technology design. These frameworks offer valuable perspectives
and methodologies for designing technology that prioritizes animal welfare and
acknowledges the intricate interactions between humans and dogs.
50
The inclusion of the animal ethics section has provided an overview of existing
ethical frameworks in animal research, emphasizing the importance of ethical
considerations and responsible research practices to minimize harm to animals. This
shed light into the responsible and ethical animal-centered design and research
practices to be followed in animal studies.
Furthermore, the review of existing technological applications for tracking and
monitoring dogs has offered valuable insights into the current landscape of dog
activity monitoring systems. This review sets the stage for exploring the potential
and possibilities of these technologies in enhancing humans' caregiving practices
through increasing their awareness of their dogs.
Overall, this chapter serves as a foundational basis for the study within this thesis,
integrating knowledge from diverse disciplines, ethical considerations, and
technological advancements. It establishes a solid groundwork for developing a
theoretical model on how dog activity monitoring systems (DAMS) for companion
dogs can mediate the human-dog relationship to improve humans' caregiving
practices.
51
CHAPTER 3
3 BEHAVIOR CHANGE AND SENSEMAKING
Caretakers’ lack of a thorough understanding related to their dogs’ health, behavior,
and needs can lead to unfavorable results regarding companion dog welfare.
Moreover, the current decline in the welfare of companion dogs in domestic settings
is directly linked to the caretakers’ unawareness of these critical aspects. Activity
monitoring systems designed for dogs may be able to promote behavior change in
humans, similar to how personal health informatics systems influence behavior.
Given that the main users of these systems are humans, they can contribute to dog
welfare by encouraging positive behavior change in caretakers and raising their
awareness about their dogs. Therefore, this chapter aims to provide an overview of
the existing behavior models, theories, and frameworks along with the technology’s
role in behavior change to understand the cognitive mechanisms behind human
behavior and behavior change.
3.1 Persuasive Role of Technology
The integration of computing technologies into human life has resulted in various
influences on our behavior. However, this relationship between humans and
technology is not one-sided. The technology can shape how people behave, but
human behavior also impacts how technology is utilized (Slob & Verbeek, 2006).
While computers were not initially designed for persuasive purposes, researchers
have recently become interested in using them to change human behavior and raise
awareness. These interactive computing systems are known as persuasive
technology and are intended to modify people’s attitudes, behaviors, or both (Fogg,
2003). In other words, persuasive technology is intentionally designed to influence
people's behaviors, and it has now taken on the role of persuasion in human life.
52
However, Fogg clearly distinguishes persuasion from compelling or deceiving
people and defines persuasion as a “voluntary change in attitude or behavior” (2003,
p.15). To better understand technology’s persuasive potential, it would be helpful to
mention the different roles that computer technologies play in human life.
On the functional triad framework (Figure 3.1), Fogg (2003) proposes that
computing technologies have three essential functions from the users’ perspective:
tools, mediums, and social actors. In their role as tools, computer technologies aim
to equip users with new capabilities, allowing them to complete actions more
efficiently and effectively. Computer technologies can have an impact on people's
attitudes and behaviors in several ways. Firstly, as tools, they can facilitate the
attainment of predetermined goals, assist people in a process or experience, or
provide motivation through calculations and measurements. Secondly, as mediums,
they can shape attitudes and behaviors by offering simulated experiences and
allowing people to explore cause-and-effect relationships. Finally, as social actors,
computer technologies can offer positive feedback, model desired attitudes and
behaviors, and provide social support to people to shape their attitudes and behaviors.
Figure 3.1. The Functional Triad: Roles Computers Play (Fogg, 2003, p.25).
