Journal of Research in Personality 69 (2017) 96–101Contents lists available at ScienceDirect
Journal of Research in Personality
journal homepage: www.elsevier .com/ locate/ j rpPutting mood in context: Using smartphones to examine how people feel
in different locationshttp://dx.doi.org/10.1016/j.jrp.2016.06.004
0092-6566/
 2016 The Authors. Published by Elsevier Inc.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
⇑ Corresponding author.
E-mail address: gsands@essex.ac.uk (G.M. Sandstrom).Gillian M. Sandstrom a,⇑, Neal Lathia b, Cecilia Mascolo b, Peter J. Rentfrow b
aUniversity of Essex, United Kingdom
bUniversity of Cambridge, United Kingdom
a r t i c l e i n f o a b s t r a c tArticle history:
Received 6 October 2015
Revised 29 March 2016
Accepted 4 June 2016
Available online 8 June 2016
Keywords:
Personality
Big five
Person-situation
Location
Context
Within-person
Mood
Positive affect
Mobile sensing
Multi-level modellingDoes personality predict how people feel in different types of situations? The present research addressed
this question using data from several thousand individuals who used a mood tracking smartphone
application for several weeks. Results from our analyses indicated that people’s momentary affect was
linked to their location, and provided preliminary evidence that the relationship between state affect
and location might be moderated by personality. The results highlight the importance of looking at
person-situation relationships at both the trait- and state-levels and also demonstrate how smartphones
can be used to collect person and situation information as people go about their everyday lives.

 2016 The Authors. Published by Elsevier Inc. This is an openaccess article under the CCBY license (http://
creativecommons.org/licenses/by/4.0/).1. Introduction
We know that personality is linked to behavior. Several studies
have shown, for example, that personality is linked to preferences
for and success in various occupations (Judge, Higgins, Thoresen, &
Barrick, 1999; Lodi-Smith and Roberts, 2007), maintaining satisfy-
ing intimate relationships (Ozer & Benet-Martinez, 2006; Roberts,
Kuncel, Shiner, Caspi, & Goldberg, 2007), and how people choose
to spend their free time (Mehl, Gosling, & Pennebaker, 2006;
Rentfrow, Goldberg, & Zilca, 2011). Our understanding of these
links is informed by interactionist theories, which argue that indi-
viduals seek out and create environments that satisfy and reinforce
their psychological needs. An implication of this argument is that
individuals experience higher positive affect and lower negative
affect when in their preferred environments. Drawing on past
research on person-environment interactions and using experience
sampling and mobile sensing technology, the present research
investigated whether personality traits moderate the associations
between state affect and locations, which are associated with
different types of situations.Personality is linked not only to behavior, but also to affect. In
one study, situational characteristics (social vs. non-social context)
and personality traits (Extraversion and Neuroticism) both pre-
dicted state positive and negative affect (Pavot, Diener, & Fujita,
1990). People experienced greater positive affect both when they
were high in Extraversion, and when they were in social situations,
but there was no interaction: people with all levels of Extraversion
experienced more positive affect in social situations. In a related
study, researchers found that positive affect (but not negative
affect) follows a diurnal rhythm, with people socializing, laughing
and singing more and more for the first 8–10 h after waking, and
then doing those activities less and less (Hasler, Mehl, Bootzin, &
Vazire, 2008). Further, this study found preliminary evidence that
this diurnal cycle of positive affect may be amplified for people
high in Extraversion. These findings have implications for within-
person variation in personality, given that affect is assumed to
mediate the influence of the situation on personality states
(Mischel & Shoda, 1995).
