Content uploaded by Gloria Mark
Author content
All content in this area was uploaded by Gloria Mark on Dec 23, 2016
Content may be subject to copyright.
Available via license: CC BY 4.0
Content may be subject to copyright.
Chen et al.
RESEARCH
Promoting Positive Affect through Smartphone
Photography
Yu Chen, Gloria Mark and Sanna Ali
Full list of author information is
available at the end of the article Abstract
With the increasing quality of smartphone cameras, taking photos has become
ubiquitous. This paper investigates how smartphone photography can be
leveraged to help individuals increase their positive affect. Applying findings from
positive psychology, we designed and conducted a four-week study with 41
participants. Participants were instructed to take one photo every day in one of
the following three conditions: a selfie photo with a smiling expression, a photo of
something that would make oneself happy, and a photo of something that would
make another person happy. After three weeks, participants’ positive affect in all
conditions increased. Those who took photos to make others happy became
much less aroused. Qualitative results showed that those in the selfie group
observed changes in their smile over time; the group taking photos to improve
their own affect became more reflective, and those taking photos for others found
that connecting with family members and friends helped to relieve stress.
Keywords: happiness; photos; smartphones; positive affect; in situ study;
positive computing; mental health
Introduction
Consistently living under stress can lead to chronic health problems such as depres-
sion, anxiety disorders, heart disease, high blood pressure, and diabetes [32]. Col-
lege students in particular are a vulnerable population that experience stress. Their
stress may come from living away from family for the first time, feeling lonely or
isolated, experiencing pressure from coursework, or worrying about finances [25, 31].
Stress is reported as one of the factors that negatively impacts students’ academic
performance, and thus can lead to depression [42]. Conventional methods to cope
with stress include medication, exercise, therapy, and seeking emotional support
[32].
Psychologists have investigated various methods of improving emotional and men-
tal well-being. For example, writing down three things that went well during the
day can significantly help people increase their level of happiness [37]. Dunn et al.
[10] found that people were happier when they spent money on others instead of
on themselves. Embodying happiness – representing happiness in a physical form
– can even relieve stress: a study showed that people became less stressed if they
adopted a smiling facial expression [23].
The last decade has witnessed the emergence of positive computing – the use of
information technology to support human well-being [6]. Researchers from diverse
fields, such as psychology, social science, psychiatry, and information science are
bringing their expertise to leverage the ever-increasing advancement of informatics
Chen et al. Page 2 of 16
to help people better manage their emotional well-being [37, 6]. At the same time,
the high adoption of smartphones and social media brings new opportunities for
measuring and sharing emotions. With the increasing quality of smartphone cam-
eras, taking photos has become ubiquitous. This trend is reflected by the widespread
popularity of photo-related social media such as Instagram, Snapchat, and Face-
book.
In this study, we investigated how we could leverage findings from positive psy-
chology to promote people’s positive affect and potentially reduce stress through
taking photos. We compared the impacts of photo-taking on well-being in three
different conditions: 1) self-perception, in which people manipulated positive facial
expressions; 2) self-efficacy, in which people did things to make themselves happy;
and 3) pro-social, in which people did things to make other people happy. We de-
veloped two Android applications as experimental platforms that prompted users
to take photos and report their mood. We then conducted a four-week in-situ study
to assess the effectiveness of taking photos to promote positive affect.
This work contributes to the field of positive computing in the following ways.
First, it empirically demonstrates the effectiveness of using smartphone photography
to promote positive affect. Second, it applies strategies that have been used in
positive computing studies. Third, it offers implications for the design of systems
that use smartphone photography to promote users’ emotional well-being.
Related Work
Emotional well-being is an essential part of mental health [35]. Positive emotions
are found to enhance cardiovascular, hormonal and immune functions, promote
healthy behaviors such as better sleep and more exercise [23], and lead to more
open-minded thinking and effective problem solving [6]. In line with the importance
of positive emotions, positive psychology emerged as a discipline with psychologists
seeking to find various methods to help people increase their emotional well-being
[5, 13, 37, 39]. Meanwhile, the advent of pervasive sensors, wearable devices, and
mobile technologies has given rise to positive computing, the use of informatics to
support mental well-being [6]. In the area of affective computing [33], researchers
have employed various sensors to detect users’ affective states from facial expres-
sions, speech, body gestures, or breath, and then have presented visualizations of
these states to users [12, 11, 3, 40]. Such methods can increase users’ awareness of
their emotions and can trigger users to self-regulate, especially if they are experi-
encing negative affective states. However, the above monitoring methods lack an
emphasis on empowering users to proactively change their emotions.
In this work, we set out to use technology to help users complete exercises designed
to increase their positive affect. We chose smartphone photography as a means to
make such practices accessible and habitual in people’s daily lives. Smartphone
photography has been used as a memory aid, such as taking snapshots of price
tags, recipes, and maps [17], as a tool to document life events [24], and as media
to communicate with friends and families [9]. Sharing photos on social media has
become widespread [28], as evidenced in the rise of Instagram, Snapchat, and photo-
sharing services on Facebook and Twitter [19]. As of 2016, Instagram has over 400
million monthly active users, and 80 million photos are uploaded every day [20].
Chen et al. Page 3 of 16
Among the photos uploaded, selfies — self-portraits made with a smartphone [36]
– are becoming a wide-spread phenomenon. We sought to investigate and leverage
findings of positive psychology to promote positive affect into the practice of taking
photos. Particularly, we apply the following three theories that have been shown to
improve people’s positive affect.
