International Journal of Medical Informatics 96 (2016) 38–50

Contents lists available at ScienceDirect

International Journal of Medical Informatics

journal homepage: www.ijmijournal.com

Making time for mindfulness

James Laurie, Ann Blandford ∗
UCLIC and Institute of Digital Health, UCL, Gower Street, London WC1E 6BT, UK

a r t i c l e i n f o a b s t r a c t

Article history: Objective: Digital mental wellbeing interventions are increasingly being used by the general public as
Received 4 February 2015 well as within clinical treatment. Among these, mindfulness and meditation programs delivered through
Received in revised form 29 January 2016 mobile device applications are gaining popularity. However, little is known about how people use and
Accepted 25 February 2016
experience such applications and what are the enabling factors and barriers to effective use. To address
this gap, the study reported here sought to understand how users adopt and experience a popular mobile-
Keywords:
based mindfulness intervention.
Mindfulness
Methods: A qualitative semi-structured interview study was carried out with 16 participants aged 25–38
Behaviour change
Usability
(M = 32.5) using the commercially popular mindfulness application Headspace for 30–40 days. All par-
User experience ticipants were employed and living in a large UK city. The study design and interview schedule were
Situated use informed by an autoethnography carried out by the first author for thirty days before the main study
began. Results were interpreted in terms of the Reasoned Action Approach to understand behaviour
change.
Results: The core concern of users was fitting the application into their busy lives. Use was also influenced
by patterns in daily routines, on-going reflections about the consequences of using the app, perceived
self-efficacy, emotion and mood states, personal relationships and social norms. Enabling factors for use
included positive attitudes towards mindfulness and use of the app, realistic expectations and positive
social influences. Barriers to use were found to be busy lifestyles, lack of routine, strong negative emotions
and negative perceptions of mindfulness.
Conclusions: Mobile wellbeing interventions should be designed with consideration of people’s beliefs,
affective states and lifestyles, and should be flexible to meet the needs of different users. Designers should
incorporate features in the design of applications that manage expectations about use and that support
users to fit app use into a busy lifestyle. The Reasoned Action Approach was found to be a useful theory
to inform future research and design of persuasive mental wellbeing technologies.
© 2016 Elsevier Ireland Ltd. All rights reserved.

1. Introduction anxiety management programs. If such technologies are to be effec-

tive vehicles for health behaviour change, then it is necessary to
The penetration of ubiquitous technologies into every corner of understand both how they are used in practice, what are the key
our lives provides excellent opportunities for new forms of health features of experience, and what are the barriers and enablers
and wellbeing interventions. [1] Among these technologies, mobile of use. A deeper understanding of these issues will contribute to
applications offer cost-effective methods of healthcare delivery. theoretical discussions about the design of e-health technologies
Mobile devices are carried with the user throughout the day, can and persuasive design (see, for example, [3,4]). The study reported
connect to the internet and can deliver interventions using fea- here investigated how people adopted one mental wellbeing app
tures such as push notifications, text based information and rich (Headspace, www.headspace.com), how they fitted use into their
media content. These features make mobile devices particularly lives, and how they experienced use over time.
suitable for the delivery of mental wellbeing interventions [2].
Consequently, there is a growing market for consumer healthcare
2. Background
applications that deliver interventions such as stress reduction and
Mindfulness is a growing area in the field of e-mental health,
within and outside clinical practice. Mindfulness is “a state of psy-
∗ Corresponding author. chological freedom that occurs when attention remains quiet and
E-mail address: a.blandford@ucl.ac.uk (A. Blandford). limber, without attachment to any particular point of view”. [5]

http://dx.doi.org/10.1016/j.ijmedinf.2016.02.010
1386-5056/© 2016 Elsevier Ireland Ltd. All rights reserved.
 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50 39

Many claims have been made for the efficacy of mindfulness for
mental wellbeing. Advocates argue that increased mindfulness is
correlated with greater self-control [6], objectivity [7], concentra-
tion [8], capacity to deal with stress [9] and empathy [10].
Kabat-Zinn [11] developed a mindfulness based stress reduction
course that taught a series of techniques to cultivate mindfulness,
such as attention focus exercises, body scans, breathing exercises
and yoga-based exercises. The purpose of these exercises is to cul-
tivate an awareness of self: to learn to pay attention to the way the
mind actually works, and to what is really going on in our bodies
and in our emotional states.
Research has explored the effectiveness of mindfulness among
clinical populations, reporting evidence that it can reduce symp-
toms of depression [12], anxiety [13] and severe mental illness
[14], and improve the quality of life of people suffering with phys-
ical health conditions such as chronic pain and fibromyalgia [4]. In
non-clinical populations, mindfulness programs have been associ-
ated with reductions in stress levels and increased wellbeing in the
workplace among doctors [15] and office workers [16].
This strong empirical basis has contributed to mindfulness med-
itation becoming a growing trend in Europe and America [17].
Mobile application developers are exploring how to deliver mind-
fulness through mobile phones and other mobile digital devices.
Plaza et al. [18] found 50 different mobile-based mindfulness
apps available for Android, IOS and Windows devices, including
Calm.com, Buddify, the Meditation App and Headspace.
Fig. 1. Timeline.
2.1. Headspace

Headspace is a commercially successful web-based application

that delivers mindfulness and meditation related content. In August
2014, Headspace had around 523,000 users in the UK and around
1.3 million users worldwide. [19] The app was chosen for the cur-
rent study because it is among the most downloaded mindfulness
apps; controlled experimental studies have reported benefits of use
(as described below); and the app developers were supportive of
our study.
Headspace delivers a course of guided mindfulness meditations,
using both audio and video files. These guided meditations are
voiced by Headspace founder Andy Puddicombe, a former Buddhist
monk. Content can be downloaded through the company’s website
or via a mobile application, available on both Android and Apple
iOS platforms. The full content comprises 365 different guided
meditations. Users are encouraged to complete a meditation each
day. Progress through the course can be viewed on a timeline that
appears on the home screen (Fig. 1). Much of the program follows
a linear pathway of daily meditations. Tapping the play button on
the timeline begins the next meditation in the series. This opens
the meditation interface (Fig. 2), which consists of a simple timer
and a play button to begin the guided meditation. Users begin with
a course of ten, 10-min, daily meditations to learn the basics. These
daily meditations increase to fifteen and then twenty minutes. After
the completion of a thirty-day foundation course, the user is given
access to advanced content related to health, relationships and per-
formance (Fig. 3). In the study reported here, participants engaged
with the foundation course. Fig. 2. Meditation interface.

2.2. Previous research into mindfulness E-Mental health symptoms of depression and anxiety followed programs of daily
interventions mindfulness exercises delivered via the Internet to desktop com-
puters in their homes. The interventions were associated with
A number of quantitative studies have demonstrated the significantly decreased symptoms of perceived stress, depression,
efficacy of mindfulness technology interventions among clinical anxiety and insomnia when compared with a control group.
populations. For example, randomised control trials have been Headspace has been identified as the highest-scoring mindful-
carried out by Glück & Maercker [20], Cavanagh et al. [21] and ness app on the Mobile Application Rating Scale (MARS) [23], and
Boettcher et al. [22] to test the effectiveness of online mindful- has been shown to be beneficial in controlled trials. Howells et al.
ness based interventions. In these studies, participants with clinical [24] conducted a randomised controlled trial with 121 ‘happiness
40 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50

dropped out for any reason, and the data does not focus on user
experience or how participants fitted use of the app into their lives.
The previous study that is closest in approach to that reported
here is a one-month field trial of a mobile app called Oiva reported
by Ahtinen et al. [28] 15 people, all employed at a university, par-
ticipated in their study. The focus of the study was on usage, user
acceptance and perceived usefulness of the app. A mixed methods
approach was taken, using quantitative measures with in-depth
interviews. The focus of the interviews was on acceptability and
perceived usefulness of the app. The main barrier to use was found
to be busyness; several participants also reported forgetting to use
it unless they had built use into their daily routine. Benefits of using
the app were found to include reduced stress and the learning of
new exercises that could be built into daily life. In the discussion,
we compare our findings to those of Ahtinen et al. [28]