Although there is agreement that features of the environment
affect how people think, feel, and behave, there is less agreement
on which aspects of the environment have psychological implica-
tions (Fleeson & Noftle, 2008; Rauthmann et al., 2014). The physi-
cal environment, including location, is one objective characteristic
G.M. Sandstrom et al. / Journal of Research in Personality 69 (2017) 96–101 97often associated with a situation (Rauthmann et al., 2014; Saucier,
Bel-Bahar, & Fernandez, 2007). The locations people regularly visit
may be consistently associated with a constellation of factors (e.g.,
affect, sociability, recreation, goal pursuit), and thus represent
types of situations that have psychological implications, and clear
links to personality For example, the recently developed DIA-
MONDS taxonomy identifies several characteristics of situations
that might feasibly be linked to locations (Rauthmann et al.,
2014); work might be high in Duty and Intellect, and social places,
such as restaurants and bars, might be high in Sociality and pOsi-
tivity. These situational factors are connected to personality-
related behaviors (Rauthmann et al., 2014), suggesting that loca-
tions should be as well.
A challenge in studying affect as it is experienced in the various
types of situations that people encounter in daily life is the
repeated collection of data on mood and situation type. Method-
ological advances such as experience sampling have made it possi-
ble to collect repeated self-reports (e.g., of affect) as people go
about their daily lives. However, to date it has been difficult to
simultaneously collect objective information about the type of sit-
uation in which people find themselves. One exception is a recent
paper that used repeated experience sampling to examine the rela-
tionship between a person’s personality traits, the types of situa-
tions they encountered, and state expressions of personality, all
of which were self-reported (Sherman, Rauthmann, Brown,
Serfass, & Jones, 2015). In this study, state personality (as mani-
fested in behavior and emotions) was independently predicted
by both personality traits and situation characteristics.
The advent of mobile sensing technology provides a potential
solution to the challenge of collecting repeated information about
both behaviors and situations: detect the type of situation using
the sensors built-into today’s ubiquitous smartphones. These
devices come equipped with location sensors, an accelerometer
that can detect a user’s physical activity, a microphone that can
detect ambient noise in the environment, and various other sen-
sors. The potential is great, but little research to date has made
use of sensed information to examine psychologically relevant
questions.
In the present research, we explored the relationship between
state affect and location, which can be thought to represent a type
of situation. Our objective was to lay a foundation of preliminary
knowledge in the under-explored domain of associations between
situation types and state affect.1 The eighth stage, related to personality, is reported in additional analyses in
Appendix 1.2. Methods
2.1. Participants
Participants were members of the general public who down-
loaded the free app from the Google Play store and installed it on
their Android phone. The analyses reported herein include all users
who provided data on the measures of interest (described below)
from February 2013, when the app was released, to July 2015,
when we began the analyses.
A total of 12,310 users provided relevant momentary self-
reports of location (i.e., they reported being at home, at work, or
in a social type of situation). Of the users who reported demo-
graphics (N = 10,889), 44% of people who reported their gender
were female, 71% of people who reported their ethnicity reported
being White, and the most common birth year ranges were
1980–1989 (38%), 1990–1999 (32%) and 1970–1979 (18%).
Given that these users may or may not have provided trait or
state self-reports of personality, and may or may not have provided
location sensor data (see Section 2.2 for why this is the case), the
analyses described in the results section include different subsetsof these users. Ethical clearance has not been granted for sharing
the data supporting this publication.
2.2. The Emotion Sense application
Emotion Sense is a smartphone application that was designed
to study subjective well-being and behavior. The app collects
self-report data through surveys presented on the phone via expe-
rience sampling. By default, the app sends two notifications at ran-
dom moments of the day between 8 AM and 10 PM, at least
120 min apart from one another. Clicking on a notification
launches a momentary assessment, which includes measures of
current affect, and measures assessing a single aspect of current
behavior or context (e.g., location, physical activity, social interac-
tions). In addition to the notification-driven surveys, the app also
collects self-initiated surveys. These included longer measures of
affect, and measures assessing multiple aspects of behavior and
context.