Smiling brings happiness. Self-perception theory states that how people be-
have will determine what they think and how they feel [2]. In one study users
became mentally stronger when they embodied body postures that were physically
expansive and implied power [8]. In another study, participants who maintained a
positive facial expression while stressed experienced less decrease in positive affect
than those in a baseline group [23]. This study also demonstrated lower heart rates
during stress recovery and the enhanced ability to endure stressful events. Kleinke et
al. [22] found that participants who engaged in positive facial expressions increased
their positive mood. The effects were greater when participants viewed themselves
in a mirror. Based on this theory, Tsujita et al. [43] designed HappinessCounter, a
device that recognizes users’ smiles, counts the number of smiles, and then provides
feedback in a mirror. Their field study showed that users became happier and smiled
more naturally after ten days. The Mood Meter [18], which also encourages smiling
of passersby in public places, consists of a camera that captures people’s facial ex-
pressions, a computer that analyzes the facial expressions and detects smiles, and
a public display that shows users’ smiles. SmileTracker [21] not only detects users’
smiles from the web camera of their computers, but also captures a screenshot of
their smile for them to reflect upon that image in the future.
Reflecting brings happiness. The “three-good-things-in-life” exercise, pro-
posed by Seligman et al. [37], asks participants to write down three things that
went well that day and their causes. Participants became happier and less de-
pressed after a one-month intervention. By implementing this exercise in online
social networks, Munson et al. [29] developed a Facebook application called Hap-
pierTogether. Another application inspired by this finding is Happier, a commercial
mobile app that aims to promote users’ positive affect by having them take pictures
to savor moments of happiness, reflect on the reason for such happiness, and then
keep the photo privately or share it with their social networks [15]. Similarly, Hap-
pify, another commercial website, also encourages users to record good things that
happened each day using gamification [16].
Giving brings happiness. An experiment conducted by Dunn et al. [10] showed
that participants in an experimental condition where they spent money on others
reported a higher degree of happiness than participants who were given instructions
to spend money on themselves. In a study of Seligman et al. [37], participants were
instructed to write and then deliver a letter of gratitude in person to someone who
had been kind to them but had never been properly thanked. The participants
significantly increased their sense of happiness, and this effect remained after six
months. Mortality was even shown to be reduced for older adults who had reported
providing instrumental and emotional support to their strong ties [4]. The above
findings suggest that giving to others can bring benefits to health and longevity.
In this section, we surveyed sensing technologies for monitoring emotional states,
practices that promote positive well-being, and technological tools that are created
Chen et al. Page 4 of 16
based on these practices. However, the practicality of the above tools (e.g. whether
they can be adopted) has not been addressed. Our interest was in investigating how
practical tools that have already been adopted could be used to enhance happiness.
We thus chose to use smartphone photography as a medium to implement findings
from research in the area of positive psychology. To our knowledge, this paper
presents the first study that explores and compares the application of theories to
promote happiness using smartphone photography.
User study
We conducted a four-week in-situ study during which the participants (college stu-
dents) carried out their normal day-to-day activities (going to class, studying, etc.).
The experiment took place at a public university on the U.S. west coast. We used a
mixed study design so that each participant served as their own baseline to account
for individual differences. The study consisted of a one-week control session followed
by a three-week intervention session. We chose a period of four weeks so that the
control and the intervention sessions spanned over the same days of the week, thus
minimizing the influence of a given day’s schedule on the users’ daily activities and
mood. To investigate how smiling, reflecting, and giving to others might impact
users’ mood, we designed three experimental conditions:
•Selfie: participants would take a selfie daily while smiling;
•Personal: participants would take a photo daily of something that makes
themselves happy;
•Other: participants would take a photo daily of something that they believe
would make another person happy and then they would send it to that person.
Materials
We developed two Android applications, SurveyApp and MettaApp, as experimen-
tal platforms for the control session and the intervention session respectively. The
SurveyApp was designed to collect users’ mood in the control session. Figure 1(a) is
the home screen of SurveyApp. The app includes five main tasks: a morning survey,
three mood surveys during the day, and an evening survey. Each task is visualized
by a colored icon. If users have finished any of the tasks, the corresponding col-
ored icon is greyed out and checked. Users could also see which day it is of the
experiment.
The MettaApp was designed to collect users’ moods and enabled them to take
photos in the intervention session. The MettaApp was built on top of the Sur-
veyApp. It extended the functions of the SurveyApp and included an additional
photo function (see Figure 1(b)). A user could take one photo per day by click-
ing the camera button. The button is then replaced by the photo that the user
has taken. Users could check all their photos by clicking the button “Click to view
photo history” (Figure 1(b)). Figure 1(c) shows the photo gallery in the timeline.
Participants
We recruited 57 participants on campus by making announcements in classes and
placing advertisements on Facebook. Ten participants withdrew from the study
during the control session due to system incompatibility issues, and six withdrew
Chen et al. Page 5 of 16
(a) (b) (c)
Figure 1: Screenshots of SurveyApp and MettaApp.
