The Quest for Happiness in Self-Tracking Mobile Technology

(1)

MA in New Media and Digital Culture Media Studies Department | Faculty of Humanities

The Quest for Happiness in Self-Tracking Mobile

Technology

Ana Crisostomo Student number 10397124 +31(0) 629 169 166 Rustenburgerstraat 354-3 1072 HD Amsterdam ana.crisostomo@gmail.com

Thesis Supervisor: Bernhard Rieder December 2013

(2)

Abstract:

The practice of self-tracking became more accessible to the general public in recent years through the widespread use of connected portable devices (in particular smartphones), improved human biometric sensors, platforms and services specifically designed for monitoring purposes, and enhanced online data storage solutions. In this context, a movement labeled Quantified Self has been gaining an increasing number of followers on a global scale, which has also propelled additional media coverage towards this specific type of personal activity.

Besides contextualizing self-monitoring practices generally considered, this study focuses on the ones in the affective domain in particular, commonly known as mood and happiness tracking. The examination aims at understanding the possible causes and potential consequences of the displacement of these experiments from an exclusively clinical and academic environment to a wide public arena, and the expansion of its focus from mental patients (on a chronic or episodic basis) and research subjects to a large population previously considered healthy and functional.

To achieve that goal, the research relies on a multi-disciplinary approach borrowing concepts and theories from fields such as Media Studies, Psychology, Philosophy, and Economics, combined with an empirical work focused both on the technological platforms and the individual practices. From the conceptual and empirical analysis emerges a phenomenon occupying a particular space framed in the intersection of technology, wellness and wellbeing, as well as science, threatening to redefine personal identity and individual behavior by expanding the limits of self-awareness and the scope for self-improvement.

Keywords: self-tracking, quantified self, affective monitoring, mood tracking, happiness

(3)

3

List of Figures

Figure 1 – Screenshot from Gary Wolf’s dashboard of personal analytics ……… 9

Figure 2 – A photo of Buckminster Fuller’s Dymaxion Chronofile ……….…..………..………... 14

Figure 3 – A plot of a third of a million email messages sent by Stephen Wolfram since 1989 …... 16

Figure 4 – Typologies of Individual Tracking ……… 22

Figure 5 – Stage-Based Model of Personal Informatics ………...……….27

Figure 6 – Computing devices as social actors ………..………... 41

Figure 7 – Screenshot from Wellness Tracker ………...……….. 59

Figure 8 – Screenshot from MebHelp Mood Tracker ……… 60

Figure 9 – Screenshot from Track Your Happiness ………...……… 63

Figure 10 – Screenshot from My Smark ………...……… 64

Figure 11 – Screenshot from Moodscope ……… 65

Figure 12 – Screenshot from MoodPanda ……….………...………….. 68

Figure 13 – Screenshot from MoodPanda (community) ……….. 62

List of Tables

Table 1 – General indicators about the QS website (November 2013) ……….………….. 113

Table 2 – Oldest QS Meetup groups (November 2013) ……….. 114

Table 3 – Top 10 QS Meetup groups by number of members (November 2013) ……….. 114

Table 4 – Top 10 QS Meetup groups by number of previous meetings (November 2013) ………. 115

Table 5 – Top 10 QS Meetup groups by number of (member) reviews (November 2013) ……….. 115

Table 6 – Wikipedia articles for “Quantified Self” (December 2013) ………. 116

Table 7 – Google Scholar results for the query “Quantified Self” (December 2013) ………. 116

Table 8 – List of general self-tracking applications ……… 117

Table 9 – Examples of mood and happiness self-tracking applications ………. 119

Table 10 – Examples of prototypes and products which infer personal mood from physiological indicators ………...………. 126

Table 11 – Examples of self-tracking experiments (from the QS Meetups) in chronological order ……… 130

(6)

6

List of Graphs

Graph 1 – Number of published articles in the QS website per year and per author (November 2013) ………...……… 19 Graph 2 – QS Meetup members by region / country (November 2013) ……….. 52 Graph 3 – QS Meetup groups by region / country (November 2013) ………..……….… 53 Graph 4 – Top 50 keywords used to describe the QS local Meetup groups (November 2013) …… 54 Graph 5 – Top 10 hashtags related to #quantifiedself (November 2013) ……….. 55 Graph 6 – Specific focus of affective self-tracking applications (November 2013) ………...… 58 Graph 7 – Usage domains of affective self-tracking applications (November 2013) ………... 58 Graph 8 – Tracking modes featured in affective self-tracking applications (November 2013) ……. 63 Graph 9 – Privacy settings of affective self-tracking applications (November 2013) ………..… 67 Graph 10 – Social sharing featured in affective self-tracking applications (November 2013) ……... 68 Graph 11 – Data comparison types featured in affective self-tracking applications (November 2013) ………... 70 Graph 12 – Goals of affective self-tracking experiments (November 2013) ………..……… 73 Graph 13 – Types of introveillance according to tracking mode and focus ………. 80

(7)

7

1. Introduction

Self-tracking is a concept which has recently gained traction, in particular in the last five years, as evidenced by the increasing number of media and academic articles published about the topic (Lupton, The Rise of the Quantified Self as a Cultural Phenomenon), and the hype surrounding consumer products and services catering for that particular need on several fronts. Forbes announced 2013 to be the year of digital health (Nosta) and several indicators seem to support that claim. In the 2013 edition of the Consumer Electronics Show (CES), an annual innovation showcase unavoidable for most of the industry professionals, one fourth of the exhibits were dedicated to health and fitness (Carroll). Still in the technological area, competitions with significant awards are being held to spur radical innovation in personal healthcare technology1_{and many startups are}

actively exploring the wellbeing and wellness market (Lebowit).

The activity of systematically logging data about oneself is not novel, is not limited to health, and does not necessarily rely on digital technology. What changed recently was a set of conditions which made self-tracking more accessible and appealing to the general public: the widespread use of connected portable devices (in particular smartphones), improved human biometric sensors, platforms and services specifically designed for monitoring purposes, and enhanced online data storage solutions. Such technological developments, combined with a favorable reception, originated specific practices under novel labels.