As well as collecting self-report data, the app also uses open-
sourced software libraries (Lathia, Rachuri, Mascolo & Roussos,
2013) to periodically collect behavioral and contextual data from
sensors in the phone. The data collected through the app is stored
on the device’s file system and then uploaded to a server when the
phone is connected to a Wi-Fi hotspot.
Emotion Sense was designed to be a tool to facilitate self-
insight, providing feedback about how participants’ mood relates
to context and activity. In an effort to maintain user engagement
over a period of weeks, participants could receive additional feed-
back by ‘‘unlocking” stages, in the same way that players can
unlock different levels of a game after achieving certain objectives.
Each stage had a particular theme (e.g., location, physical activity)
that determined which behavior and context questions (e.g.,
‘‘Where are you right now?”, ‘‘Compared to most days, how phys-
ically active have you been today?”) were asked in the self-report
surveys. The second stage, related to location, is the only stage
reported in these results.1
2.3. Measures
2.3.1. Trait personality
Users reported their personality on the Ten-Item Personality
Inventory (Gosling, Rentfrow, & Swann, 2003). They rated the
extent to which ten pairs of words (e.g., ‘‘Extraverted, Enthusias-
tic”) applied to them, on a scale from 1 = Disagree strongly to
7 = Agree strongly.
2.3.2. Affect
Emotion Sense allows users to track and quantify their psycho-
logical well-being in various ways. On each self-report survey,
whether notification-driven or self-initiated, users indicate their
current feelings by tapping on a two-dimensional affect grid (see
Fig. 1), where the x-axis denotes valence, from negative to positive,
and the y-axis denotes arousal, from sleepy to alert (Russell, Weiss,
& Mendelsohn, 1989).
2.3.3. Location
One way Emotion Sense assesses current location is through
place self-reports. Users respond to the question ‘‘Where are you
right now?”, indicating whether they are at ‘‘Home,” ‘‘Work,”
‘‘Family/Friend’s House,” ‘‘Restaurant/Café/Pub,” ‘‘In transit,” or
‘‘Other” (see Fig. 2). We treated both ‘‘Family/Friend’s House” and
‘‘Restaurant/Café/Pub” as social types of situations, and did not
Fig. 1. Measuring mood in Emotion Sense using the affect grid: Users select a point
on a grid that quantifies valence (horizontally) and arousal (vertically).
Fig. 2. An example of how users self-report their location.
98 G.M. Sandstrom et al. / Journal of Research in Personality 69 (2017) 96–101examine ‘‘In Transit” or ‘‘Other” reports. These locations were
expected to be ones where people spent much of their time, and
visited over and over again.
Location is also sensed via the phone’s location sensors. For
15 min before each survey notification and at various intervals
throughout the day, Emotion Sense determines the location of
the phone by collecting the latitude, longitude, accuracy, speed,
and bearing of the device from its location sensors. This is a coarse
measure of location, with accuracies ranging from the tens to hun-
dreds of meters, but measuring location more accurately consumes
far more battery power.
A three-step process allowed us to match self-report data (e.g.,
state affect and state personality) to places (i.e., home, work,
social). First, we mapped each geographical location to a place by
connecting the location sensor data to the location self-reports.
Next, we mapped each self-report (e.g., state affect and state per-
sonality) to a geographic location by connecting the self-report to
the temporally closest location sensor data. Finally, we mapped
these self-report geographic locations to places using the mappings
created in the first step. (See Appendix 1, Section 2.4 for more
details.)3. Results
We investigated whether a person’s mood is related to the loca-
tion they are in, and how that relationship is moderated by trait
personality.3.1. How is affect related to location?
