Social Comparison in mHealth: The Role of Similar Others and Feelings of Envy

Abstract

To foster ambitious goal setting, mHealth app developers increasingly implement social comparison features such as leaderboards. However, extant research does not sufficiently look at affective consequences of such features and their impact on goal- setting behavior. We focus on two aspects of social comparison to better address this issue: (1) the similarity of comparison targets and (2) the affective consequence of envy. We distinguish between two similarity dimensions (performance and related attributes) and two distinct emotions of envy (benign and malicious). In an experimental study, we find that comparing to targets similar on related attributes (age and gender) determines the relevance of the comparison and positively impacts benign and malicious envy. We further show that comparing to targets similar in performance (step count) decreases malicious envy and increases benign envy, based on appraisals of perceived control. Moreover, benign and malicious envy differentially impact goal-setting behavior.

Full text

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
1
Social Comparison in mHealth: The Role of
Similar Others and Feelings of Envy
Completed Research Paper
Monica Fallon
University of Mannheim
L15, 1-6 68161 Mannheim, Germany
fallon@uni-mannheim.de
Manuel Schmidt-Kraepelin
Karlsruhe Institute of Technology
76131 Karlsruhe, Germany
manuel.schmidt-kraepelin@kit.edu
Scott Thiebes
Karlsruhe Institute of Technology
76131 Karlsruhe, Germany
scott.thiebes@kit.edu
Simon Warsinsky
Karlsruhe Institute of Technology
76131 Karlsruhe, Germany
simon.warsinsky9@kit.edu
Ali Sunyaev
Karlsruhe Institute of Technology
76131 Karlsruhe, Germany
sunyaev@kit.edu
Abstract
To foster ambitious goal setting, mHealth app developers increasingly implement social
comparison features such as leaderboards. However, extant research does not
sufficiently look at affective consequences of such features and their impact on goal-
setting behavior. We focus on two aspects of social comparison to better address this
issue: (1) the similarity of comparison targets and (2) the affective consequence of envy.
We distinguish between two similarity dimensions (performance and related attributes)
and two distinct emotions of envy (benign and malicious). In an experimental study, we
find that comparing to targets similar on related attributes (age and gender) determines
the relevance of the comparison and positively impacts benign and malicious envy. We
further show that comparing to targets similar in performance (step count) decreases
malicious envy and increases benign envy, based on appraisals of perceived control.
Moreover, benign and malicious envy differentially impact goal-setting behavior.
Keywords: Social Comparison, Envy, mHealth, Gamification, Leaderboards
Introduction
Physical inactivity is a major risk factor for global mortality (6% of deaths globally) with it being the
principal cause for approximately 25% of breast and colon cancer burden, 27% of diabetes, and 30% of the
ischemic heart disease burden (World Health Organization 2020). Despite evidence supporting improved
health outcomes from regular physical activity (World Health Organization 2014), population levels of
physical activity remain low (Guthold et al. 2018). Mobile Health (mHealth) technology has the potential
to impact physical activity behavior. Yet, the extent to which individuals actively use mHealth is often
limited to few initial interactions (Levy 2014) and it is unclear how mHealth use impacts behavior change
(Fallon et al. 2019). For mHealth to be effective, it is not only important that people use mHealth, but also
that they stay motivated over a sustained period of time and set ambitious goals for themselves.
Extant research has shown that setting ambitious goals is associated with higher levels of effort and
performance in physical activity (Shilts et al. 2004). To foster ambitious goal setting, mHealth app
developers increasingly implement social comparison features such as leaderboards that illustrate one’s

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
2
own performance in comparison to others (Schmidt-Kraepelin, Thiebes, Schöbel, et al. 2019). Social
comparison features are based on the idea that social comparison information can have an impact on
people's behavior; an effect described by Festinger in his seminal work on social comparison theory (SCT)
(Festinger 1954). According to SCT, the existence of a discrepancy on a comparison dimension leads to
discrepancy-reducing actions. One example is goal-setting behavior, where people first set a goal for
themselves (Locke and Latham 2006). Then, depending on how their present state compares to their goal
(Carver and Scheier 1982; Locke and Latham 2002), maintain their original goal, lower their goal or adopt
an even more challenging goal (Bandura 1991). Given the case that mHealth users are confronted with social
comparison information after they have initially set a physical activity goal for themselves (e.g., through
recommendations for sufficient levels of physical activity in the media), social comparison information may
likely lead to the adjustment of these goals (Liu et al. 2019).
Researchers have reported inconclusive evidence for the effectiveness of social comparison features on goal
setting behavior in mHealth. While some research shows that social comparison features can have both
positive and negative effects on goal-setting (e.g., Consolvo et al. 2006; Liu et al. 2019), the inner workings
are mostly hidden and we know only little about the underlying reasons why such features can lead to
contradicting effects. Accordingly, it remains largely unclear how to leverage features that enhance the
positive effects and mitigate the negative effects, which is crucial when it comes to designing mHealth apps
with the goal to positively influence users' health behavior. To overcome this lack of knowledge and shine
light into the black box of social comparison features in mHealth, we focus on two different aspects of social
comparison: (1) the similarity of comparison targets and (2) the affective consequence of envy. It is widely
presumed that individuals who are similar are the best comparison targets because they determine
likelihood for success (Wheeler et al. 1997). However, the cognitive, affective and behavioral consequences
of comparing to similar or dissimilar others have not yet been fully investigated. In addition, several
researchers in domains other than mHealth have shown that social comparison information can evoke
feelings of envy (Krasnova et al. 2015; van de Ven 2017). Traditionally, envy has been conceptualized as a
unitary construct describing an unpleasant emotion in which one feels inferior, resentful or even hostile
(Smith and Kim 2007). More recent research suggests that two distinct forms of envy exist (i.e., benign envy
and malicious envy) that result in distinct behavioral consequences (van de Ven 2016). We follow this line
of research and explore how who users compare themselves to impacts the two distinct feelings of envy and
how this may influence goal-setting behavior. Specifically, we ask the following research questions:
RQ1: What is the role of similar comparison targets on feelings of benign and malicious envy?
RQ2: How do benign and malicious envy differentially impact goal increase behavior?
In order to answer our research questions, we conducted an online experiment among 285 potential users
of mHealth apps for physical activity. Within our online experiment, we followed a between-subject design
and provided participants with social comparison information of comparison targets with different levels
of similarity. Our results indicate that comparing to similar comparison targets can increase feelings of both
benign and malicious envy and that benign and malicious envy have differential impacts on goal increase
behavior. The study contributes to information systems (IS) literature in three key ways. First, we
contribute to the growing body of literature that explores affective factors influencing user behaviors (Stein
et al. 2015). Second, we extend our knowledge concerning the circumstances that determine when social
comparison features lead to positive user experiences and when they yield negative outcomes (Schmidt -
Kraepelin, Thiebes, Stepanovic, et al. 2019). Third, we take a closer look at how social comparison features
can impact self-regulation of behavior in mHealth (Fallon et al. 2019).
This paper proceeds as follows. In the next section, we briefly introduce the landscape of extant research on
social comparison in mHealth and describe the theoretical foundations of SCT and envy. Afterwards, we
develop our hypotheses and present our research model. Then, we outline the applied research method,
including our experimental design. Subsequently, we present our results. We outline implications of our
findings, limitations of the study, and opportunities for future research in the discussion section, before we
briefly conclude our paper in the last section.