The term ‘self-tracking’ is often employed in association with other expressions, such as ‘personal analytics’2_{or ‘personal metrics’ (information based on personal data), ‘personal informatics’}3_(the

technology used to collect, manage and visualize personal data), and ‘the quantified self’. The latter is a designation coined by Wired magazine editors Gary Wolf and Kevin Kelly in 2007, to label the belief that the answers to many fundamental questions in life reside within the individual, and that improvement can only be achieved through measurement (Kelly, What is the Quantified Self?). Rather than announcing a future trend, the terminology merely labeled a situation which was already a

1_{See the Qualcomm Tricorder XPRIZE <}_{http://www.qualcommtricorderxprize.org/}_{> (a designation inspired}

by the tricorder device from the fictional science fiction TV series Star Trek), a $10 million competition in this field.

2_{This term has been popularized by the experiments of the scientist Stephen Wolfram (see}_{section 2.2}_{) and his}

Wolfram Alpha Personal Analytics tool for Facebook <http://www.wolframalpha.com/facebook/>.

3_{This is commonly attributed to the researcher Ian Li (see}_{section 4.1}_{) who has also created the website}

(8)

8

reality within their network of acquaintances (Wolf, Know Thyself: Tracking Every Facet of Life, from Sleep to Mood to Pain, 24/7/365). Their website <http://quantifiedself.com/> developed into a central platform for a movement which rapidly expanded, virtually and physically, beyond the Silicon Valley area to become truly global4_.

The embracing spirit of the movement generated an informal community5_{which is open to any}

self-tracker, or individual interested in the monitoring process, and encompasses all types of tracking experiments. For the above reasons, it would be difficult to approach the topic of current self-tracking practices without referring to this group, which by no means implies that self-tracking practices do not occur outside the Quantified Self (QS) domain. In fact, one would have to operationalize the concept of self-tracking in order to identify practices which fall outside the scope of the definition. Self-tracking can be understood as the individual practice of systematically gathering data in the personal life domain for a certain period of time with a specific goal. Within this definition, practices can be distinguished according to the type of awareness involved (conscious or non-conscious), and type of initiative (self-initiated or mandated by other). While personal data monitoring may be a byproduct of many daily routines involving digital technology (i.e. web browsing), this study will only focus on voluntary, conscious and self-initiated experiments such as the ones where individuals track their mood or measure their happiness levels on a daily basis. While these activities can be carried out by most individuals with access to basic technology (which ultimately can be the ‘pen and paper’ type), it is likely that the most active and involved self-trackers, as well as the most diverse and innovative experiments, will be found among the QS group. Since there is, at the moment, no other organization or established movement assembling the above characteristics, this collective is considered as a prime source for the empirical investigation in this study.

In general terms, self-tracking activities are conducted in categories such as nutrition, fitness, sleep, health, cognition and mood, either in an isolated or in an integrated fashion - see an example of monitored personal indicators in Figure 1.

5_{In a recent article, Sociology PhD student Whitney Erin Boesel argues that Quantified Self represents already}

something more stable than a movement and can be referred to as a community (Boesel, Data Occupations). However, considering how recent the phenomenon is, how diverse the practices it entails are, and the little research that has been done on the matter, I will employ in this study the term ‘group’ instead of ‘community’.

(9)

9

Figure 1 – Screenshot from Gary Wolf’s dashboard of personal analytics

Source: <http://www.wired.com/medtech/health/magazine/17-07/lbnp_knowthyself>

These divisions do not exhaust all possibilities and individual observations can be classified under other categories, such as relationships and lifestyle. Experiments which deal directly with physical indicators appear to be more common than the ones which are concerned with cognitive and affective

(10)

10

dimensions6_{. One possible reason may be that in the case of physical and behavioral tracking, it has}

been more clearly established ‘what’ needs to be measured and ‘how’, as well as the reason ‘why’ might be considered more conventional and, therefore, more easily accepted. Monitoring cognitive and affective states can involve a higher level of complexity and uncertainty, which may imply that not all individuals are willing, or interested, in performing this type of self-tracking. In the case of affective logging, the process can also become rather sensitive, as it deals with information associated to emotions and moods which can be perceived as more intimate.

Since self-tracking encompasses such a wide variety of fields and practices, it becomes more valuable to direct this research to one specific area, taking into account that studies on self-monitoring of affective dimensions appear to be academically under-represented in comparison to ones on physical health7_{. This focus also allows a more precise delineation of the field of study for the empirical stage,}

eventually leading to more specific results.

While the current research will naturally examine some elements and properties of self-tracking practices in general terms, as a required contextualization for the topic, I will try to direct its scope, as much as possible, to affective monitoring – an area in which QS experiments on mood and happiness can be found. It is relevant to highlight this ‘tentative’ nature, since many affective experiments display a holistic character involving other indicators so, in some cases, it might not be feasible to completely disentangle affective monitoring from other types of tracking.

The guiding research question for the present study is then following: how can current self-tracking practices be defined and contextualized from a technological and social perspective?

That overarching interrogation will be then supported by the following three sets of sub-questions: 1. What types of self-tracking experiments are currently being undertaken and

by whom? What does the process of self-tracking entail? (sections 3, 4 and 7) 2. What are examples of affective self-tracking practices and technologies? To

which extent do practices of affective self-tracking aided by mobile technology impact self-perception and individual behavior? (sections 6 and 7)

6_{As a reference, in a 2012 U.S. survey conducted by the Pew Research Center, less than 1% of the health apps}

downloaded by smartphone owners was related to mood (Fox, Mobile Health 2012 14).

7_{When analyzing literature on ‘digital health’ and ‘mobile health’ the vast majority of the examples provided}

(11)

11

3. What are the identifying features of these particular practices and technologies? What are their possible causes and their potential impact from an ideological and social point of view? (sections 5 and 7)

Self-tracking practices aided by mobile technology, framed in the context of this recent movement, are a multi-faceted phenomenon lacking a formal definition and delimitation and, for that reason, I considered beneficial to present relevant sets of concepts within their disciplinary domain first. These theories and models, initially described independently, inform then different empirical approaches producing specific results. It is in the subsequent discussion stage that all elements become truly integrated and that their connection produces additional insights.

This study is structured into six main topical sections: some presenting broader conceptual perspectives and others focused on more specific and pragmatic approaches.

The first section will include a brief historical introduction to self-tracking experiments, and the second one will provide a social contextualization of these practices by introducing and describing the Quantified Self group.

The third one will present a functional approach to self-tracking describing the types and stages of the self-tracking practice.

The fourth section will refer to conceptual approaches which grant different entry points to the self-tracking theme, including ideas related to topics such as healthism, quantification, surveillance, posthumanism, and technology as a social actor.