We created two sets of dummy codes: one set with Home as the
reference group (to compare mood at Home vs. at Work, and at
Home vs. in a Social type of situation), and the other set with Work
as the reference group (to compare mood at Work vs. in a Social
type of situation). We ran hierarchical linear modelling, using the
lme4 package in R (Bates, Maechler, Bolker, & Walker, 2014), pre-
dicting mood from the place dummy codes (level 1), grouped by
user (level 2). This analysis was carried out for both the location
self-reports, and the locations sensed by the phone sensors. We
report results for grid valence, but see Appendix 1 for analyses
examining grid arousal and alternate measures of high and low
arousal positive affect and negative affect. Given that degrees of
freedom are not provided by the lme4 package, we approximated
them from the number of level 2 units (i.e., users) minus the
number of predictors in the level 2 equation, not including the
intercept, minus 1 (Raudenbush & Bryk, 2002).3.1.1. Self-reported location
Of the 12,310 users who provided momentary self-reports of
location, 6759 reported their mood in at least two of the three
targeted locations and were used in these analyses. These users
provided, on average, 23 reports of location (from 2 to 305 reports,
median = 18).
The results from our analyses indicated that participants expe-
rienced a more positive mood when they reported being in a social
type of situation (M = 0.34, SD = 0.18) vs. at home (M = 0.20,
SD = 0.31), b = 0.29, t(6,758) = 38.84, p < 0.001, and when they
reported being at home vs. at work (M = 0.18, SD = 0.25),
b = 
0.08, t(6,758) = 
14.18, p < 0.001. (NOTE: Means and standard
deviations were computed for each type of location for each per-
son, and then averaged across people. They are unstandardized.)
For more nuanced results, showing differences between high
arousal and low arousal positive and negative emotions, see
Appendix 1.
G.M. Sandstrom et al. / Journal of Research in Personality 69 (2017) 96–101 993.1.2. Sensed location
Of the 12,310 users who provided momentary self-reports of
location, 7582 also provided relevant momentary self-reports of
mood that could be linked up to location readings from the loca-
tion sensor with some degree of accuracy (see Appendix 1, Sec-
tion 2.4 for details). Of these, 3646 users reported their mood in
at least two of the three targeted locations and were used in these
analyses. These users provided, on average, 111 reports of mood
(from 2 to 1117 reports, median = 88).
Consistent with the results for self-reported locations, users
reported a more positive mood when the phone sensors detected
that they were in a social type of situation (M = 0.34, SD = 0.17)
vs. at home (M = 0.23, SD = 0.32), b = 0.20, t(3,645) = 37.22,
p < 0.001, and when the phone sensors detected that they were
at home vs. at work (M = 0.20, SD = 0.27), b = 
0.11, t(3,645)
= 
33.97, p < 0.001. For more nuanced results, showing differences
between high arousal and low arousal positive and negative
emotions, see Appendix 1.
3.2. Does trait personality moderate the relationship between affect
and location?
Next we tested whether an individual’s personality traits mod-
erated the relationship between their location and their state
affect. For example, do people high in Extraversion experience
more positive affect in social types of situations compared to peo-
ple low in the trait? We created two datasets: (1) a dataset that
had only reports from home and work, and (2) a dataset that had
only reports from home and in social types of situations. Then
we ran regressions predicting state affect from the interactions
between location (dummy coded as before) and each of the big five
personality traits (standardized across users), entered simultane-
ously. When we found significant interactions involving personal-
ity, we examined the simple slopes for low and high values of the
personality trait at 
1 and +1 SD from the mean.
3.2.1. Self-reported location
Of the 6759 users who provided momentary self-reports of
location and reported their mood in at least two of the three tar-
geted locations, 1434 also self-reported trait personality and were
included in these analyses.
The more positive mood associated with being at home vs.
being at work was moderated by Openness, b = 
0.04, t(1,428)
= 
2.70, p = 0.01, and Agreeableness, b = 0.04, t(1,428) = 3.08,
p = 0.002 (see Appendix 1 for full model statistics). The mood ben-
efits associated with being at home vs. being at work were larger
for people who were high in Openness, b = 
0.14, t(1,428) =

7.24, p < 0.001, than for people low in Openness, b = 
0.06,
t(1,428) = 
2.92, p = 0.004. The mood benefits associated with
being at home vs. being at work were larger for people who were
low in Agreeableness, b = 
0.14, t(1,428) = 
7.16, p < 0.001, than
for people who were high in Agreeableness, b = 
0.06, t(1,428)
= 
3.32, p < 0.001.