The main concern of behavior change through technology is motivating people to
perform a target behavior. According to Fogg (2003), technology can persuade
individuals to modify their behavior by motivating, guiding, and providing positive
53
feedback to achieve the desired behavior. Also, Fogg (2003) states that the
information and feedback provided via interactive technologies are essential
motivators for people to perform a behavior. Thus, people's decisions to engage in
an activity are influenced mainly by the information and feedback provided by
technology. Similarly, Lilley (2009) posits that technology has three main ways of
influencing behavior. First, it can provide feedback on the results of a particular
behavior, which can help guide future actions. Second, it can encourage people to
behave in certain ways by designing technology with specific affordances and
constraints. Finally, technology can sustain a certain behavior by using persuasive
methods to change people's thinking and actions. However, although information
and feedback offered by technology are crucial to motivate users, motivation by itself
is often not enough for a behavior to be performed. Thus, several behavior change
strategies are also applied in the design of persuasive technologies.
One example of persuasive technology is personal health informatics systems, which
allow individuals to modify their behavior by analyzing self-monitoring data to
accomplish a specific goal. These systems analyze the user's data, present it in an
understandable way, and offer feedback to assist users in achieving their desired
behavior (Fogg, 2003). Furthermore, persuasive technologies use different methods
to influence people's behavior change, in addition to offering self-monitoring data.
Fogg (2003) describes seven types of behavior change strategies included in
persuasive technologies. These include;
Reduction: Technology should make it easier to achieve the target behaviors
by reducing the required effort to perform them. The less perceived effort to
achieve the desired behavior would presumably result in increased
motivation to be engaged in it.
Tunneling: Technology should guide users within an experience through a
sequence of pre-defined actions/steps. This guidance can also increase the
chance of providing further opportunities for persuasion along the way.
54
Tailoring: Technology should provide users with information tailored to their
individual needs.
Suggestion: Technology should offer suggestions to users at the right
moment.
Self-monitoring: Technology should allow users to self-monitor to adjust
their behaviors or attitudes to achieve the desired outcome. Self-monitoring
aims to reduce the effort required to track one’s performance, make it easier
for users to know their status while performing a specific behavior, and give
feedback.
Surveillance: Technology should enable users to observe others’ behaviors
to increase the likelihood of achieving the desired outcome.
Conditioning: Technology should support users in changing behaviors or
turning them into habits using positive reinforcement.
Building on Fogg’s persuasive design principles, Oinas-Kukkonen and Harjumaa
(2008) propose the following strategies for designing computer systems to improve
the computer-human dialogue;
Praise: A system/technology should use praise for providing positive
feedback to the users.
Rewards: Technology should reward users to encourage them to perform the
desired behavior.
Reminders: Technology should remind users of the target behavior.
Suggestion: Technology should provide users with suggestions, i.e.,
suggestions to choose healthier foods instead of others to promote healthy
eating habits.
Similarity: Technology should imitate users in specific ways, i.e., using a
particular language familiar to a target user group.
Liking: Technology should appeal to its target users regarding its look and
feel.
Social role: Technology should take on a social role.
55
Oinas-Kukkonen and Harjumaa (2008) expand these strategies by suggesting other
design techniques for technology to motivate users with a particular focus on social
support;
Social learning: Technology should enable users to observe others who
perform a target behavior to promote social learning.
Social comparison: Technology should enable comparison between users.
Normative influence: Technology should enable users with similar goals to
come together.
Social facilitation: Technology should allow users to find others performing
the target behavior/have similar goals.
Cooperation: Technology should enable cooperation among users.
Competition: Technology should enable competition between users.
Recognition: Technology should allow users performing a target behavior to
be recognized.
Moreover, there are other behavior change techniques used in activity tracking
systems such as giving credit, social influence, providing personal awareness
(Consolvo et al., 2006), goal setting (Consolvo et al., 2009; Munson & Consolvo,
2012), (Consolvo et al., 2006), and visual displays of personal data (Consolvo et al.,
2008a; Consolvo et al., 2008b).