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
3
Theoretical Background
Social Comparison in mHealth
Social comparison features have been widely implemented in mHealth physical activity apps,
predominantly in the form of social networking features (e.g., messages including social norms, or group
competitions), leaderboards and features allowing the deliberate sharing of physical activity (e.g., sharing
your daily step count with a group of friends). However, research has yielded inconclusive findings
regarding the effects of social comparison features. For example, whereas social comparison features can
increase engagement with mHealth (e.g., Anderson et al. 2007), they are also associated with avoidance of
an app because the comparison is perceived as forced and unwanted (Jia et al. 2017; Miller and Mynatt
2014). Similar contrasts are seen regarding the effects of social comparison on physical activity motivation.
Whereas social comparison can increase physical activity motivation by increasing awareness of others’
physical activity levels, it can also decrease motivation if people are constrained in converting the increased
awareness into actual physical activity (Anderson et al. 2007; Jia et al. 2017; Wu et al. 2015).
When investigating effects of social comparison features, extant research largely treats social comparison
as a black box and focuses on measuring its direct effect on physical activity behavior (e.g., Chen et al. 2016;
Lee and Lim 2015) or usability of an mHealth app (e.g., Middelweerd et al. 2015; Zuckerman and Gal-Oz
2014). To the best of our knowledge, there are only few studies that aim to open this black box by
investigating cognitive or affective consequences of social comparison features. Table 1 provides a selection
of these studies that help us understand the causal links between social comparison and physical activity
behavior. The results indicate that there are in fact important factors that explain the effect of social
comparison on physical activity, such as perceived competitive climate (Wu et al. 2015), self-efficacy (Miller
and Mynatt 2014) or goal-setting (Arigo 2015; Chen et al. 2017; Liu et al. 2019). Another stream of research
is concerned with investigating the impact of different characteristics of the social comparison information
(e.g., with whom individuals compare themselves to or how they are placed on a leaderboard) on physical
activity motivation. For example, extant research indicates that social comparison to individuals that are
perceived as highly dissimilar can lead to decreased motivation (Arigo 2015) and that social comparison to
foreign individuals may result in more long-term motivational benefits (Fritz et al. 2014). Furthermore,
after manipulating individuals’ position on a fitness leaderboard, Jia et al. (2017) found that individuals
enjoyed the social comparison most when being placed at the bottom, which may translate to a heightened
sense of physical activity motivation. Overall, the largely inconsistent findings show that it is crucial to
consider cognitive and affective consequences of social comparison features in mHealth physical activity
apps as well as the effects of different social comparison information.
Study
Social comparison
features
Affective, cognitive or goal-related
outcome
Overall
effects
of SC
Manipula
tion of SC
elements
Social
networking
Leader-
boards
Social
sharing
Anderson et al. (2007)
X
Qualitative analysis of SC outcomes
positive
No
Arigo et al. (2015)
X
X
Goal setting, negative responses to SCs
mixed
No
Chen et al. (2017)
X
Goal setting
positive
No
Fritz et al. (2014)
X
Qualitative analysis of SC outcomes
mixed
No
Jia et al. (2017)
X
Enjoyment with app, motivation for PA
mixed
Yes
Liu et al. (2019)
X
Goal setting, goal attainment
mixed
No
Miller & Mynatt (2014)
X
Self-efficacy
positive
No
Patel et al. (2016)
X
Goal achievement
mixed
Yes
Tong et al. (2018)
X
Qualitative analysis of SC outcomes
mixed
No
Wu et al. (2015)
X
Attitude towards PA
mixed
No
This study
X
Benign and malicious envy, goal setting
?
Yes
Table 1. Overview of Studies Measuring Affective, Cognitive or Goal-related Outcomes of SC

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
4
Social Comparison Theory
Social comparison is defined as “the process of thinking about information about one or more people in
relation to the self” (Wood 1996). First proposed by Festinger (1954), SCT describes that there is a natural
drive within us to compare upwards to better performing individuals. When there is a discrepancy between
an individual and a superior other, the individual uses the information to self-evaluate their standing and
abilities, which leads to action to reduce that discrepancy (Festinger 1954). Since the seminal work from
Festinger, researchers have focused on the cognitive and affective consequences of social comparison
(Gerber et al. 2017). In summary, this research has found that people generally choose to compare with
people who are superior to them in some way, which has subsequent consequences on a person’s cognitive
appraisals about their abilities and emotion.
Who user’s compare to is especially important for obtaining accurate self-evaluations about their abilities.
Festinger (1954) proposes that individuals choose similar others as comparison targets. While his original
theoretical statement emphasizes similarity regarding performance (e.g., comparing to others who take a
similar number of steps per day), some of his discussion focuses on similarity regarding related attributes
(e.g., gender and age). Consequently, researchers were troubled with operationalizing similarity. The proxy
model of social comparison aimed to overcome this problem by distinguishing between similarity on two
dimensions - the dimension of performance and the dimension of related attributes (Wheeler et al. 1997).
Performance refers to how similar the comparison others have performed at the task at hand (e.g., do the
comparison targets take a similar number of steps per day as me?). Related attributes refer to how similar
the others are on attributes known to influence performance (e.g., are the comparison targets similar to me
regarding age and gender?). The proxy model of social comparison proposes that individuals who are
similar on the performance dimension and on the related attributes dimension will be the best comparison
targets because they determine likelihood for success (Wheeler et al. 1997).
Benign and Malicious Envy
Several researchers have found envy to be a prominent affective consequence of social comparison (e.g.,
Krasnova et al. 2015; van de Ven 2017). Originally, envy was described as an unpleasant and painful emotion
in which one feels inferior, resentful, and hostile (Smith and Kim 2007). However, more recent research
suggests that two distinct forms of envy with distinct behavioral patterns prevail. On the one hand, benign
envy is an emotion that leads to positive improvement for oneself through a moving-up motivation. On the
other hand, malicious envy leads to hostile feelings toward the envied person through motivations aimed
at pulling-down the other from the superior position (van de Ven et al. 2009). Social comparison has the
potential to yield both forms of envy (van de Ven 2017; van de Ven et al. 2009).
From a theoretical standpoint, there are two reasons to distinguish between benign and malicious envy.
The first is grounded in appraisal theory (Roseman 1996), which states that specific emotions are caused
by a specific mix of appraisal perceptions of the situation. Emotions with different appraisals are considered
distinct emotions. Perceived control of the situation is a crucial appraisal when considering the subsequent
effect on the feeling of envy. Perceived control refers to one’s perceived ability to do something about the
situation (van de Ven 2016). Some researchers even argue that low perceived control is a necessary
condition for envy to occur (Ortony et al. 1988; Smith 1991). In fact, perceived control is an appraisal that
can distinguish between feelings of benign and malicious envy (van de Ven et al. 2012). The second reason
to distinguish between the two types of envy is grounded in a functional approach to emotions (Cosmides
and Tooby 2000). This functional approach comes from Arnold (1960), who defined emotions as felt action
tendencies, and Frijda (1993), who argued that changes in action readiness are the distinguishing factor of
emotions. The functional approach to emotions implies that very distinct action tendencies are unlikely to
be caused by the same emotion. For example, the action tendency to improve oneself through a moving-up
motivation, which is associated with benign envy or the action tendency to pull-down others from the
superior position, which is associated with malicious envy.
The research summarized in this section suggests that distinguishing between benign and malicious envy
is a suitable theoretical basis for understanding the impact of social comparison features in mHealth. The
distinct action tendencies as a result of benign and malicious envy are likely to play a key role in goal-setting
behavior. We propose that using this theoretical basis could explain the mixed results of social comparison
features in mHealth research. As opposed to looking at the direct effect of social comparison on physical