The more positive mood associated with being in a social type
of situation vs. being at home was moderated by Neuroticism,
b = 0.03, t(1,428) = 2.02, p = 0.04. The mood benefits associated
with being in a social type of situation vs. being at home were lar-
ger for people who were high in Neuroticism, b = 0.35, t(1,428)
= 15.42, p < 0.001 than for people who were low in Neuroticism,
b = 0.28, t(1,428) = 11.37, p < 0.001.
3.2.2. Sensed location
Of the 3646 users who reported their mood in at least two of the
three targeted locations and provided momentary self-reports of
mood that could be linked to location readings from the locationsensor, 651 also self-reported trait personality and were included
in these analyses.
As with the self-reported locations, the more positive mood
associated with being at home vs. being at work was moderated
by Openness, b = 0.05, t(645) = 3.64, p < 0.001. Openness showed
the opposite pattern to the self-reported location data: the mood
benefits associated with being at home vs. being at work were
larger for people who were low in Openness, b = 
0.18,
t(645) = 
10.15, p < 0.001, than for people high in Openness,
b = 
0.09, t(645) = 
4.73, p < 0.001.
The more positive mood associated with being in a social type
of situation vs. being at home was moderated by Extraversion,
b = 
0.04, t(645) = 
1.98, p = 0.05, and Agreeableness, b = 
0.05,
t(645) = 
2.34, p = 0.02. The mood benefits associated with being
in a social type of situation vs. being at home were larger for people
who were low in Extraversion, b = 0.23, t(645) = 8.04, p < 0.001
than for people who were high in Extraversion, b = 0.15, t(645)
= 5.60, p < 0.001. The mood benefits associated with being in a
social type of situation vs. being at home were larger for people
who were low in Agreeableness, b = 0.24, t(645) = 8.04, p < 0.001,
than for people who were high in Agreeableness, b = 0.15,
t(645) = 5.70, p < 0.001.4. General discussion
A person’s momentary mood fluctuated in relation to their loca-
tion (i.e., at home, at work, and in social types of situations. We
used both self-reported location, and location sensed via smart-
phone sensors and found remarkably consistent results (on grid
valence, which was reported in the main text, but also on alternate
measures of high and low arousal positive and negative affect,
which were reported in Appendix 1). People reported more posi-
tive affect in social types of situations than at home or work. Fur-
ther, high arousal positive and negative affect were reported more
at work than at home, whereas low arousal positive and negative
affect were reported more at home than at work.
We found some evidence that the more positive mood associ-
ated with being in a social type of situation vs. being at home
was moderated by personality. The difference in mood reported
in social types of situations vs. at home was greater for people
who were low in Agreeableness, low in Extraversion or high in
Neuroticism. These results are consistent with the finding that
Extraversion and Agreeableness are positively related to high qual-
ity social relationships, whereas Neuroticism is negatively related
to high quality social relationships (Lopes, Salovey, & Straus,
2003; Lopes et al., 2004). Perhaps social types of situations are
associated with greater mood gains for those who generally strug-
gle more with their social relationships.
We also found some evidence that the more positive mood
associated with being at home vs. being at work was moderated
by personality. The difference in mood reported at home vs. at
work was greater for people who were low in Agreeableness. This
might suggest that people who are low in Agreeableness struggle
more in their forced interactions with colleagues at work
(thus experiencing more positive affect at home vs. at work),
whereas they thrive in their chosen interactions with friends in
social types of situations (thus experiencing more positive affect
in social types of situations vs. at home).
