Content uploaded by Juho Hamari
Author content
All content in this area was uploaded by Juho Hamari on Oct 12, 2017
Content may be subject to copyright.
‘‘Working out for likes’’: An empirical study on social influence
in exercise gamification
Juho Hamari, Jonna Koivisto
⇑
Game Research Lab, School of Information Sciences, FIN-33014 University of Tampere, Finland
article info
Article history:
Keywords:
Gamification
Social networking
Social influence
Continued use
eHealth
mHealth
abstract
Today, people use a variety of social and gameful (mobile) applications in order to motivate themselves
and others to maintain difficult habits such as exercise, sustainable consumption and healthy eating.
However, we have yet lacked understanding of how social influence affects willingness to maintain these
difficult habits with the help of gamification services. In order to investigate this phenomenon, we mea-
sured how social influence predicts attitudes, use and further exercise in the context of gamification of
exercise. Our results show that people indeed do ‘‘work out for likes’’, or in other words, social influence,
positive recognition and reciprocity have a positive impact on how much people are willing to exercise as
well as their attitudes and willingness to use gamification services. Moreover, we found that the more
friends a user has in the service, the larger the effects are. Furthermore, the findings of the empirical study
further provide new understanding on the phenomenon of social influence in technology adoption/use
continuance in general by showing, in addition to subjective norms, how getting recognized, receiving
reciprocal benefits and network effects contribute to use continuance.
Ó2015 Elsevier Ltd. All rights reserved.
1. Introduction
In their daily lives, people are often ridden with a tendency to
favour short-term rewards instead of long-term rewards. This cog-
nitive bias, titled hyperbolic discounting (Ainslie, 1975), leads us to
sometimes neglect behaviours that would be beneficial to us and
consequently causes us to, for example, procrastinate, skip exer-
cise, smoke, and overconsume. When trying to break these cycles,
a strong willpower is not always enough, and therefore, people are
constantly seeking for novel ways to motivate themselves. During
the last couple of years, new technological approaches for these
motivational problems have been introduced. For example applica-
tions for fitness (Fitocracy; Zombies, Run!), housekeeping (Chore
Wars), and even keeping up with one’s aspirations in life
(Mindbloom) all attempt to motivate people by restructuring rela-
tively long-term goals by providing the users with short-term
goals, activities, rewards and social support.
This emerging technological approach for motivating people
toward different types of beneficial behaviours draws from the
design of social network services as well as games and has com-
monly been titled as gamification which refers to implementation
of elements familiar from games to create similar experiences as
games commonly do (Deterding, Dixon, Khaled, & Nacke, 2011;
Hamari, Huotari, & Tolvanen, 2015). Such features have thus far
been implemented in various contexts (Hamari, Koivisto, & Sarsa,
2014). Furthermore, very positive views and perhaps even unwar-
ranted expectations regarding gamification have been expressed
(see e.g. IEEE, 2014). However, doubts have also been cast on the
concept and its effectiveness in truly motivating people (Gartner,
2012). Thus far, meta-studies indicate that most studies do report
positive findings from gamification implementations. However,
understanding over what kind of gamification works, which psy-
chological aspects mediate the effects, and in which contexts the
approach can be beneficial is not yet sufficient (Hamari et al.,
2014). Nevertheless, the amount of research on the topic is rapidly
increasing (Hamari et al., 2014), and to further highlight the time-
liness of these developments, business analyses by Gartner (2011)
and IEEE (2014) have reported predictions that the number of gam-
ification endeavours will be increasing considerably in the coming
years.
Moreover, common to many such motivational applications is
the attempt to employ social influence through a user community
in order to entice people to maintain their sustainable behaviour.
The generally increased use of social features in technologies can
also be observed elsewhere. People adopt technologies increas-
ingly through word-of-mouth (Cheung & Thadani, 2012) or differ-
ent kinds of recommendation systems (Li, Wu, & Lai, 2013; Stibe,
Oinas-Kukkonen, & Lehto, 2013; Xiao & Benbasat, 2007) as well
as consume socially (Zhou, Zhang, & Zimmermann, 2013).
http://dx.doi.org/10.1016/j.chb.2015.04.018
0747-5632/Ó2015 Elsevier Ltd. All rights reserved.
⇑
Corresponding author. Tel.: +358 50 318 73 63.
E-mail address: jonna.koivisto@uta.fi (J. Koivisto).
Computers in Human Behavior 50 (2015) 333–347
Contents lists available at ScienceDirect
Computers in Human Behavior
journal homepage: www.elsevier.com/locate/comphumbeh
Moreover, social networking services frequently expose people to
opinions and attitudes of others, which may further influence the
behaviour of the participants (see e.g. Zhou, 2011). While the num-
ber of technological approaches invoking social influence and
related psychological phenomena for steering human behaviour
towards sustainable, healthy, and otherwise beneficial behaviours
is growing, research-based knowledge on whether these
technological solutions with social features are able to actually
motivate people to pick up and continue with the encouraged
behaviours is still somewhat scarce.
Therefore, in this study we investigate how social influence aids
people in continuing and maintaining the beneficial behaviours
promoted by the gamification technology. We specifically focus
on one category of beneficial behaviour; namely physical exercise
and the gamification service devised to encourage such behaviour.
In particular, in this study we seek to magnify ‘social influence’ and
investigate how several social factors work in parallel to increase
willingness to use gamification and continue exercising. We com-
pose the social influence from four factors: (1) subjective norms,
(2) recognition from accepting the social influence, and (3) per-
ceived reciprocal benefits. As an antecedent to social influence
we measure (4) network effects (in order to investigate on which
aspects of social influence having a larger network affects). The
theorization expands upon the traditional measurement of social
influence by extending the widely employed models, the theories
of reasoned action (TRA) and planned behaviour (TPB). The study
employs data gathered through an online survey from the users
of an exercise-related gamification service.
2. Theory and background
2.1. Gamification
So far, the gamification approach (Deterding et al., 2011;
Hamari et al., 2015) has been harnessed and studied, for example,
in the domains of education (e.g. Bonde et al., 2014; Christy & Fox,
2014; de-Marcos, Domínguez, Saenz-de-Navarrete, & Pagés, 2014;
Denny, 2013; Domínguez et al., 2013; Farzan & Brusilovsky, 2011;
Filsecker & Hickey, 2014; Hakulinen, Auvinen, & Korhonen, 2013;
Simões, Díaz Redondo, & Fernández Vilas, 2013), commerce
(Hamari, 2013, 2015), intra-organizational communication and
activity (Farzan et al., 2008a, 2008b; Thom, Millen, & DiMicco,
2012), government services (Bista, Nepal, Paris, & Colineau,
2014), public engagement (Tolmie, Chamberlain, & Benford,
2014), environmental behaviour (Lee, Ceyhan, Jordan-Cooley, &
Sung, 2013; Lounis, Pramatari, & Theotokis, 2014), marketing and
advertising (Cechanowicz, Gutwin, Brownell, & Goodfellow, 2013;
Terlutter & Capella, 2013), and activities such as crowdsourcing
(Eickhoff, Harris, de Vries, & Srinivasan, 2012; Ipeirotis &
Gabrilovich, 2014), to name a few.
In addition to the above mentioned domains, several studies
have examined gamification in the context of this study: health
and exercise. Table 1 outlines findings made in the area of gam-
ification of exercise and health. The results of the studies indicate
positive effects from gamification, for example, on physical activity
(e.g. Chen & Pu, 2014; Chen, Zhang, & Pu, 2014), healthy eating
habits (e.g. Jones, Madden, & Wengreen, 2014), as well as willing-
ness to continue using the health-related system (e.g. Cafazzo,
Casselman, Hamming, Katzman, & Palmert, 2012; Elias, Rajan,
McArthur, & Dacso, 2013). However, some studies suggest that
novelty effects might affect the perceptions of benefits from the
gamification approaches (Koivisto & Hamari, 2014). Furthermore,
the findings from studies conducted in the domain of gamification
of health and exercise are somewhat in line with the results from
literature examining the use video games for health benefits. A
review by Biddiss and Irwin (2010) reported that inconclusive
results were found in terms of significantly increasing physical
activity. However, potential for effects from short-term interven-
tions was noted indicating that, similarly to gamification imple-
mentations, benefits may be derived especially in the short-term
due to novelty factors, but indications of long-term benefits are
still scarce.
Regarding social aspects in gamification of health and exercise
particularly, for example, Chen and Pu (2014) and Chen et al.
(2014) studied social features in an exercise gamification context
with the aim of increasing physical activity. They experimented
with social conditions of cooperation, competition and a hybrid
setting with features of both of the previous. In their studies, the
social conditions did increase physical activity when compared to
exercising alone. Of the conditions, the cooperation setting lead
to most positive effects. Furthermore, the findings of the controlled
trial study by Allam, Kostova, Nakamoto, and Schulz (2015) indi-
cated that a combination of gamification and social support fea-
tures implemented in their web-based intervention increased
physical activity. Therefore, the studies by Chen and Pu (2014),
Chen et al. (2014), and Allam et al. (2015) suggest that social
aspects and especially supportive social interactions could have
an important effect in motivating users towards behaviours with
gamification.
2.2. Social influence in social psychology
Human beings have a psychological need for experiencing relat-
edness (Deci & Ryan, 2000; Ryan & Deci, 2000), which refers to the
needs of belonging and being connected with others. When these
needs are fulfilled in a given context, the experienced relatedness
may increase intrinsic motivations toward activities related to that
context (Ryan & Deci, 2000; Zhang, 2008). In other words, for
example in the information technology context, experiencing relat-
edness through the use of a system potentially makes the user
more willing to engage with the system and continue using it
(Zhang, 2008).
One method for creating such senses of relatedness is by
organizing into groups. Groups form around mutual goals and
through different stages of group formation usually develop to
share mutual norms which are an important antecedent for group
cohesion (Tuckman, 1965). The process of group cohesion is
revealed in the tendency of a group of people to stay together
and pursue some instrumental objectives, and thus, reciprocally
benefit from the social community (Carron & Brawley, 2000).
Becoming a member of a social community may thus lead to
individuals being affected by the social influence from others.
Depending on whether the individual wishes to become part of
the social community, he or she may accept the social influence,
for example, the diffusion of the behavioural norms of the commu-
nity that are communicated through the process of signalling the
norms (Ajzen, 1988, 1991; Fishbein, 1979). Depending on whether
the individual accepts the norms, the social community provides
feedback to the individual on his or her behaviour (Hernandez,
Montaner, Sese, & Urquizu, 2011; Lin, 2008). In case the individual
has accepted the social influence and has received positive feed-
back from the relevant community, this may further lead to
satisfaction for the individual who is conforming and complying
with the norms (Kelman, 1958; Lin, 2008; Lindenberg, 2001).
Theories on group formation (Tuckman, 1965), relatedness
regarding emergence of intrinsic motivations (Ryan & Deci, 2000)
and social influence (Cialdini, Green, & Rusch, 1992; Kelman,
1958) suggest that social influence also includes the affective
experience derived from gaining recognition from accepting and
conforming with the social influence. The fulfilment of needs of
relatedness essential to intrinsic motivation requires a supportive
334 J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347
environment, where the individual becomes recognized and
accepted (Ryan & Deci, 2000). Furthermore, the recognition by
the social community emerges from acceptance, which may moti-
vate an individual to conform to the community’s expectations.