The fifth section will introduce the affective component by describing attempts to gauge well-being at a collective level and, more importantly, by presenting several theories and models of examination and assessment of affective states on an individual basis.

The empirical work, incorporated in the sixth section, will include three different types of observation. The first one will be dedicated to the QS group (with a brief analysis of its website and Show&Tell groups) with the goal of contextualizing the phenomenon from a social perspective; the second one will be focused on the monitoring platforms (with the examination of a sample of 25 applications dedicated to mood and happiness tracking) aiming at providing a technological contextualization; and the third and last one will be focused on the self-tracking practices (with a comparative analysis of 20 QS presentations on self-tracking experiments in this area) with the objective of situating these practices both from a social and technological perspective.

(12)

12

The goal of the above framework is to facilitate the collection, interpretation and correlation of meaningful material, which will then be translated into a valid contribution to the present thematic field.

(13)

13

2. An Historical Overview of Self-Tracking Practices

The practice of systematically self-tracking some type of personal data does not require technology and can ultimately rely solely on human memory. However, if one considers only the written evidence of such practice, then there are a few historical cases worth referring as examples of self-experimentation and self-monitoring.

2.1 – Analog logging

In the sixteenth century, the Italian physician and professor Sanctorius Sanctorius was already keeping a personal record of his weight, before and after every meal, as well as ingested food and excrements for 30 years, in order to analyze the energy expenditure of a human being (Neuringer 79). Curiously, a similar personal experiment is being currently undertaken by Computer Science researcher Larry Smarr8_{in an attempt to gather a more accurate insight about his personal health,}

but in this case using state of the art technology. The amount of information and level of detail between the two is incomparable, as the Italian physician, unlike the north-American researcher, could not have possibly conceived that, for instance, “human stool has a data capacity of 100,000 terabytes of information stored per gram” (Bowden).

Nevertheless, personal monitoring does not mandate a quantitative approach or a health interest. On a more qualitative level, the first records of personal diaries used in a systematic manner for a significant period of time, are also dated from the sixteenth century (Samuel Pepys is referred to as being the earliest well-documented diarist). In the eighteenth century, Benjamin Franklin devised a system to track his daily behavior according to thirteen human virtues he believed to lead to an ideal life (Houston). In the subsequent centuries, several eminent figures such as Queen Victoria, Sigmund Freud, Virginia Woolf, Anaïs Nin, among many others, reflected their daily routine and internal impressions into written memories (Blythe). In some cases, it was precisely the personal account of a particular type of existence that brought attention to an individual’s life, as it happened with the

8_{Larry Smarr is often referred as an example to illustrate a highly detailed and scientific type of self-tracking}

in the health domain. He has been the subject of numerous interviews in the media and has also given a TEDMED talk on his experiments in 2013: <http://www.tedmed.com/talks/show?id=18018>.

(14)

14

posthumously published Diary of Anne Frank, possibly one of the most read personal diaries worldwide.

The American visionary architect, entrepreneur, inventor and author Buckminster Fuller is said to have the most well-documented human life in history: starting in 1920 and for the subsequent 63 years, conceiving his own life as an experiment, he documented his daily existence resorting to physical records ranging from notes to letters, from sketches to bills and receipts in a personal project he labelled the Dymaxion Chronofile9_{(Krausse and Lichtenstein 14) (see Figure 2).}

Resembling this enterprise in format, was Andy Warhol’s experiment with Time Capsules: a collection of 612 cardboard boxes containing all sorts of personal items which he systematically filed, sealed and stored for over a decade until his death in 198710_.

Figure 2 – A photo of Buckminster Fuller’s Dymaxion Chronofile

Source: <http://www.bavc.org/sam-green-talks-buckminster-fuller>

9_{This collection occupies a linear extension of more than 350 meters and is currently available at the Stanford}

University Library <http://library.stanford.edu/collections/r-buckminster-fuller-collection>.

10_{This collection takes approximately 2.500 square meters and currently resides at The Andy Warhol Museum,}

Pittsburgh: <http://www.warhol.org/collection/archives/>. It is also possible to explore the content of one of the boxes online through an interactive application in the Museum’s website.

(15)

15

2.2 – Digital monitoring

The range and level of detail of such physical collections is soon seriously challenged by individuals who adopt technology to support the collection of their personal information. Already in 1945, in an article for The Atlantic Monthly, Vannevar Bush presented the Memex concept, a device where personal information (such as books, records and communications) would be stored with the goal of supporting individual human memory (Bush). In the 1980s and 1990s, the first experiments related to lifelogging11_{appeared, based on more widely available portable and wearable technology}12_,

followed then by initiatives on several fronts.

In the early 2000s, Microsoft announced the company’s investigation efforts towards a project entitled MyLifeBits, directly inspired in the Memex for which the subject, their researcher Gordon Bell, had already started collecting personal data13_{(Scheeres). Initially in close relation to that}

project, a series of workshops on the topic of Continuous Archival and Retrieval of Personal Experiences (CARPE) were organized from 2004 to 2006, attracting academic and corporate researchers working in the field.

In many instances, the inward gaze starts assuming a public dimension. Projects of lifecasting (video broadcasting one’s life through digital media) arise as art experiments, such as Quiet: We Live in Public in 1999 by Josh Harris14_{, and as television shows (the most notorious example being Big}

Brother). In 2003, a military research proposal connected to individual surveillance is presented by the U.S. based Defense Advanced Research Projects Agency (DARPA). The project, then under the name of LifeLog, aimed at mapping all relationships, memories, events and experiences of an individual, but it was suspended the following year, probably due to public privacy concerns (Shachtman).

11_{Lifelogging can be defined as “a comprehensive archive of an individual's quotidian existence created with}

the help of pervasive computing technologies” (Allen 48).

12_{During these two decades, the Canadian professor and researcher Steve Mann designs, builds and wears}

several versions of computerized eyewear which allowed the recording of events as seen by his eyes (Mann,

My “Augmediated” Life).

13_{A book on the experiment and related considerations has been published by Gordon Bell and Jim Gemmell}

under the title Total Recall: How the E-Memory Revolution Will Change Everything.

14_{A documentary on Josh Harris and this particular experience in which the 100 volunteers agreed to live}

together for 30 consecutive days in a closed and fully (video) surveyed environment was released in 2009 <http://weliveinpublic.blog.indiepixfilms.com/>.