(a) SurveyApp Homepage; (b) MettaApp homepage; (c) Photo history on
MettaApp.
during the intervention session due to personal reasons. In the end, 41 participants
completed the entire study, including 14 in the Selfie condition, 14 in the Personal
condition, and 13 in the Other condition. All participants were undergraduate or
graduate students who used an Android phone as their primary phone. The par-
ticipants, 13 males and 28 females, were between 18 and 36 years old. They were
assigned randomly to one of the three conditions: Selfie, Personal, and Other. Table
1 shows the distribution of participants’ gender and major by experimental condi-
tions. We categorize their majors by STEM (science, technology, engineering and
mathematics) and others. At the end of the study, participants were compensated
with $25.
Table 1: Demographic information of participants.
Condition Gender Major
Male Female STEM Other
Selfie 5 9 5 9
Personal 4 10 6 8
Other 4 9 5 8
Total 13 28 16 25
Procedure
Before the study, we invited participants to the laboratory for an informational
meeting, to fill out a general survey, and to sign informed consent. We assisted
them with installing the SurveyApp at the beginning of the control session and the
MettaApp at the beginning of the intervention session on their own Android phones
from a given link. During the study, users reported their mood during each day, three
times per day. In the evening survey, we also asked participants to indicate whether
there was any significant event that happened to them that day at work or at home
Chen et al. Page 6 of 16
Figure 2: Mood sampling page
that affected their mood or stress level. If so, we asked them to briefly describe it.
The above tasks were completed on the SurveyApp during the control session and
on MettaApp during the intervention session. Starting from the beginning of the
intervention session, participants used the MettaApp to take photos according to
the condition to which they were assigned. At the end of the study, participants
returned to our laboratory for an exit interview.
Mood sampling
We obtained users’ moods using a visual representation of Russell’s Circumplex
model [34]. This model measures users’ mood in two dimensions that are orthogo-
nal: valence (i.e., how positive one feels) and arousal (i.e., how intense the feeling is).
Even though the initial goal was to improve users’ positive affect, we assessed users’
mood in both valence and arousal, since we wanted to gain a more nuanced under-
standing of the effect that our interventions might have on users’ positive affect.
We instructed participants on the meaning of the measures during the pre-study
informational meeting. During the control session, the mood sampling requests were
triggered via a notification on the SurveyApp three times during each day: in the
morning (approximately 10 am), in the afternoon (approximately 2 pm), and in the
evening (approximately 7 pm). During the intervention session, the mood sampling
requests were triggered on the MettaApp three times after a photo was taken: 5
minutes, 1 hour and 3 hours after the photo. Figure 2 shows the interface from
where participants input their mood using two sliding bars. On the upper sliding
bar, participants indicated the valence of their feeling “right now” using a range of
–50 (negative) to +50 (positive). On the lower sliding bar, they selected a value for
their arousal between –50 (arousal low) and +50 (arousal high). We chose a large
range between –50 and +50 to to maximize the opportunity to capture nuanced
responses given the constraints of the limited screen space on smartphones. We
logged the time when they answered the probes.
Chen et al. Page 7 of 16
Taking photos
Participants took one photo every day using MettaApp during the three-week inter-
vention session from Week 2 to Week 4. They took photos following the instructions
they received: the Selfie group took photos while smiling, the Personal group took
photos of things that made themselves happy, and the Other group took photos of
things that would make other people happy and then they sent the photos to oth-
ers. For the Other group, they could choose their preferred methods to send photos,
e.g., text message, email, or social media apps. Participants were shown the photo
they took that day every time they opened the app, and they were also able to view
all the photos they had taken in the previous days. The photos were uploaded and
backed up to a secure server, and were only accessible to the participant and the
research team.
Exit interview
After the study, each participant returned to the laboratory for an individual exit
interview. During the semi-structured interview, we asked participants their daily
frequency of using the application, and their experiences while using the MettaApp.
We also reviewed the photos with the participants and asked them to show the pho-
tos with which they felt most happy and to discuss the reasons. Then we encouraged
them to contribute ideas about designing technology for emotion intervention. Fi-
nally, participants were instructed to uninstall the experimental applications from
their smartphones. After uninstallation, the applications were deactivated and could
no longer collect any data from participants. Each interview took about 25 minutes,
and we audio-recorded all the interviews.
Results
We collected the following types of data: 1) the valence and arousal obtained from
daily mood sampling from the phone during the control and intervention sessions, 2)
photos uploaded during the intervention session, and 3) interview data which were
then transcribed. This section reports both quantitative and qualitative results. We
present results of quantitative analyses on the intervention effects on mood and
the comparison of intervention effects among the three conditions. We then present
our findings of how the photo-taking made the participants happy by qualitatively
analyzing their interview transcripts. Finally, we compared the participants’ photos
in the three conditions through visual inspection and coding.
Intervention effects on mood
We collected 2,897 mood measures from experience sampling. The mean valence was
15.37 (SD=19.8, Max=50, Min=−50) and the mean arousal was −2.76 (SD=25.1,
Max=50, Min=−50). To examine the effects of the intervention of taking photos,
we conducted a Linear Mixed-Effects Model (LMM) analysis in SPSS. LMM han-
dles random and fixed effects and was used since our participants had repeated
measures over days of their mood responses, from the mood sampling. We averaged
the valence responses at 5-minutes, 1-hour, and 3-hours for each day and simi-
larly, arousal responses at 5-minutes, 1-hour, and 3-hours for each day. We thus
had 985 total responses. Our dependent variables were Valence (the daily averaged
Chen et al. Page 8 of 16
valence responses) and Arousal (the daily averaged valence responses), analyzed
in separate models. With Condition (Selfie/Personal/Other) as a between-subjects
variable, and Intervention (before/after the intervention) as a within-subjects vari-
able, we entered an interaction term of Condition x Intervention. These variables
were entered as fixed effects. Participants were entered as random effects.