These moderation results should be interpreted with caution for
several reasons. First, though significant, these effects were small, |
b|’s 6 0.05. Second, the results were not always the same for the
self-reported locations as they were for the sensed locations. For
example, the difference in mood reported in social types of
situations vs. at home was larger for people high in Openness when
the location was self-reported, but larger for people low in
100 G.M. Sandstrom et al. / Journal of Research in Personality 69 (2017) 96–101Openness when the location was sensed. Finally, these results con-
flict with past research, which found additive effects of trait per-
sonality and situation characteristics on expressed behavior and
emotion, but no interactions (Sherman et al., 2015). Further
research is necessary to determine whether these results are
reliable.
The current results are consistent with an average layperson’s
intuition: when people are asked to reflect on the factors that affect
their expression of traits, they often mention the location (Saucier
et al., 2007). Although the physical characteristics of a location may
be directly related to the expression of traits (e.g., people may
express more Neuroticism in unfamiliar environments, or less
Extraversion in green spaces), it is likely that the psychological
characteristics that become associated with a particular location
are actually driving the relationship. For instance, being at home
may be associated with spending time with family, and taking on
a more communal role, whereas being at work may be associated
with competition, and taking on a more agentic role. Rauthmann
et al. (2014) proposed eight situational factors with psychological
consequences: Duty, Intellect, Adversity, Mating, pOsitivity, Nega-
tivity, Deception, Sociality. We expect that the three locations we
examined varied in many of these factors (e.g., Intellect may be
more of a factor at work than at home or in social types of situa-
tions, whereas Sociality may be more of a factor in social types of
situations than at work or at home); these assumptions could, of
course, be tested in a future study. Future work is needed to deter-
mine which characteristics of a situation are active in the locations
that we examined, but also in locations more broadly. This knowl-
edge about what types of locations are likely to present situations
high in each type of characteristic would allow future mobile sens-
ing studies to sample a more complete set of locations. The useful-
ness of mobile sensing for testing questions about person-
environment interactions will depend on its ability to sample from
locations with a wide range of situational characteristics, and its
ability to accurately label locations according to their characteris-
tics with as little user intervention and training as possible.
Smartphone sensors provided a powerful means of passively
detecting the type of situation a user is encountering, and reducing
the burden of self-reports in the current study. Their usefulness in
future studies may depend on how capable they are of detecting
other psychologically relevant situational factors. In this study,
we focused on using location sensors to learn the semantics of
places, so that we could examine relationships between place,
affect, and personality. Collecting data from other sensors could
augment this analysis, as well as provide other signals that may
characterize the type of situation that a user is encountering. For
example, collecting Wi-Fi scans can be used for finer-grained,
indoor localisation (Gao et al., 2011), and bluetooth sensors can
be used to detect co-located bluetooth devices of other study par-
ticipants (Eagle & Pentland, 2006). Smartphones’ microphones can
be used to measure ambient noise (Lathia, Rachuri, Mascolo, &
Rentfrow, 2013); with sufficient training data, these can even be
used to analyze participants’ speech (Rachuri et al., 2010). Recent
work has shown that the ambiance of a place can also be derived
from photographs taken there (Redi, Quercia, Graham, & Gosling,
2015). Work is needed to establish the validity of data from various
sensors to assess psychologically meaningful situational
characteristics.
One way to establish the validity of sensor data is to compare it
to self-reports, as we did in this study. However, various decisions
need to be made in order to sample sufficient data from the sensors
at the right times and at the right level of detail. In this study, the
place label (home, work, or social) given to a particular geographic
location (latitude, longitude pair) was inferred by matching
self-reported location to location sensor data (see Appendix,
Section 2.4). This process has a number of limitations. First, if aparticipant has never self-reported their location in a particular
geographic area, we cannot make any inferences about that place.