(16)

16

In the second half of the 2000s, with the popularization of social media platforms, and alongside a growing interest in the field of information visualization, many projects incorporating personal metrics emerge15_{. Two names which are often referred to in the context of personal analytics are}

Stephen Wolfram and Nicholas Felton. The first started consciously gathering information about his email messages back in 1989 (aspects such as volume, date and time – see Figure 3), incorporating afterwards indicators on keystrokes, calendar events, and phone calls, having compiled more than one million data points which he then visually represented in chronological graphs where life patterns became visible (Wolfram). The latter began publishing an Annual Report of his life in 2005 and has continued to do so on a yearly basis, consolidating statistics on the usage of time, books read, photos taken, places visited, food ingested, among many other indicators (Felton). In this case, the emphasis is put not only in the different life indicators which might be tracked every year, but also on the visual representation of the information – features which lead to all of his yearly reports being sold out.

Figure 3 – A plot of a third of a million email messages sent by Stephen Wolfram since 1989

Source: <http://blog.stephenwolfram.com/2012/03/the-personal-analytics-of-my-life/>

15_{See as examples, the entries for the 2008 competition on Personal Information Visualization by FlowingData}

(17)

17

Although quite diverse in nature and format, these examples illustrate the possibilities that lie within the personal logging domain. The goals attributed to these activities can range from self-discovery to (posthumous) self-preservation, from self-improvement to self-creation.

Currently, the rapid expansion of the smartphone market and the development of new consumer connected devices and wearables16_{, as well as growing media awareness of self-monitoring}

experiences, has sparked curiosity among the general public17_{about self-tracking possibilities, and}

has encouraged individuals who had already engaged in such activities to share their experiences more widely. The following section will then describe the social context of these present practices through the examination of the QS group.

16_{In 2011 the number of Internet connected devices (9 billion) surpassed already the world human population}

(approximately 7 billion), and two thirds of those devices fell under the mobile category with estimates pointing to 12 billion connected mobile devices in 2020 (Swan, Sensor Mania! The Internet of Things, Wearable

Computing, Objective Metrics, and the Quantified Self 2.0 218).

17_{It has also created reactions within the artistic community. As an example, see The Monthly Sculptures}

Determined by the Daily Quantification Records by British artist Ellie Harrison. The referred sculptures derived

from a project in which she tracked, on a daily basis, fourteen different aspects of her life for one year: <http://www.ellieharrison.com/index.php?pagecolor=3&pageId=project-monthlysculptures>.

(18)

18

3. A Social Contextualization of Current Self-Tracking Practices

3.1 – The emergence of the Quantified Self (QS) group

Wired magazine editors Gary Wolf and Kevin Kelly introduced in 2007 the concept of ‘Quantified Self’ (QS)18_{to designate the personal monitoring and measuring practices they observed in their direct}

network of acquaintances (Wolf, Know Thyself: Tracking Every Facet of Life, from Sleep to Mood to Pain, 24/7/365). Collaborating with one of the leading publications in the technology industry, which some proclaim to advocate techno-libertarian values (Willis), both names are active figures in identifying innovative trends emerging in the technological landscape. Kelly was actually one of the founding members of the magazine in 1993 and is considered to be a reputable figure in the technological sphere. He has also published numerous articles and books that span beyond the topic of technology, and founded two non-profit organizations (Kelly, Biography). Wolf is similarly a prolific writer19 and is currently working on a book under the title The Quantified Self. He is interested

in the topic of self-knowledge but on a larger scale, and in that domain he invokes the term ‘macroscope’ to refer to a “technological system that radically increases our ability to gather data in nature, and to analyze it for meaning” (Wolf, QS & The Macroscope).

The idea of personal insights combined with measurements is also patent in the QS motto “Self knowledge through numbers” visible on their website <http://quantifiedself.com> which serves as an important platform in a collaborative movement attracting users from all over the world. In a period of six years (from September 2007 to October 2013) more than 800 articles were published in that website by 34 authors (see Graph 1). However, the nuclear publishing team consists of Gary Wolf, Kevin Kelly, previous Director Alexandra Carmichael20_{, and current Program Director Ernesto}

18_{The ‘Quantified Self’ concept attracted more mainstream attention through a TED talk given by Gary Wolf in}

2010. In nearly three years, the video <http://www.ted.com/talks/gary_wolf_the_quantified_self.html> has gathered approximately 400.000 visualizations.

19_{In his personal website, he refers that one of his favorite articles is about the supermemo}

<http://www.wired.com/medtech/health/magazine/16-05/ff_wozniak>, a learning system that uses spaced repetition to seal knowledge in memory devised by the Polish researcher Piotr Wozniak. This system could be ultimately classified as a tool for cognitive self-improvement.

20_{Alexandra Carmichael is co-founder of the collaborative health research site CureTogether, a Research}

Affiliate at the Institute for the Future, and a regular blogger on personal data topics (Wolf, Welcome Alexandra

(19)

19

Ramirez21_{. The web articles range from presentations of personal projects to suggested literature}

related to the self-tracking topic, from summaries of previous QS Show&Tell events to interviews with tool makers, including also any relevant updates on new devices, platforms, and upcoming gatherings.

Graph 1 – Number of published articles in the QS website per year and per author (within the top 4 publishing authors) (November 2013)

Besides sharing information online, the QS group is also engaged in regular face-to-face interaction: the list of events includes already five Global Conferences (the last one held in San Francisco listed more than 400 participants), and more than 600 meetings organized by approximately 100 local groups in cities in all continents22_.

21_{Ernesto Ramirez is a PhD candidate in Public Health at the University of California, San Diego (Carmichael,}

Welcome Ernesto Ramirez!).

22_{These Show&Tell meetings can be initiated by active users in any country and differ from the Global}

Conferences which are directly organized by the social enterprise created by the founders Gary Wolf and Kevin Kelly to support the QS movement designated by QS Labs (also responsible for the website). However, this central group provides recommendations for local initiatives and is also willing to contribute with financial or logistic support. More detailed information can be found on this page < http://quantifiedself.com/how-to-start-your-own-qs-showtell/>. 0 20 40 60 80 100 120 140 160 180 200 2007 2008 2009 2010 2011 2012 2013

(20)

20

The movement is open to all types of self-tracking and encourages users to share their experiences in the areas of health, nutrition, sleep, fitness, cognition, mood and happiness, focusing on the methodology used, as well as the results achieved. At a first glance, the objective does not seem to greatly differ from the one established by the aforementioned CARPE (Continuous Archival and Retrieval of Personal Experiences) workshops held until 2006 (see section 2.2), but while those accepted only professional researchers (either from the academic or corporate spheres) and had a formal structure, the QS initiative is open to anyone who has engaged in some meaningful type of self-monitoring project and is usually conducted with a certain degree of informality.