Table 2 shows the mean Valence and Arousal in the three conditions before and
after the intervention. For Valence, we found a significant main effect of Interven-
tion: F(1,926)=10.03, p=.002, Mean Before=13.64, SE=1.64; Mean After=16.65,
SE=1.54. The main effect of Condition and the interaction between Intervention
and Condition were not significant. Thus, participants in all three conditions rated
their valence higher with the photo interventions.
For Arousal, we found a trend of significant Condition x Intervention interaction:
F(2, 922)=2.63, p=.072. Participants in the Other condition reported lower arousal
after the photo intervention (see Table 2). The main effect of Intervention and the
main effect of Condition were not significant. Based on Russell’s circumplex model
of mood mapping [34], we refer to the lower arousal scores as reflecting a calmer
mood. Thus, taking photos is associated with participants in the Other condition
becoming calmer compared with the Personal condition.
An LMM analysis of the temporal effects of rating Valence and Arousal 5 minutes,
1 hour, and 3 hours after the photo-taking showed no significant difference with time
of response.
Table 2: Mean and Std. Deviation of participants’ valence and arousal in
the control session and the intervention session.
Condition Session Valence Arousal
Mean SD Mean SD
Selfie Control 12.16 13.48 −2.30 17.47
Intervention 13.55 16.28 −3.60 21.09
Personal Control 12.62 15.29 2.97 17.17
Intervention 16.00 18.55 3.45 21.67
Other Control 16.72 14.80 −5.74 22.40
Intervention 19.84 15.66 −11.05 21.89
Qualitative results: How do photos make people happy?
We then analyzed the transcripts of the exit interviews to understand why and
how taking different types of photos influenced people’s mood. Three researchers
transcribed the audio recordings of the interviews. We then used grounded theory
[41] to analyze the interview data. Table 3 summarizes the main themes derived
from the interview data.
Table 3: Themes of coded qualitative data.
Condition Themes Number
Selfie Changed mood, due to feeling more confident, comfortable, or creative in smiles 5
Constraints: brought more stress; inconvenient; repetitive smiles became boring 4
Personal Became more mindful, reflective, and appreciative 9
Became aware that things around them served as important sources of happiness 5
Other Receiving responses from the recipients of the photos made participants happy 7
Helped the participants communicate their current situation 6
Took photos of things that embedded shared memories 4
Connecting with strong ties reduced their stress 6
Chen et al. Page 9 of 16
Selfie condition. Five out of the fourteen participants in the Selfie condition
observed changes in their smile and mood over the course of the three-week photo
intervention session. Some participants felt more confident over time about them-
selves taking selfies, such as P29: “As days went on, I got more comfortable taking
photos of myself. If you feel good about yourself, then [a] selfie would be a way to
capture that.” P46 reported that he became better at taking smiling selfies and
noticed less stress on his face. P12 looked back on her selfies from time to time and
became more creative in her photos by making gestures while taking the selfie. P40
reported that she sometimes reflected on the moment when she smiled. Two partic-
ipants reported that even fake smiles lifted their mood up. As P29 said, “It made
me feel good, thinking, ‘this is probably how I look like for the rest of the day.’...
It’s a way of telling me that I could get through the day no matter what happens.
One of the photos was taken when I found out my friend passed away. That was a
fake smile. I was depressed. I figured [that] if I can see myself smiling in the picture,
things would be okay for the day.”
Participants also reported some constraints in taking the photos. First, fake and
forced smiles sometimes brought them stress (N=4). A previous study [30] shows
that the emotions induced during intervention should match users’ genuine affect
e.g., happy/energetic, upset/subdued. This might explain some participants’ neg-
ative feedback of smiling selfies. Second, some participants found it inconvenient
to find a private place to take a smiling selfie (N=3). “Sometimes it was difficult
because I was not comfortable taking photos of myself in the public places. Not easy
to fit into my day”, said P21. Third, for participants who always took the selfie at
the same place, repeatedly taking the same photo became boring (N=2).
Personal condition. In the Personal condition, most participants became more
mindful, reflective, and appreciative by taking photos. The most frequently reported
reason (N = 9) for being happier after the three-week photo session was that the
photos helped them to be reflective. They thought more carefully about the source
of their happiness. “I do not use a social media app to reflect on something hap-
pen[ing] on a particular day. Using this app made me think of something [that]
made me happy, reminding me of things that made me happy”, said P27. A theme
that emerged in the data was that participants started to realize happiness could
come from things in their lives that they usually take for granted. For example, P31
commented, “They just open my eyes and made me realize what makes me happy.
Those are simple things that I never thought about before. Just like everyday objects
and places in my room. They are places that made me content and stress-free at that
time. Not big, but it does have an impact.” For P51, he realized that he was happy
because of social connections and experiences. “All the photos had special meanings
for me: hanging out with friends, socializing with people I care about, enjoying the
experience, like coffee or a movie. I took one immediately after watching a movie
with my roommate.”
Some participants started to pay attention to their family members. As P28 said,
“Instead of going routinely and mechanically during the day, I stop and look around
for something that makes me smile. I didn’t consciously do that before. I find that
happiness is close to me. A lot of them are my family and my pet. For my family, I
didn’t think of them as a daily source of happiness. I usually took them for granted.”