Moreover, recognition as a form of acceptance may increase the
cohesiveness of the group by, for example, increasing the attrac-
tiveness of the group or other group members (Lott & Lott, 1965;
McCauley, 1989), i.e. the positive attitudes the group members
have toward each other (Lott & Lott, 1965). A positive attitude
toward the group and its members evidently makes it more pleas-
ant to be part of the group, and in case of groups or communities
formed around a common interest, some level of attraction can
be expected to exist between the members of the group (Bonner,
1959; Lott & Lott, 1965). Members of cohesive groups, who have
positive attitudes toward each other, tend to conform to group
standards and produce uniform conduct with the other group
members (Hogg, 1992).
Consequently, positive attitudes toward other group members
may be strengthened by the received recognition (Cialdini &
Goldstein, 2004; Hogg, 1992), which can further lead to willingness
to reciprocate, to signal appreciation and acceptance toward others
in the group (Cialdini & Goldstein, 2004). Reciprocity, that is,
returning a favour or a positive action with another (Cialdini &
Goldstein, 2004; Cialdini et al., 1992), is a social drive, which has
Table 1
Works on gamification in the health and exercise domain.
Work Domain Context Results
Allam et al. (2015) Exercise/
health
Web-based intervention with gamification
and social features for rheumatoid arthritis
patients
– Compared to control group, physical activity increased over time for
experimental group with access to social support sections and gamification
– Compared to control group, health care utilization decreased significantly for
experimental group with access to social support features and group with access
to both social support features and gamification
– Compared to control group, experimental group with access to either social
support or the gamification gained more empowerment
– Compared to control group, experimental group with a gamified experience
used the website more often than the ones without gamification
Brauner, Calero Valdez,
Schroeder, and Ziefle
(2013)
Exercise Kinect-based exergame – Game performance not affected by performance motivation, but by gamer type
– Positive effect on perceived pain
Cafazzo et al. (2012) Health Gamified mHealth application for diabetes – Increase of 50% in daily average frequency of blood glucose measurement
– 88% (14/16 participants) stated that they would continue using the system
Chen and Pu (2014) Exercise Gamified mobile fitness application – 15% increase in physical activities when using the application compared to
exercising alone
– Cooperation and hybrid conditions in the group setting outperformed
competition
– More messages sent in cooperation condition
– Positive correlation between physical activities and number of messages
exchanged
Chen et al. (2014) Exercise Gamified mobile fitness application – The social settings (competition, accountability and hybrid) help users to
persist more in physical activity compared to baseline
– Hybrid setting more likely to motivate users to walk more and actively help
others
Elias et al. (2013) Health Gamified mobile application for asthma-care – Survey of children with asthma (N= 9) indicated they would play games as the
one studied if they involved breathing into a spirometer
– 6/9 preferred the game-based system over a spirometer alone, while 3/9
preferred having both
Hamari and Koivisto
(2014)
Exercise App/web-based gamification service for
exercise
– Components of flow pertaining to autotelic experience, clear goals, feedback,
control and challenges were most important
Hamari and Koivisto
(2015)
Exercise App/web-based gamification service for
exercise
– Utilitarian and social motivations have a positive direct association with
attitude towards exercise gamification
– Utilitarian motivations’ association with continued use of exercise
gamification is mediated by attitude towards it
– Hedonic motivations have a positive direct association with continued use of
exercise gamification
Jones et al. (2014) Health Gamification of fruit and vegetable
consumption
– Significant increase in fruit and vegetable consumption on intervention days
Koivisto and Hamari
(2014)
Exercise App/web-based gamification service for
exercise
– Perceived enjoyment and usefulness of gamification decline with use
– Women found to report greater social benefits from the use of gamification
Riva, Camerini, Allam,
and Schulz (2014)
Health Internet-based self-management with
interactive sections for chronic back pain
– Availability of interactive sections significantly increased patient
empowerment and reduced medication misuse in the intervention group
– Decrease in frequency of physical exercise and pain burden was reported, but
the decrease was equal in control and intervention groups
Thorsteinsen, Vittersø,
and Svendsen (2014)
Exercise Online, interactive physical activity
intervention
– Intervention group had significantly more minutes of physical activity on
registration weeks
– Intervention group had more intense physical activity on one registration
week
– No significant differences between control and intervention groups at the end
of the study
Watson, Mandryk, and
Stanley (2013)
Exercise Classroom exergame – Exercise and game components improved enjoyment and player experiences
of classroom activities
– Game elements improved proportion of correct answers on a retention test
J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347 335
been argued to exist in all human cultures (Burger, Sanchez,
Imberi, & Grande, 2009; Gouldner, 1960). When receiving positive
feedback, people often feel obligated by the social norms to return
the favour and reciprocate. Therefore, reciprocal benefits may arise
from group interactions, when receiving recognition is recipro-
cated by providing recognition. Through this type of process the
shared norms of a group may be further diffused and strengthened.
In addition to the processes of recognition and reciprocity, the
size of the relevant community (see e.g. Lin & Bhattacherjee,
2008) is likely to be an important factor for social influence, since
it is proportional to the amount of influence a person can be
exposed to. According to the theory of network externalities, the
network effects (i.e., the value from the network) arise when the
benefits from using the service depend on the number of relevant
other users (Katz & Shapiro, 1985; Lin & Bhattacherjee, 2008).
In sum, based on the theorizations in social psychology regard-
ing social influence, we propose conceptualizing the social influ-
ence via three main factors: (1) subjective norms, (2) recognition,
and (3) the resulting further reciprocal benefits. Further, the social
influence can be positively influenced by the proportion to which
the individual is exposed to other people in the relevant
community.
2.3. Social influence in technology adoption
Theories applied in studies investigating technology adoption,
use and attitudes towards systems, commonly employ social influ-
ence as one of the predictors for the outcome behaviour or beha-
vioural intention. Such theories are, for example, the theory of
reasoned action (TRA) and the theory of planned behaviour (TPB)
which was developed from the TRA. In the basic structures of the
TRA and TPB, the behavioural intentions to engage in a behaviour
are investigated by measuring determinants for the intentions:
attitude toward the behaviour, and subjective norms.
1
Using these
theories, research into social influence in technology adoption has
often operationalized this influence as subjective norms. These
norms refer to the beliefs and perceptions of how important the rele-
vant others of an individual regard a given behaviour and whether
they expect one to perform it (Ajzen, 1988, 1991; Fishbein, 1979;
Fishbein & Ajzen, 1975). The attitude toward the behaviour refers
to the behavioural beliefs of the outcomes or other features of the
behaviour that the individual attributes to it (Ajzen, 1991). These
outcomes are evaluated as positive or negative. Together, the beliefs
and their evaluations form the attitude, either positive or negative,
that the individual projects on the performance of the behaviour
(Ajzen, 1991).
Prior information systems literature has investigated the social
influence in a variety of contexts, such as organizational knowledge
sharing (Bock, Zmud, Kim, & Lee, 2005; Lewis, Agarwal, &
Sambamurthy, 2003), social networking services (Cheung, Chiu, &
Lee, 2011; Cheung & Lee, 2010), e-learning (Hernandez et al.,
2011), blogs (Hsu & Lin, 2008), and e-commerce (Hamari, 2013).
Table 2 reviews studies on social influence in technology adoption.
While most of the studies have employed the basic TRA/TPB models,
some studies have also extended the social influence with addi-
tional variables. Most of these extensions have enhanced the mea-
surement of mere subjective norms with variables concerned with
Table 2
Works on social influence in technology adoption literature.
Work Context Variables of social influence Outcome variables
Several studies of studies (e.g. Bock et al., 2005; Çelik, 2011; Hsieh
et al., 2008; Lewis et al., 2003; Pavlou & Fygenson, 2006;
Venkatesh & Davis, 2000; Venkatesh et al., 2003)
Various Subjective norms/social influence Attitudinal/behavioural
outcomes
Cheung et al. (2011) Online social
networks
Subjective norm, group norm, social identity We-intention
Cheung and Lee (2010) Online social
networking
site
Subjective norm, group norm, social identity
(SI), cognitive SI, affective SI, evaluative SI
We-intention to use an online
social networking site
Shen, Cheung, and Lee (2013) Instant
messaging
Subjective norm, group norm, social identity We-intention
Baker and White (2010) Social
networking
site
Subjective norm, group norm Engagement in frequent social
networking site use
Hernandez et al. (2011) E-learning Sense of community, social influence,
altruism, recognition by peers, recognition by
the instructor
Attitude toward, usage of, and
continuance intention to use
ICT interactive tools
Hsu and Lin (2008) Blog Social norms, community identification Intention to blog
Hamari (2013) E-commerce Social comparison Service use
Lin (2008) Virtual
communities
Trust, social usefulness Sense of belonging, member
loyalty
Zhou (2011) Online
community
Subjective norm, social identity (SI), cognitive
SI, affective SI, evaluative SI, group norm
Participation intention,
participation behaviour
Dholakia et al. (2004) Virtual
communities
Group norms, mutual agreement, mutual
accommodation, social identity (SI), cognitive
SI, affective SI, evaluative SI
Desires, we-intentions,
participation behaviour
Hsu and Lu (2004) Online game Social norms, critical mass Attitude toward playing an
online game, intention to play
an online game
Hsu and Chiu (2004) World Wide
Web
Interpersonal norm, social norm Intention, e-service usage
Hsiao and Chiou (2012) Virtual
community
Social norms, trust, social capital Intention to stay in a group
Mäntymäki and Riemer (2014) Social virtual
world
Interpersonal influence, secondary sources of
info, perceived network size
Continuous use intentions
1
The TPB includes also measurement of perceived behavioural control. Gamifica-
tion may be implemented in both, non-organizational or organizational contexts. Use
of gamification in a non-organizational context, which is investigated in this study, is
most likely a volitional act and not considered to be affected by control issues to a
significant degree (Ajzen, 1988, 1991). Thus, in this study, the control beliefs are not
included. Instead, the research model builds upon the subjective norm and attitude as
antecedents for behavioural intentions.
336 J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347
social identity and/or group norms. Social identity variables are
most often used to examine identification with the group, whereas
group norm variables investigate the sharing of common goals
within the group (e.g. Cheung & Lee, 2010; Cheung et al., 2011;
Dholakia, Bagozzi, & Klein Pearo, 2004). These extensions consider
social influence from the point of view of group formation pro-
cesses. Some studies have shown that social identity (e.g. Cheung
& Lee, 2010; Zhou, 2011) and group norms (e.g. Cheung et al.,
2011; Dholakia et al., 2004; Zhou, 2011) positively predict, for
example, we-intentions and intentions to participate in the services.
As the effects of social influence towards the target behaviour
are of interest in this study, the core of the research model draws
from the TRA/TPB (Ajzen, 1988, 1991; Fishbein, 1979) similarly
to several previous works on social influence in information sys-
tems. In this study, however, we theorize that the phenomenon
of social influence should also capture the processes that are
induced by being exposed to the subjective norms. These are espe-
cially relevant in the context studied here: technologies that aim at
motivation of the users (e.g. gamification), in which the social
dimension is heavily utilized.