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
5
activity behavior, we aim to better understand the theoretical reasons why this occurs. We do this by
identifying two similarity dimensions and theorizing about their impact on benign and malicious envy. We
propose that these two emotions have unique consequences on goal-setting behavior.
Research Model and Hypotheses
In light of the existing gaps in the literature, we aim to understand social comparison and its influence on
goal-setting behavior through benign and malicious envy. We rely on literature from SCT and the affective
consequence of envy to develop our research hypotheses. Based on these hypotheses, we have
conceptualized our research model as depicted in Figure 1.
Figure 1. Research Model
The proxy model of social comparison proposes that the comparer assesses how similar they are on the
performance dimension in order to determine likelihood for success or attainability (Wheeler et al. 1997).
Consistent with this, we propose that when there is a small discrepancy regarding the performance
comparison information on a leaderboard (e.g., higher similarity in step-count), achieving the same step-
count as better performing others will seem attainable. This appraisal of the situation is important for self-
evaluation and determining what one is capable of doing (Festinger 1954). If the comparer perceives
similarity on the performance dimension, they will also appraise the situation as being attainable and
therefore experience higher perceived control of the situation (van de Ven et al. 2012). Conversely, if there
is a large discrepancy regarding the performance comparison information on a leaderboard (e.g., lower
similarity in step-count), achieving the same step-count as the better performing others can seem almost
impossible and therefore the individual will appraise the situation as having low ability to control it.
Consequently, we hypothesize:
H1: Similarity on the performance dimension will be positively associated with appraisals of perceived
control.
Appraisals that lead to certain emotions are subjective perceptions of the situation (Scherer et al. 2001). In
our context, perceived control is an appraisal of the situation regarding one’s standing in comparison to
others and the perceptions of their ability to attain the same step count. It is the appraisal of the situation
itself that leads to specific emotions (Roseman 1996; van de Ven et al. 2011). We expect that depending on
the appraisal of perceived control of the situation individuals will experience one of two distinct emotions
– benign or malicious envy. We propose that individuals that perceive high control over attaining a higher
step count will also feel more benign envy because there is an opportunity to improve, which seems
attainable (van de Ven et al. 2011, 2012). The key distinction is not the attainability of the position of the

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
6
superior other, but rather whether the person who compares upwards perceives the situation to be within
their control. The opportunity is appraised as within one’s control and therefore individuals will experience
benign envy, characterized as motivations to improve oneself. Conversely, we propose that individuals that
perceive low control over attaining a higher step count will experience more malicious envy because they
do not have the opportunity to act constructively (Smith 1991). Analogue to benign envy, it is not the lack
of attainability of the position of the superior other, but rather whether the person who compares upwards
perceives the situation to be within their control. The opportunity to improve is appraised as being out of
one’s control and therefore individuals will experience malicious envy characterized as hostile feelings
towards the individuals performing better. Moreover, if the opportunity to improve is appraised as out of
one’s control, they will feel little motivation to improve themselves. Therefore, we hypothesize:
H2a: Appraisals of high perceived control will lead to more feelings of benign envy.
H2b: Appraisals of low perceived control will lead to more feelings of malicious envy.
Festinger (1954) proposes that comparing to others similar in age and gender is a more relevant comparison
target and people actually compare themselves more often when they are similar on related attributes. The
reason behind this is that information regarding such similar comparison targets reveals more information
about one’s own performance. Since users have only limited access to relevant information such as athletic
ability or effort when using a leaderboard, they draw on related attributes to put their own performance in
relation to that of others (Wheeler et al. 1997). In this regard, related attributes are seen as factors that help
to determine sufficient levels of performance. If a comparison target that is similar with regard to related
attributes has shown better performance, this likely produces the feeling that one’s own performance is not
sufficient. However, if the comparison target is different, information on that target’s performance yields
less information about one’s own performance since other levels of physical activity might be considered to
be more appropriate for them. As a consequence, comparing to similar others on related attributes has a
higher likelihood of provoking an emotional response than comparing with dissimilar others. Thus, it can
potentially increase feelings of both benign and malicious envy (van de Ven et al. 2009). This makes the
similarity of related attributes a double-edged sword. On the one hand, if individuals compare their step
count to others on a leaderboard that are the same age and gender, they will make more comparisons
(Festinger 1954) and are likely to be confident that they can also achieve the same step count (Wheeler et
al. 1997). In this case, comparing with similar others will lead to feelings of benign envy, characterized as
the motivation to improve oneself because similar others accomplished it. On the other hand, if individuals
compare their step count to others on a leaderboard that are the same age and gender, this may lead to the
comparer doubting that their abilities really match individuals with the same related attributes (Salovey
and Rodin 1984). In this case, they will begin searching for factors that might mitigate the success of others.
For example, the comparer will recall all negative aspects of individuals performing better and will perceive
them as being of little worth (Salovey and Rodin 1984). This results in the consequence of malicious envy,
characterized by hostile feelings aimed at pulling the other down. Consequently, we hypothesize:
H3a: Similarity on the related attributes dimension will be positively associated with benign envy.
H3b: Similarity on the related attributes dimension will be positively associated with malicious envy.
Benign envy is associated with the distinct action tendency to improve oneself (van de Ven et al. 2009). This
is evident in that individuals experiencing benign envy focus their attention on the object they envy the
person for (e.g., having a high step count) (Crusius and Lange 2014). Focusing on the object of envy is what
motivates one to improve. People experiencing benign envy actually work longer on tasks, perform better
and plan to study more (van de Ven et al. 2011). Thus, when given the opportunity to increase their goals,
we propose that individuals experiencing benign envy will also be more likely to increase their step count
goal because their attention is focused on a higher step-count. In this scenario, the social comparison
information acts as social feedback, in which individuals can evaluate their abilities in comparison to others,
appraise the situation as being within their control and experience the emotion of benign envy or the
motivation to improve themselves. This will be evident in a goal increase, in which they set higher and more
ambitious goals for themselves. Therefore, we hypothesize:
H4: Feelings of benign envy will be positively associated with a goal increase.
Malicious envy is associated with the distinct action tendency to pull others down or wishing for others to
fail (van de Ven et al. 2009). This is evident in that individuals experiencing malicious envy focus their