Chen et al. Page 10 of 16
Some became mindful of small things around them. For example, P25 started to
consciously notice something that was nice even if it was in the background that
she would not have noticed otherwise. She photographed mostly flowers that she
walked by during the day and took two photos of her cat.
Realizing that things around them served as an important source of happiness,
some participants reported that they became more appreciative (N=5). As P36 said,
“They make me appreciate the small things in my life –things that I would normally
not notice or take for granted. There are some photos of family members, reminding
me of a reason to live for and making me happy. Sometimes I took pictures of my
laptop. It helps me do well in school and brings a lot of convenience to my life. It
made me happy. I don’t get excited, but feel grateful. It’s good that I have one.”
P23 and P24 reported that they started to cherish the time with their friends or
significant others and felt grateful for their company.
Other condition. In the Other condition, 95% of photos were sent to strong ties,
i.e., family members, friends, and significant others. Most participants reported
that they thought more of, and felt more connected with strong ties during the
photo intervention period than before. As P20 mentioned, “I don’t talk to my dad
every day. But when I sent the photo to him, it made him happy, as a way of
communication.” P18 also reported reflecting and appreciating more of her life. “I
feel taking the photos made me realize lots of simple things not only made other
people happy but also made myself happy.”
Seven participants mentioned receiving a response from the recipients of the pho-
tos. The participants became more satisfied because they became aware that they
made the person who received the photo happy. “It was fun to send stuff to my
girlfriend to make her laugh. Seeing her reactions will always make me smile,” said
P44. Similarly, P43, who often sent pictures to her boyfriend reported, “I usually
send photos of what I was doing or watching, or something that happened that day,
for example, an advertisement or a flyer for a show. He always responded: ‘that’s
really cute!’ ‘That’s awesome, can we see the show?’ That made me happy and
showed how supportive he was and always had the same amount of excitement as
I had.” P16 sent a photo to her friend as a birthday gift. “She has a crush on
someone, and I took the photo on her birthday. I messaged her this photo greeting
her happy birthday, and she said that made her day. I was really happy.”
Many participants reported that by taking and sending pictures of their present
moment, they made their strong ties happy. The photos helped the participants
communicate their current situation, e.g., how they were feeling, what they were
working on, and what environment they were in (N=6). As P35 said, “I was at the
library and decided to show my mom how hard I was working. So, I took a picture
of my notes and textbooks and then sent it to her. It made her happy knowing the
effort I was putting in.” P37 took most of the photos for her mother and sister,
who were in a different country: “For my mom, it’s mostly what I’m doing. Some
pictures might look boring, but she was happy knowing what I was doing.”
Participants also took photos of things that embedded shared memories (N=4).
For example, P43 intentionally took pictures to make his girlfriend happy; “There
was something we joked about before. It was the personal connection that gave the
meaning. I was not taking pictures [that would be] super meaningful for others.”
Chen et al. Page 11 of 16
P30, who usually took photos of her mother’s favorite things, said, “It was nice to
have something to send to somebody every day. I usually sent [them] to my mom.
Sometimes she laughed at the pictures: ‘thanks for thinking of me today’. . . It let
her know something reminded me of her and that I was missing her.”
Participants also mentioned that connecting with strong ties reduced their stress.
“People can be comforted by these sort of photos. If someone is feeling depressed,
the first thing they need is connection,” described P15. This trend is more visible
for participants who are international students and whose family is physically far
from them (N=4). “Just the action of sending a photo already made my parents
happy, because they feel more assured about my studies and my life, or because
I’m thinking about them. When I felt stressed with my studies, the intimacy from
interacting dispelled the loneliness, making me appreciative and relieved. That takes
me away from the stress,” explained P10. Connecting with strong ties may be one
explanation of why participants in the Other condition reported feeling much less
aroused after taking and sending photos.
Analyzing the photos
We collected a total of 692 photos from participants, comprised of 271 in the Selfie
condition, 227 in the Personal condition, and 194 in the Other condition. Two
researchers independently coded the photos in the three conditions. For the Selfie
condition, they coded the locations where the selfies were taken, e.g. home and
school. For the Personal and the Other conditions, they coded the content of the
photos, e.g., friends and food. With an agreement rate of 96.1% on the coded labels
initially, the two researchers then reached a consensus on the rest of the photos
mediated by a third researcher.
Figure 4 summarizes the locations and their distribution. For the selfies, most of
them were taken at home (65.3%). The rest were taken in cars (10.0%), at study
areas (3.1%), or at restaurants (2.1%). Why were most smiling selfies taken at
home? The interview data reveal that many participants started a day by making a
smiling selfie at home (N=4) or signaled the end of a day with a selfie when arriving
at home (N=3). Others mentioned that they tried to take smiling selfies in private
places rather than public places such as in the classroom or at workplaces in order
not to be embarrassed. As P14 said, “I always make sure no one is around, and I
look presentable.” Always taking smiling selfies with the same facial expression at
the same place could explain why some participants felt bored by taking the selfies.
By contrast, some participants preferred to take selfies with a background that
embedded particular meanings, such as at a banquet, before a wedding, and after a
satisfying haircut. This suggests that encouraging users to smile during meaningful
events and at a variety of occasions can help reduce the perception of boredom in
the smiling selfie exercise.