Consequently, we extend the social influence by inclusion of the
effects and benefits an individual may derive from the community
and social activity, regardless of the state of identification with the
group. These derived benefits such as feedback in the form of recog-
nition as well as reciprocity have not been examined in the previous
studies in a concise manner. Therefore, the theoretical development
in this study consists of 3 main refinements: (1) recognition as an
indicator of feedback from conforming to social influence, (2) per-
ceived reciprocal benefit obtained from the process of conforming
and the feedback received on it, and (3) the effect of the size of
the community on the factors of social influence (see Fig. 1).
2.4. Hypotheses for the extended social influence
When considering social influence in the context of technology
use, subjective norms reflect the user’s perceptions of how other
users perceive the use of the service (Ajzen, 1991; Fishbein,
1979). By participating in the community in the service, a user is
likely to become exposed to the influence of others. Furthermore,
in services that incorporate social interaction, e.g. ‘‘liking’’ and
commenting, a user can receive recognition on his or her activities
from other users. In the context of an information technology with
social elements, such recognition could be considered to represent
the social feedback a user receives on their behaviour (Cheung &
Lee, 2010; Cheung et al., 2011; Lin, 2008) or on accepting the social
influence and the normative expectations of the community.
Based on the discussions on subjective norms and recognition,
we hypothesize that the more strongly a person believes others
to expect and support a certain behaviour, the more positively
the recognition from conducting the behaviour and thus conform-
ing to those expectations will be perceived by the individual (see
e.g. Kelman, 1958). In line with Bock et al. (2005), Lewis et al.
(2003), and Venkatesh and Davis (2000), we propose that the
subjective norm affects attitude directly as well as behavioural
intentions mediated by attitude.
Accordingly, we hypothesise the following pertaining to the
influence of subjective norms.
H1a. Subjective norms positively influence the impact of recogni-
tion received from conforming to subjective norms.
H1b. Subjective norms positively influence the attitude toward the
technology.
Furthermore, receiving recognition from relevant others can cre-
ate reciprocal behaviour (Cialdini & Goldstein, 2004; Cialdini et al.,
1992). The reciprocal interaction can promote a form of social use-
fulness of the system – i.e., receiving benefit from and, in turn, con-
tributing to the social community (Lin, 2008; Preece, 2001;
Wellman & Wortley, 1990). In other words, receiving recognition
potentially increases the perceived mutual benefits received from
the use of the system (see e.g. Chiu, Hsu, & Wang, 2006). We opera-
tionalize the measurement of this construct as reciprocal benefits
(see e.g. Hsu & Lin, 2008; Lin, 2008) and hypothesize that a positive
relationship can be expected to exist between receiving recognition
and perceived reciprocal benefits derived from the system use.
Furthermore, we hypothesise that simply receiving recognition
has also a direct positive effect on attitude towards the use of the
service, since receiving positive recognition in general is considered
a positive experience, and thus, is likely to have a positive relation-
ship with attitude. Similarly, as the reciprocal activities are expect-
edly regarded as positive, they are likely to create further positive
attitude towards the service as well (see e.g. Cialdini et al., 1992).
Consequently, we hypothesize a positive relationship to exist also
between reciprocal benefits and attitude towards the system use.
In conclusion, the following hypotheses are suggested pertain-
ing to receiving recognition and reciprocal benefits:
H2a. Getting recognition positively influences the experience of
reciprocal benefits within the system.
H2b. Recognition positively influences attitude toward the system.
H3. Perceived reciprocal benefits positively influence the attitude
toward the system.
In previous studies, network exposure has occasionally been
considered interchangeable with subjective norms (Hsieh, Rai, &
Keil, 2008). However, the number of peers in a system is con-
ceptually distinct from the norms or the expectations one believes
others to share and hold (Hsieh et al., 2008; Lin & Bhattacherjee,
2009). Specifically, the network exposure considers one’s percep-
tions of the size of the relevant network (Hsieh et al., 2008; Lin &
Bhattacherjee, 2009). Therefore, we measure the size of the rele-
vant peer-group within the system with the construct of network
exposure (see Lin & Bhattacherjee, 2008).
In the context of information technology with social elements,
the size of the relevant network of an individual within a system
is likely to affect the amount of social activity the individual can
partake. Furthermore, the number of people using a service has,
indeed, been deemed important for services that are centred on
social interaction (Baker & White, 2010; Lin & Lu, 2011;
Sledgianowski & Kulviwat, 2009). Many studies have investigated
the role of network exposure in the information technology
Extended social influence
Subjecve
norms Recognion Mutual
benefits
Atude &
Behavior
Exposure to
the
community
Fig. 1. Conceptual model.
J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347 337
context, however, the studies have not considered that the effect
would in fact be mediated through other social factors (see
Table 2). However, we argue that the effect of network exposure
to attitude and use intentions is not direct, since the mere number
of users has no intrinsic value. Instead, we hypothesize that the
network effects on attitude are (fully) mediated by social factors
operationalized in this study: subjective norms, recognition and
reciprocal benefits. Through the social interaction and benefits
derived from it, the network exposure can translate into socially
valuable content (see e.g. Lin, 2008; Preece, 2001), and conse-
quently, further affect attitude towards the system as well as its
use. The more relevant others there are in the system, the more
normative information the user is likely to be exposed to. In the
same vein, the more users a person is exposed to, the more recog-
nition and reciprocal benefits he or she is likely to be exposed to.
Therefore, also the network exposure is partially mediated through
social influence to recognition and further reciprocal benefits.
We propose that all of the factors within our extended social
influence (subjective norms, recognition and reciprocal benefits)
mediate the effects of network exposure. Therefore, we hypothe-
sise the following:
H4a. Network exposure positively influences subjective norms.
H4b. Network exposure positively influences perceived recogni-
tion.
H4c. Network exposure positively influences perceived reciprocal
benefit.
H4d. The influence of network exposure on attitude is fully and
positively mediated by the extended social influence (NEXP has a
positive direct association with SUBJN, RECOG and RECIPB but
not with ATT).
H4e. The influence of network exposure is partially and positively
mediated within the causal chain of social factors
(NEXP ?SUBJN ?RECOG ?RECIPB).
In this study, attitude towards system use refers to the overall
evaluation of the system’s usage, be it favourable or unfavourable
(Ajzen, 1991; Fishbein & Ajzen, 1975). A strong relationship
between attitude and use intentions has been confirmed in several
studies (see, e.g. Baker & White, 2010; Bock et al., 2005; Lin &
Bhattacherjee, 2009).
Word-of-mouth (WOM) refers to a person’s willingness to
recommend a system to others. In the context of continued use
intention (Bhattacherjee, 2001), it reflects the satisfaction of the
user with the system in question, and his or her willingness to
recommend the service to other people (Cheung & Thadani,
2012; Kim & Son, 2009; Srinivasan, Anderson, & Ponnavolu, 2002).
As the instrumental purpose of the investigated technology is to
promote continued exercise, it is also important to measure
whether the continued use of the technology (Ajzen, 1991) does
indeed co-exist with the intention to continue exercising. We
expect to find that users who are more likely to continue the sys-
tem use, will also be more likely to pursue the behaviour that is
promoted by the system. Thus, the intentions to continue the beha-
viour are hypothesized to rise as a behavioural outcome.
Accordingly, we suggest the following hypotheses:
H5. Attitude positively influences intentions to continue using the
system.
H6. Attitude positively influences intentions to recommend the
system to others (i.e. WOM).
H7. Intentions to continue using the system positively influence
intentions to continue exercising.
See Fig. 2 for the research model and hypotheses.
3. The empirical study
3.1. Data
The data was gathered via an online questionnaire from the
users of a service called Fitocracy, an online service that gamifies
exercise. The motivational design of the system consists mainly of
affordances corresponding to achievement and competence as well
as social influence and relatedness (see Zhang, 2008 on motiva-
tional affordances). More specifically, the service enables tracking
of one’s exercise. The user enters the exercise details into the
Theory of planned behavior* and use
connuance
*SUBJN now conceptually included as a part of social influence
Network effects Extended social
influence
NEXP
SUBJN
RECIPB
ATT
WOM
CU
CE
H1b
H6
H5
H7
RECOG
H4a
H4c H3
H2b
H4b
H2a
H1a
NEXP = Network exposure, SUBJN = Subjecve norms, RECOG = Recognion, RECIPB = Reciprocal benefits, ATT = Atude,
WOM = Word-of-mouth intenons, CU = Connued use, CE = Connued exercise
Fig. 2. Research model.
338 J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347
system as the service does not contain automatic tracking. Thus, the
service relies on the self-reported data logged by the user.
Furthermore, the service incorporates motivational design in the
form of gamification. Based on the exercises logged by the user,
the system enables the users to gain points, level-ups, and achieve-
ments (on badges, see Hamari & Eranti, 2011). For example, when a
user logs an activity, the system calculates the point value that the
user gains with the exercise. The point value is adjusted based on
applicable details, such as amount, distance, time, intensity or
weights, provided by the user. The user can also complete quests
by performing and tracking an exercise corresponding to a given
set of conditions, or challenge other users into duels. Moreover,
other users can give feedback on achievements, level-ups and sta-
tuses by ‘liking’ or commenting the updates. The service holds simi-
larities with social networking services in that it creates a venue for
social activity such as group-forming and communication, incorpo-
rates profile-building, and the possibility of sharing content (Baker
& White, 2010; Boyd & Ellison, 2007; Ellison, Steinfield, & Lampe,
2007; Lin & Lu, 2011; Pfeil, Arjan, & Zaphiris, 2009).
The survey was conducted by posting a description of the study
and the survey link to the discussion forum and groups within the
service. The questionnaire was therefore accessible only for users
of the service. The survey was conducted at the end of 2012. At
the time of the gathering of data, the service could be used with
an iPhone application or via a Web browser. An Android applica-
tion was released while the data gathering neared completion.
The survey respondents were entered in a prize draw for one $50
Amazon gift certificate.
Table 3 outlines the demographic details of the respondents. As
can be seen from Table 3, the gender divide of the sample was
fairly equal. The ages between 20 and 29 are more represented
in the data than other age groups, however, the age distribution
is wide with respondents in all the categories. The lengths of
experience with the service reported by the respondents are dis-
tributed rather evenly. Table 3 also describes the details of
amounts of exercise reported by the respondents. Furthermore,
respondents were asked to report how they use the service: on a
mobile device (mobile phone or tablet) and/or on a computer. Of
the 200 respondents, 186 (93%) stated that they used the service
on a computer. The options were not mutually exclusive, and the
respondents could choose both mobile options and the computer
option. In fact, of the 186 computer users, 101 (54%) reported also
using the application on some type of a mobile device. Therefore,
no effects of the devices used are examined.
3.2. Measurement
All variables included 4 items and were measured with 7-point
Likert scales. All operationalizations of psychometric constructs
were adapted from previously published sources. Table 4 reports
the number of items in each construct as well as the sources from
which each of the constructs was adapted. One item was omitted
from the CE construct because the loading was rather low
(0.589), although it was still higher than with any other construct.
See Appendix A for the survey items, loadings and construct
sources.
3.3. Validity and reliability
All of the model-testing was conducted via component-based
PLS-SEM in SmartPLS 2.0 M3 (Ringle, Wende, & Will, 2005).
Compared to co-variance-based structural equation methods (CB-
SEM), the key advantage of the component-based PLS (PLS-SEM)
estimation is that it is non-parametric, and therefore, makes no
restrictive assumptions about the distributions of the data.