Users do not have to comply with the established rules of the scientific method, but the group considers the results of this (researcher) citizen science (Cornell, Making citizen scientists) or personal science (Roberts) to be valuable and relevant to the scientific community. The recognized benefits of research centered on one single individual (the n=1 type of studies are also present in science23_{) can}

include the existence of repeated, longitudinal data, and customized treatments, while the potential risks comprise aspects such as mortality, history, maturation and treatment fidelity (Carmichael, Daniel Gartenberg: The Role of QS in Scientific Discovery). The concern with strict scientific validity is not a driving force in most experiments, since the goal does not relate to generalizing the results to a population, but understanding their meaning for the individual and eventually to inspire others to undertake an analogous type of examination. Similarly to scientific practice, these experiments cannot deliver certainty, but only methodically explore a range of possibilities with the prospect of meaningful results.

The participants create their own experiments and try to document them as well as possible. A personal presentation, often video recorded and then posted online24_{, is structured according to the}

three QS prime questions (Wolf, Our Three Prime Questions): 1) What did you do?, 2) How did you do it?, and 3) What did you learn?.

The experiences do not have necessarily to rely on the latest technological devices – the users can resort to simple spreadsheets, basic word processing software, or a combination of both basic and complex techniques - and the results do not have to be purely expressed in numerical values, which might constitute a surprise for those who are less familiarized with the group’s activities taking into consideration its slogan (“Self knowledge through numbers”). In fact, the words “quantified” and

23_{On this matter, see the 1981 article by Allen Neuringer “Self-experimentation: A Call for Change”.}

24_{In October 2013, the Quantified Self group in Vimeo <}_{https://vimeo.com/groups/quantifiedself}_{> included}

more than 500 videos from presentations at the Global Conference and the local Show&Tell meetings covering a time period of four years.

(21)

21

“numbers” should not be taken literally (see section 5.2 on the theme of quantification), since they serve mainly to emphasize aspects related to experimentation, systematization, data and, to a certain extent, technology. In that sense, a more accurate version of that sentence could be “Self knowledge through data”. The focus of the movement is placed in the sharing and learning features, so customized self-tracking methodologies trying to establish unusual correlations between different datasets, or using DIY or ‘hacked’ devices, are welcome. The ‘Quantified Self’ moniker has also inspired reactions from other communities, such as the artistic one with at least two art exhibitions25

under that title organized so far.

3.2 – The QS group within the self-tracking spectrum

The QS group is only the visible side of a larger group of self-trackers. The following illustration, proposed by Sociology Ph.D. student Whitney Erin Boesel, categorizes the activity of individual monitoring according to criteria such as personal intention and awareness, and helps position the QS group within the wider tracking spectrum.

25_{The first art exhibition was held in 2011 at the LAB Gallery in Dublin, Ireland (see}

<http://www.dublincity.ie/RecreationandCulture/ArtsOffice/TheLAB/PreviousExhibitions/Pages/Quantifie dSelf.aspx>) and the second one in 2012 at the Gallery Project in Detroit, Michigan, U.S, (see <http://www.annarbor.com/entertainment/gallery-project-quantified-self/>).

(22)

22

Figure 4 – Typologies of Individual Tracking (dimensions not to scale)

Source: <http://thesocietypages.org/cyborgology/2013/05/22/what-is-the-quantified-self-now/>

The wider circle refers to all forms of monitoring, including the ones which are performed at a macro level, and therefore related to issues of societal surveillance which I briefly touched upon in section 5.3, but are not the focus of this study. Situated within that wider circle is the area of individual tracking which includes conscious and non-conscious monitoring. The latter refers, for instance, to the user’s digital trail or information captured without the individual being aware of it (i.e. logging aspects of the user’s online behavior, such as visited websites for profiling purposes). Regarding what is then considered to be voluntary self-tracking, it is possible that this activity is either performed upon request from another individual or organization (i.e. the request from a physician for medical reasons26_{) or self-initiated. The boundaries between these typologies are not always so clearly}

26_{Members of the medical community and the health industry are regular attendees of the QS Global}

Conferences and many are excited with the possibilities offered by this type of technology contemplated under the ‘digital health’ or ‘m-health’ (mobile health) scope (Lupton, M-health and Health Promotion: The Digital

(23)

23

defined in reality, and the terminology used may not be the most accurate, but the benefit of such a scheme is to facilitate an initial approach to these categories and understand the distinction between groups. As previously stated, this study is located within the sphere of voluntary, conscious and self-initiated experiments.

Following the above classification, some questions naturally impose themselves. The first one is: beyond the QS group, how many people are performing an activity which might fall under the ‘self-tracking’ category27_{? The definition of the term may vary depending on the source cited, but the one}

used for this study has been operationalized in the Introduction section.

Early in 2013, the Pew Research Center divulged the results of a survey on the current status of health self-tracking in the U.S. and, even though 70% of the respondents admitted to track some type of health indicator, nearly 50% of those did not take note of the values and, from those who did, only 21% did it with the use of technology. Furthermore, it was found that the act of self-monitoring is closely linked with chronic conditions, since only 19% of the self-trackers claimed not to have any chronic disease (Fox, Tracking for Health). In the previous year, the same organization published a report on mobile health where 19% of smartphone owners (45% of the U.S. population) had downloaded at least one health app on their phone. More than 80% of these health apps pertained to the exercise, diet and weight categories (Fox, Mobile Health 2012 11).

In January 2013, Forrester published the findings of their market research study on health tracking devices where a mere 4% of the U.S. adult population is estimated to match the profile of a consumer who would be interested in purchasing a fitness wearable (Colella). The perception of active self-monitoring is here also associated with chronic conditions, a very specific health goal, or an obsessive type of personality. Even if by 2012 figures, more than 500 companies in the health industry were developing self-tracking tools (Swan, The Quantified Self: Fundamental Disruption in Big Data Science and Biological Discovery 86), apps and wearable devices do not seem to be extensively popular within the mainstream consumer market. At least, not yet.