For the Personal and the Other conditions, we coded the content of the photos
to investigate what kinds of things participants indicated as making themselves
or other people happy. Table 5 lists the themes of the two conditions ranked by
proportion. Food ranks top among all photo themes, 19.4% in the Personal condition
and 22.2% in the Other condition. It seems that food made participants themselves
happy, as well as their strong ties. A social theme, which includes family, friends, and
Chen et al. Page 12 of 16
Table 4: Distribution of the locations where the smiling selfies were taken.
Location Number Percent of total
At home 190 65.3
In car 29 10.0
Outdoors 22 7.6
At work/study place 9 3.1
In restaurant 6 2.1
Miscellaneous 15 5.1
Total 271 100
significant others, is another common theme in both the Personal condition (17.2%)
and the Other condition (8.2%). Photos of personal theme were frequently taken by
participants, 16.3% in the Personal condition and 20.6% in the Other condition. The
personal theme includes personal spaces where people live, study, and work, as well
as personal items such as toys, pictures, figurines, and ornaments. Photos of this
theme were frequently sent by participants in the Other condition to their strong
ties. They were used as a communication channel to inform the receiver about their
everyday life, and thus increase mutual awareness and intimacy between the sender
and the receiver. By contrast, participants in the Personal condition took photos
of places where they live, study, and work, which could serve to remind them that
happiness exists in their surroundings – the simple and tiny things around them.
For the entertainment photos, such as video games, Youtube videos, and Netflix,
participants in the Personal condition took more than the Other condition (15.0%
vs. 9.8%). For nature themes, such as flowers, the sea, and tress, the Other condition
has a slightly larger share than the Personal condition (18.0% vs. 14.1%). Some
selfies were taken by participants in the Other condition and these were mainly
sent to significant others (5.2%).
Table 5: Distribution of photo themes in the Personal and Other condi-
tions.
Personal condition Other condition
Theme Number Percent of total Theme Number Percent of total
Food 44 19.4% Food 43 22.2%
Social 39 17.2% Personal 40 20.6%
Personal 37 16.3% Nature 35 18.0%
Entertainment 34 15.0% Entertainment 19 9.8%
Nature 32 14.1% Social 16 8.2%
Pet 20 8.8% Technology 11 5.7%
Technology 10 4.4% Selfie 10 5.2%
Urban 5 2.2% Beauty 6 3.1%
Art 3 1.3% Art 5 2.6%
Beauty 2 0.9% Spiritual 4 2.1%
Spiritual 1 0.4% Urban 3 1.5%
Selfie 0 0.0% Pet 2 1.0%
Total 227 100.0% Total 194 100.0%
Discussion
The results suggest that any photo-taking with the intent to increase one’s happi-
ness can increase positive affect, specifically photos intended to promote happiness
via smiling self-expression (selfies), those taken of things to make ones’ self happy,
or those intended to make others happy. Moreover, sending photos to others makes
people less aroused. As described earlier, based on Russell’s circumplex model of
Chen et al. Page 13 of 16
mood mapping [34], we refer to the lower arousal scores in the Other condition as
participants becoming calmer. Humans are social creatures. Connecting with strong
ties helps people become calmer, especially for those who tend to cope with stress
through emotional support [7]. In fact, most of the photos taken were of things that
connect the sender and the receiver, for example, those that document the current
state of their life or embed shared memories. Seemingly small things can increase
the intimacy of strong ties in online communication [1]. On the other hand, taking
photos that make people close to them happy further requires users to think beyond
themselves to benefit others. As the research of Seppala and Tomasello [38] shows,
depression and anxiety are linked to self-focus. When people make an effort to in-
crease the happiness of other people, they are broadening their perspective beyond
themselves. Other-focused attention and thinking about others has been shown to
trigger a decrease in heart rate and skin conductance [6, 14].
We also asked participants in the interviews to compare MettaApp with pho-
tography apps on social media. Most participants (N=22) mentioned the photos
with MettaApp were mainly for themselves. They felt more comfortable expressing
themselves in the pictures without being disturbed by external factors, such as im-
pression management, or how others will perceive them. By contrast, photos posted
on Instagram, Facebook, and Snapchat are mainly targeted for their social circle,
which is sometimes hundreds of people. Participants would keep their audience in
mind, take into account the likes and comments on social media, and try to make
the photos presentable and look perfect. For P36 in the Personal condition, taking
a moment to stay centered in his life without social influence helped him rediscover
the source of happiness in his life.
Further, we encouraged participants in the interviews to suggest future technolo-
gies that could enhance their happiness using photography based on their experience
in this study. One recommendation that surfaced often from participants is to design
technologies to help people review photos of happy moments in the past (N=14).
Such a technology could display the photos of happy moments to people when are
experiencing a bad mood. Participants also imagined tools that could help them
review happy moments at the end of the day for a better sleep, or at the beginning
of the next day to start a day with positive energy. Reflecting emotions, especially
positive emotions, is shown to help improve users’ mental well-being [21, 27]. Some
participants suggested technologies that could pop up “happiness” photos at ran-
dom times of the day to give them a surprise of positive reminiscence. Participants
also suggested “smart” photography that detects mood automatically. With perva-
sive sensors and wearable devices that track users’ mood, future technology may
send users “happiness” photos when it detects their negative mood (P27). Mean-
while, such technology could also recommend that a user take a photo to record the
moment if the sensors have detected an increase in a user’s positive affect (P31).