2.3. The reasoned action approach

Hekler et al. [29] argue that macro theories of behaviour devel-

oped within the mainstream psychology community can provide
useful conceptual frameworks to guide the evaluation of health
and wellbeing behaviour change technologies. For instance, Grimes
& Grinter [30] used the stages of change theory from the Trans-
theoretical Model [31] to structure their analysis of interview data
when assessing a diet change game. Similarly, Hsu and Blandford
Fig. 3. Performance pack. [32] drew from the cost−reward construct of Decision Theory to
analyse interviews of people trying to lose weight.
RAA is the latest in a family of theories of human behaviour
seekers’, 57 of whom used Headspace over a period of ten days (64 developed by Fishbein and Ajzen [33]. Earlier versions of this
participants were in a control condition). Their study showed statis- approach include the Theory of Reasoned Action (TRA) [34] and
tically significant changes in positive affect and reduced depressive the Theory of Planned Behaviour (TPB) [35]. RAA describes how
symptoms. Bostock & Steptoe [16] carried out a controlled trial of environmental factors, skills and abilities affect the likelihood of
the Headspace app with 238 workers from two large UK companies. intentions becoming actions. The three primary conditions are
The intervention was associated with increased subjective ratings intention to act, ability to perform the action and suitable environmen-
of well-being and reductions in job strain in comparison with a tal conditions to act. Intention to act is, in turn, predicted by attitude
wait-list control group. There was also a non-significant trend for and beliefs about the behaviour, belief in one’s ability to perform the
lower blood pressure after eight weeks. However, neither of these behaviour (self-efficacy), and the influence of social norms upon the
studies investigated user behaviour or the user experience of the actor. Beliefs about behaviour include both beliefs about what the
application. outcome of a behaviour will be, and whether the outcome is of value
Few qualitative studies have been conducted into mobile-based to the person. Perceived behaviour control, or self-efficacy, is the
mindfulness interventions, so very little has been written about extent to which a person believes that they are capable of carrying
the user experience of these applications. Morris et al. [25] report out a required behaviour. Social norms are the subjective interpre-
a series of case studies for a mobile application that delivered tations and reflections upon social attitudes. This can include norms
mood tracking and thought control exercises. They found that inherited from significant others as well as the wider culture.
participants quickly understood and internalized the principles of The Reasoned Action Approach (RAA [33]) was chosen as a con-
mindfulness, which they were able to apply to difficult situations in ceptual framework to guide data gathering and analysis in this
their lives. However, the study only reports outcomes that are clin- study because it synthesises research on thought processes that
ically relevant and therefore provides little insight about the user contribute to behaviour change. As described below, the RAA was
experience. Boggs et al. [26] conducted an interview-based qualita- adopted when it became clear that themes emerging from the user
tive study of an online intervention designed to deliver mindfulness interviews corresponded closely with the model.
based cognitive therapy (MBCT) to depressed patients. Participants
reported that they wanted greater flexibility of content to make it
easier to fit the intervention into their lives, more personal contact 3. Objectives
with clinicians and access to content through their mobile devices.
The intervention was not available on mobile devices and this was In order to develop effective mental wellbeing technologies and
a highly structured clinical intervention for patients with depres- improve current interventions for behaviour change, it is essential
sion, so findings may not generalise to a non-clinical population or to understand the relationship between users and the technology.
to a commercial application such as Headspace. This involves understanding complex psychological phenomena
Chittaro and Vianello [27] conducted a 4-week evaluation of an and behaviours occurring over time, to deliver design recommen-
interactive app developed by their own team (AEON) that focused dations for future development of such technologies. As noted
on “de-centering”. Their focus was on the degree to which par- above, although studies have demonstrated the effectiveness of
ticipants developed mindfulness practices; for participants who mindfulness and meditation delivered through technology, little
completed the trial (2.2% of those who started), there was also qualitative work has been done to gain insight into the user experi-
a qualitative questionnaire; responses to this were largely very ence of these interventions. Therefore, there is little understanding
positive, but there is no corresponding data for participants who of the enablers and barriers to effective use of mobile mindfulness
 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50 41

interventions, which might inform the design of future interven- Table 1

Persuasion Context addressed in this study.
tions.
In order to understand the user experience of one mindfulness The Intent To facilitate mindfulness meditation
app, Headspace, autoethnography and semi-structured interviews Persuader The key persuader is the developer and
were used. These reflexive and in-depth research methods are distributor of Headspace (Andy Puddicombe),
effective ways to understand user experience over an extended but ‘ownership’ of the change is with the
period of time. Data was initially analysed inductively (as described participant who is choosing to use the app (or
not).
below) and subsequently interpreted in terms of the Reasoned
Intended Outcome/Change Effective engagement with mindfulness
Action Approach [33]. practices in daily life, primarily through
ongoing use of the app. However this
engagement also includes developing
4. Preparatory study: autoethnography mindfulness practices to use outside app use.
Designer Bias The designer has developed the app based on
Autoethnography is a research method where the researcher his own beliefs about the efficacy of
becomes a participant in order to enter into an empathic rela- mindfulness practices based on those of
Kabat-Zinn.
tionship with users. For instance, O’Kane et al. [36] present an
The Event A daily ten to fifteen minute guided meditation
autoethnography of using a wrist blood pressure monitor over a delivered through an audio file, accessed
three month period to better understand the experience of using through the user timeline. Every three to four
a device to help manage a long-term health condition. Krizek [37] days there are short videos to supplement the
audio content.
argues that researchers must recognise that autoethnography pro-
Use Context Participants had a stated desire to engage with
vides a limited, subjective insight into the user experience. As a mindfulness practices.
consequence, the findings may not generalise to another person User Context Participants were all adults, in full-time work,
or context. The purpose of this autoethnography was to inform the fitting mindfulness practices into their daily
design of the main study and to provide early insights to be explored lives.
Technology Context The app is available and was used by
during subsequent interviews.
participants on both the IOS and Android
platform.
4.1. Method The Strategy Daily guided meditations designed to walk the
user through a meditation.
Message Daily meditations are supplemented with
The first author used the Headspace application on his personal teaching about the benefits, principles and
smartphone for a thirty-day period during which he carried out techniques of mindfulness meditation.
eighteen daily meditations. Throughout the autoethnography, data Route The content is primarily a direct message type.
However, information is provided and
was collected through the application ‘Day Journal’, which was
principles are explained that encourage the
accessed at any point throughout the day. A short description of user to reflect on their experience.
the experience of using Headspace was made, along with any other
reflections that occurred at the time. Thematic analysis [38] was
conducted on the diary entries to draw out major themes. This A third theme that emerged was the importance of context of
allowed the author to build up a compositional picture of his expe- use. Despite the fact that meditation is highly introspective, the
rience of use, including the sensations, emotions and interpretative diary entries indicate the importance of context of use, such as time
reflections of the experience of use. and location. The final theme that emerged from the autoethnog-
raphy focussed on the application design itself and the experience
4.2. Results of using a mobile device during mindfulness sessions.
These four themes informed the generation of an initial inter-
The first author had initially set out to use the app on a daily view schedule for the main study exit interviews, as described in
basis. However, as the study continued, patterns of use became more detail below.
irregular and were characterised by periods of consecutive daily use
followed by a number of days without use. A number of questions 5. Semi-structured interview study methods
emerged at this stage including, ‘Why does my use of the app follow
this particular pattern? What is the difference between days when A semi-structured interview study was carried out to explore
I have used it and days when I have not used it?’ the user experience of Headspace. Three-month download codes
These questions generated a series of observations about the were donated by Headspace to give participants full access to the
role played by certain factors in how the app was used. These factors content. Participants downloaded the application onto their smart-
included beliefs about effects of use, the role of the study itself, phones and signed up for their free three-month period of use.
the disruptive effects of busy routines and the quality of emotional They were asked to use the application in a self-directed way. Semi-
experiences before, during and after using the application. structured interviews were conducted prior to the trial period to
A key theme that emerged through the study period concerned explore expectations and intentions. An exit interview was con-
routine and priority. Diary entries indicate there was an ongo- ducted thirty to forty days after initial download to assess and
ing struggle about the extent to which the application should be explore the user experience, as described below. The relationship
prioritised over other uses of time. Another theme that emerged of this study to the themes of the special issue are summarised in
concerned the consequences of using the application. Every diary Tables 1 and 2.
entry contained an evaluative statement about the experience.
These evaluations usually comprised emotional, affective quali- 5.1. Participants
ties. These ranged from positive: “The world feels softer and more
pleasant and my mind is just still”, to more negative evaluations: Sixteen participants were recruited, five males and eleven
“Troubled meditation. Couldn’t really concentrate as I was worrying females. The age range of participants was 25–38 with a mean age of
about things”. Even when there was no clear consequence of using 32.5 years. All participants had been living in London, United King-
the application, there was still an evaluation, based on expectations. dom for at least a year. Participants were recruited by sending an
42 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50