In 2013, a report from IMS Research estimated that installation of sports and fitness apps on smartphones would grow 63% from 2012 to 2017 (IHS Electronics and Media Press Release). It is relevant to clarify, taking into consideration the apparently conflicting information, that the purchase of a device or the installation of an app does not imply its regular use. As alluded by some observers,

27_{The aforementioned Sociology Ph.D. student Whitney Boesel published in 2013 an article exploring in more}

detail the topic of exclusion from the QS community and definition of its membership status (Boesel, You, Me,

(24)

24

the predicament with these types of devices lies precisely in the lack of sustainable use (Swan, Sensor Mania! The Internet of Things, Wearable Computing, Objective Metrics, and the Quantified Self 2.0 240). The above observations deserve two brief notes. The first is that self-tracking does not require a smartphone or a wearable device, so many reports fail to account for such cases. The second is that if one would want to be accurate in the definition of self-tracking, monitoring aspects related to behavior and lifestyle would also have to be included. In such scenario, it would not be possible to propose realistic figures regarding the number of people committed to practices of self-tracking. Following the question of volume, comes one of characterization: are there particular features which distinguish active self-trackers from the remainder of the population? In a QS website post dating from 2010, the ex-NASA engineer Matthew Cornell proposes the potential attributes of the ‘data-driven personality’ of a self-tracker which can be summed up as follows: insatiable curiosity, willingness to take risks and continuously change, skepticism, problem solving mentality, and early adoption of gadgets (Cornell, Is There a Data-Driven Personality?). It is naturally an insider’s perspective which can be conflicting with the external image of individuals with a compulsive or obsessive personality, as referred to in the results of the study previously mentioned, or with a narcissist disposition – a matter tackled empirically by Gary Wolf in 2009. A survey based on the questions from an approved narcissist psychological assessment test was distributed among the QS group, and no correlation was found between conducting self-tracking activities and levels of narcissism above average (Wolf, Are Self-Trackers Narcissists?). The results could eventually be contested, since the sample considered was relatively small and not necessarily representative, but the main objective was to highlight the fact that if there are particular traits that differentiate self-trackers from non-self-self-trackers, then self-centeredness is not one of them.

The subsequent section moves beyond contextualization efforts to focus more specifically on the self-monitoring practice per se and the particular process it entails.

(25)

25

4. A Functional Analysis of Self-Tracking Practices

In order to further clarify the notion of self-tracking, it is fruitful to analyze these monitoring practices from a functional perspective and examine the several elements and stages involved in the process. The studies published in this domain can provide specific terminology and a structured approach which may be of assistance in the empirical stage.

4.1 – A definition of Personal Informatics (PI) and a taxonomy for

self-tracking

The term personal informatics28_{, defined as systems which “help people collect personally relevant}

information for the purpose of self-reflection and gaining self-knowledge” (Li, Dey and Forlizzi, A Stage-Based Model of Personal Informatics Systems 558), can comprise functions related to personal information management29_{, social networking}30_{, coordination}31_{, and memory}32_{(Li, Dey and Forlizzi,}

Understanding My Data, Myself: Supporting Self-Reflection With Ubicomp Technologies 408), besides the ones related to health and wellbeing referred previously. While the first are relevant functions, it is important to state that they are not directly examined in this research, since their nature is rather distinct from the one under analysis.

28_{The website <}_{http://personalinformatics.org}_{> created by Ian Li, who published a PhD thesis on “Personal}

Informatics & Context: Using Context to Reveal Factors That Affect Behavior” in 2011, seems to be a central platform for several resources in this field, ranging from lists of personal informatics tools to papers on the topic.

29_{This category can include standard and popular tools such as calendars <}_{http://www.google.com/calendar}_>,

contact lists <http://www.plaxo.com/>, mind maps <http://www.mindmup.com>, notes <http://evernote.com/>, reminders <http://www.rememberthemilk.com/>, among many others.

30_{In this case, taking note of one’s habits or preferences might be only the means to the goal of establishing or}

reinforcing social contact. Users can then dutifully record, for instance, their listening habits <http://www.last.fm/>, reading selection <http://www.goodreads.com/>, or places visited <https://apps.facebook.com/tripadvisor/> as a means to promote social networking.

31_{This item can be closely intertwined with the personal information management function and it can also}

include tools which are commonly used in a professional environment.

32_{On the subject of technological devices primarily conceived to aid individual memory, read the 2006 article}

“iRemember - A Personal Long Term Memory Prosthesis” reporting an experiment conducted by Sunil Vemuri, Chris Schmandt, Walter Bender from the MIT Media Lab.

(26)

26

The classification of a self-tracking project varies depending on the criteria employed. In the first MA thesis on the QS group published in 2012, Anthropology student Adam Butterfly conducted an ethnographical study within the QS collective and identified three axes according to which these personal experiments could be categorized: 1) degree of technological involvement, 2) level of complexity, and 3) goal type (ranging from driven or exploratory) (Butterfly).

Such taxonomy brings to life a three-dimensional spectrum of possibilities within self-tracking experiments and, while the axes are independent, they can at times be closely intertwined. For instance, the device choice may be associated with the goal established. The terms persuasive technology and mindful or reflective technology (Munson) can be used to differentiate between tools which try to steer the user’s behavior towards a certain direction and tools which focus on insights based on individual behavior33_{. So the mere choice of one device over another can influence,}

consciously or not, the development of an experiment and the opposite can also happen: the formulation of a certain goal determining the choice of technology. Simultaneously, projects which started as being mostly exploratory and considering many variables can become more concentrated on particular goals with a reduced number of variables, or vice-versa. It is a fluid field where the position of the experiment can continuously shift under the guidance of its author. In order to better understand the possibilities offered within those three axes, the following section will examine the steps of the self-tracking process.

4.2 – The stages of the self-tracking process

The Stage-Based Model of Personal Informatics proposed by Li, Dey and Forlizzi in 2010 provides a supportive scheme on the self-tracking process. In the model the authors identified five consecutive stages guiding a self-monitoring procedure: 1) preparation, 2) collection, 3) integration, 4) reflection, and 5) action as illustrated in Figure 5.

33_{These terms should not be mistaken for the dichotomy between fast technology and slow technology (Hallnäs}

and Redström 201) where the first is based on efficiency and performance, and the second one on contemplation and reflection. One recent example of contemplative technology is the Decelerator Helmet by the German designer and artist Lorenz Potthast <http://www.lorenzpotthast.de/deceleratorhelmet/>.