Limitations
This study focuses on three exercises instead of covering an exhaustive list from
positive psychology. It may be worth exploring interventions that combine these
conditions, such as taking selfies with strong ties in the photo or sending selfies to
strong ties. It is also possible that the period of the photo intervention coincided
Chen et al. Page 14 of 16
with a period where our participants were more positive, or changes over time could
have played a role in the results. However, the intervention occurred towards the
end of the academic quarter when students generally experience more stress. So the
fact that valence increased and arousal decreased for some people is contrary to
what we would expect without any intervention, given the time when the study was
conducted. Moreover, since people were tested over a period of time, experiencing
different environments, the environment should play less of a role in influencing
the results. In this study we did a within-subjects design, where each participant
served as their own control. In future studies, to rule out changes over time that
could affect the results, we could include a control group to further validate the
findings of this study.
Conclusions
We aimed to leverage the prevalence of smartphone photography along with theories
of positive psychology to help college students become happier and reduce stress.
To this end, we conducted a four-week study with 41 participants to investigate
the effects of taking daily photos using their smartphones in three conditions: the
Selfie condition in which participants took a smiling selfie, the Personal condition in
which participants took a photo of something that made themselves happy, and the
Other condition in which participants took and sent a photo of something to make
another person happy. Quantitative and qualitative results show that participants
in all three conditions became more positive after taking their assigned type of
photo daily for three weeks. Some participants in the Selfie condition observed a
more natural smile over time; participants in the Personal condition became more
reflective, and some participants reported that the photos led them to be more
appreciative of the little things in their lives that made them happy. Participants
in the Other condition became much less aroused (i.e., calmer) with photo-taking,
and some reported the increased intimacy and connection with strong ties as an
important factor that can reduce anxiety, serve to pacify themselves, and lead them
to become more positive. Compared to photos posted on social media, participants
felt more comfortable, conscious, and reflective when taking the photos. They also
suggested future technology that could help them take and review photos of happy
moments using mood-tracking sensors.
This paper provides empirical support on the feasibility of increasing users’ happi-
ness by applying positive psychology to smartphone photography. It also contributes
to the emerging field of positive computing by presenting reasons for how conducting
exercises to promote happiness using mobile technology could help people enhance
their mood. The findings can offer insights for designers to create systems that
enhance emotional well-being.
Author’s contributions
YC took a leading role in designing and conducting the study, implementing the system, analyzing the data, and
writing of the manuscript. GM made significant contributions in designing the study, analyzing the data, and writing
the manuscript. SA participated in designing the study, conducting interviews, and writing the manuscript.
Acknowledgements
We are grateful for anonymous reviewers for their valuable feedback. We also thank Ding Ye for contributing time in
developing the applications and Jackeline Maldonado and Cindy Chen for assisting in the project.
Chen et al. Page 15 of 16
Funding
This work is sponsored by the Swiss National Science Foundation under the grant #158933 and the NSF under
grant #1218705.
Competing interests
The authors declare that they have no competing interests.
References
1. Bales, E., Li, K. A., & Griwsold, W. (2011). CoupleVIBE: mobile implicit communication to improve awareness
for (long-distance) couples. In Proceedings of the ACM 2011 conference on Computer supported cooperative
work (pp. 65-74). ACM.
2. Bem, D. J. (1973). Self-perception theorys. Advances in experimental social psychology, 6, 1–62.
3. Bernhardt, D., & Robinson, P. (2007). Detecting affect from non-stylised body motions. In Affective
Computing and Intelligent Interaction (pp. 59-70). Springer Berlin Heidelberg.
4. Brown, S. L., Nesse, R. M., Vinokur, A. D., & Smith, D. M. (2003). Providing social support may be more
beneficial than receiving it results from a prospective study of mortality. Psychological Science, 14(4), 320-327.
5. Biswas-Diener, R., & Dean, B. (2010). Positive psychology coaching: Putting the science of happiness to work
for your clients. John Wiley & Sons.
6. Calvo, R. A., & Peters, D. (2014). Positive computing: Technology for wellbeing and human potential. MIT
Press.
7. Cohen, S., & McKay, G. (1984). Social support, stress and the buffering hypothesis: A theoretical analysis.
Handbook of psychology and health, 4, 253-267.
8. Cuddy, A. (2012). Your body language shapes who you are. TED Global. Retrieved on September 23, 2015
from http://www.ted.com/talks/amy cuddy your body language shapes who you are?language=en.
9. Cui, Y., Kangas, J., Holm, J., & Grassel, G. (2013). Front-camera video recordings as emotion responses to
mobile photos shared within close-knit groups. In Proceedings of the SIGCHI Conference on Human Factors in
Computing Systems (pp. 981-990). ACM.
10. Dunn, E. W., Aknin, L. B., & Norton, M. I. (2008). Spending money on others promotes happiness. Science,
319(5870), 1687-1688.
11. Kisacanin, B., Pavlovic, V., & Huang, T. S. (Eds.). (2005). Real-time vision for human-computer interaction.
Springer Science & Business Media.
12. Fernandez, R., & Picard, R. W. (2005). Classical and novel discriminant features for affect recognition from
speech. In Interspeech (pp. 473-476).
13. Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: The broaden-and-build theory
of positive emotions. American psychologist, 56(3), 218.
14. Goetz, J. L., Keltner, D., & Simon-Thomas, E. (2010). Compassion: an evolutionary analysis and empirical
review. Psychological bulletin, 136(3), 351.