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
7
attention on individuals who are performing better (not on the object they envy as is the case with benign
envy) (Crusius and Lange 2014). Because malicious envy is characterized with an attentional focus on others
and the wish for others to fail, it has been thought of as the more undesirable emotion that can deplete self-
regulatory resources (Crusius and Lange 2014; Hill et al. 2011). In this scenario, the social comparison
information acts as social feedback, in which individuals evaluate their abilities in comparison to others,
appraise the situation as being out of their control and experience the emotion of malicious envy or the
tendency to pull others down. This focus on others and the experience of negative emotions will deplete
self-regulatory resources and be evident in a negative effect on goal increase. Therefore, we hypothesize:
H5: Feelings of malicious envy will be negatively associated with a goal increase.
Methods
Data Collection
We employed an experimental approach to test the research hypotheses. Data were collected using an
online survey administered by a third party organization (Chandler and Shapiro 2016). In line with
suggestions in literature, we restricted participation to users with a high reputation (at least 99% approval
ratings and at least 5,000 conducted tasks) in order to ensure high data quality (Peer et al. 2014). We also
included an attention check question to remove responses of participants who were not reading the
questions and simply clicking an answer choice (Thomas and Clifford 2017). Because we collected data
through an online survey, procedural methods were used to control for common method bias (CMB). We
applied ex ante recommendations of Podsakoff, MacKenzie, Lee, and Podsakoff (2003) to control for CMB,
including instructing participants that answers are fully anonymized, that they should take their time to
carefully and honestly answer the questions and that no right and wrong answers exist; counterbalancing
question orders; using existing, reliable measures; randomizing items; and proximal separation (i.e.,
different pages) of measurements for independent and dependent variables.
After answering demographic questions, participants were randomized to receive one of four different
leaderboards. The leaderboard was framed as being part of an existing mHealth app (see Figure A-1 in the
Appendix). It included ten rank positions, with the participant always being placed on rank five.
Participants were placed on rank five to ensure they could compare both upwards and downwards (Wu et
al. 2015). The social comparison information shown to each group differed with respect to low or high
similarity of other people on the leaderboard in terms of performance (step count) and related attributes
(age and gender). An overview of the leaderboard manipulation is shown in Table 2.
Similarity of performance
low
high
Similarity of
related attributes
low
Group 1
Group 3
high
Group 2
Group 4
Table 2. Leaderboard Manipulation
Participants in groups with high similarity on the performance dimension were shown a leaderboard where
the top and bottom rank had a step count that is 10% higher or lower than theirs, respectively. In groups
with low similarity on the performance dimension the top rank had a step count that is 80% higher than
that of the participant. These values were determined based on a pre-test on the extent to which the
leaderboard manipulations impacted perceived similarity on the performance dimension . The step count
increments between ranks were consistent above and below the user rank respectively. To infuse realism,
the increments were multiplied with a random factor between .95 and 1.05 and rounded to the nearest
whole number.
Participants in groups with high similarity on the related attributes dimension were shown a leaderboard
where everyone was of the same gender and also of an age not differing more than five years from theirs.
Contrary, the leaderboard for the groups with low similarity on the related attributes dimension included
only people that are of the different gender and differing at least 10 years in age. The genders of the other
people on the leaderboard were not explicitly shown. Rather, they could be derived from their names, as we
took caution to only use names that could be unambiguously assigned to either gender. As an example,

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
8
Figure A-1 in the appendix shows the four leaderboards corresponding to the four different groups for a
participant that is a 40-year-old male and has an average daily step count of 10,000. Participants then
completed questions related to the other variables in our model.
Operationalization of Variables
When possible, we adapted measures from prior studies and used multi-item scales to improve reliability
and validity. We illustrated in a pre-test that the different leaderboard manipulations described above
impact mental states (e.g., perceived similarity). Lench et al. (2014) argue that it is better to show that a
specific mental state (e.g., perceived similarity) produces an outcome rather than to simply show that the
manipulation produces both the mental state and the final outcome. Consistent with this reasoning, we
used perceived similarity of performance and perceived similarity of related attributes as our independ ent
variables. This allows individual variation in responses to the different leaderboards to be included in the
experimental approach rather than a source of error that should be minimized in between-subject designs
(Lench et al. 2014). The items used to measure perceived similarity of performance and related attributes
are specified in the Table A-1 in the appendix. Additionally, all other items (benign envy, malicious envy
and perceived control) were adapted measures from prior studies and can also be found in the Table A-1 in
the appendix. Goal increase was measured as a binary variable. If a participant’s goal increased after viewing
the social comparison information they were coded as 1.
The survey also captured demographic variables including age and gender. We controlled for the potential
influence of other variables that are not central to this study but that are likely to influence envy or goal -
setting behavior based on prior research. This includes dispositional benign envy, dispositional malicious
envy (Lange and Crusius 2015) and social comparison orientation (Gibbons and Buunk 1999). Besides the
emotion of envy, which we focus on in our study, researchers have found use for measuring dispositional
benign and malicious envy, which measures individual differences in the tendency to experience envy (a
personality trait) (Lange and Crusius 2015). Additionally, we control for social comparison orientation,
which is a stable personal trait that describes the extent to which people generally make comparisons about
their opinions, abilities and general aspects of themselves (Gibbons and Buunk 1999). We used the
measures from prior studies to measure these variables, which can be found in Table A-1 in the appendix.
Results
Overall, we gathered responses from 285 participants residing in the U.S. 42 responses were excluded from
analysis because participants failed to correctly answer the attention-check question. Of the 243 remaining
participants, 96 were female and 147 were male. On average, participants were 36.55 years old (min. 20
years, max. 69 years), had taken 5,579 steps per day during the last seven days and reported to have a daily
step count goal of 6,950 steps. 148 participants reported that they had assessed their step counts via a
smartphone application, 66 used a dedicated activity tracker, 26 participants self-assessed their step counts
and 3 participants used another (undefined) means to keep track of their step counts.
The survey results were used to validate both, the construct measurement scales and the proposed
theoretical relationships. First, we assessed the measurement scales for validity and reliability (Table 3).
Two indicators of the social comparison orientation construct were eliminated from analysis since they
displayed outer loadings below .6. Furthermore, two indicators of the construct had loadings between .6
and .7, but we retained the indicators due to the explorative nature of our study (Hair et al. 2013), and
because the construct’s composite reliability and Cronbach’s Alpha were above the recommended threshold
of .7 (Hair et al. 2011). Besides that, all indicators fulfilled the minimum loading requirements between the
indicator and its corresponding underlying factor, showing convergent validity. All constructs showed
sufficient composite reliability (CR) values greater than .7 and, thus, exceeded the suggested thresholds
(Nunnally 1978). The average variance extracted (AVE) for each construct was greater than the suggested
minimum of .5 (Fornell and Larcker 1981) and the square root of each construct’s AVE exceeded the inter-
construct correlations, demonstrating adequate discriminant validity (Table 3). In addition, the heterotrait-
monotrait (HTMT) ratio of correlations showed only values below the threshold of .85 for all con structs,
also suggesting no discriminant validity problems (Henseler et al. 2014).

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
9
Construct
CR
AVE
Inter-construct correlations
1)
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
(1) Similarity (Related Attributes)
.975
.907
.953
(2) Similarity (Performance)
.923
.858
.142
.926
(3) Perceived Control
1
1
.068
.266
1
(4) Malicious Envy
.936
.831
.166
-.146
-.109
.911
(5) Benign Envy
.933
.776
.288
.185
.342
.282
.881
(6) Goal Increase
1
1
.042
-.030
.160
.106
.207
1
(7) Dispositional Malicious Envy
.938
.752
.059
-.086
-.168
.459
-.082
.083
.867
(8) Dispositional Benign Envy
.927
.719
.268
.082
.258
.355
.401
.198
.101
.848
(9) Social Comparison Orientation
.922
.568
.107
.114
.148
.468
.227
.088
.314
.528
.754
(10) Age
1
1
.072
.004
-.011
-0.122
.006
-.055
-.251
.006
-.163
1
(11) Gender
1
1
-.072
-.099
-.019
.089
-.022
-.006
.096
-.022
.007
-.088
1
Table 3. Construct Reliability and Correlations
The effects of the leaderboard manipulations on perceived similarity were as expected and consistent with
the pre-test. Those who received a high similarity of performance leaderboard perceived a higher degree of
performance similarity (Group 3, M =5.8, SD = 1.05) (Group 4, M =5.9, SD = .91) than those who received
the low similarity leaderboard (Group 1, M = 4.5, SD = 1.33) (Group 2, M =5.1, SD = .97). There was a
statistically significant difference between the high group means and the low group means as determined
by one-way ANOVA (F (3,238) = 20.35, p < .00). Those who received the high similarity of related attributes
leaderboard perceived a higher degree of related attributes similarity (Group 2, M =6.0, SD = .91) (Group
4, M =6.1, SD =.89) than those who received the low similarity leaderboard (Group 1, M =1.8, SD = 1.19)
(Group 3, M =2.0, SD =1.26). There was a statistically significant difference between the high group means
and the low group means as determined by one-way ANOVA (F (3,238) = 293.63, p < .00).
To test our hypotheses and assess our model, we adopted PLS-SEM and used the SmartPLS software,
version 3.3.0 (Ringle et al. 2015). Given the exploratory nature, our PLS-SEM is considered an appropriate
data analysis approach (Urbach and Ahlemann 2010). The significance of the structural path estimates was
assessed using bootstrapping with 5,000 subsamples and bias-corrected, accelerated confidence intervals
(Ringle et al. 2015). We tested the structural model by evaluating the direct effects and the explained
variances (R²). While assessing the model, we controlled for dispositional benign envy, dispositional
malicious envy, social comparison orientation, age and gender. Figure 2 shows the identified direct effects.
Figure 2. Analysis Results (Structural Model)
*Significant at 5%; ∗∗ significant at 1%; *** significant at .1% or lower