Table 2 a minimum throughout the trial period to minimise influence on

Persuasive software features addressed in this study.
usage. Participants were asked to use the app in a self-directed way,
Primary task support Reduction, tunneling and rehearsal: These and were not asked to keep a diary, in order to maximise ecological
methods are used in the introduction and validity.
delivery of the guided meditations. Self
monitoring: This is encouraged through
continuous self reflection during the
5.4. Exit interviews
meditations.
Computer-human dialogue A timeline style interface for users to access Semi Structured exit interviews were conducted within thirty
support audio and video content. to forty days of initial download. Each lasted 15–30 min. Interviews
Credibility support Trustworthiness, expertise, authority: It is
began with a discussion of patterns of use before moving on to an
mentioned in the marketing of the application
and in the content itself that Andy Puddicombe exploration of the participant’s personal reflections about why they
is a former Buddhist monk. Third party had used the app in that way. Guiding questions in the later stages of
endorsements: During sign up, users read the interview were informed by the themes of the autoethnography
quotes from journalists and celebrities. Surface
as outlined above.
credibility: The quality of the design of the app
supports the notion of credibility.
After the first few interviews were completed, it became clear
Social support Social comparison, normative influence, social that there was a lot of variability in user experience between par-
facilitation, cooperation, competition: The app ticipants. Consequently, subsequent interviews became more open
has a built in buddy feature, allowing users to and participant-directed. Towards the end of each interview, fur-
team up with others. However, no participant
ther questions were raised that related to specific themes or the
reported using this feature.
concepts underpinning the Reasoned Action Approach that had not
yet been discussed. This had the benefit of allowing the participant
email about the study through social networks and via email. This to initially discuss their experience freely, before being engaged in
email explained that the purpose of the study was to explore the discussions that were of theoretical interest to the authors.
personal experiences of people using the Headspace app with the
intention to uncover insights that will improve the design of similar 5.5. Data analysis
technologies. Since we wished to focus on the experiences of busy
people who had some routine to their lives, the email invitation A key aim of the current study was to use theoretical reflection
stated inclusion criteria: participants should be between 25 and to deepen insights into the phenomena under observation. Furniss
40 years old and in full-time work. Potential participants had to be et al. [39] argue that ‘extant theory can be a source of creativity and
interested in using an app to help them meditate and practice mind- insight’. However, the danger in applying theory in the research
fulness with no prior experience of using Headspace. The reward process is that it can restrict data gathering and analysis because the
for taking part in the study would be a three-month subscription to researcher selectively attends to data that confirms their theories
the application. The study was approved by the departmental Ethics [29]. Consequently, the approach developed in this research project
Committee (ref. UCLIC/1213/015). Each participant gave informed was to initially engage with the empirical world using inductive
consent and understood that they could leave the study at any time. methods, followed by a process of theoretical reflection.
Data analysis was informed by Grounded Theory coding meth-
5.2. Materials ods [40]. However, there were a number of divergences from
Grounded Theory. Firstly, participants were recruited at the begin-
The Headspace application was downloaded from the Android ning of the study to reflect a population of urban professionals, so
Play or Apple iOS Application store onto each participant’s mobile theoretical sampling was not used in this study. A second diver-
smartphone. The application was accessed on these mobile devices. gence was that theoretical reflection was used in the later stages of
Participants also had to access the Headspace website via a desktop data analysis to enrich the understanding of the interviews.
computer or laptop to sign up. Data analysis proceeded through three stages. In stage one,
interviews were transcribed and analysed immediately using a
5.3. Procedure method of open coding, without theoretical reflection. This allowed
for the identification of concepts from the interview transcripts.
An initial semi structured interview was conducted with par- In stage two, themes and concepts that had been identified
ticipants before they began using the application. The interview during stage one were fed back into subsequent analysis to seek
sought to explore current mobile habits, prior experience of mind- verification and to explore their validity. This enabled the forma-
fulness and meditation, as well as expectations and intentions of tion of key categories. At this stage, a clear core concern of users
use. Questions were designed to avoid influencing perceptions of emerged from the interviews, namely to what extent and how to fit
the application. This interview lasted 10–15 min. app use into their lives. A further process of axial coding sought to
Immediately following the initial interview, an instruction email find the causal conditions, context and intervening conditions that
was sent detailing how to sign up and download the application. affect this core concern. Towards the end of the interview process,
The email explained that they would be contacted in approximately an affinity diagram of concepts was created to express the themes,
one month to arrange a follow-up interview to discuss their expe- concepts and relationships that were identified both from the data
rience of the app. All participants responded to the email within itself and from early theoretical reflections.
twenty-four hours. In stage three, the emerging phenomena were interpreted in
During the autoethnography, it had become clear that the study terms of the Reasoned Action Approach [33]. Each category was
itself could have an effect on how participants used and expe- considered in the context of the theory. The theory offered insights
rienced the application, so participants were asked to use the into the core concern of the users and illuminated many aspects
application in an entirely self-directed way. Participants were of their experiences. However, as discussed below, other themes
informed during recruitment that the purpose of the study was to (such as participants’ emotional experiences and their responses to
investigate real usage. Therefore, if they tried the app and decided the voice of the narrator) are not directly addressed by the theory.
they did not want to continue using it, this by itself would provide The following section presents the results of stages 1 and 2 of
valuable insights. Secondly, contact with participants was kept to this process. Care was taken to avoid shoehorning the results into
 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50 43