Secondly, PLS-SEM is considered to be a more suitable method
for prediction-oriented studies, while co-variance-based SEM is
Table 3
Demographic information of the respondents: gender, age, time using the service and exercise information of the respondent data.
Frequency Percent Frequency Percent
Gender Length of experience
Female 102 51 Less than 1 month 24 12
Male 98 49 1–3 months 38 19
Age (mean = 29.5, median = 27.5) 3–6 months 29 14.5
Less than 20 9 4.5 6–9 months 26 13
20–24 51 25.5 9–12 months 33 16.5
25–29 54 27 12–15 months 38 19
30–34 41 20.5 15–18 months 7 3.5
35–39 22 11 More than 18 months 5 2.5
40–44 16 8 Exercise sessions per week (mean = 5.3, median = 5.0)
45–49 3 1.5 1–4 83 41.5
50 or more 4 2 5–9 106 53.0
10–14 6 3.0
15 or more 5 2.5
Exercise hours per week (mean = 7.2, median = 6.0)
1–4 51 25.5
5–9 99 49.5
10–14 40 20.0
15 or more 10 5.0
Table 4
Measurement instruments.
Construct Name Included/
Total
items
Adapted from
ATT Attitude 4/4 Ajzen (1991)
CU Continuance
intentions for
system use
4/4 Bhattacherjee (2001)
CE Continuance
intentions for
exercise
3/4 Bhattacherjee (2001)
NEXP Network
exposure
4/4 Lin and Bhattacherjee (2008)
RECIPB Reciprocal
benefits
4/4 Hsu and Lin (2008), Lin (2008)
RECOG Recognition 4/4 Hernandez et al. (2011), Hsu and
Lin (2008), Lin and Bhattacherjee
(2010), Lin (2008)
SUBJN Subjective
norms
4/4 Ajzen (1991), Fishbein (1979),
Venkatesh et al. (2003)
WOM Word-of-mouth
intentions
4/4 Kim and Son (2009)
J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347 339
better suited for testing which models best fit the data (Anderson &
Gerbing, 1988; Chin, Marcolin, & Newsted, 2003).
Convergent validity (see Table 5) was assessed with three met-
rics: average variance extracted (AVE), composite reliability (CR),
and Cronbach’s alpha (Alpha). All of the convergent validity metrics
were clearly greater than the thresholds cited in relevant literature
(AVE should be greater than 0.5, CR greater than 0.7 (Fornell &
Larcker, 1981), and Cronbach’s alpha above 0.7 (Nunnally, 1978)).
Only well-established measurement items were used (see
Appendix A). Furthermore, there were no missing data; therefore,
no imputation methods were used. We can, therefore, conclude that
the convergent validity and reliability requirements are met.
Discriminant validity was assessed, firstly, through comparison
of the square root of the AVE of each construct to all of the correla-
tion between it and other constructs (see Fornell & Larcker, 1981),
where all of the square root of the AVEs should be greater than any
of the correlations between the corresponding construct and
another construct (Chin, 1998; Jöreskog & Sörbom, 1996).
Secondly, in accordance with the work of Pavlou, Liang, and Xue
(2007), we determined that no inter-correlation between con-
structs was higher than 0.9. Thirdly, we assessed discriminant
validity by confirming that every item had the highest loading with
its corresponding construct (Appendix B). All three tests indicate
that the discriminant validity and reliability are acceptable.
In addition, in order to reduce the likelihood of a common
method bias, we randomized the order of the measurement items
in the survey to limit the respondent’s ability to detect patterns
between the items (Cook, Campbell, & Day, 1979). The common
method bias refers to a situation where there is ‘‘variance that is attri-
butable to the measurement method rather than to the constructs the
measures represent’’ (Podsakoff, MacKenzie, Lee, & Podsakoff,
2003). Nevertheless, we tested whether common method bias
existed in our data by ‘‘controlling for the effects of an unmeasured
latent methods factor’’ as proposed by Podsakoff et al. (2003) in the
same manner as practically demonstrated in a PLS-SEM environ-
ment by Liang, Saraf, Hu, and Xue (2007). According to Williams,
Edwards, and Vandenberg (2003), if the loadings of the ‘‘method fac-
tor’’ are statistically insignificant and/or considerably low in com-
parison to indicator loadings of the substantive factors, there is no
evidence of the common method bias. In addition, the square of
the loadings represents the percentage of the variance explained.
As reported in Appendix C, we found a few significant loadings on
the ‘‘method factor’’, however, they explain a negligibly small share
of the variance (0.006 on average), whereas the indicators explain
0.762 variance on average in substantive factors. Therefore, we can
be confident that common method bias is not likely to be an issue.
The sample size satisfies the guidelines stating that the PLS-SEM
minimum sample size should be equal to the larger of the follow-
ing: ten times the largest number of formative indicators used to
measure one construct or ten times the largest number of struc-
tural paths directed at a particular latent construct in the structural
model (Chin & Newsted, 1999; Hair, Ringle, & Sarstedt, 2011). Also,
data should contain 150 observations for models with three or
more indicators on constructs (Anderson & Gerbing, 1984).
3.4. Results
The research model (see Fig. 3) could account for 43.1% of the
continued use intention as well as 56.8% of intention to recom-
mend the service to other people. Furthermore, the model accounts
for 44.1% of continued intention to exercise. The perceived social
benefits accounted for 61.9% of the variance of attitudes toward
the use of a gamified service. In addition, the model also accounted
for 16.7% of the variance in social influence, 31.7% of recognition,
and finally 44.9% of the variance of perceived reciprocal benefit.
As can be seen in Fig. 3 all of the direct paths in the research model
Table 5
Convergent and discriminant validity.
AVE CR Alpha ATT CU CE NEXP RECIPB RECOG SUBJN WOM
ATT 0.795 0.939 0.914 0.892
CU 0.734 0.917 0.880 0.657 0.857
CE 0.654 0.847 0.719 0.462 0.664 0.809
NEXP 0.855 0.959 0.944 0.372 0.228 0.268 0.925
RECIPB 0.700 0.903 0.857 0.713 0.492 0.431 0.472 0.837
RECOG 0.804 0.943 0.919 0.587 0.390 0.332 0.483 0.644 0.897
SUBJN 0.735 0.917 0.879 0.666 0.494 0.442 0.408 0.582 0.461 0.857
WOM 0.756 0.925 0.893 0.754 0.617 0.385 0.428 0.651 0.654 0.660 0.869
Square roots of AVEs are reported in bold in the diagonal.
NEXP
SUBJN
R2=0.167
RECIPB
R2=0.449
ATT
R2=0.619
WOM
R2=0.568
CU
R2=0.431
CE
R2=0.441
0.358***
t=5.509
0.754***
t=19.095
0.657***
t=13.103
0.664***
t=13.977
RECOG
R2=0.317
0.408***
t=6.610
0.210***
t=3.230
0.398***
t=5.152
0.165**
t=2.205
0.354***
t=5.619
0.543***
t=6.386
0.316***
t=4.301
* < 0.1, ** < 0.05, *** < 0.01
NEXP = Network exposure, SUBJN = Subjecve norms, RECOG = Recognion,
RECIPB = Reciprocal benefits, ATT = Atude, WOM = Word-of-mouth intenons,
CU = Connued use, CE = Connued exercise
Fig. 3. Path model results.
340 J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347
were positive and statistically significant. Therefore, the model
supports the hypotheses H1a–H4c and H5–H7.
To further investigate the relationship of network effects and
attitude (H4d), in addition to the direct effect, we modelled a
mediated effect by the social factors in the model in accordance
with our hypotheses. When modelling only a single direct effect
between these variables and at the same time omitting the social
factors, a significant positive association existed (.372
⁄⁄⁄
).
However, when the social factors were added so that direct effect
and mediation were both modelled, the direct effect was now
insignificant at a close to zero value (.060). This suggests, that
the effect between network exposure and attitude is fully
mediated through the extended social influence. The resulting total
effect between the variables was .414
⁄⁄⁄
as mediated by the social
factors, which is very close to the value of the uncontrolled direct
effect. Furthermore, in addition to the direct effect between net-
work exposure and both recognition and reciprocal benefits, we
also modelled a mediation by the social factors (subjective norms
mediating recognition; subjective norms and recognition mediat-
ing reciprocal benefits) in accordance with our hypotheses (H4e).
While the direct effect from network exposure to both recognition
and reciprocal benefits was significant (recognition = .354
⁄⁄⁄
, reci-
procal benefits = .210
⁄⁄⁄
), the total effect mediated by the social
factors showed an increase in the effects (recognition = .483
⁄⁄⁄
,
reciprocal benefits = .472
⁄⁄⁄
). See Table 6 for the total effects. See
Table 7 for summary of hypotheses.
In addition to the hypotheses presented in the research model,
we controlled also for the effects of age, gender and length of experi-
ence with the service on all of the dependent variables. Only one of
the tests provided statistically significant results: gender had a
slight effect on continued intentions to exercise (R
2
of CE increased
by 0.017), indicating that women reported somewhat higher inten-
tions to continue with exercising in the future (t= 2.470
⁄⁄
).
Even though in the original TRA/TPB the subjective norm and
attitude are both modelled as direct predictors for behavioural
intentions, prior literature has considered that the modelled
relationships of these variables may potentially vary based on
the social influence processes (internalization, identification, or
compliance (Kelman, 1958)) and consequently, for example, the
voluntariness of use (Davis, Bagozzi, & Warshaw, 1989). People
adopt the case service voluntarily and therefore we can expect that
it is the individual’s own attitudes that are likely to affect the deci-
sions on use and therefore the subjective norms would be
mediated through attitude rather than being a direct predictor.
Indeed, in our data the relationship between subjective norms
and continued use is mediated by attitude which is indicated by
the following tests: (a) SUBJN ?CU is positive and significant
when the relationship ATT ?CU is omitted, but (b) there is no sig-
nificant direct association between SUBJN ?CU when ATT ?CU is
modelled, (c) there is a positive and significant association
between SUBJN ?ATT and (d) the total effect indicated by the path
model shows a significant total effect between SUBJN and CU in the
mediation model (Fig. 3 and Table 6).
4. Discussion
In this study, we investigated the role of social influence in
gamified exercise with the aim of examining how the social aspects
affect use intentions of the technology, intentions to recommend it
to others as well as the intentions to continue exercise, i.e. the
behaviour that is being supported by the technology. Moreover,
motivated by the fact that social influence as a process is clearly
a manifold phenomenon in contexts such as gamification services,
this study sought to expand measurements and conceptualizations
of social influence.
4.1. Theoretical and practical implications
While in prior studies social influence in technology use has
commonly been investigated by only measuring subjective norms
with the theories of reasoned action and planned behaviour
(Ajzen, 1991; Fishbein, 1979), in our research we theorized that
merely measuring subjective norms does not take into account
the feedback and benefits received through the social influence
process which clearly seem to be important aspects in systems
Table 6
Total effects as mediated according to the path model.