(27)

27

Figure 5 - Stage-Based Model of Personal Informatics

Source: Li, Dey and Forlizzi, A Stage-Based Model of Personal Informatics Systems 561

Once the motivation to conduct a self-tracking experiment is set, the individual enters the preparation phase of the process translating the initial intention into a goal which can be rather abstract (as in the case of an exploratory study), or quite specific34_{. Then follow decisions regarding}

the type of data to collect, along with the methodology and regularity of the procedure. The data can include physiological indicators (i.e. heart rate, body temperature, skin galvanic response), physical activity (i.e. steps taken), affective conditions (i.e. mood), behavior (i.e. hours spent executing certain activities), and these categories are not mutually exclusive. Some authors claim that some of the most surprising and meaningful experiments derive from the combination of high valence (i.e. mood) and low valence (i.e. heart rate) human values which create more actionable results (Swan, Sensor Mania! The Internet of Things, Wearable Computing, Objective Metrics, and the Quantified Self 2.0 239). The data type definition will then impact the technology in the collection stage, which can range from being user-driven (also labeled as active collection) to system-driven (or passive collection), with several possibilities within that spectrum depending on the complexity and goals of the initiative. Manual operations are commonly deemed as more demanding, as they depend on the individual’s motivation and discipline. On the other hand, automated collection can also bring about

34_{According to some theories, personal goals can be classified within a hierarchical scale ranging from very}

abstract to very specific including the following four levels respectively: system concept, principle level, program level, and sequence level (Li, Dey and Forlizzi, Understanding My Data, Myself: Supporting

(28)

28

disadvantages, especially when a high volume of data is being harvested in an exploratory study where the correlations between the indicators has not been clearly established up-front35_.

Another important choice relates to the collection frequency, which can either be continual (hourly, daily, weekly), or episodic (only when a particular event happens). These decisions can precede the choice of gadget or, if the device is actually the guiding element of the experiment, be a byproduct of the technology selected. Currently, there is a wide variety of products and services in this area which allow the choice between platforms catering for highly specific needs or supporting a generic purpose36_.

The third stage presented – integration – refers to the act of processing the data into a structured visual output, and its duration is determined by the answers to the initial questions. If the data collection is user-driven and manual, then the user is responsible for producing the information visualization directly. When technology is driving the operational side of the experience, this step can be relatively short since the visualization is usually automatically generated. Without delving too deeply into the information visualization field37_{, it might be relevant to refer that several studies have}

been conducted to examine how different elements of personal data visualization impact the user’s subsequent behavior38_{. A number of tools offer the possibility of personal customization within a}

pre-established range of options, even if some authors argue that personal data should be matched with deeply customized visualizations for additional meaning (Aseniero, Carpendale and Tang).

It is possible that amidst the self-tracking experiment, the user decides to change the technological platform or device used (the reasons can be connected to inconvenient data collection, complexity of the technology involved, issues with data visualization, among others), which then raises questions related to the interoperability of the data39_{. In cases where the data migration is not possible, or it}

35_{As claimed by some authors, the success of passive lifelogging depends on establishing relationships between}

captured items and focusing on the truly relevant ones (Gemmell et al. 54).

36_{Some authors associate more comprehensive approaches with multi-faceted systems and targeted ones with}

uni-faceted systems (Li, Dey and Forlizzi, A Stage-Based Model of Personal Informatics Systems 564). See Table 8for some examples in the generic category.

37_{In the second half of the 2000s some authors categorized information visualization projects dealing with}

individual data for personal consumption under the label ‘casual information visualization’. For more information, see the 2007 article “Casual Information Visualization: Depictions of Data in Everyday Life”.

38_{In the 2013 paper “Persuasive Performance Feedback: The Effect of Framing on Self-Efficacy”, the authors}

study the impact of three types of framing effects on individual behavior: valence of performance, presentation type, and data unit (Choe et al.).

39_{The topic of data portability is also discussed within the QS group and self-trackers are advised to consider}

this aspect prior to running the experiments (Plattel). To tackle this problem, as well as facilitating the simultaneous use of data from different devices, several products and services dedicated to API integration

(29)

29

implies a level of technical knowledge which the user does not possess, then the monitoring process needs to be re-initiated40_{. This is one of the risks of what is designated by the barriers cascade}

property of the model, where initial complications trickle down to ensuing phases.

In the reflection stage, the user approaches the gathered personal data critically. According to another study conducted by Li, Dey and Forlizzi, the user can then ask questions fitting into one or more of the following six categories: 1) status (focusing on the present), 2) history (analyzing the data longitudinally), 3) goals (what still needs to be achieved or which targets should be set), 4) discrepancies (examining the difference the current status and future goals), 5) context (concentrating on secondary elements related to the main indicators collected), and 6) factors (understanding correlation and establishing causality between elements) (Li, Dey and Forlizzi, Understanding My Data, Myself: Supporting Self-Reflection With Ubicomp Technologies 408). The boundaries between those categories are not always clearly distinguishable, as one type of interrogation may naturally lead to another one, but some might be more common in an exploratory experiment (which Li describes as the discovery phase), and others in a program with specific objectives (maintenance phase).

When the data is derived from an automated, or semi-automated, system working in a continuous mode, volume can become a challenging factor in the interpretation phase. Some authors refer in this context the materialization of new data flows which demand a fine-tuned ability to identify patterns41_{, anomalies, and establish correct correlations at a faster pace (Swan, Sensor Mania! The}

Internet of Things, Wearable Computing, Objective Metrics, and the Quantified Self 2.0 235). The obstacle does not usually rely on the harvesting of the data itself, but on the following sense making stage42_.

The interpretation of the individual data can then lead to behavioral change, even though the experiment does not necessarily have to achieve the action stage, and can remain solely as a personal

management have emerged (some examples: Fluxstream <https://fluxtream.org/>, Healthgraph <http://developer.runkeeper.com/healthgraph>, Sense <http://open.sen.se/>, Singly <http://singly.com/>, Sympho<http://sympho.me/>).

40_{Even when the data migration is a feasible possibility, some authors point to the (de)contextualization of the}

data captured by a certain piece of technology as one of the challenges faced by personal informatics tools (Brubaker, Hirano and Hayes).

41_{In a QS post from 2010, Matthew Cornell provides some basic strategies to pursue meaningful patterns in}

personal data (Cornell, Patterns.)