the theoretical model at this stage as it was important to let the The remaining six users reported that they used the app more reg-
data speak for itself before interpretation within the RAA model. ularly at one particular time of the day. Among these six were four
The findings are related to the Reasoned Action Approach in the of the most regular meditators. Two participants reported using
discussion section below. the app during the night to help them with sleep. Seven partici-
pants expressed a dislike for using the app either first thing in the
morning or last thing at night, due to tiredness. Conversely, four
6. Results
participants reported that these times were the best times to use
the app, as they would not be distracted.
Most participants began the study with the intention to use the
app on a regular basis. However, four participants came to the con-
clusion within three uses that they did not want to use the app. 6.2. Context of use
This was explored in their exit interviews and their reasons are
presented below. Location and environment were important themes for users.
In presenting the findings, we begin by summarising the Participants reported large differences in experience when using
reported patterns of use. In subsequent sections, we present the the app in different contexts. Most app usage occurred at home,
core theme of fitting use of the app into everyday life, followed in a place where they would not be distracted, somewhere safe,
by other factors that shaped the experience, namely: the role of quiet and comfortable; usually in their bedroom. Three participants
context (time and location) of use, routines and busy lifestyles, found quiet places in their workplace to use the app. However, most
the influence of other people, the changing roles of intentions and reported a struggle to find somewhere where they could use the
self-disappointment, the perceived consequences of using the app, app. Four participants reported using the app on public transport.
the changing role of expectations, the role played by emotions and This proved to be less successful than when using it in an environ-
mood, and the particular experiences of using the app on a mobile ment with less stimulation. However, there were beneficial effects,
smartphone device. especially on longer journeys.
Barriers to effective use included difficulty in finding time, nega-
tive expectations about mindfulness, negative experiences of using
the app, uncomfortable emotions, disappointment at self for not 6.3. Busy lives and routines
using the app more and a sense that the app ‘was not right for
the person’. Enabling factors included the development of routines, A central concern for users was how to fit their use of the app into
finding a time or place to conduct a meditation, positive expecta- their busy city lifestyles. Participants explained that their busy lives
tions about usage, simplicity of design of the app and the personal were a significant factor in limiting their app usage. One participant
way in which the content was delivered. perceived that it was not simply having a busy life that stopped her
All participants’ names have been changed to maintain confi- using it−it was the psychological effect that this had on her.
dentiality. “I think that when I feel that I have to do too many things, I feel
that I cannot fit the space to do it. If it’s just 10 minutes then I
6.1. Frequency and patterns of use feel that I can do this. But if it’s 20, then I’m thinking about it too
much.” Julie
Six participants reported using the app between one and three As well as being able to concentrate for the required time,
times. Four of these six reported that they had tried the application participants also expressed a belief that they could only use the
with an open and inquisitive mind but had decided after a couple app regularly if they had a routine. One participant had initially
of tries that the app would not suit them. Two other users tried expressed a strong intention to use the app. However, there was a
the app three times but reported that throughout the month, their two-week period without use in the middle of the 30-day period.
motivation had not reached high enough levels to facilitate routine When asked why this was, she answered,
use. One suggested this was mainly due to her busy lifestyle and
social issues. Another had only used the app on two occasions: both “Because of the fact that I haven’t had any kind of routine in my
times when he felt stressed and turned to the app to help him deal life.” Katie
with this. Many participants expressed a strong belief that once a routine
Six participants reported using the app between five and ten was started, it would generate a momentum which would motivate
times. Of these six participants, all reported that they had struggled them to use the app. One participant reflected on a run of days when
to fit the app into their busy lifestyles and had often turned to com- she used the app every day. Rather than taking personal responsi-
peting activities. All reported an intention to use the app more than bility for this pattern, she attributed it to the fact that she had ‘got
they had; five users expressed disappointment with themselves into a routine’.
that they had not developed more regular habits of use. Among
these users, there were four reports of using the app for several “I wasn’t more busy or less busy than any other time, I was just
consecutive days followed by a run of days with no use. more aware that I wanted to do it and I was more motivated
Two participants had used the app between eleven and twenty because I had got into a routine, I guess.” Mary
times. These users were naïve to meditation before the study, Once participants had established using Headspace as part of
but both expressed a high level of motivation at the outset. They their regular routine, there was always a risk that this routine
reported beneficial effects from using the app but both felt some would be disturbed and they would find themselves a few days
disappointment with themselves that they had not developed a later having to re-establish this routine. One participant described
daily routine. this phenomenon as ‘peaks and troughs’:
The final two participants had used the app between twenty and
thirty times. These users were both prior (although not highly expe- “I found it kind of went in peaks and troughs. I would get into
rienced) meditators who had included the app into their regular the routine of it for several nights a few nights and I would do it
meditation practices. every day and then I would get out of it and there would be
There was a wide variety of usage patterns throughout the day. a whole week where I wouldn’t have used it and I’d be like
Ten participants reported little uniformity in their usage patterns. ‘dammit I want to use it more’.” Mary
44 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50

Nine participants expressed disappointment with themselves However, two participants reported using the app more than
about not being able to use the app to the extent that they had they had initially intended. For instance, one participant began the
intended. However, there was also a strong theme of the impor- study with little intention to use the app with any kind of regu-
tance of self-directed use, with three participants stating that they larity. She had merely expressed a vague desire to train herself to
specifically chose not to establish a routine, as they wanted to keep be more mindful. During this initial interview she also expressed
meditation as a leisure activity, rather than a work-like activity a low self-belief in her capacity to meditate. However, during the
which they had to do. exit interview, she expressed surprise at the amount of meditation
that she had done. She also expressed surprise that she had actually
“I guess at the moment, I need to feel that I want to do it, rather
enjoyed using the app.
than that I have to do it, cause then I would probably give up. I
think there is so much in my life that I have to do that I wouldn’t “I quite enjoyed it, I really felt it was kind of like, it felt good to
want to make it scheduled, that I HAVE to do it, to fit it into all be doing it, instead of feeling weird, of feeling it was a waste of
my other stuff that I HAVE to do.” Rosie time. I thought this is quite nice.” Rosie
A number of users spoke about how the app gave them ‘permis- Finally she expressed a strong increase in self-efficacy with
sion’ to take some time out of the day, to take a break from their regards to meditating. She attributed this change in self-efficacy
routines, to switch off from constantly engaging with social media, to the way in which the app had taught her practically how to
emails, entertainment, socialising or work. meditate.
“I felt like I was being allowed to just have some time out and
just think of nothing. And that was good.” Rosie
Although the study had been designed to minimise the effects 6.5. Social context
of the study itself on the user experience, five users reported being
influenced by the fact that they were taking part in a study. In all As well as individual factors, participants reported that other
cases, this effect was reported as being beneficial. Many expressed people influenced their use of the app. One reported being influ-
gratitude that they had been ‘allowed’ or encouraged to meditate enced in both positive and negative ways by the opinions of friends.
by partaking in the study. One even jokingly asked that she be The opinions were of a conflicting nature: one narrative encour-
threatened to meditate more often. aged a cynical attitude towards mindfulness; the other encouraged
embracing mindfulness. She had not yet made up her mind about
“I intended to use it every day. Because I wanted the fact that I these differing attitudes. This caused an inner tension or confu-
was doing the study to push me to do it every day.” Georgina sion:
“Can you threaten me?” Samantha “Like I said, my friends do it and I have friends who do yoga and
This led participants to reflect on their own intentions and how stuff and they manage to fit it in every day and I often struggle
they were living up to their own expectations. to find the time.” Georgina
“Yes I think half of my friends were kind of sceptical
about it. Which kind of played into my scepticism. . .”
Georgina
6.4. Intentions and disappointment with self
Another participant reported that she had been unsure what to
Many participants reported that they had not used the applica- make of the app but finally made up her mind after demonstrating
tion as much as they had wanted to. This often resulted in feelings the app to two colleagues who subsequently responded in the same
of disappointment with self or guilt. One participant expressed this way as her, reinforcing her opinion:
as a constant battle between thinking she needed to use the app
“I thought maybe it’s just me. But I played it to two of the girls
and being distracted from using it. She began to chastise herself
that I sit with and they said the exact same thing.” Amanda
and suggest that she lacked self-discipline. She then pointed out
that if there is a consequence to not using the app, a punishment, One participant explained that she had had a single conversa-
she is more likely to use it. tion about the fact that she was using the app with a friend. Before
the conversation, she had only experienced frustration and bewil-
“I lack discipline. I let my routine slide a bit· · ·If there is a neg-
derment with using the app. The friend’s response fundamentally
ative consequence of not doing something if I don’t do it, then
affected the way that she experienced and thought about the app
I’m more likely to do it.” Samantha
from that moment onwards, informing the narrative by which she
This participant had not expressed a strong intention to use it rationalized her continued use.
regularly during her initial interview, so her self-disappointment
“I only spoke briefly about meditation with a friend of mine who
had emerged during the period of use. Other users also showed a
does it. And he told me. . . I told him that I wasn’t feeling any
change of intention during their times of use. For instance, during
difference. . . and he told me that it’s not an immediate thing
the initial interviews, one participant expressed an intention to use
and I need to train my brain to do it, but after I’ve done it for
the app every day. However, this participant only used the app on
a while I’ll find that I can even meditate when I’m doing the
five occasions throughout the trial. During the exit interview she
dishes.” Sofia
explained that she would like to continue using the app but with
very different intentions. There had clearly been a reduction in her A number of users reported that other people had enabled them
belief that she was capable of meditating regularly: to use the app. One of the most prolific users attributed almost all
her motivation to the fact that her partner also meditates every
“I don’t think I would expect to do it every day now. I think I
day.
would be a bit more realistic. But I would want to try to do it.
If I could try and do it· · ·I mean it’s only ten minutes, isn’t it? If “My partner meditates. When I’m with him I mediate with him.
I could do it 4 or 5 times a week, I would be happy with that.” He meditates twice a day. So doing it once a day is slacking. So
Sally I have an environmental factor that affects my behaviour.” Jess
 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50 45