RECOG RECIPB ATT WOM CU CE
NEXP 0.483
⁄⁄⁄
0.472
⁄⁄⁄
0.414
⁄⁄⁄
0.312
⁄⁄⁄
0.272
⁄⁄⁄
0.181
⁄⁄⁄
(direct,
SUBJN)
a
(direct,
SUBJN,
RECOG)
(SUBJN,
RECOG,
RECIPB)
(SUBJN,
RECOG,
RECIPB,
ATT)
(SUBJN,
RECOG,
RECIPB,
ATT)
(SUBJN,
RECOG,
RECIPB,
ATT, CU)
SUBJN 0.172
⁄⁄⁄
0.479
⁄⁄⁄
0.361
⁄⁄⁄
0.314
⁄⁄⁄
0.209
⁄⁄⁄
(RECOG) (RECOG,
RECIPB)
(RECOG,
RECIPB,
ATT)
(RECOG,
RECIPB,
ATT)
(RECOG,
RECIPB,
ATT, CU)
RECOG 0.381
⁄⁄⁄
0.287
⁄⁄⁄
0.251
⁄⁄⁄
0.166
⁄⁄⁄
(RECIPB) (RECIPB,
ATT)
(RECIPB,
ATT)
(RECIPB,
ATT CU)
RECIPB 0.300
⁄⁄⁄
0.261
⁄⁄⁄
0.174
⁄⁄⁄
(ATT) (ATT) (ATT, CU)
ATT 0.436
⁄⁄⁄
(CU)
Significance levels reported as
⁄
<0.1,
⁄⁄
<0.05,
⁄⁄⁄
<0.01.
a
Listed in parentheses are the different paths through which the total effects is
composed. Direct effects are reported in Fig. 3.
Table 7
Confirmation of hypotheses.
H# IV ?DV Hypothesis Supported
H1a SUBJN ?RECOG Subjective norms positively influence
the impact of feedback from
conforming to subjective norms
Yes
H1b SUBJN ?ATT Subjective norms positively influence
the attitude toward the technology
Yes
H2a RECOG ?RECIPB Getting recognition positively
influences the experience of
reciprocal benefits within the system
Yes
H2b RECOG ?ATT Recognition positively influences
attitude toward the system
Yes
H3 RECIPB ?ATT Perceived reciprocal benefits
positively influence the attitude
toward the system
Yes
H4a NEXP ?SUBJN Network exposure positively
influences subjective norms
Yes
H4b NEXP ?RECOG Network exposure positively
influences perceived recognition
Yes
H4c NEXP ?RECIPB Network exposure positively
influences perceived reciprocal
benefit
Yes
H4d NEXP ?ATT The influence of network exposure
on attitude is fully and positively
mediated by the extended social
influence
Yes
H4e NEXP ?RECOG,
RECIPB
The influence of network exposure is
partially and positively mediated
within the causal chain of social
factors
Yes
H5 ATT ?CU Attitude positively influences
continued use intention
Yes
H6 ATT ?CE Attitude positively influences
intentions to recommend the service
(i.e., WOM)
Yes
H7 CU ?CE Intentions to continue using the
system positively influence
intentions to continue exercising
Yes
J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347 341
such as gamification. Therefore, in this study, we sought to increase
theoretical knowledge and contribute to the conceptualizations of
the extended social influence, that is, the role of the network expo-
sure on social influence, the role of social influence on IT use con-
tinuance, and further role in behaviour continuance. Guided by the
theoretical developments in social psychology (Cialdini &
Goldstein, 2004; Cialdini et al., 1992), we theorized that social
influence is not confined only to the perceptions of the individual
about the beliefs of relevant others. Instead, it also includes the
resultant positive recognition from signalling acceptance to those
norms. We expanded theories of reasoned action and planned
behaviour with recognition, i.e. the degree of positive feedback
from conforming to subjective norms. Furthermore, we theorized
that the reciprocal influence-compliance (Cialdini & Goldstein,
2004) with the community’s norms can promote perceived bene-
fits (the perceived increase in mutual benefits derived from the
use). We hypothesized also that the size of the immediate commu-
nity within the system would have a positive effect on all of the
social aspects measured within the study.
Our empirical study does indeed support these theorizations
and the hypotheses regarding them: according to the findings, the
social factors consisting of subjective norms, perceived recognition,
and perceived reciprocal benefit all had a positive relationship with
use and exercise continuance as well as intention to recommend
the technology to others as mediated by attitude. Furthermore,
the continued use intentions were positively associated with the
continued exercise intentions. Therefore, the findings of this study
suggest that the social factors are an important antecedent for sus-
tained behaviour and continued use intentions of motivational
technologies. The findings of the study are in line with previous
research on the importance of social aspects in gamification, and
particularly gamification of exercise. For example, the findings of
the studies by Chen and Pu (2014) and Chen et al. (2014) indicated
that social features in gamification in general increased physical
activity. Especially, when users cooperate together for exercise
results they get the most positive effects. Similarly, in the study
by Allam et al. (2015) gamification and social support features
implemented into their intervention system increased physical
activity of the users. However, the previous studies did not investi-
gate the psychological aspects regarding the social aspects, but
rather examined the relationships between the system elements
and behaviour directly. In our study, we had an opposite approach
as we particularly examined the process of social influence in the
gamification of exercise but did not measure which particular ele-
ments affected social influence. Therefore, this study has attempted
to fill this gap in the understanding of the social aspects in gamifica-
tion. In synergy, the present study and the earlier studies begin to
provide a more comprehensive view of the social influence from
gamification elements to psychological aspects and onto behaviour.
The results of the present study, in combination with the findings of
Chen and Pu (2014), Chen et al. (2014) as well as Allam et al. (2015),
suggest that the social support and benefits derived from the social
interaction with the other users are important for continuing the
behaviour supported by the system.
Furthermore, an important theoretical finding of this research
was that the association between attitude and network effects is
mediated through the extended social influence. Our empirical
study supports the theorization that the social factors do indeed
fully positively mediate the effect of network exposure on attitude.
Moreover, another theoretical contribution pertains to the com-
pounding nature of social factors and network effects. The larger
the network, the larger the social influence becomes, the more
recognition one receives and the more people benefit from the
reciprocal social behaviour. Therefore, the results of the path
model support the conclusion that the social factors positively
and partially mediate network exposure on each other.
Moreover, interesting findings regarding the relationships of the
factors within the extended social influence and the dependent
variables were made in the study. Firstly, the direct effect of recog-
nition on attitude was significant, but rather small, while the total
effect (direct effect + effect mediated by reciprocal benefits) was
considerably larger. This finding suggests that merely getting
recognized does not necessarily translate into positive attitudes
towards the service. However, when the user also perceives receiv-
ing reciprocal benefits from the service in addition to the recogni-
tion, it leads to further positive attitudes towards the service.
Therefore, the nature of the interaction with the other users seems
to be important instead of simply being acknowledged by the com-
munity. Secondly, in the data, subjective norms had a strong posi-
tive relationship with attitude, however, a portion of its influence
was mediated by recognition and reciprocal benefits. This is an
interesting finding as the common theories used to explain beha-
vioural intentions, the TRA and TPB, consider both attitude and sub-
jective norms as antecedents to behavioural intentions. However,
as noted, in our research the relationship between subjective norms
and behavioural intentions was not significant, while there was a
significant, partially mediated, positive relationship between sub-
jective norm and attitude (on the relationships of subjective norm,
attitude and behavioural intention, see e.g. Davis et al., 1989).
In this study we investigated the phenomenon of social influ-
ence in the context of gamification. As the system studied here
has been designed to have affordances that support social interac-
tion (Zhang, 2008), it might be intuitive that factors pertaining to
social influence play an important role. Nevertheless, more tradi-
tional information technologies can equally benefit from the use
of similar design. The results of this research suggest that in order
to support adoption and use of information technologies, the pro-
cess of social influence could be harnessed in the design in several
ways. Firstly, affording features that enable users/community to
signal norms within the community in the system enables the dif-
fusion of norms, and thus, enables creation and strengthening of the
community. Secondly, providing features such as sharing functions
and badges (see Hamari, 2013; Hamari & Eranti, 2011; Zhang, 2008)
affords users to communicate or make visible their behaviour
related to accepting the social influence. Thirdly, providing features
such as ‘‘liking’’ and commenting enables users to give feedback on
other users’ activities (and thus enables recognition and further
support for emergence of relatedness related intrinsic motivations
towards use of the system). Fourthly, the formation of such social
communities in the information system context can support the
continued social interaction and further use as well as the forma-
tion of reciprocal benefits through increased cooperation. Our data
also show that the size of a person’s network in a community within
a given system is an important predictor for other social benefits
that can be derived from the use. Therefore, the integration of
new users to the community seems to be crucial. Furthermore,
efforts in the information system design towards connecting new
and old users should prove to be highly beneficial. The diffusion
of the community’s norms was especially pertinent in the context
of this study (gamification), particularly since the system’s function
was behaviour change. In similar systems, design aiming to pro-
gress the formation of a strong community and dissemination of
the community’s norms should prove to be highly effective.
4.2. Future research directions
Future studies should seek to investigate the difference in
effects of social influence depending on the system type, for exam-
ple, depending on the degree of possible exposure to social interac-
tion within the system. Although the type of motivational
technologies investigated in this study seems to bring social influ-
ence to the forefront, the same phenomenon is also highly relevant,
342 J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347
for example, in the context of utilitarian information systems. For
instance, in adoption of corporate systems, such as enterprise
resource planning systems, it would be highly relevant to investi-
gate how in the post-adoption phase the aspects of influence
acceptance, recognition and benefits from reciprocity can influence
satisfaction with the system, and further, attitude formation and
intentions. Future work could also consider differences in the
impact of social influence depending on the level of suggestibil-
ity/conformity of the users (see e.g. Lascu & Zinkhan, 1999).
In the motivational system/gamification contexts, further stud-
ies could also benefit from considering differences stemming from
the users’ orientation toward gameful interactions (Hamari &
Tuunanen, 2014; Yee, 2006). Moreover, as this study has explored
the role of social influence in gamification and motivational systems,
further studies could investigate hedonistic (Hirschman & Holbrook,
1982; van der Heijden, 2004; Webster & Martocchio, 1992) and utili-
tarian motivations (e.g., Davis, 1989) in such environments.
4.3. Limitations
As is commonplace with research conducted by online surveys,
the data of this study is self-reported and the respondents are self-
selected. Use of self-reported data may affect the results as the
users responding are most presumably actively engaged with the
service and eager to participate in activities related to it.
Therefore, the results possibly represent perceptions and intentions
of active users of the service and disregard less active and unen-
gaged users. The perceptions of less active users could be addressed
in future studies as well as reasons for not being or becoming
involved in the service. Future research would also benefit from
combining survey data with actual usage data as well as random-
ized experiments in order to diminish the effects of self-reported
and selected data.
Furthermore, it is also commonplace with quantitative studies
that the results are reductionist and geared towards generalizable
overall indications of the phenomenon. Therefore, the study does
not investigate all possible ways of using the system nor all possi-
ble motivations behind using it. Therefore, it is expected and likely
that the phenomenon can be more complex if we were to investi-
gate the phenomenon on a more granular level by using e.g. qual-
itative methods. As with any system, ultimately it is up to the users
how they eventually interact with a given system. Considering an
exercise gamification system such as Fitocracy examined in this
study, it is possible that some users may be motivated simply by
the tracking features of the system and pay less attention to the
gameful and social aspects. In contrast, some users may be moti-
vated simply by the gamification and disregard the social features.