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
10
The results show that similarity (performance) positively influences perceived control (.266; p < .01; [.103;
.414]), supporting H1. Furthermore, perceived control showed a positive effect on benign envy (.256; p
< .001; [.129; .380]), supporting H2a and a negative effect on malicious envy (-.120; p <.05; [-.217; -
.030]), supporting H2b. We also found that similarity (related attributes) positively influences both,
benign envy (direct effect: .196; p < .01; confidence interval [.077; .312]) and malicious envy (.116; p < .05;
[.007; .230]), supporting H3a and H3b. Benign envy positively influences goal increase (.171; p < .01
[.058; .285]), supporting H4, but malicious envy showed no significant effect on goal increase (-.018; p
= .822; [-.172; .137]), not supporting H5. Among the five control variables, only dispositional benign
envy showed an association with goal increase. As expected, benign envy was influenced by dispositional
benign envy and malicious envy was influenced by dispositional malicious envy. Furthermore, social
comparison orientation influenced malicious envy. In terms of explained variances (R²), our model
explains 7% of perceived control, 26% of benign envy, 35% of malicious envy, and 7% of goal increase.
To provide deeper insights into the impact of similarity and the mediating role of envy and perceived
control, we conducted a mediation analysis. Again, we used bootstrapping with 5,000 subsamples and bias-
corrected, accelerated confidence intervals while controlling for dispositional benign envy, dispositional
malicious envy, social comparison orientation, age and gender. The results reveal that the effect of
similarity (performance) on benign envy is fully mediated by perceived control (specific indirect effect:
.063; p < .05; [.018; .136]; direct effect: .071; p = .346; [-.069; .229]) and that the effect of perceived control
on goal increase is partially mediated by benign envy (specific indirect effect: .038; p < .05: [.010; .085];
direct effect: .122; p < .05; [.015; .218]). For the overall influence of similarity (performance) on goal
increase mediated via perceived control and benign envy, no significant effect was found (specific indirect
effect: .010; p = .106; [.002; .029]; direct effect: -.084; p = .233; [-.216; .057]). No significant effect was
found for the influence of similarity (performance) on goal increase mediated via benign envy (specific
indirect effect: .030; p = .073; [.005; .075]; direct effect: -.039; p = .575; [-.169; .106]), although the p-value
was close to the threshold of .05. Also, for the effect of similarity (performance) on malicious envy no
significant mediation via perceived control was found (specific indirect effect: -.022; p = .163; [-.063; 0];
direct effect: -.163; p < .01; [-.277; -.045]).
Discussion
Principal Findings
In this work, we aimed to better understand the inconclusive findings on the extent to which social
comparison features impact goal-setting behavior by focusing on the similarity of comparison targets and
the affective consequences of benign and malicious envy. Our findings show that the impact of social
comparison features on goal-setting behavior is to some extent driven by two distinct feelings of envy and
that envy in mHealth is driven by the degree of similarity to comparison targets. Our study yields three key
findings, which are summarized in Table 4.
Previous research gaps
Key Findings
Extant research shows that social comparison can increase as
well as decrease individuals’ motivation towards physical
activity. However, we know little about how the design of social
comparison information influences these outcomes. More
research is needed to understand how mHealth apps can evoke
positive emotional responses to social comparison (e.g., benign
envy) and avoid negative ones (e.g., malicious envy).
Similarity on the related attributes dimension is a
double-edged sword, which can increase feelings
of both benign and malicious envy.
Similarity on the performance dimension impacts
feelings of benign and malicious envy via two
unique pathways related to perceptions of
perceived control.
Extant research has treated social comparison in mHealth
mainly as a black box and produced contradicting findings. First
studies started to investigate the influence of affective and
cognitive outcomes of social comparison information. However,
more research is needed to understand why social comparison
information leads to positive as well as negative outcomes.
Benign and malicious envy have differential
impacts on goal increase. While our results
indicate a positive relationship between benign
envy and goal increase, we did not find a
significant influence of malicious envy.
Table 4. Summary of Key Findings
By distinguishing two types of similarity (i.e., similarity on related attributes and similarity on performance)
and investigating their impact on the emotional responses of benign and malicious envy, we contribute to

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
11
research regarding the impact of social comparison information characteristics on affective and cognitive
outcomes. Thereby, we find that similarity on the related attributes dimension may be beneficial as well as
detrimental for individuals since it can increase feelings of both benign and malicious envy. Similarity on
the performance dimension, on the other hand, may increase benign envy and decrease malicious envy
mediated by perceptions of perceived control. Furthermore, our research shows that benign and malicious
envy have differential impacts on goal increase behavior. While we found a positive relationship between
benign envy and goal increase, our study results did not show a significant influence of malicious envy.
Theoretical and Practical Contributions
Our research findings have several theoretical contributions. First, our research distinguishes between two
unique pathways in which performance similarity impacts benign and malicious envy. We find that
performance similarity positively impacts appraisals of perceived control and benign envy. Conversely, low
perceptions of performance similarity have adverse effects on appraisals of perceived control, which results
in malicious envy. We provide theoretical arguments for why it is important to distinguish between the two
types of envy and hypothesize why different perceptions of performance similarity will have differential
impacts on benign and malicious envy. So far, IS research has focused on envy as a negative emotion in the
context of social comparison (Krasnova et al. 2015). Our findings also find support for the effect of social
comparison on the positive emotion of benign envy. Toward this end, we further contribute by providing
reasons why IS researchers should consider benign and malicious envy as a result of social comparison.
Second, we illustrate that benign and malicious envy have differential impacts on goal-setting behavior.
This is evident in that benign envy positively predicts goal increase behavior. While the hypothesized
negative effect of malicious envy on goal increase was insignificant, it is still evident that malicious envy
does not result in increased goal-setting. We attribute these effects to the distinct action tendencies
associated with benign and malicious envy. This finding helps explain previous inconclusive findings in
mHealth on the extent to which social comparison impacts goal-setting behavior. Third, we identified two
similarity dimensions that can be represented in mHealth leaderboards, which also impact users perceived
similarity on the associated dimensions. Based on the proxy model of social comparison, we identified
similarity on the performance dimension and similarity on the related attributes dimension. We show that
manipulating the design of leaderboards on these two dimensions (performance and related attributes) also
impacts associated perceptions of similarity. To the best of our knowledge, the separate effects of these two
similarity dimensions has not been explored in mHealth research. While researchers have implied that
similarity of users may be important to consider when implementing social comparison features, so far only
qualitative findings have been presented (Tong et al. 2018).
Our study yields practical implications for how to design mHealth in a way that leads to positive user
experiences and avoids negative outcomes (Liu et al. 2017; Schmidt-Kraepelin, Thiebes, Stepanovic, et al.
2019). First, we show that benign envy can play a vital role in fostering mHealth users to set more ambitious
goals for themselves. Thus, developers of mHealth apps for physical activity should aim to evoke feelings of
benign envy in their users in order for them to enhance their physical activity behavior. Our findings suggest
developers may achieve this by carefully selecting comparison targets that are similar with regard to related
attributes and performance and, thus, yield relevant social comparison information. Conversely, our results
also show that higher levels of similarity with regard to related attributes are associated with higher levels
of malicious envy which makes this dimension of similarity a double-edged sword for mHealth app
developers. Although our results do not indicate an effect of malicious envy on goal increase behavior,
extant research has shown that malicious envy is not a desirable affective response to social comparison
information since it largely results in negative consequences, such as reduced cognitive and affective well-
being (Krasnova et al. 2015) and decreased motivation (Utz and Muscanell 2018). Our findings also
highlight the vital role of perceived control to increase levels of benign envy and decrease levels of malicious
envy. Thus, developers of mHealth apps that integrate leaderboard functionalities should implement
measures that increase levels of perceived control in their users. This may be achieved by providing social
comparison information of others that are similar with regard to their performance. However, developers
should also bear in mind that the social comparison information should leave sufficient space for users to
adjust their goals to a more ambitious (but still realistic) one.