6.6. Perceived consequences of using the app which the app helped them to deal with their emotions, while
three stated that emotions had little impact on their continued use.
There was a wide variety of experiences, ranging from dis- These sections of the interviews often revealed that the partici-
appointment to visceral sensations of calm. Probably the most pant had internalized some of the teaching used in the app about
commonly mentioned effect was that participants emerged from mindfulness.
using the app feeling calm and relaxed:
“Yes I was using it as a tool to step away from the emotions and
“Each time I did it, I felt better afterwards. . .at some times, real not get too involved in it. It calmed me down and made me think
peace, at other times, slightly less agitated.” Robert and make me look at these emotions non-judgmentally.” Mary
A number of participants simply enjoyed the fact that using the Eight participants reported that difficult emotional feelings
app enabled them to access a different mind-set to normal life. They were a barrier to use, particularly when people were highly stressed
saw a value in simply being ‘in the moment’. Accessing this alterna- or agitated. At these times, users reported frustration that they were
tive mind-set was in itself a reward, particularly as it allowed them unable to control their thoughts. Users frequently reported that the
to view their life from a different perspective. app was more difficult to use when stressed.
“You know it’s part of my job being a PA, I’m always looking at “If you’re feeling quite stressed or there’s tension in your body,
a calendar for December. I’m always ‘ahead’, ‘ahead’, ‘ahead’, all probably using the app is going to be harder than if you’re
the time and so it was nice to feel, ‘I’m right in the moment here’. already quite relaxed.” Craig
When it says ‘just listen to the sounds’, I was thinking, ‘I never
Three participants responded to these negative emotions by
just sit and listen to the seagulls above, or cars driving past.”’
attempting to maintain a regular disciplined routine, regardless
Rosie
of how they were feeling. This was discussed by the two most
Other participants reported that the positive consequence of experienced meditators, both of whom spoke of the importance
using the app was that it was helping them learn about medita- of developing a discipline that transcended their moods.
tion and develop mindfulness practices during their day-to-day life, Participants who reported the most intense emotional experi-
especially the practice of ‘returning to the breath’. ences during the previous month explained that when they were
agitated, mindfulness was frustrating and did little for them. These
“It actually did what I wanted it to do, which was to take me
users advocated other behaviours during these times, which would
back to the basics of meditation.” Jess
enable them to escape or express their emotional turbulence, rather
However, five participants reported that they experienced lit- than become more aware of it. Most of these behaviours had an
tle or no positive effects from use. Two of those who were unable active quality to them, such as exercising or socialising. They were
to report clear positive benefits of using the app stopped using it convinced that while there was a value to mindfulness practices in
altogether. The other participants rationalized their continued use usual situations, at times of heightened emotion this was not what
by stating that they believed that there would be long term effects was needed.
if they continued to use the app. Two of those stated that they felt
“I think meditation really makes you notice what’s going on for
it was wrong to look for a reward, as mindfulness was not about
you and that’s actually quite hard and I want to distract myself
seeking rewards.
with TV or the internet cos that’s much easier than sitting there
for 10 minutes and noticing that I am maybe quite anxious or
6.7. Expectations
stressed.” Sally
Each of the above themes reveals the importance of user expec-
tations, and the interplay between expectations and behaviour.
These expectations were initially influenced by a variety of fac- 6.9. The context of the specific app design
tors revealed in the initial interviews. These included beliefs and
worldviews related to meditation, previous personal experiences When asked about design features, all participants commented
of mindfulness, current social relationships such as partner and that the app was well designed and that it had helped them to
housemate meditation habits and attitudes, and popular cultural meditate. For more naïve meditators, the guided meditations had
and media narratives of mindfulness. There was also an ongoing been particularly helpful to focus their meditations and learn about
influence of expectations during the study, which was shown by meditation.
the way use was affected by reflections about the app and about “I do think that if I did not have the app I would not be able to
mindfulness during the period of the study. All users approached do it at all. Because I would just completely get lost in my own
the app at least once during the study with an attitude that mind- thoughts and there would be nothing to bring me back.” Sofia
fulness could offer a quick fix to a problem they were experiencing.
When they did not experience the immediate results that they were However, more experienced meditators responded more criti-
expecting, there was a loss of motivation. Those with more realis- cally. Two expressed frustration that they were forced to start at a
tic expectations about mindfulness and a greater knowledge of the beginner’s level. Three users questioned whether the app was suit-
benefits and challenges of mindfulness practices were less likely to able for them, as it did not follow the particular style of meditation
report this loss of motivation. practice that they had used previously.
“I do kind of guided visualization, rather than a kind of ‘empty
6.8. Emotions your mind’ meditation style. I approach it in a totally different
kind of way. So I found the whole business of. . . listening to your
Emotional experiences were reported to be an important aspect breath and counting, counting to ten in circles, I found really
of the user experience and were reported by some to be a barrier frustrating.” Jon
to use. Twelve of the sixteen participants spoke about the role of
emotions in their experience of using the app. Seven users specif- One participant thought deeper about designing technologies
ically reported times when they turned to the app for help with to facilitate meditation, suggesting that she might benefit from a
difficult emotional situations. Nine users reflected on the way in tactile quality in meditation technology.
46 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50

“You’ve got the different learning styles for people. I think if you As summarised above, patterns of use varied across participants;
can incorporate that sort of thing. . . Not everyone is auditory, the RAA predicts that much of this variance can be explained by dif-
not everyone is visual, not everyone is kinaesthetic· · ·I’m tactile. ferent levels of intention, both at the beginning and throughout the
I like to touch and smell and taste and all that. And having an study. Three participants described wanting to use the application
online meditation seems a bit. . . for me, it wasn’t a tactile type occasionally, in a self-directed way, as a leisure activity, rather than
of thing.” Samantha as part of a routine. All other users attempted to establish a daily, or
at least regular, routine of using the app. Although a small number
Five participants made reference to small design features of the
of users came close to achieving this, none used the app every day.
app that had presented problems. These were relatively insignifi-
The majority of users who began the study with strong intentions
cant problems that users were able to work around. However, they
struggled to establish a regular pattern of use. However, other users
demonstrate that there are particular problems that need to be
began the study with low intentions to use the app but went on to
overcome when delivering meditation on a mobile device.
establish regular routines. Stated intention to act only provides a
“When I started the session, there was a clock saying how much limited explanation of the variance in behaviour. One characteristic
time there was left and it was kind of distracting me from the of the RAA is the low importance it gives to emotions and non-
meditation because I couldn’t help but look at how much time cognitive phenomena, and the high importance it gives to rational
there was left. And also the colour is bright and orange. I couldn’t decisions. As discussed below in the section on emotions, users do
ignore it.” Jo not always behave according to rational utility. In the following
sections, we discuss the role that attitudes, self-efficacy and social
“He leaves long gaps and sometimes I’m worried that I have run
norms played in intention to act, before discussing environmental
out of battery on my phone.” Jess
factors, expectations, emotions and design features that affected
Headspace delivers its content through spoken guided medi- behaviour and experience.
tations and video tutorials that are voiced by the founder, Andy
Puddicombe. During the interviews, it became clear that users 7.2. Attitudes and beliefs
were not simply responding to abstract information or even to
an impersonal detached voice. Rather, users were, to some extent, The RAA suggests that attitudes, beliefs and expectations
experiencing contact with a person. Users differed widely in their predict intentions. The study revealed an interplay between
response to this. Thirteen users reported that they liked the voice users’ expectations and their subsequent use of the app. Four
guidance; they spoke positively about Andy, with five specifically users began the study with the belief that meditation would
referring to him as a guru character. Other responses were more have little immediate effect and should be used for a longer
ambiguous. One participant reported that he did not like being told period of time before beneficial effects were seen. All other
by Andy that he had to meditate the next day and this put him users expected at least some immediate sensations of relaxation
off using the app. This particular participant had reported that he and mental calm. For most users, apart from the most expe-
had been forced to meditate by his mother when he was a child. rienced meditators, attitudes changed as a result of using the
Three participants reported strong negative responses, suggesting app.
that their dislike of the voice was their main reason for stopping Some users began the study with high expectations and were
using the app. E.g.: disappointed by their experiences, leading to a trailing off of use.
Many of these users began the study with a belief that the app could
“I didn’t feel comfortable. And I thought maybe it’s just me. I was
offer a quick fix for stress and other mental phenomena. However,
talking to Kelly about it and she said ‘I really like the voice’. And
seven users reported that practising meditation had been difficult
so I realized it was like a personal preference. Maybe I’ve had
or frustrating when stressed or emotional. These negative experi-
nightmares about someone with a similar voice or something.
ences changed their attitude about mindfulness, so that four users
I don’t know. . . I think it would’ve been really different if there
came to the view that mindfulness was not what they needed on
was a different voice.” Elisa
stressful occasions. Interviews revealed that while the app content
“I’m going to be honest and tell you that I hate it. I just can’t was well designed for naïve and less experienced meditators, those
bear Andy’s voice. Perhaps his tone has caused some sort of with more experience often found it restrictive and inflexible. Two
abreaction in me.” Sue stopped using it because they felt it did not fit with their ideas
about meditation. However, eleven users reported positive con-
sequences to use. These included feelings of relaxation and calm,
7. Discussion learning to meditate, and learning to be more mindful in life. These
positive consequences functioned as rewards, increasing positive
In this section, we re-visit the findings, relating them to the judgments of app use. Three users reported believing that, although
Reasoned Action Approach (RAA) to behaviour change. The study they had not received any strong benefits from use, there would
illuminates ways in which conditions affect users of the applica- be rewards if they continued using the app for longer periods.
tion, and how the design of the technology facilitates and prohibits This justified their continued use. More experienced meditators,
mindfulness meditation practices; we therefore also discuss impli- or those who had read about mindfulness previously and therefore
cations for design, and also limitations of our study. had more realistic and in-depth understanding of the purpose and
outcomes of mindfulness practices, were better able to establish
the patterns of use that they had originally intended.
7.1. Interpreting the findings in terms of RAA
7.3. Self-efficacy
As a descriptive model of behaviour, RAA provides empiri-
cally verified constructs with which behaviour can be analysed. The RAA suggests that self-efficacy predicts intention. There is
The theory directs the researcher to reflect upon the attitudes a clear relationship between the struggles with routines and emo-
of participants about the behaviour under investigation and the tions reported in the interviews and expressed self-efficacy. Some
role of social norms and self-efficacy in shaping the patterns of users expressed surprise and frustration that they were not able
use. to establish routine patterns of use: their self-efficacy had been
 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50 47