15. Happier Inc. Happier official site. Retrieved on September 22, 2015 from https://www.happier.com.
16. Happify Inc. Happify official site. Retrieved on September 22, 2015 from http://www.happify.com.
17. H¨akkil¨a, J., Huhtala, J., Sarjanoja, A. H., & Schmidt, A. (2012). Price tags, maps, recipes: mobile phone
photos for functional purposes. In Proceedings of the 7th Nordic Conference on Human-Computer Interaction:
Making Sense Through Design (pp. 41-44). ACM.
18. Hernandez, J., Hoque, M. E., Drevo, W., & Picard, R. W. (2012). Mood meter: counting smiles in the wild. In
Proceedings of the 2012 ACM Conference on Ubiquitous Computing (pp. 301-310). ACM.
19. Hu, Y., Manikonda, L., & Kambhampati, S. (2014). What We Instagram: A First Analysis of Instagram Photo
Content and User Types. In ICWSM.
20. Instagram. Instagram statistics. Retrieved September 21, 2015 from https://instagram.com/press/.
21. Jaques, N., Chen, W. V., & Picard, R. W. (2015). SmileTracker: Automatically and Unobtrusively Recording
Smiles and their Context. In Proceedings of the 33rd Annual ACM Conference Extended Abstracts on Human
Factors in Computing Systems (pp. 1953-1958). ACM.
22. Kleinke, C. L., Peterson, T. R., & Rutledge, T. R. (1998). Effects of self-generated facial expressions on mood.
Journal of Personality and Social Psychology, 74(1), 272.
23. Kraft, T. L., & Pressman, S. D. (2012). Grin and bear it the influence of manipulated facial expression on the
stress response. Psychological science, 23(11), 1372-1378.
24. Lehtim¨aki, K., & Rajanti, T. (2008). Documenting the ordinary: mobile digital photography as an agent of
change in people’s practices concerning storing and sharing of photography. In Proceedings of the 5th Nordic
conference on Human-computer interaction: building bridges (pp. 499-502). ACM.
25. Mark, G., Wang, Y., & Niiya, M. (2014). Stress and multitasking in everyday college life: an empirical study of
online activity. In Proceedings of the SIGCHI conference on human factors in computing systems (pp. 41-50).
ACM.
26. McCrae, R. R., & Costa Jr, P. T. (1999). A five-factor theory of personality. Handbook of personality: Theory
and research, 2, 139-153.
27. McDuff, D., Karlson, A., Kapoor, A., Roseway, A., & Czerwinski, M. (2012). AffectAura: an intelligent system
for emotional memory. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp.
849-858). ACM.
28. Miller, A. D., & Edwards, W. K. (2007). Give and take: a study of consumer photo-sharing culture and practice.
In Proceedings of the SIGCHI conference on Human factors in computing systems (pp. 347-356). ACM.
29. Munson, S. A., Lauterbach, D., Newman, M. W., & Resnick, P. (2010). Happier together: integrating a
wellness application into a social network site. In Persuasive Technology (pp. 27-39). Springer Berlin Heidelberg.
30. Nass, C., Jonsson, I. M., Harris, H., Reaves, B., Endo, J., Brave, S., & Takayama, L. (2005). Improving
automotive safety by pairing driver emotion and car voice emotion. In CHI’05 Extended Abstracts on Human
Factors in Computing Systems (pp. 1973-1976). ACM.
Chen et al. Page 16 of 16
31. National Institute of Mental Health. Depression of College Students. Retrieved September 21, 2015 from
http://www.nimh.nih.gov/health/publications/depression-and-college-students/index.shtml.
32. National Institute of Mental Health. Fact Sheet on Stress. Retrieved September 21, 2015 from
http://www.nimh.nih.gov/health/publications/stress/index.shtml.
33. Picard, R. W., & Picard, R. (1997). Affective computing (252). Cambridge: MIT press.
34. Russell, J. (1980). A circumplex model of affect. Journal of personality and social psychology 39(6): 1161.
35. Ryff, C. D., & Keyes, C. L. M. (1995). The structure of psychological well-being revisited. Journal of
personality and social psychology, 69(4), 719.
36. Saltz, J. (2014). Art at arm’s length: A history of the selfie. New York Magazine, 47(2), 71-75.
37. Seligman, M. E., Steen, T. A., Park, N., & Peterson, C. (2005). Positive psychology progress: empirical
validation of interventions. American psychologist, 60(5), 410.
38. Seppala, E., & Tomasello, M. (2013). The compassionate mind. The Observer, 26(6).
39. Sin, N. L., & Lyubomirsky, S. (2009). Enhancing well-being and alleviating depressive symptoms with positive
psychology interventions: A practice-friendly meta-analysis. Journal of clinical psychology, 65(5), 467-487.
40. Spire. Retrieved September 22, 2015 from https://www.spire.io.
41. Strauss, A., & Corbin, J. (1994). Grounded theory methodology. Handbook of qualitative research, 273-285.
42. The American College Health Association. 2014. National College Health Assessment. Retrieved September 21,
2015 from
http://www.acha-ncha.org/docs/ACHA-NCHA-II ReferenceGroup ExecutiveSummary Spring2014.pdf.
43. Tsujita, H., & Rekimoto, J. (2011). Smiling makes us happier: enhancing positive mood and communication
with smile-encouraging digital appliances. In Proceedings of the 13th international conference on Ubiquitous
computing (pp. 1-10). ACM.