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
12
Limitations and Future Research
We acknowledge several limitations of our study that provide avenues for future research. First, we address
the issue of the low R² of perceived control. This may be explained by other dimensions of social comparison
that are important for impacting perceived control (e.g., comparing to known others, advantaged others or
disadvantaged others). Additionally, there may be certain personality traits that inherently make
individuals more likely to appraise a situation as being controllable (e.g., agreeableness or
conscientiousness). We did not consider such personality factors in our study. Consistent with prior
research, we measured perceived control with a single item (Van Dijk & Zeelenber 2002). While PLS-SEM
has shown to be able to easily handle single-item constructs, a multiple item measures for perceived control
could enhance scale reliability (Hair et al. 2016). Future research should consider measuring perceived
control with a multiple item scale. Second, we address the low R² of goal increase. There may be other
important cognitive-affective factors to consider besides benign and malicious envy that impact goal
increase. Envy has been found to be a salient consequence of social comparison and our aim was to
distinguish between two distinct feelings of envy to better explain positive and negative impacts on goal
increase. Thus, addressing other potential factors did not fit the scope of this study. Additionally, other
aspects of goal-setting behavior besides goal increase (e.g., goal-attainment or goal disengagement) might
be better explained by benign and malicious envy. Because we used an online sample to test our hypotheses,
we were not able to capture goal-attainment or goal disengagement over a longer period of time. Future
research can explore these avenues to better understand the impact of social comparison on goal-related
behavior. mHealth research proposes that a variety of features can impact goal-setting behavior. For
example, social comparison, social support, self-incentives and extrinsic incentives are all expected to
contribute to goal-setting behavior (Fallon et al. 2019). We only looked at social comparison and, thus, other
mHealth features were not the focus of this study. Future research should explore other mHealth features,
affective consequences, and the impact on goal-setting behavior. Third, we address the non-significant
negative effect of malicious envy on goal increase. We hypothesized that the feeling of malicious envy will
deplete self-regulatory resources and be evident in a negative effect on goal increase. While the negative
effect on goal-increase was non-significant, the effect was in the expected direction. Similar to the reasoning
explained above, we only focused on one aspect of goal-setting behavior (goal increase). The negative effects
of malicious envy could be better explained with other goal-setting behavior outcomes, such as goal
disengagement or goal attainment. Future research can capture such outcome measures with a long-term
study design to address the impact of malicious envy on other goal-setting behavior over time. Despite the
discussed limitations, our study shows that benign and malicious envy are relevant consequences of social
comparison and to some extent predictors of goal-setting behavior for physical activity. Further research
can consider other systems and domains besides the mHealth context to evaluate if there are generalizable
characteristics of the technology and implementation context (e.g., efficiency numbers of employees), which
influence the degree of benign and malicious envy.
Conclusion
To sum up, this study aimed to understand the impact of social comparison on the affective consequence of
envy and goal-setting behavior. We find that comparing to targets with similar related attributes (age and
gender) determines the relevance of the comparison and positively impacts benign and malicious envy. We
contribute to mHealth and social comparison research by showing that comparing to targets who are similar
on the performance dimension (step count) decreases malicious envy and increases benign envy, based on
appraisals of perceived control. That is, if users compare to targets with a similar step count, they appraise
the situation as being controllable, whereas if they compare to targets who have a significantly higher step
count, they appraise the situation as uncontrollable. Moreover, benign and malicious envy differentially
impact goal-setting behavior. While benign envy results in increased goal-setting behavior, malicious envy
does not. This study yields promising insights and addresses multiple gaps in general IS literature by
affirming existing calls on affective factors that influence user behaviors, providing reasons to distinguish
between benign and malicious envy in an mHealth context and increasing our understanding of how social
comparison features impact goal-setting behaviors. The obtained findings illustrate the relevance for
further research on benign and malicious envy, its antecedents and effects as well as on integrating the
concepts of benign and malicious envy into the cumulative tradition of IS research.