Further studies could investigate how different user factors might
moderate the effects between gamification element use and psy-
chological aspects. The different ways in which individual respon-
dents ultimately perceive and interact with the system is out of
reach of this study due to the chosen method and approach.
Relatedly, in this study it was not directly measured how differ-
ent gamification elements affect aspect of social influence and use
but rather investigated the formation and process of social influ-
ence midst gamification service users. Further studies could be
conducted where the linkages between service features and psy-
chological responses were recorded.
Acknowledgements
The research has been partially supported by individual study
grants for both authors from the Finnish Cultural Foundation as
well as carried out as part of research projects (40134/13, 40111/
14, 40107/14) funded by the Finnish Funding Agency for
Innovation (TEKES). Both authors have contributed to this article
equally.
Appendix A. Survey items, loadings, and construct sources
Construct name and source Indicator Survey item Loading
Attitude
Ajzen (1991) ATT1 All things considered, I find using Fitocracy to be a wise
thing to do
0.868
ATT2 All things considered, I find using Fitocracy to be a good idea 0.914
ATT3 All things considered, I find using Fitocracy to be a positive
thing
0.898
ATT4 All things considered, I find using Fitocracy to be favourable 0.885
Continuance intentions for system use
Bhattacherjee (2001) CU1 I predict that I will keep using Fitocracy in the future at least
as much as I have used it lately
0.883
CU2 I intend to use Fitocracy at least as often within the next
three months as I have previously used
0.820
CU3 I predict that I will use Fitocracy more frequently rather
than less frequently
0.831
CU4 It is likely that I will use Fitocracy more often rather than
less often during the next couple months
0.891
Continuance intentions for exercise
Bhattacherjee (2001) CE1 I plan to increase the amount of exercise rather than to
decrease it
0.900
CE2 I predict that I will exercise more frequently within the next
three months
0.880
CE3 I think I will keep exercising in the near future at least as 0.615
(continued on next page)
J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347 343
Appendix A (continued)
Construct name and source Indicator Survey item Loading
much as I have during the last few months
Network exposure
Lin and Bhattacherjee (2008) NEXP1 I have a lot of friends on Fitocracy who follow my activities 0.908
NEXP2 Many people follow my activities on Fitocracy 0.941
NEXP3 I follow many people on Fitocracy 0.915
NEXP4 I have many friends in Fitocracy 0.935
Reciprocal benefits
Hsu and Lin (2008), Lin (2008) RECIPB1 I find that participating in the Fitocracy community can be
mutually helpful
0.834
RECIPB2 I find my participation in the Fitocracy community can be
advantageous to me and other people
0.842
RECIPB3 I think that participating in the Fitocracy community
improves my motivation to exercise
0.811
RECIPB4 The Fitocracy community encourages me to exercise 0.860
Recognition
Hernandez et al. (2011), Hsu and Lin (2008), Lin
(2008), Lin and Bhattacherjee (2010)
RECOG1 I feel good when my achievements in Fitocracy are noticed 0.882
RECOG2 I like it when other Fitocracy users comment and like my
exercise
0.900
RECOG3 I like it when my Fitocracy peers notice my exercise reports 0.936
RECOG4 It feels good to notice that other user has browsed my
Fitocracy feed
0.868
Subjective norms
Ajzen (1991), Fishbein (1979), Venkatesh, Morris,
Davis, and Davis (2003)
SUBJN1 People who influence my attitudes would recommend
Fitocracy
0.782
SUBJN2 People who are important to me would think positively of
me using Fitocracy
0.888
SUBJN3 People who I appreciate would encourage me to use
Fitocracy
0.895
SUBJN4 My friends would think using Fitocracy is a good idea 0.861
Word-of-mouth intentions
Kim and Son (2009) WOM1 I would recommend Fitocracy to my friends 0.860
WOM2 I will recommend Fitocracy to anyone who seeks my advice 0.907
WOM3 I will refer my acquaintances to Fitocracy 0.825
WOM4 I will say positive things about Fitocracy to other people 0.883
Appendix B. Cross-loadings
ATT CE CU NEXP RECIPB RECOG SUBJN WOM
ATT1 0.868 0.420 0.530 0.339 0.636 0.438 0.638 0.603
ATT2 0.914 0.444 0.641 0.330 0.636 0.478 0.585 0.651
ATT3 0.898 0.393 0.591 0.323 0.611 0.576 0.536 0.724
ATT4 0.885 0.392 0.577 0.336 0.659 0.592 0.617 0.704
CE1 0.419 0.900 0.601 0.199 0.331 0.299 0.414 0.343
CE2 0.398 0.880 0.547 0.204 0.366 0.226 0.421 0.304
CE3 0.289 0.615 0.451 0.261 0.358 0.284 0.211 0.285
CU1 0.580 0.668 0.883 0.213 0.449 0.332 0.536 0.552
CU2 0.568 0.483 0.820 0.194 0.428 0.387 0.318 0.517
CU3 0.557 0.432 0.831 0.230 0.409 0.373 0.361 0.590
CU4 0.550 0.660 0.891 0.152 0.403 0.262 0.454 0.471
NEXP1 0.317 0.254 0.198 0.908 0.387 0.373 0.380 0.371
NEXP2 0.285 0.187 0.153 0.941 0.353 0.413 0.317 0.343
NEXP3 0.399 0.264 0.242 0.915 0.532 0.550 0.392 0.457
NEXP4 0.354 0.277 0.234 0.935 0.441 0.419 0.410 0.390
344 J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347
Appendix B (continued)
ATT CE CU NEXP RECIPB RECOG SUBJN WOM
RECIPB1 0.521 0.310 0.385 0.384 0.834 0.526 0.494 0.582
RECIPB2 0.591 0.374 0.403 0.447 0.842 0.630 0.535 0.542
RECIPB3 0.660 0.412 0.417 0.330 0.811 0.435 0.471 0.512
RECIPB4 0.614 0.348 0.441 0.412 0.860 0.553 0.446 0.547
RECOG1 0.476 0.267 0.324 0.413 0.592 0.882 0.349 0.531
RECOG2 0.503 0.235 0.267 0.380 0.532 0.900 0.383 0.564
RECOG3 0.608 0.321 0.403 0.475 0.634 0.936 0.445 0.635
RECOG4 0.507 0.360 0.392 0.457 0.547 0.868 0.467 0.610
SUBJN1 0.520 0.321 0.372 0.352 0.447 0.302 0.782 0.496
SUBJN2 0.573 0.377 0.445 0.318 0.536 0.403 0.888 0.584
SUBJN3 0.593 0.408 0.447 0.374 0.506 0.424 0.895 0.576
SUBJN4 0.593 0.404 0.426 0.358 0.504 0.439 0.861 0.600
WOM1 0.665 0.291 0.487 0.355 0.528 0.486 0.562 0.860
WOM2 0.715 0.348 0.603 0.343 0.568 0.638 0.600 0.907
WOM3 0.540 0.312 0.459 0.431 0.531 0.508 0.547 0.825
WOM4 0.680 0.386 0.583 0.376 0.636 0.632 0.585 0.883
Bolded loadings represent the items belonging to the corresponding construct.
Appendix C. Common method bias test
Method factor loading t-value Variance explained Substantive factor loading t-value Variance explained
RECIPB1 0.086 1.279 0.007 0.928
⁄⁄⁄
16.453 0.861
RECIPB2 0.089 1.024 0.008 0.883
⁄⁄⁄
11.230 0.780
RECIPB3 0.058 0.796 0.003 0.792
⁄⁄⁄
11.981 0.627
RECIPB4 0.115 1.453 0.013 0.745
⁄⁄⁄
10.161 0.555
WOM1 0.095 1.233 0.009 0.805
⁄⁄⁄
10.638 0.648
WOM2 0.067 0.917 0.004 0.891
⁄⁄⁄
14.436 0.794
WOM3 0.097 1.709 0.009 0.826
⁄⁄⁄
16.672 0.682
WOM4 0.137 1.903 0.019 0.965
⁄⁄⁄
17.279 0.931
CE1 0.199
⁄⁄
2.436 0.040 0.467
⁄⁄⁄
4.788 0.218
CE2 0.074 1.506 0.005 0.939
⁄⁄⁄
29.406 0.882
CE3 0.057 1.372 0.003 0.942
⁄⁄⁄
36.960 0.887
CU1 0.112
⁄
2.174 0.013 0.962
⁄⁄⁄
29.108 0.925
CU2 0.090 1.417 0.008 0.807
⁄⁄⁄
15.551 0.651
CU3 0.034 0.581 0.001 0.820
⁄⁄⁄
14.526 0.672
CU4 0.010 0.151 0.000 0.838
⁄⁄⁄
15.532 0.702
ATT1 0.090 1.274 0.008 0.810
⁄⁄⁄
11.375 0.656
ATT2 0.010 0.125 0.000 0.863
⁄⁄⁄
11.433 0.745
ATT3 0.060 1.233 0.004 0.964
⁄⁄⁄
24.611 0.929
ATT4 0.039 0.667 0.002 0.930
⁄⁄⁄
17.409 0.865
RECOG1 0.024 0.539 0.001 0.902
⁄⁄⁄
21.243 0.814
RECOG2 0.071 1.584 0.005 0.814
⁄⁄⁄
19.742 0.663
RECOG3 0.100
⁄⁄⁄
2.850 0.010 0.973
⁄⁄⁄
32.956 0.947
RECOG4 0.050 1.429 0.003 0.899
⁄⁄⁄
30.818 0.808
NEXP1 0.086
⁄⁄
2.493 0.007 0.993
⁄⁄⁄
68.324 0.986
NEXP2 0.013 0.408 0.000 0.929
⁄⁄⁄
47.059 0.863
NEXP3 0.028 0.843 0.001 0.930
⁄⁄⁄
31.040 0.865
NEXP4 0.101
⁄⁄
2.588 0.010 0.851
⁄⁄⁄
29.084 0.724
SUBJN1 0.056 0.754 0.003 0.822
⁄⁄⁄
13.457 0.676
SUBJN2 0.006 0.134 0.000 0.891
⁄⁄⁄
22.538 0.794
SUBJN3 0.046 0.741 0.002 0.826
⁄⁄⁄
12.592 0.682
SUBJN4 0.001 0.028 0.000 0.890
⁄⁄⁄
22.773 0.792
Average 0.006 0.762
J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347 345
References
Ainslie, G. (1975). Specious reward: A behavioral theory of impulsiveness and
impulse control. Psychological Bulletin, 82(4), 463–496.
Ajzen, I. (1988). Attitudes, personality, and behaviour. Chicago, IL: Dorsey Press.
Ajzen, I. (1991). The theory of planned behaviour. Organizational Behavior and
Human Decision Processes, 50(2), 179–211.
Allam, A., Kostova, Z., Nakamoto, K., & Schulz, P. J. (2015). The effect of social
support features and gamification on a web based intervention for rheumatoid
arthritis patients: Randomized controlled trial. Journal of Medical Internet
Research, 17(1), e14.
Anderson, J. C., & Gerbing, D. W. (1984). The effect of sampling error on
convergence, improper solutions, and goodness-of-fit indices for maximum
likelihood confirmatory factor analysis. Psychometrika, 49(2), 155–173.
Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A
review and recommended two-step approach. Psychological Bulletin, 103(3),
411–423.
Baker, R. K., & White, K. M. (2010). Predicting adolescents’ use of social networking
sites from an extended theory of planned behaviour perspective. Computers in
Human Behavior, 26, 1591–1597.