While there are methods that attempt to infer place semantics
based on other signals (e.g., time of day/day of week), we limited
our analysis to those places that were explicitly labelled. We
matched location samples with place self-reports based on a fixed
temporal threshold of 15 min; reducing this constant may improve
the accuracy of location inferences, at the expense of reducing the
number of labelled self-reports. Finally, we assumed that a user
was in the same place if the two location samples were less than
1600 m from one another. In highly dense urban environments,
this may not necessarily be the case.
Even if phone sensors are shown to measure situational charac-
teristics with some level of validity, work is needed to understand
the capabilities and limitations of the sensors. We used location
data that was passively captured from participant’s devices in
order to infer where they were at the time of reporting their state
personality or their mood. While location sensors are broadly reli-
able (indeed, they are used for mapping applications), they do not
work when a user has disabled their device’s location services, and
they do not work in all types of situations, such as underground or,
in some cases, indoors. Obtaining a highly accurate and recent
location sample is possible, but it is a time- and battery-
consuming task, since it requires turning on the GPS sensor, wait-
ing for it to obtain a satellite fix, and then collecting the data.
Instead, we resorted to collecting coarse-grained data, which is
easier and quicker to sample and more energy-efficient, but may
be less accurate.
The current results bolster the call for further research of
within-person variation in affect and personality. We found evi-
dence that people experience fluctuations in state affect when they
are in different locations. As research continues to identify the fac-
tors of a situation that are psychologically meaningful, future work
will be able to investigate the interactive effects of various situa-
tional factors on people’s feelings and behavior. The combination
of experience sampling and mobile sensing allowed us to collect
large amounts of within-person data in the current work, and
could prove invaluable to further study, to the extent that phone
sensors can objectively measure psychologically active situational
variables. The current study describes how people feel and behave
in different locations, but as a field we are just beginning to learn
about the full extent to which feelings and behavior vary by type
of situation.
Acknowledgements
We thank K. Rachuri for his contributions to the open source
software (accessible via https://github.com/emotionsense), and G.
Harari for discussing plans for analysis. This research was sup-
ported by the UK Engineering and Physical Sciences Research
Council (Ubiquitous and Social Computing for Positive Behaviour
Change, Ref. EP/I032673/1).
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jrp.2016.06.004.
References
Bates, D., Maechler, M., Bolker, B., & Walker, S. (2014). lme4: Linear mixed-effects
models using Eigen and S4. R package version 1.1-10.
Eagle, N., & Pentland, A. (2006). Reality mining: Sensing complex social systems.
Personal and Ubiquitous Computing, 10(4), 255–268.
Fleeson, W., & Noftle, E. (2008). The end of the person–situation debate: An
emerging synthesis in the answer to the consistency question. Social and
Personality Psychology Compass, 2(4), 1667–1684.
G.M. Sandstrom et al. / Journal of Research in Personality 69 (2017) 96–101 101Gao, Y., Yang, Q., Li, G., Chang, E. Y., Wang, D., Wang, C., Qu, H., ... Zhang, F. (2011).
XINS: The anatomy of an indoor positioning and navigation architecture. In
Proceedings of the 1st international workshop on mobile location-based service
(pp. 41–50). ACM.
Gosling, S. D., Rentfrow, P. J., & Swann, W. B. (2003). A very brief measure of the Big-
Five personality domains. Journal of Research in Personality, 37(6), 504–528.
Hasler, B. P., Mehl, M. R., Bootzin, R. R., & Vazire, S. (2008). Preliminary evidence of
diurnal rhythms in everyday behaviors associated with positive affect. Journal of
Research in Personality, 42(6), 1537–1546.
Judge, T. A., Higgins, C. A., Thoresen, C. J., & Barrick, M. R. (1999). The big five
personality traits, general mental ability, and career success across the life span.
Personnel Psychology, 52(3), 621–652.
Lathia, N., Rachuri, K. K., Mascolo, C., & Rentfrow, P. J. (2013). Contextual
dissonance: Design bias in sensor-based experience sampling methods. In
Proceedings of the 2013 ACM international joint conference on pervasive and
ubiquitous computing (pp. 183–192). ACM. September.