Fogg (2009b) also categorizes behavior change types in the “Behavior Grid”
framework. Considering that there are various types of human behavior, strategies
for the design interventions should also vary depending on these behavioral
differences. Fogg (2009b) proposes a Behavior Grid that categorizes 35 different
types of behavior based on behavior change type and scheduling/timing. According
to Fogg, new behaviors are approached differently than familiar ones, so different
strategies should be employed to motivate new behaviors. The scheduling/timing of
a behavior can range from a one-time action to a habitual behavior. This difference
is crucial for persuasion because people are more likely to perform a behavior once
rather than committing to future tasks, which can be more challenging. For instance,
56
playing with a dog once is entirely different from adopting a dog. In other words,
many aspects come into play when designing for persuasion. Thus, the persuasion
strategies should differ considering this variation in behavior types.
In addition to the strategies for the design of persuasive technology, Fogg (2003)
also underlines the importance of timing and context to influence users’ attitudes and
behaviors. He states, “new computing capabilities, most notably networking, and
mobile technologies, create additional potential for persuading people at the optimal
time and place” (p. 184). In other words, increased connectivity and mobility enable
products to intervene at the right time and place, thus, enhancing their abilities to
motivate and persuade users.
As can be seen, the design of persuasive technologies involves various factors to
consider. If the strategies mentioned above are applied correctly, these technologies
have the potential to increase people's awareness and motivation towards performing
desired behaviors. For instance, a study on wearable fitness trackers, which are a
form of health informatics system, demonstrated that offering health-related personal
insights through these devices can lead to long-term behavior changes (Choe, Lee,
Munson, Pratt & Kientz, 2013). Another study examining the effects of using a
fitness tracker on people’s activity levels showed that the device use resulted in a
significant increase in participants’ activity levels (Cadmus-Bertram et al., 2015).
Additionally, additional research suggests that activity-monitoring devices assist
users in acquiring a deeper understanding of their actions and conduct within the
context of the data provided by these devices (Fritz, Murphy & Zimmermann, 2014).
Understanding the persuasive role that technology can play in people's lives through
the proper application of various persuasive design principles and behavior change
strategies is essential to understanding the potential impact of technology on
behavior. However, technology is not the only determinant of human behavior. To
fully grasp the impact of technology on behavior, it is also necessary to understand
the psychological mechanisms underlying human behavior. The next section
57
provides an overview of behavior change models, theories, and frameworks in the
psychological literature.
3.2 Behavior Change Models, Theories, and Frameworks
There is a considerable amount of research in the field of HCI focusing on behavior
change through technology, particularly via personal informatics systems, to
promote positive behavior change. It is important to understand first the
psychological mechanisms behind human behavior and behavior change to
comprehend the potential and use of persuasive technologies. Several models and
theories related to behavior change in the psychology literature explain the
determinants of human behavior. The following sections present an overview of the
existing behavior change models, theories, and frameworks within the psychology
literature to help us understand human behavior and how interactive technologies
can be utilized to encourage behavior change.
As suggested by Kuru (2013), the four most prominent theories adopted in the
personal informatics and health behavior domain, especially in the physical activity
context, are the Transtheoretical Model of Behavior Change (TTM) (Prochaska,
Johnson, and Lee, 1998), The Theory of Planned Behavior (Ajzen, 1991), The
Theory of Reasoned Action (Ajzen & Fishbein, 1980), and Social Cognitive Theory
(Bandura, 2001) (Buchan et al., 2021). Buchan et al. (2012) distinguish two kinds of
physical activity interventions in their review: stage-based models and social
cognitive models. While stage-based models suggest that people go through stages
when adopting a new behavior, social cognitive models assume that behavior is
mainly controlled by cognitive processes.
The most popular stage-based model within the personal informatics domain, the
Transtheoretical Model of Behavior Change (Figure 3.2) devised by Prochaska and
Velicer (1997), proposes that a behavioral change process occurs in six distinct
stages: pre-contemplation, contemplation, preparation, action, maintenance, and
58
termination (Prochaska & Velicer, 1997). These are called the stages of change. In
the pre-contemplation stage, people do not intend to perform a target behavior,