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
13
References
Anderson, I., Maitland, J., Sherwood, S., Barkhuus, L., Chalmers, M., Hall, M., Brown, B., and Muller, H.
2007. “Shakra: Tracking and Sharing Daily Activity Levels with Unaugmented Mobile Phones,”
Mobile Networks and Applications (12:2–3), pp. 185–199.
Arigo, D. 2015. “Promoting Physical Activity among Women Using Wearable Technology and Online
Social Connectivity: A Feasibility Study,” Health Psychology and Behavioral Medicine (3:1), pp. 391–
409.
Arnold, M. 1960. Emotion and Personality, New York: Columbia University Press.
Bandura, A. 1991. “Social Cognitive Theory of Self-Regulation,” Organizational Behavior and Human
Decision Processes (50), pp. 248–287.
Carver, C. S., and Scheier, M. F. 1982. “Control Theory: A Useful Conceptual Framework for Personality-
Social, Clinical, and Health Psychology,” Psychological Bulletin (92:1), pp. 111–135.
Chandler, J., and Shapiro, D. 2016. “Conducting Clinical Research Using Crowdsourced Convenience
Samples,” Annual Review of Clinical Psychology (12:1), pp. 53–81.
Chen, Y., Randriambelonoro, M., Geissbuhler, A., and Pu, P. 2016. “Social Incentives in Pervasive Fitness
Apps for Obese and Diabetic Patients,” Proceedings of the ACM Conference on Computer Supported
Cooperative Work, CSCW (26-Februar), pp. 245–248.
Chen, Yu, Chen, Yunan, Randriambelonoro, M., Geissbuhler, A., and Pu, P. 2017. “Design Considerations
for Social Fitness Applications: Comparing Chronically Ill Patients and Healthy Adults,” in
Proceedings of the 2017 ACM Conference on Computer Supported Cooperative Work and Social
Computing.
Consolvo, S., Everitt, K., Smith, I., and Landay, J. A. 2006. “Design Requirements for Technologies That
Encourage Physical Activity,” Conference on Human Factors in Computing Systems - Proceedings
(1), pp. 457–466.
Cosmides, L., and Tooby, J. 2000. “Evolutionary Psychology and the Emotions,” in Handbook of
Emotions (2nd ed.), New York, NY: Guilford, pp. 91–115.
Crusius, J., and Lange, J. 2014. “What Catches the Envious Eye? Attentional Biases within Malicious and
Benign Envy,” Journal of Experimental Social Psychology (55), Elsevier Inc., pp. 1–11.
Fallon, M., Spohrer, K., and Heinzl, A. 2019. “Deep Structure Use of MHealth: A Social Cognitive Theory
Perspective,” in European Conference on Information Systems (ECIS), Stockholm.
Festinger, L. 1954. “A Theory of Social Comparison Processes,” Human Relations (7:2), pp. 117–140.
Fornell, C., and Larcker, D. F. 1981. “Evaluating Structural Equation Models with Unobservable Variables
and Measurement Error,” Journal of Marketing Research (18:1), pp. 39–39.
Frijda, N. H. 1993. “The Place of Appraisal in Emotion,” Cognition & Emotion (7:3–4), pp. 357–387.
Fritz, T., Huang, E. M., Murphy, G. C., and Zimmermann, T. 2014. “Persuasive Technology in the Real
World: A Study of Long-Term Use of Activity Sensing Devices for Fitness,” Conference on Human
Factors in Computing Systems - Proceedings, pp. 487–496.
Gerber, J. P., Wheeler, L., and Suls, J. 2017. “A Social Comparison Theory Meta-Analysis 60+ Years On,”
Psychological Bulletin (144:2), pp. 177.
Gibbons, F. X. F., and Buunk, B. 1999. “Individual Differences in Social Comparison: Development of a
Scale of Social Comparison Orientation.,” Journal of Personality and Social Psychology (76:1), pp.
129–142.
Hair, J. F., Ringle, C. M., and Sarstedt, M. 2011. “Pls-Sem: Indeed a Silver Bullet,” Journal of Marketing
Theory and Practice (19:2), pp. 139–151.
Hair, J., Ringle, C., and Sarstedt, M. 2013. “Partial Least Squares Structural Equation Modeling: Rigorous
Applications, Better Results and Higher Acceptance,” Long Range Planning (46:1–2), pp. 1–12.
Hair, J., Hult, G., Ringle, C., and Sarstedt, M. 2016. “A Primer on Partial Least Squares Structural
Equation Modeling (PLS-SEM), SAGE Publications, Inc.
Henseler, J., Ringle, C. M., and Sarstedt, M. 2014. “A New Criterion for Assessing Discriminant Validity in
Variance-Based Structural Equation Modeling,” Journal of the Academy of Marketing Science (43:1),
pp. 115–135.
Hill, S. E., Delpriore, D. J., and Vaughan, P. W. 2011. The Cognitive Consequences of Envy : Attention ,
Memory , and Self-Regulatory Depletion A Sociofunctional Perspective on Envy, (101:4), pp. 653–
666.
Jia, Y., Liu, Y., Yu, X., and Voida, S. 2017. “Designing Leaderboards for Gamification: Perceived

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
14
Differences Based on User Ranking, Application Domain, and Personality Traits,” Proceedings of the
2017 CHI Conference on Human Factors in Computing Systems, pp. 1949–1960.
Krasnova, H., Widjaja, T., Buxmann, P., Wenninger, H., and Benbasat, I. 2015. “Why Following Friends
Can Hurt You: An Exploratory Investigation of the Effects of Envy on Social Networking Sites among
College-Age Users,” Information Systems Research (26:3), pp. 585–605.
Lange, J., and Crusius, J. 2015. “Dispositional Envy Revisited: Unraveling the Motivational Dynamics of
Benign and Malicious Envy,” Personality and Social Psychology Bulletin (41:2), pp. 284–294.
Lee, Y., and Lim, Y. K. 2015. “Understanding the Roles and Influences of Mediators from Multiple Social
Channels for Health Behavior Change,” CSCW 2015 - Proceedings of the 2015 ACM International
Conference on Computer-Supported Cooperative Work and Social Computing, pp. 1070–1079.
Lench, H. C., Taylor, A. B., and Bench, S. W. 2014. “An Alternative Approach to Analysis of Mental States
in Experimental Social Cognition Research,” Behavior Research Methods (46:1), pp. 215–228.
Levy, D. 2014. “Emerging MHealth: Paths for Growth,” MHealth Team for PwC.
(https://www.pwc.com/gx/en/healthcare/mhealth/assets/pwc-emerging-mhealth-full.pdf).
Liu, C.-W., Gao, G. G., and Agarwal, R. 2019. “Unraveling the ‘Social’ in Social Norms: The Conditioning
Effect of User Connectivity,” Information Systems Research (30:4), pp. 1272–1295.
Liu, D., Santhanam, R., and Webster, J. 2017. “Research Commentary Toward Meaningful Engagement: A
Framework for Design and Research of Gamified Information Systems,” MIS Quarterly (Vol. 41), pp.
1011–1034.
Locke, E. A., and Latham, G. P. 2002. “Building a Practically Useful Theory of Goal Setting and Task
Motivation: A 35-Year Odyssey,” American Psychologist (57:9), pp. 705–717.
Locke, E. A., and Latham, G. P. 2006. “New Directions in Goal-Setting Theory,” Current Directions in
Psychological Science (15:5), pp. 265–268.
Martin, R., Suls, J., and Wheeler, L. 2002. “Ability Evaluation by Proxy: Role of Maximal Performance
and Related Attributes in Social Comparison,” Journal of Personality and Social Psychology (82:5),
pp. 781–791.
Middelweerd, A., van der Laan, D. M., van Stralen, M. M., Mollee, J. S., Stuij, M., te Velde, S. J., and Brug,
J. 2015. “What Features Do Dutch University Students Prefer in a Smartphone Application for
Promotion of Physical Activity? A Qualitative Approach,” International Journal of Behavioral
Nutrition and Physical Activity (12:1), pp. 1–11.
Miller, A. D., and Mynatt, E. D. 2014. “StepStream: A School-Based Pervasive Social Fitness System,” in
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 2823–2832.
Nunnally, J. C. 1978. Psychometric Theory, (2d ed.), New York: McGraw-Hill.
Ortony, A., Clore, G., and Collins, A. 1988. The Cognitive Structure of Emotions, New York: Cambridge
University Press.
Patel, M. S., Volpp, K. G., Rosin, R., Bellamy, S. L., Small, D. S., Fletcher, M. A., Osman-Koss, R., Brady, J.
L., Haff, N., Lee, S. M., Wesby, L., Hoffer, K., Shuttleworth, D., Taylor, D. H., Hilbert, V., Zhu, J.,
Yang, L., Wang, X., and Asch, D. A. 2016. “A Randomized Trial of Social Comparison Feedback and
Financial Incentives to Increase Physical Activity,” American Journal of Health Promotion (30:6),
pp. 416–424.
Peer, E., Vosgerau, J., and Acquisti, A. 2014. “Reputation as a Sufficient Condition for Data Quality on
Amazon Mechanical Turk,” Behavior Research Methods (46:4), pp. 1023–1031.
Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., and 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), pp. 879–903.
Ringle, C. M., Wende, S., and Becker, J. ‑M. 2015. “SmartPLS 3 (Version 3),” SmartPLS GmbH,
Boenningstedt, Germany. (www.smartpls.com).
Roseman, I. J. 1996. “Appraisal Determinants of Emotions: Constructing a More Accurate and
Comprehensive Theory,” Cognition & Emotion (10:3), pp. 241–278.
Salovey, P., and Rodin, J. 1984. “Some Antecedents and Consequences of Social-Comparison Jealousy,”
Journal of Personality and Social Psychology (47:4), pp. 780–792.
Scherer, K., Schorr, A., and Johnstone, T. 2001. Appraisal Processes in Emotion: Theory, Research,
Application, New York: Oxford University Press.
Schmidt-Kraepelin, M., Thiebes, S., Schöbel, S., and Sunyaev, A. 2019. “Users’ Game Design Element
Preferences in Health Behavior Change Support Systems for Physical Activity: A Best-Worst-Scaling
Approach,” in Proceedings of the 40th International Conference on Information Systems (ICIS),