Bhattacherjee, A. (2001). Understanding information systems continuance: An
expectation-confirmation model. MIS Quarterly, 25(3), 351–370.
Biddiss, E., & Irwin, J. (2010). Active video games to promote physical activity in
children and youth: A systematic review. Archives of Pediatrics and Adolescent
Medicine, 164(7), 664–672.
Bista, S. K., Nepal, S., Paris, C., & Colineau, N. (2014). Gamification for online
communities: A case study for delivering government services. International
Journal of Cooperative Information Systems, 23(2).
Bock, G.-W., Zmud, R. W., Kim, Y.-G., & Lee, J.-N. (2005). Behavioral intention
formation in knowledge sharing: Examining the roles of extrinsic motivators,
social-psychological forces, and organizational climate. MIS Quarterly, 29(1),
87–111.
Bonde, M. T., Makransky, G., Wandall, J., Larsen, M. V., Morsing, M., Jarmer, H., et al.
(2014). Improving biotech education through gamified laboratory simulations.
Nature Biotechnology, 32(7), 694–697.
Bonner, H. (1959). Group dynamics: Principles and applications. New York: Ronald
Press.
Boyd, D. M., & Ellison, N. B. (2007). Social network sites: Definition, history, and
scholarship. Journal of Computer-Mediated Communication, 13, 210–230.
Brauner, P., Calero Valdez, A., Schroeder, U., & Ziefle, M. (2013). Increase physical
fitness and create health awareness through exergames and gamification: The
role of individual factors, motivation and acceptance. In A. Holzinger, M. Ziefle,
M. Hitz, & M. Debevc (Eds.). Lecture notes in computer science (Vol. 7946,
pp. 349–362). Berlin, Heidelberg: Springer.
Burger, J. M., Sanchez, J., Imberi, J. E., & Grande, L. R. (2009). The norm of reciprocity
as an internalized social norm: Returning favors even when no one finds out.
Social Influence, 4(1), 11–17.
Cafazzo, J. A., Casselman, M., Hamming, N., Katzman, D. K., & Palmert, M. R. (2012).
Design of an mHealth app for the self-management of adolescent type 1
diabetes: A pilot study. Journal of Medical Internet Research, 14(3), e70.
Carron, A. V., & Brawley, L. R. (2000). Cohesion: Conceptual and measurement
issues. Small Group Research, 31(1), 89–106.
Cechanowicz, J., Gutwin, C., Brownell, B., & Goodfellow, L. (2013). Effects of
gamification on participation and data quality in a real-world market research
domain. In Proceedings of gamification ’13, Stratford, Ontario, Canada, October 2–4
(pp. 58–65).
Çelik, H. (2011). Influence of social norms, perceived playfulness and online
shopping anxiety on customers’ adoption of online retail shopping: An
empirical study in the Turkish context. International Journal of Retail &
Distribution Management, 39(6), 390–413.
Chen, Y., & Pu, P. (2014). HealthyTogether: Exploring social incentives for mobile
fitness applications. In Proceedings of Chinese CHI ‘14, Toronto, ON, Canada, April
26–27 (pp. 25–34).
Chen, Y., Zhang, J., & Pu, P. (2014). Exploring social accountability in pervasive fitness
apps. In Proceeding of the UBICOMM2014, Rome, Italy, August 24–28 (pp. 221–226).
Cheung, C. M. K., Chiu, P.-Y., & Lee, M. K. O. (2011). Online social networks:
Why do students use Facebook? Computers in Human Behavior, 27, 1337–1343.
Cheung, C. M. K., & Lee, M. K. O. (2010). A theoretical model of intentional social
action in online social networks. Decision Support Systems, 49(1), 24–30.
Cheung, C. M. K., & Thadani, D. R. (2012). The impact of electronic word-of-mouth
communication: A literature analysis and integrative model. Decision Support
Systems, 54(1), 461–470.
Chin, W. W., Marcolin, B. L., & Newsted, P. R. (2003). A partial least squares latent
variable modeling approach for measuring interaction effects: Results from a
Monte Carlo simulation study and an electronic-mail emotion/adoption study.
Information Systems Research, 14(2), 189–217.
Chin, W. W. (1998). The partial least squares approach for structural equation
modelling. In G. A. Marcoulides (Ed.), Modern methods for business research
(pp. 295–336). London: Lawrence Erlbaum Associates.
Chin, W. W., & Newsted, P. R. (1999). Structural equation modeling analysis with
small samples using partial least squares. In R. H. Hoyle (Ed.), Statistical strategies
for small sample research (pp. 307–342). Thousand Oaks, CA: Sage.
Chiu, C.-M., Hsu, M.-H., & Wang, E. T. G. (2006). Understanding knowledge sharing
in virtual communities: An integration of social capital and social cognitive
theories. Decision Support Systems, 42(3), 1872–1888.
Christy, K. R., & Fox, J. (2014). Leaderboards in a virtual classroom: A test of
stereotype threat and social comparison explanations for women’s math
performance. Computers & Education, 78, 66–77.
Cialdini, R. B., & Goldstein, N. J. (2004). Social influence: Compliance and
conformity. Annual Review of Psychology, 55, 591–621.
Cialdini, R. B., Green, B. L., & Rusch, A. J. (1992). When tactical pronouncements of
change become real change: The case of reciprocal persuasion. Journal of
Personality and Social Psychology, 63(1), 30–40.
Cook, T. D., Campbell, D. T., & Day, A. (1979). Quasi-experimentation: Design &
analysis issues for field settings. Boston: Houghton Mifflin.
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance
of information technology. MIS Quarterly, 13(3), 319–340.
Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer
technology: A comparison of two theoretical models. Management Science,
35(8), 982–1003.
Deci, E. L., & Ryan, R. M. (2000). The ‘‘what’’ and ‘‘why’’ of goal pursuits: Human needs
and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268.
de-Marcos, L., Domínguez, A., Saenz-de-Navarrete, J., & Pagés, C. (2014). An
empirical study comparing gamification and social networking on e-learning.
Computers & Education, 75, 82–91.
Denny, P. (2013). The effect of virtual achievements on student engagement. In
Proceedings of CHI 2013: Changing perspectives, Paris, France, April 27–May 2,
2013 (pp. 763–772).
Deterding, S., Dixon, D., Khaled, R., & Nacke, L. (2011). From game design elements
to gamefulness: Defining gamification. In Proceedings of the 15th international
academic MindTrek conference: Envisioning future media environments, Tampere,
Finland, September 28–30 (pp. 9–15).
Dholakia, U. M., Bagozzi, R. P., & Klein Pearo, L. (2004). A social influence model of
consumer participation in network- and small-group-based virtual
communities. International Journal of Research in Marketing, 21(3), 241–263.
Domínguez, A., Saenz-de-Navarrete, J., de-Marcos, L., Fernández-Sanz, L., Pagés, C., &
Martínez-Herráiz, J.-J. (2013). Gamifying learning experiences: Practical
implications and outcomes. Computers & Education, 63, 380–392.
Eickhoff, C., Harris, C. G., de Vries, A. P., & Srinivasan, P. (2012). Quality through flow
and immersion: Gamifying crowdsourced relevance assessments. In Proceedings
of the 35th international ACM SIGIR conference on research and development in
information retrieval, Portland, Oregon, USA, August 12–16, 2012 (pp. 871–880).
Elias, P., Rajan, N. O., McArthur, K., & Dacso, C. C. (2013). InSpire to promote lung
assessment in youth: Evolving the self-management paradigms of young people
with asthma. Medicine 2.0, 2(1), e1.
Ellison, N. B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook ‘friends:’
Social capital and college students’ use of online social network sites. Journal of
Computer-Mediated Communication, 12, 1143–1168.
Farzan, R., & Brusilovsky, P. (2011). Encouraging user participation in a course
recommender system: An impact on user behavior. Computers in Human
Behavior, 27(1), 276–284.
Farzan, R., DiMicco, J. M., Millen, D. R., Brownholtz, B., Geyer, W., & Dugan, C.
(2008a). When the experiment is over: Deploying an incentive system to all the
users. In Symposium on persuasive technology, Aberdeen, Scotland, April 2008.
Farzan, R., DiMicco, J. M., Millen, D. R., Brownholtz, B., Geyer, W., & Dugan, C. (2008b).
Results from deploying a participation incentive mechanism within the
enterprise. In Proceedings of the twenty-sixth annual SIGCHI conference on human
factors in computing systems, Florence, Italy, April 5–10, 2008 (pp. 563–572).
Filsecker, M., & Hickey, D. T. (2014). A multilevel analysis of the effects of external
rewards on elementary students’ motivation, engagement and learning in an
educational game. Computers & Education, 75, 136–148.
Fishbein, M. (1979). A theory of reasoned action: Some applications and
implications. Nebraska Symposium on Motivation, 27, 65–116.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An
introduction to the theory and research. Reading, MA: Addison-Wesley.
Fornell, C., & Larcker, D. (1981). Structural equation models with unobservable
variables and measurement error. Journal of Marketing Research, 18(1), 39–50.
Gartner (2011). Gartner says by 2015, more than 50 percent of organizations that
manage innovation processes will gamify those processes.<http://
www.gartner.com/it/page.jsp?id=1629214> Accessed 25.11.14.
Gartner (2012). Gartner says by 2014, 80 percent of current gamified applications will
fail to meet business objectives primarily due to poor design. <http://
www.gartner.com/newsroom/id/2251015> Accessed 25.11.14.
Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American
Sociological Review, 25(2), 161–178.
Hair, J. F., Ringle, C. M., & Sarstedt, S. (2011). PLS-SEM: Indeed a silver bullet. Journal
of Marketing Theory and Practice, 19(2), 139–151.
Hakulinen, L., Auvinen, T., & Korhonen, A. (2013). Empirical study on the effect of
achievement badges in TRAKLA2 online learning environment. In Proceedings of
learning and teaching in computing and engineering (LaTiCE) conference, Macau,
March 21–24, 2013 (pp. 47–54).
Hamari, J. (2013). Transforming homo economicus into homo ludens: A field
experiment on gamification in a utilitarian peer-to-peer trading service.
Electronic Commerce Research and Applications, 12(4), 236–245.
Hamari, J. (2015). Do badges increase user activity? A field experiment on effects of
gamification. Computers in Human Behavior.
Hamari, J., & Eranti, V. (2011). Framework for designing and evaluating game
achievements. In Proceedings of the DiGRA 2011 conference: Think design play,
Hilversum, The Netherlands, September 14–17, 2011.
Hamari, J., Koivisto, J., & Sarsa, H. (2014). Does gamification work? A literature
review of empirical studies on gamification. In Proceedings of the 47th
346 J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347
Hawaii international conference on system sciences, Hawaii, USA, January 6–9,
2014.
Hamari, J., Huotari, K., & Tolvanen, J. (2015). Gamification and economics. In S. P.
Walz & S. Deterding (Eds.), The gameful world: Approaches, issues, applications.
Cambridge, MA: MIT Press.
Hamari, J., & Koivisto, J. (2014). Measuring flow in gamification: Dispositional flow
scale-2. Computers in Human Behavior, 40, 133–143.
Hamari, J., & Koivisto, J. (2015). Why do people use gamification services.
International Journal of Information Management.