challenged between the start and finish of the study. Others therefore curious to reflect that many users in this study reported
reported increased self-efficacy: they reported that using the app a sense of futility in their attempts to regulate strong emotions.
had taught them that they were able to meditate. This is likely As discussed above, those who approached the application with
to be the reason why users reported that routines generated a an expectation of experiencing a ‘quick fix’, may have been disap-
momentum. If the application is used for a number of days in a pointed at not seeing immediate results. Many people use mobile
row, the self-efficacy of the user increases, as the evidence con- phones to get a ‘quick fix’; checking their messages and playing
vinces them that they can set aside the time to use the application. games while they are waiting or travelling. This ‘fiddling’ behaviour
This corresponds to the predictions of the RAA, which states that is characteristic of anxiety regulation strategies. Mindfulness medi-
as self-efficacy increases, behaviour becomes more likely. tation does not offer this quick fix, or anxiety avoidance. Kabat-Zinn
[11] suggests that mindfulness is a lengthy learning process of
7.4. Social norms growth: mindfulness is something that is cultivated over time.
The process can only happen successfully by taking the time to
The RAA suggests that social norms predict intention to act. experience and come to terms with difficult emotions, rather than
The interviews demonstrate a relationship between social norms avoiding them.
and attitudes towards the behaviour, intentions to use, and actual Finally, the interviews revealed a strong consequence of hav-
behaviour. This is particularly so when the social influence comes ing a person at the forefront of content delivery. Seven participants
from those closest to the participant. However, social influence was expressed a growing connection to the person instructing the medi-
commented upon by a minority of participants, so it may not play tations. This personal connection was expressed especially by those
a strong role in mindfulness practice, which is a personal activity participants who had used the application more regularly. How-
for which social judgments are of little relevance to many users. ever, conversely, three stopped using the app because they did not
Indeed, a number of users denied that other people had had any feel comfortable with the voice. The force with which this reaction
influence on their use whatsoever. was expressed by some users suggests a strong emotional compo-
nent that is not well accounted for within RAA; this is an area for
7.5. Environment and ability to act further investigation.

The RAA suggests that intention to act will only be translated 7.7. Design recommendations for mindfulness technologies
into action if the user is able to perform the action and envi-
ronmental conditions are suitable for performance. Environmental The Reasoned Action Approach provides a series of higher-
conditions included busy routines and distracting events but also level guidelines for the development of technologies to facilitate
included locations and environments. If the user had a very busy mindfulness practices. To effect behaviour change, the technology
day, or lacked a formal routine, this would reduce the likelihood must facilitate high levels of intention. This can be achieved by
of using the app. If the user could not find somewhere quiet, with- managing beliefs about the outcome of mindfulness behaviours,
out distraction, this decreased likelihood of using the app. Users managing and utilising the role of social norms, and increasing
attempted to manage these conditions by finding quiet spaces and self-efficacy. The technology must also teach skills and facilitate
quiet periods during the day when they could take time out to use environmental factors that increase the likelihood of intention
the application. becoming action.
The design features of the app could be conceived as ‘environ- As reported above, expectation and beliefs about the conse-
mental factors’ that inhibit or facilitate the participant to carry out quences of using the app were both an enabling factor and a barrier
meditations. There were a number of minor issues with using the to use. These conclusions are broadly in agreement with the PSD
app on a mobile device that could potentially distract users dur- model of Oinas-Kukkonen and Harjumaa [4], which states that
ing the exercises. These included receiving texts and emails during adoption of a persuasive technology is dependent on high levels of
meditations and other factors discussed above. perceived credibility and perceived persuasiveness. In particular,
those with an accurate understanding of the benefits and practices
7.6. Emotions of mindfulness were more likely to succeed in using the applica-
tion as they intended over the period of the study. It is therefore
Within the RAA model, emotion is a small factor that plays important that designers of mindfulness applications seek to man-
a role through the way it influences behavioural beliefs, norma- age expectations and beliefs. Early content should provide a clear
tive beliefs and control beliefs. However, the interviews reveal that explanation about the benefits, consequences and proper use of
emotion plays a substantial role in use patterns. Eight participants mindfulness meditations. Particular attention should be directed
reported that emotions and mood were at times a powerful factor towards ensuring that people do not approach the app simply as
that completely stopped their intention to act from becoming an a ‘quick fix’ or distraction from certain mental states. The focus
action. Seven participants reported that they used the app to help of this content should be directed towards enabling the user to
them deal with difficult emotions. These participants explained contextualise and gain perspective on difficult mental states and
that using the app would enable them to achieve a greater objec- gain an understanding of the potential struggles and benefits of
tivity about the cause of their emotional state. However, it was also use.
reported that once agitation reached a certain threshold, medita- In this study, many users reported that they found it difficult
tion would become much more difficult. This could reduce their to fit use of the app into their busy schedules. A key problem to
desire to use the app on future occasions. Four users also reported overcome in adoption of this technology is that successful medi-
that emotional states acted as a distraction from their attempt to tation requires setting aside time during the day to practice. The
establish a healthy balanced routine. Those who reported the high- PSD model states that unobtrusiveness is an important factor in
est states of agitation reported that they turned to a more physical whether an intention to use a persuasive technology becomes
activity on these occasions, such as exercise or drinking alcohol. actual use (the behaviour-intention gap). Unobtrusiveness is a con-
One of the key aims of mindfulness meditation is to improve struct that describes how a persuasive technology fits with the
emotion regulation [41]. Studies such as Goldin & Gross [42] report user’s environment and task flow [4]. However, the construct of
significant reductions in anxiety and depression during mindful- unobtrusiveness has some limitations for a mindfulness applica-
ness sessions, using both objective and subjective measures. It is tion because meditation is meant to be a disruption from the tasks
48 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50