Lathia, N., Rachuri, K., Mascolo, C., & Roussos, G. (2013). Open source smartphone
libraries for computational social science. In Proceedings of the 2013 ACM
conference on Pervasive and ubiquitous computing adjunct publication
(pp. 911–920). ACM. September.
Lodi-Smith, J., & Roberts, B. W. (2007). Social investment and personality: A meta-
analysis of the relationship of personality traits to investment in work, family,
religion, and volunteerism. Personality and Social Psychology Review, 11(1),
68–86.
Lopes, P. N., Brackett, M. A., Nezlek, J. B., Schütz, A., Sellin, I., & Salovey, P. (2004).
Emotional intelligence and social interaction. Personality and Social Psychology
Bulletin, 30(8), 1018–1034.
Lopes, P. N., Salovey, P., & Straus, R. (2003). Emotional intelligence, personality, and
the perceived quality of social relationships. Personality and Individual
Differences, 35(3), 641–658.
Mehl, M. R., Gosling, S. D., & Pennebaker, J. W. (2006). Personality in its natural
habitat: Manifestations and implicit folk theories of personality in daily life.
Journal of Personality and Social Psychology, 90(5), 862–877.
Mischel, W., & Shoda, Y. (1995). A cognitive-affective system theory of personality:
Reconceptualizing situations, dispositions, dynamics, and invariance in
personality structure. Psychological Review, 102(2), 246–268.Ozer, D. J., & Benet-Martinez, V. (2006). Personality and the prediction of
consequential outcomes. Annual Review of Psychology, 57, 401–421.
Pavot, W., Diener, E., & Fujita, F. (1990). Extraversion and happiness. Personality and
Individual Differences, 11(12), 1299–1306.
Rachuri, K. K., Musolesi, M., Mascolo, C., Rentfrow, P. J., Longworth, C., & Aucinas, A.
(2010). EmotionSense: A mobile phones based adaptive platform for
experimental social psychology research. In Proceedings of the 12th ACM
international conference on ubiquitous computing (pp. 281–290). ACM.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and
data analysis methods (Vol. 1) : . Sage.
Rauthmann, J. F., Gallardo-Pujol, D., Guillaume, E. M., Todd, E., Nave, C. S., Sherman,
R. A., ... Funder, D. C. (2014). The Situational Eight DIAMONDS: A taxonomy of
major dimensions of situation characteristics. Journal of Personality and Social
Psychology, 107(4), 677–718.
Redi, M., Quercia, D., Graham, L. T., & Gosling, S. D. (2015). Like partying? Your face
says it all. Predicting the ambiance of places with profile pictures arXiv
preprintarXiv:1505.07522.
Rentfrow, P. J., Goldberg, L. R., & Zilca, R. (2011). Listening, watching, and reading:
The structure and correlates of entertainment preferences. Journal of Personality,
79(2), 223–258.
Roberts, B. W., Kuncel, N. R., Shiner, R., Caspi, A., & Goldberg, L. R. (2007). The power
of personality: The comparative validity of personality traits, socioeconomic
status, and cognitive ability for predicting important life outcomes. Perspectives
on Psychological Science, 2(4), 313–345.
Russell, J. A., Weiss, A., & Mendelsohn, G. A. (1989). Affect grid: A single-item scale
of pleasure and arousal. Journal of Personality and Social Psychology, 57(3),
493–502.
Saucier, G., Bel-Bahar, T., & Fernandez, C. (2007). What modifies the expression of
personality tendencies? Defining basic domains of situation variables. Journal of
Personality, 75(3), 479–504.
Sherman, R. A., Rauthmann, J. F., Brown, N. A., Serfass, D. G., & Jones, A. B. (2015). The
independent effects of personality and situations on real-time expressions of
behavior and emotion. Journal of Personality and Social Psychology, 109(5), 872.