Hamari, J., & Tuunanen, J. (2014). Player types: A meta-synthesis. Transactions of the
Digital Games Research Association, 1(2), 29–53.
Hernandez, B., Montaner, T., Sese, F. J., & Urquizu, P. (2011). The role of social
motivations in e-learning: How do they affect usage and success of ICT
interactive tools? Computers in Human Behavior, 27, 2224–2232.
Hirschman, E. C., & Holbrook, M. B. (1982). Hedonic consumption: Emerging
concepts, methods and propositions. Journal of Marketing, 46, 92–101.
Hogg, M. A. (1992). The social psychology of group cohesiveness: From attraction to
social identity. New York: Harvester Wheatsheaf.
Hsiao, C.-C., & Chiou, J.-S. (2012). The effect of social capital on community loyalty
in a virtual community: Test of a tripartite-process model. Decision Support
Systems, 54(1), 750–757.
Hsieh, J. J. P.-A., Rai, A., & Keil, M. (2008). Understanding digital inequality:
Comparing continued use behavioral models of the socio-economically
advantaged and disadvantaged. MIS Quarterly, 32(1), 97–126.
Hsu, M.-H., & Chiu, C.-M. (2004). Internet self-efficacy and electronic service
acceptance. Decision Support Systems, 38(3), 369–381.
Hsu, C.-L., & Lin, J. C.-C. (2008). Acceptance of blog usage: The roles of technology
acceptance, social influence and knowledge sharing motivation. Information &
Management, 45, 65–74.
Hsu,C.-L., & Lu, H.-P. (2004).Why do people playon-line games?An extended TAM with
social influencesand flow experience. Information & Management, 41(7), 853–868.
IEEE (2014). Everyone’s a gamer – IEEE experts predict gaming will be integrated into
more than 85 percent of daily tasks by 2020.<http://www.ieee.org/about/news/
2014/25_feb_2014.html> Accessed 06.06.14.
Ipeirotis, P. G., & Gabrilovich, E. (2014). Quizz: Targeted crowdsourcing with a
billion (potential) users. In Proceedings of WWW ’14, Seoul, Korea, April 7–11 (pp.
143–154).
Jones,B. A., Madden,G. J., & Wengreen,H. J. (2014). TheFIT game: Preliminaryevaluation
of a gamification approachto increasing fruitand vegetableconsumption in school.
PreventiveMedicine.http://dx.doi.org/10.1016/j.ypmed.2014.04.015.
Jöreskog, K. G., & Sörbom, D. (1996). LISREL 8 user’s reference guide. Scientific Software.
Katz, M. L., & Shapiro, C. (1985). Network externalities, competition, and
compatibility. The American Economic Review, 75(3), 424–440.
Kelman, H. C. (1958). Compliance, identification, and internalization: Three
processes of attitude change. Journal of Conflict Resolution, 2, 51–60.
Kim, S. S., & Son, J.-Y. (2009). Out of dedication or constraint? A dual model of post-
adoption phenomena and its empirical test in the context of online services. MIS
Quarterly, 33(1), 49–70.
Koivisto, J., & Hamari, J. (2014). Demographic differences in perceived benefits from
gamification. Computers in Human Behavior, 35, 179–188.
Lascu, D., & Zinkhan, G. (1999). Consumer conformity: Review and applications for
marketing theory and practice. Journal of Marketing Theory and Practice, 7(3),
1–12.
Lee, J. J., Ceyhan, P., Jordan-Cooley, W., & Sung, W. (2013). GREENIFY: A real-world
action game for climate change education. Simulation & Gaming, 44(2–3),
349–365.
Lewis, W., Agarwal, R., & Sambamurthy, V. (2003). Sources of influence on beliefs
about information technology use: An empirical study of knowledge workers.
MIS Quarterly, 27(4), 657–678.
Li, Y.-M., Wu, C.-T., & Lai, C.-Y. (2013). A social recommender mechanism for e-
commerce: Combining similarity, trust, and relationship. Decision Support
Systems, 55(3), 740–752.
Liang, H., Saraf, N., Hu, Q., & Xue, Y. (2007). Assimilation of enterprise systems: The
effect of institutional pressures and the mediating role of top management. MIS
Quarterly, 31(1), 59–87.
Lin, H.-F. (2008). Determinants of successful virtual communities: Contributions
from system characteristics and social factors. Information & Management, 45,
522–527.
Lin, C.-P., & Bhattacherjee, A. (2008). Elucidating individual intention to use
interactive information technologies: The role of network externalities.
International Journal of Electronic Commerce, 13(1), 85–108.
Lin, C.-P., & Bhattacherjee, A. (2009). Understanding online social support and its
antecedents: A socio-cognitive model. The Social Science Journal, 46(4), 724–737.
Lin, C.-P., & Bhattacherjee, A. (2010). Extending technology usage models to
interactive hedonic technologies: A theoretical model and empirical test.
Information Systems Journal, 20(2), 163–181.
Lin, K.-Y., & Lu, H.-P. (2011). Why people use social networking sites: An empirical
study integrating network externalities and motivation theory. Computers in
Human Behavior, 27, 1152–1161.
Lindenberg, S. (2001). Intrinsic motivation in a new light. Kyklos, 54, 317–342.
Lott, A. J., & Lott, B. E. (1965). Group cohesiveness as interpersonal attraction: A
review of relationships with antecedent and consequent variables. Psychological
Bulletin, 64(4), 259–309.
Lounis, S., Pramatari, K., & Theotokis, A. (2014). Gamification is all about fun: The
role of incentive type and community collaboration. In Proceedings of ECIS 2014,
Tel Aviv, Israel, June 9–11 (pp. 1–14).
Mäntymäki, M., & Riemer, K. (2014). Digital natives in social virtual worlds: A multi-
method study of gratifications and social influences in Habbo Hotel.
International Journal of Information Management, 34(2), 210–220.
McCauley, C. (1989). The nature of social influence in groupthink: Compliance and
internalization. Journal of Personality and Social Psychology, 57(2), 250–260.
Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill.
Pavlou, P. A., & Fygenson, M. (2006). Understanding and predicting electronic
commerce adoption: An extension of the theory of planned behavior. MIS
Quarterly, 30(1), 115–143.
Pavlou, P. A., Liang, H. G., & Xue, Y. J. (2007). Understanding and mitigating
uncertainty in online exchange relationships: A principal-agent perspective.
MIS Quarterly, 31(1), 105–136.
Pfeil, U., Arjan, R., & Zaphiris, P. (2009). Age differences in online social networking:
A study of user profiles and the social capital divide among teenagers and older
users in MySpace. Computers in Human Behavior, 25, 643–654.
Podsakoff, P. M., MacKenzie, S. B., Lee, J., & Podsakoff, N. P. (2003). Common method
biases in behavioral research: A critical review of the literature and
recommended remedies. Journal of Applied Psychology, 88(5), 879–903.
Preece, J. (2001). Sociability and usability in online communities: Determining and
measuring success. Behaviour & Information Technology, 20(5), 347–356.
Ringle, C. M., Wende, S., & Will, A. (2005). SmartPLS 2.0 M3.<http://www.smartpls.
de/> Accessed 06.06.14.
Riva, S., Camerini, A. L., Allam, A., & Schulz, P. J. (2014). Interactive sections of an
internet-based intervention increase empowerment of chronic back pain
patients: Randomized controlled trial. Journal of Medical Internet Research, 16(8).
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of
intrinsic motivation, social development, and well-being. American Psychologist,
55(1), 68–78.
Shen, X.-L., Cheung, C. M. K., & Lee, M. K. O. (2013). Perceived critical mass and
collective intention in social media-supported small group communication.
International Journal of Information Management, 33(5), 707–715.
Simões, J., Díaz Redondo, R., & Fernández Vilas, A. (2013). A social gamification
framework for a K-6 learning platform. Computers in Human Behavior, 29(2),
345–353.
Sledgianowski, D., & Kulviwat, S. (2009). Using social network sites: The effects of
playfulness, critical mass and trust in a hedonic context. The Journal of Computer
Information Systems, 49(4), 74–83.
Srinivasan, S. S., Anderson, R., & Ponnavolu, K. (2002). Customer loyalty in e-
commerce: An exploration of its antecedents and consequences. Journal of
Retailing, 78, 41–50.
Stibe, A., Oinas-Kukkonen, H., & Lehto, T. (2013). Exploring social influence on
customer engagement: A pilot study on the effects of social learning, social
comparison, and normative influence. In 2013 46th Hawaii international
conference on system sciences (HICSS) (pp. 2735–2744).
Terlutter, R., & Capella, M. L. (2013). The gamification of advertising: Analysis and
research directions of in-game advertising, advergames, and advertising in
social network games. Journal of Advertising, 42(2–3), 95–112.
Thom, J., Millen, D., & DiMicco, J. (2012). Removing gamification from an enterprise
SNS. In Proceedings of the ACM 2012 conference on computer supported cooperative
work, Seattle, Washington, USA, February 11–15, 2012 (pp. 1067–1070).
Thorsteinsen, K., Vittersø, J., & Svendsen, G. B. (2014). Increasing physical activity
efficiently: An experimental pilot study of a website and mobile phone
intervention. International Journal of Telemedicine and Applications.
Tolmie, P., Chamberlain, A., & Benford, S. (2014). Designing for reportability:
Sustainable gamification, public engagement, and promoting environmental
debate. Personal and Ubiquitous Computing, 18(7), 1763–1774.
Tuckman, B. W. (1965). Developmental sequence in small groups. Psychological
Bulletin, 63(6), 384–399.
van der Heijden, H. (2004). User acceptance of hedonic information systems. MIS
Quarterly, 28(4), 695–704.
Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology
acceptance model: Four longitudinal field studies. Management Science, 46,
186–204.
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of
information technology: Toward a unified view. MIS Quarterly, 27(3), 425–478.
Watson, D., Mandryk, R. L., & Stanley, K. G. (2013). The design and evaluation of a
classroom exergame. In Proceedings of gamification ’13, Stratford, ON, Canada,
October 02–04 (pp. 34–41).
Webster, J., & Martocchio, J. J. (1992). Microcomputer playfulness: Development of a
measure with workplace implications. MIS Quarterly, 16(2), 201–226.
Wellman, B., & Wortley, S. (1990). Different strokes from different folks:
Community ties and social support. American Journal of Sociology, 96(3),
558–588.
Williams, L., Edwards, J., & Vandenberg, R. (2003). Recent advances in causal
modeling methods for organizational and management research. Journal of
Management, 29(6), 903–936.
Xiao, B., & Benbasat, I. (2007). E-commerce product recommendation agents: Use,
characteristics, and impact. MIS Quarterly, 31(1), 137–209.
Yee, N. (2006). Motivations of play in online games. Journal of Cyberpsychology and
Behavior, 9, 772–775.
Zhang, P. (2008). Motivational affordances: Reasons for ICT design and use.
Communications of the ACM, 51(11), 145–147.
Zhou, T. (2011). Understanding online community user participation: A social
influence perspective. Internet Research, 21, 167–181.
Zhou, L., Zhang, P., & Zimmermann, H. D. (2013). Social commerce research: An
integrated view. Electronic Commerce Research and Applications, 12(2), 61–68.
J. Hamari, J. Koivisto / Computers in Human Behavior 50 (2015) 333–347 347