42_{Curiously, Vannevar Bush alerted for a similar issue already in his 1945 article “As We May Think”: “The}

difficulty seems to be, not so much that we publish unduly (…), but rather that publication has been extended far beyond our present ability to make real use of the record” (Bush).

(30)

30

exploration. There is a set of cognitive and behavioral theories which are usually presented in research related to personal change. One of them is the Trans-Theoretical Model of Behavior Change43_{, which proposes change as a sequential operation incorporating five stages}

(precontemplation, contemplation, preparation, action, and maintenance), and ten different types processes (under the experiential and behavioral categories) (Velicer et al.). Other studies refer the Social Cognitive Theory (Bandura 1) which also emphasizes the external social context of the individual. Other theoretical frameworks presented to examine the topic of intentional behavioral change, include the Cognitive Dissonance Theory (Festinger), focusing on the establishment of internal consonance, and the Presentation of Self Theory (Goffman) building a metaphor between regular human interaction and a theatrical performance 44_{. While it would be interesting to explore}

behavioral approach, this study will not focus directly on the action stage due to its specific scope. From the above description, it is important to retain that the collection and reflection stages are particularly important as being the ones which can be user-driven or system-driven – a useful distinction to bear in mind when empirically analyzing self-tracking experiments. Additionally, it will be useful to verify if some of the issues reported above, such as data interoperability and information overload, are commonly faced by self-trackers.

However, prior to the empirical part, it is relevant to examine the self-tracking practices also from a conceptual point of view in order to characterize the social, cultural and technological context in which they occur.

43_{Some health focused studies present a critical perspective towards this model, claiming that it focuses more}

on attitude than behavior, and that it has its limitations when addressing long-term goals (Maitland et al. 2).

44_{One study, building upon the premises of most of the above theories, and complementing it with empirical}

research, proposed the following eight properties when designing a self-tracking app or device leading to successful behavioral change: it should be 1) abstract and reflective, 2) unobtrusive, 3) public, 4) aesthetic, 5) positive, 6) controllable, 7) trending / historical, and 8) comprehensive (Consolvo, McDonald and Landay 408). Other studies underline aspects such as usability, goal consonance, and understanding of the underlying technology as important elements (Andrew, Borriello and Fogarty).

(31)

31

5. A Conceptual Analysis of Self-Tracking Practices

The goal of self-knowledge and the desire for self-improvement have informed numerous theories and movements throughout human history. In Ancient Greece, the philosopher Socrates constantly provoked his fellow citizens, encouraging them to have a critical stance regarding what they considered to be their self-knowledge. In this regard, one of the famous sentences attributed to him - “the unexamined life is not worth living for a human being” (Plato) - illustrates his belief in the individual practice of systematic inquisition45_.

For several centuries, at least in the western world, the pursuit of knowledge was situated in the realm of the transcendental, and genuine insight, whether about oneself or the universe, would only be obtained through religion. In the seventeenth century the focus started shifting from the contemplation of the divine towards the analysis of the terrestrial and humane, with rationalist thinkers such as Descartes (“I think, therefore I am”), and towards the observational with empiricist authors such as Locke (with the concept of the human being as a blank slate). With the advent of the Enlightenment in the eighteenth century, science and secular education emerged as fundamental sources of knowledge, a situation which for some was still compatible with religious faith, while for others it implied an abrupt rupture with tradition, leading to impactful events as the French Revolution. Another development worth referring is related to the notion of quantification applied to the social and individual spheres, which is materialized in the utilitarian theories advocated by Bentham and Stuart Mill, focusing on the maximization of happiness and the calculus of pleasure. In the modern period, thinkers from a diversity of fields held views which may be of interest to briefly invoke in the light of self-tracking emotional states before zooming into contemporary theories which already integrate technology as a central element. Self-knowledge as a constitutional individual concern is emphasized by Kierkegaard (“one must first learn to know oneself before knowing anything else” (Kierkegaard 10), who was heavily influenced by Socrates. Nonetheless, the path to attain such knowledge was by no means consensual. Some thinkers, such as Nietzsche and Emerson, argued that focusing on the past would be detrimental for the individual, and that the ability to forget was essential for personal happiness. Others, namely Freud, claimed that only through understanding the past was one able to gather meaningful insights and reduce personal suffering (a distinct proposition from the one aiming at maximizing happiness).

45_{For an in-depth analysis of the hermeneutics of the self in the Greco-Roman philosophy and its comparative}

(32)

32

Additional discussions can revolve around the fact that self-discovery is secondary to the capacity of personal development, as stated by Foucault: “Modern man, is not the man who goes off to discover himself, his secrets and his hidden truth; he is the man who tries to invent himself” (Foucault, The Foucault Reader 42). On a more structural level, one could argue whether this capacity for invention would ultimately lead to self-improvement and satisfaction, or even if happiness itself, which most human beings claim to tirelessly pursue, is altogether a hypocritical category, as polemically argued by Žižek, since it drives the individual to dream about things he does not really want (Žižek 60). This section will then refer to specific contemporary theories supporting the social, cultural and political contextualization of self-tracking practices with the purpose of understanding the causes and the potential consequences of this phenomenon.

5.1 – The intensified inward gaze, healthism and the pursuit of the perfect

self

The contemporary uncertainty “predisposes the postmodern self to take uneasy refuge in the most basic shelter of all: his or her own body” (Chrysanthou 470). The outward gaze in the quest for knowledge and purpose (i.e. in religion, in social community), shifts towards the self and gradually zooms into every aspect of the individual existence, amplifying its weaknesses and revealing the unfulfilled potential. This is the privileged ground for many of the self-monitoring activities under study, whether they are concerned with fitness, particular health aspects or mood and happiness. Self-monitoring activities are usually conducted in the spirit of gathering self-knowledge, which will ultimately lead to self-improvement. This procedure stresses the notion of the human being as an inherently flawed figure, but aspiring to a model of perfection which is believed to be achieved through an iterative and conscious process. Currently, the attribute of excellence resides, first and foremost, within the individual, an idea clearly illustrated by Chrysanthou’s statement: “perfectibility is displaced from the political sphere to the personal” (Chrysanthou 471). This goal can be accomplished in several fronts, but particular emphasis is placed on the physical, intellectual and emotional wellbeing.

According to the same author, health has become a new ideology, and within this healthism movement, intensified through the means of connected mobile technology, the onus is also transferred from the public and collective dimension, to the private and personal one (Crawford

The Quest for Happiness in Self-Tracking Mobile Technology