of everyday life. Those users who tried to fit the app into the task sphere for women in a secure psychiatric unit. This sphere changed
flows of their life, such as using the app while travelling on the train, colour in rhythm with the heart of the person who was holding the
reported less satisfying effects. ball. The sphere was used as a meditation tool to help the user
Fitting use into a busy schedule is a skill that can be learned. become more aware of their body.
Therefore, a mindfulness application should teach new meditators A number of e-health technologies deliver mindfulness with-
how they can fit this practice into their busy lives. These messages out any human presence in the technology, for example through
can be delivered within the content of the application. For instance written instructions followed by a timed meditation, or through
an introduction session could be delivered to users that teaches a mood reflection application that asks users to enter their mood.
methods of setting aside time, and of time management, such as However, Fogg [45] argues that technologies that use human pres-
alarm setting and implementation intention strategies. Persuasive ence, or play a role as a social actor, are able to deliver a richer,
technologies can use methods such as asking users to sign a contract more engaging experience. This is clearly the case for Headspace, as
of use, or using social networking features where goals, successes reported by users. However, during the study, three users stopped
and failures are shared with other users. A number of users who using the app because they did not feel comfortable with the voice,
wanted to establish regular routines reported that they would have even though they liked almost everything else about the app. One
preferred a more prominent scheduling, reminder or nudge facil- possible way of designing around this problem could be to enable
ity. A benefit of a mobile device is that it offers the capacity to users to choose from a selection of guides, both male and female,
link the application with diaries, alarm clocks and other features to take them through the program, so that they can find someone
that can assist the user to establish routines. Context aware tech- they feel comfortable with.
nologies are fast improving in their capacity to sense our affective These design recommendations overlap with, but are not iden-
states, locations and other contexts. We can imagine a future where tical to, those of Ahtinen et al. [28] Their main recommendations
an application reminds us to meditate when we reach a certain are to:
affective state, when we are in a particular location, or during a
period of time when the device sensors show decreased move- • find a proper place and time for challenging content; this finding
ment. is strongly supported by our study.
Eight users reported that when they experienced higher levels of • guide without restricting choice; this finding is supported by our
emotional arousal, they found it more difficult to meditate or even study, particularly in the responses of the more experienced med-
reported that it became a negative experience. These users reported itators.
that it was easier to do something more active, as sitting still and • provide (active) exercises for everyday life; this is consistent with
focusing on themselves simply heightened their awareness of their the finding that many participants found the lack of physical
anxiety. These experiences can be managed by guiding users to activity promoted by Headspace difficult to deal with if they were
avoid ‘quick fix’ thinking about mindfulness, and to help them at all agitated.
understand how mindfulness should be used as an emotion reg- • focus on intrinsic rather than extrinsic rewards; this did not
ulation strategy, as discussed above. In one of the original courses emerge as a strong theme in our study.
of mindfulness developed by Kabat-Zinn [11], yoga was incorpo- • provide a flexible tool for self-reflection; again, this did not
rated; yoga enables a bodily, active articulation or expression of our emerge as strong theme in our study, highlighting differences in
internal states which often functions to diffuse internal emotional findings caused by differences in study design and in the design
energies. Some users who reported needing something more active of the mindfulness apps that were the focus of study.
to help them deal with difficult emotional states might have found
a solution in yoga. How yoga practices could be delivered through
7.8. Limitations
a mobile app, alongside the current meditations, is an interesting
design problem.
As becomes evident when comparing our findings with those
A number of users reported that the study itself was an enabling
of Ahtinen et al. [28], the nature of Headspace (which focuses
factor. These users expressed gratitude that they were ‘being given
on meditation rather than physical activities) shaped participants’
permission’ to meditate. Users are enabled by an external figure,
experiences significantly. The study design also influenced what
especially one to whom they may feel a sense of accountability.
data was gathered: data was gathered from participants at the
This might be a clinician in clinical services; in consumer applica-
beginning and end of the study. This had the benefit of not dis-
tions, this may be a friend. Headspace has a ‘buddy system’ that
rupting users during the study, ensuring higher levels of ecological
can be used on the application. However, no participant used this
validity. However, it also meant that experiences during the study
feature and it was only mentioned by two participants during the
were not sampled. Further, despite attempting to control for the
exit interviews, so there may be room for improvements with this
effects of the study, it is clear from the interviews that some
feature, to bring it to the forefront and to alert users to its potential
users were influenced by the study. In addition, the study only
value.
explores user experience of people beginning to use the app; the
Hawkins et al. [43] found that technologies with high levels of
user experience is likely to change once the user develops their
interactivity were more successful at engaging users and changing
own practice and moves into the maintenance stage. Finally, all
behaviour. However, the user of Headspace has a passive role. For
were busy, employed city-dwellers, and none had reported mental
the first thirty days users are not able to choose which meditation
health problems.
they listen to and their interaction consists of pressing play and lis-
tening. This has the benefit of creating a simple and seamless user
experience. However, this linear pathway can be off-putting or frus- 8. Conclusion
trating for those who are more experienced meditators. This ties in
closely with the theme of the importance of self-directed learning In this study, the Reasoned Action Approach was found to be
and flexibility of use reported in other studies of e-mental health a valuable theoretical framework for analysing and understanding
(e.g., [43]). Making the application more interactive would result behaviour and experience.
in a more engaging experience. Thieme et al. [44] offer one possi- To be truly effective, behaviour change technologies need
ble solution of increasing interactivity through a more embodied to work with people’s lifestyles, values and expectations. This
style of technology aided meditation: they developed a meditation research revealed that the primary concern of users of a mobile
 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50 49

mindfulness application was fitting use of the application into their [12] W. Kuyken, S. Byford, R. Taylor, E. Watkins, E. Holden, K. White, B. Barrett, R.
lives. User responses to this concern were influenced by a number of Byng, A. Evans, E. Mullan, J. Teasdale, Mindfulness-based cognitive therapy to
prevent relapse in recurrent depression, J. Consult. Clin. Psychol. 76 (6) (2008)
interrelated factors. These were the demands of their routines, the 966–978.
perceived consequences of using the application, beliefs about their [13] K.W. Chen, C.C. Berger, E. Manheimer, D. Forde, J. Magidson, L. Dachman, C.W.
ability to fit the app into their lives, the influence of other people Lejuez, Meditative therapies for reducing anxiety: a systematic review and
meta-analysis of randomized controlled trials, Depress. Anxiety 29 (7) (2012)
and social norms, and quality of their mood and emotional states. 545–562.
The study also revealed factors that stop people from using the [14] L. Davis, S. Kurzban, Mindfulness Based treatment for people with severe
app from the beginning. These include issues with the design and mental illness: a literature review, Am. J. Psychiatric Rehabil. 15 (2) (2012)
202–232.
delivery of the program, especially in terms of not being comfort-
[15] H.B. Beckman, M. Wendland, C. Mooney, M.S. Krasner, T.E. Quill, A.L.
able with program content or methods of delivery. Future research Suchman, R.M. Epstein, The impact of a program in mindful communication
and design should explore the ways in which technologies can a) on primary care physicians, Acad. Med. 87 (6) (2012) 815–819.
[16] S.K. Bostock, A. Steptoe, Can finding headspace reduce work stress?
encourage people to reflect on how to make time and space for
Randomised controlled workplace trial of mindfulness app, Psychosom. Med.
mindfulness; b) manage expectations about behavioural outcomes; 75 (3) (2013) 36–A37.
c) increase the self-efficacy of users; and d) improve environmen- [17] JWT, 10 Trends that will shape our world in 2014. (2014) (Retrieved from
tal conditions to make digital health interventions more usable and http://www.jwt.com/blog/consumer insights/10-trends-that-will-shape-our-
world-in-2014-and-beyond/).
engaging. [18] I. Plaza, M.M.P. Demarzo, P. Herrera-Mercadal, J. García-Campayo,
Mindfulness-Based mobile applications: literature review and analysis of
current features, JMIR mhealth and uhealth 1 (2) (2013) 24.
Conflict of interest
[19] D. Derbyshire, Should we be mindful of mindfulness? The Observer. (2014)
(Retrieved from http://www.theguardian.com/society/2014/feb/23/should-
No conflicts of interest to declare. we-be-mindful-of-mindfulness-nhs-depression).
[20] T.M. Glück, A. Maercker, A randomized controlled pilot study of a Brief
web-based mindfulness training, BMC psychiatry 11 (1) (2011) 175.
Contributions [21] K. Cavanagh, C. Strauss, F. Cicconi, N. Griffiths, A. Wyper, F. Jones, A
randomised controlled trial of a Brief online mindfulness-based intervention,
This paper reports on a study completed by the first author as Behav. Res. Ther. 51 (9) (2013) 573–578.
[22] J. Boettcher, V. Åström, D. Påhlsson, O. Schenström, G. Andersson, P. Carlbring,
part of his Masters research. The second author is the corresponding Internet-based mindfulness treatment for anxiety disorders: a randomized
author. controlled trial, Behav. Ther. 45 (2) (2014) 241–253.
The contributions of the authors are as follows: [23] M. Mani, D.J. Kavanagh, L. Hides, S.R. Stoyanov, Review and evaluation of
mindfulness-based iPhone apps, JMIR mHealth and uHealth 3 (3) (2015) e82.
[24] A. Howells, I. Ivtzan, F.J. Eiroa-Orosa, Putting the ‘app’ in happiness: a
• James Laurie: conducted the study, including all data collection randomised controlled trial of a smartphone-based mindfulness intervention
and primary data analysis; drafted first version of this paper. to enhance wellbeing, J. Happiness Studies (2014) 1–23.
[25] M.E. Morris, Q. Kathawala, T.K. Leen, E.E. Gorenstein, F. Guilak, M. Labhard, W.
• Prof Ann Blandford: primary supervisor for Laurie’s MSc; oversaw
Deleeuw, Mobile therapy: case study evaluations of a cell phone application
the design of the study; gave guidance on data collection and for emotional self-awareness, J. Med. Internet Res. 12 (2) (2010).
analysis; edited the paper (through several iterative cycles). [26] J.M. Boggs, A. Beck, J.N. Felder, S. Dimidjian, C.A. Metcalf, Z.V. Segal,
Web-Based intervention in mindfulness meditation for reducing residual
depressive symptoms and relapse prophylaxis: a qualitative study, J. Med.
Both authors have reviewed the submitted version and Internet Res. 16 (3) (2014).
approved the content. [27] L. Chittaro, A. Vianello, Evaluation of a mobile mindfulness app distributed
through on-line stores: a 4-week study, Int. J. Hum. Comput. Stud. 86 (2016)
63–80.
Acknowledgments [28] A. Ahtinen, E. Mattila, P. Välkkynen, K. Kaipainen, T. Vanhala, M. Ermes, R.
Lappalainen, Mobile mental wellness training for stress management:
feasibility and design implications based on a one-month field study, JMIR
We are grateful to Dr David Cox and Headspace for their gen- mHealth and uHealth 1 (2) (2013).
erosity in providing access to Headspace for participants, and to all [29] E.B. Hekler, P. Klasnja, J.E. Froehlich, M.P. Buman, Mind the theoretical gap:
participants for their time, energy and thoughtful input. interpreting, using, and developing behavioral theory in HCI research, In
Proceedings of the SIGCHI Conference on Human Factors in Computing
Systems (2013) 3307–3316 (ACM.).
References [30] A. Grimes, R. Grinter, Designing persuasion: health technology for low-income
African American communities, Persuasive Technol. 4744 (2007) 24–35.
[1] R.S. Weinstein, A.M. Lopez, B.A. Joseph, K.A. Erps, M. Holcomb, G.P. Barker, E.A. [31] J.O. Prochaska, W.F. Velicer, The transtheoretical model of health behavior
Krupinski, Telemedicine, telehealth, and mobile health applications that change, Am. J. Health Promot. 12 (1) (1997) 38–48.
work: opportunities and barriers, Am. J. Med. 127 (3) (2014) 183–187. [32] A. Hsu, A. Blandford, Designing for psychological change: individuals’ reward
[2] S.R. Steinhubl, E.D. Muse, E.J. Topol, Can mobile health technologies transform and cost valuations in weight management, J. Med. Internet Res. (2014).
health care? JAMA 310 (22) (2013) 2395–2396. [33] M. Fishbein, I. Ajzen, Predicting and Changing Behavior: The Reasoned Action
[3] S.M. Kelders, R.N. Kok, H.C. Ossebaard, J.E. Van Gemert-Pijnen, Persuasive Approach, Taylor & Francis, 2011.
system design does matter: a systematic review of adherence to web-based [34] M. Fishbein, I. Ajzen, Belief, Attitude, Intention, and Behavior: An Introduction
interventions, J. Med. Internet Res. 14 (6) (2012). to Theory and Research, Addison-Wesley, Reading, MA, 1975.
[4] H. Oinas-Kukkonen, M. Harjumaa, Persuasive systems design: key issues, [35] I. Ajzen, The theory of planned behavior, Organ. Behav. Hum. Decis. Process.
process model, and system features, Commun. Assoc. Inform. Syst. 24 (1) 50 (2) (1991) 179–211.
(2009) 28. [36] A.A. O’Kane, Y. Rogers, A.E. Blandford, Gaining empathy for non-routine
[5] J.R. Martin, Mindfulness: a proposed common factor, J. Psychothe. Integr. 7 (4) mobile device use through autoethnography, In Proceedings of the 32nd
(1997) 291–312. Annual Acm Conference on Human Factors in Computing Systems (2014)
[6] E.J. Masicampo, R.F. Baumeister, Relating mindfulness and self-regulatory 987–990 (ACM).
processes, Psychol. Inquiry 18 (4) (2007) 255–258. [37] R. Krizek, Ethnography as the excavation of personal narrative, in: R.P. Clair
[7] M.R. Leary, E.B. Tate, The multi-faceted nature of mindfulness, Psychol. (Ed.), Expressions of Ethnography: Novel Approaches to Qualitative Methods,
Inquiry 18 (4) (2007) 251–255. SUNY Press, New York, 2003, pp. 141–152.
[8] S. Young, The science of enlightenment. Boulder, CO : Sounds True. (1997). [38] V. Braun, V. Clarke, Using thematic analysis in psychology, Qual. Res. Psychol.
[9] P. Grossman, L. Niemann, S. Schmidt, H. Walach, Mindfulness-based stress 3 (2) (2006) 77–101.
reduction and health benefits: a meta-analysis, J. Psychosom. Res. 57 (2004) [39] D. Furniss, A. Blandford, P. Curzon, Confessions from a grounded theory PhD:
35–43. experiences and lessons learnt, In Proceedings of the SIGCHI Conference on
[10] P.R. Fulton, Mindfulness as clinical training, in: C.K. Germer, R.D. Siegel, P.R. Human Factors in Computing Systems (2011) 113–122 (ACM).
Fulton (Eds.), Mindfulness and Psychotherapy, Guilford Press, New York, [40] A. Strauss, J.M. Corbin, Basics of Qualitative Research: Grounded Theory
2005, pp. 55–72. Procedures and Techniques, Sage Publications Inc, 1990.
[11] J. Kabat-Zinn, Full catastrophe living: the program of the stress reduction [41] N.A. Farb, A.K. Anderson, Z.V. Segal, The mindful brain and emotion regulation
clinic at the University of Massachusetts Medical Center, New York: Delta in mood disorders, Can. J. Psychiatry 57 (2) (2012) 70, Revue Canadienne De
(1990). Psychiatrie.
50 J. Laurie, A. Blandford / International Journal of Medical Informatics 96 (2016) 38–50

[42] P.R. Goldin, J.J. Gross, Effects of mindfulness-based stress reduction (MBSR) on [44] A. Thieme, J. Wallace, P. Johnson, J. McCarthy, S. Lindley, P. Wright, P. Olivier,
emotion regulation in social anxiety disorder, Emotion 10 (1) (2010) 83. T.D. Meyer, Design to promote mindfulness practice and sense of self for
[43] R.P. Hawkins, J.Y. Han, S. Pingree, B.R. Shaw, T.B. Baker, L.J. Roberts, vulnerable women in secure hospital services, In Proceedings of the SIGCHI
Interactivity and presence of three eHealth interventions, Comput. Human Conference on Human Factors in Computing Systems (2013) 2647–2656
Behav. 26 (5) (2010) 1081–1088. (ACM.).
[45] B.J. Fogg, Persuasive technology: using computers to change what we think
and do, Ubiquity 2002 (December) (2002) 5.