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
15
Munich, Germany, pp. 1–17.
Schmidt-Kraepelin, M., Thiebes, S., Stepanovic, S., Mettler, T., and Sunyaev, A. 2019. “Gamification in
Health Behavior Change Support Systems-A Synthesis of Unintended Side Effects,” in Proceedings of
the 14th International Conference on Wirtschaftsinformatik (WI 2019), Siegen, Germany.
Shilts, M. K., Horowitz, M., and Townsend, M. S. 2004. “Goal Setting as a Strategy for Dietary and
Physical Activity Behavior Change: A Review of the Literature,” American Journal of Health
Promotion (19:2), pp. 81–93.
Smith, R. 1991. “Envy and the Sense of Injustice,” in The Psychology of Envy and Jealousy, New York:
Guilford Press, pp. 79–99.
Smith, R. H., and Kim, S. H. 2007. “Comprehending Envy,” Psychological Bulletin (133:1), pp. 46–64.
Stapel, D. A., and Koomen, W. 2001. “I, We, and the Effects of Others on Me: How Self-Construal Level
Moderates Social Comparison Effects,” Journal of Personality and Social Psychology (80:5), pp.
766–781.
Stein, M.-K., Newell, S., Wagner, E. L., and Galliers, R. D. 2015. “Coping with Information Technology:
Mixed Emotions, Vacillation, and Nonconforming Use Patterns,” MIS Quarterly (39:2), pp. 367–392.
Thomas, K. A., and Clifford, S. 2017. “Validity and Mechanical Turk: An Assessment of Exclusion Methods
and Interactive Experiments,” Computers in Human Behavior (77), pp. 184–197.
Tong, H. L., Coiera, E., and Laranjo, L. 2018. “Using a Mobile Social Networking App to Promote Physical
Activity: A Qualitative Study of Users’ Perspectives,” Journal of Medical Internet Research (20:12),
pp. e11439–e11439.
Urbach, N., and Ahlemann, F. (2010). “Structural Equation Modeling in Information Systems Research
Using Partial Least Squares,” Journal of Information Technology Theory and Application (11:2), pp.
5–40.
Utz, S., and Muscanell, N. L. 2018. “Your Co-Author Received 150 Citations: Pride, but Not Envy,
Mediates the Effect of System-Generated Achievement Messages on Motivation,” Frontiers in
Psychology, pp. 1–12.
Van Dijk, W. W., and Zeelenberg, M. 2002. “Investigating the Appraisal Patterns of Regret and
Disappointment,” Motivation and Emotion (26:4), pp. 321–331.
van de Ven, N. 2016. “Envy and Its Consequences: Why It Is Useful to Distinguish between Benign and
Malicious Envy,” Social and Personality Psychology Compass (106), pp. 1–26.
van de Ven, N. 2017. “Envy and Admiration: Emotion and Motivation Following Upward Social
Comparison,” Cognition and Emotion (31:1), pp. 193–200.
van de Ven, N., Zeelenberg, M., and Pieters, R. 2009. “Leveling Up and Down: The Experiences of Benign
and Malicious Envy,” Emotion (9:3), pp. 419–429.
van de Ven, N., Zeelenberg, M., and Pieters, R. 2011. “Why Envy Outperforms Admiration,” Personality
and Social Psychology Bulletin (37:6), pp. 784–795.
van de Ven, N., Zeelenberg, M., and Pieters, R. 2012. “Appraisal Patterns of Envy and Related Emotions,”
Motivation and Emotion (36:2), pp. 195–204.
Wheeler, L., Martin, R., and Suls, J. 1997. “The Proxy Model of Social Comparison for Self-Assessment of
Ability,” Personality and Social Psychology Review (1:1), pp. 54–61.
Wood, J. V. 1996. “Social Comparison and How We Study It,” Personality and Social Psychology Bulletin
(22:5), pp. 520–537.
World Health Organization. 2014. “Global Status Report on Noncommunicable Diseases.”
(http://apps.who.int/iris/bitstream/handle/10665/148114/9789241564854_eng.pdf;jsessionid=475
C8301FE69F9C42EC1CD090B0CABB2?sequence=1).
World Health Organization. 2020. “Global Strategy on Diet, Physical Activity and Health,” WHO, World
Health Organization. (http://www.who.int/dietphysicalactivity/pa/en/, accessed April 30, 2020).
Wu, Y., Kankanhalli, A., and Huang, K.-W. 2015. “Gamification in Fitness Apps: How Do Leaderboards
Influence Exercise?,” in Thirty Sixth International Conference on Information Systems, Fort Worth,
pp. 1–12.
Zuckerman, O., and Gal-Oz, A. 2014. “Deconstructing Gamification: Evaluating the Effectiveness of
Continuous Measurement, Virtual Rewards, and Social Comparison for Promoting Physical Activity,”
Personal and Ubiquitous Computing (18:7), pp. 1705–1719.

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
16
Appendix
ID
Item
Definition
References
α
PERCEIVED SIMILARITY (RELATED ATTRIBUTES)
SM_1
People on the leaderboard are similar to me in terms of gender.
The extent to which one
perceives that
comparison targets are
similar regarding
attributes known to
influence performance.
Self-developed based
on Martin et al.
(2002); Stapel &
Koomen (2001)
.966
SM_2
People on the leaderboard are similar to me in terms of age.
SM_3
People on the leaderboard are a relevant comparison because
they are similar to me in terms of age.
SM_4
People on the leaderboard are a relevant comparison because
they are similar to me in terms of gender.
PERCEIVED SIMILARITY (PERFORMANCE)
SM_5
People on the leaderboard are similar to me in terms of their
average daily step count.
The extent to which one
perceives that
comparison targets are
similar regarding daily
step count.
Self-developed based
on Martin et al.
(2002); Stapel &
Koomen (2001)
.835
SM_6
People on the leaderboard take a similar amount of average
steps per day as me.
PERCEIVED CONTROL
PC_1
There is nothing I could do about increasing my step count (1)
there is something I could do about increasing my step count
(9)
The extent to which one
perceives that they have
the ability to do
something about a
situation.
Van Dijk &
Zeelenberg, (2002);
Van de Ven et al.
(2012); Roseman
(1996)
1
BENIGN ENVY
EN_1
I felt inspired to increase my step count.
The extent to which one
experiences feelings that
lead to positive
improvement for oneself
through a moving-up
motivation.
Utz & Muscanell
(2018); Van de Ven et
al. (2009); Van de
Ven et al. (2017)
.903
EN_2
I wanted to put in effort to increase my step count.
EN_3
I wanted to increase my step count.
EN_4
I thought about what it would be like to increase my step count.
MALICIOUS ENVY
EN_5
It was frustrating that others perform more steps than me.
The extent to which one
experiences hostile
feelings and motivations
aimed at pulling-down
others from a superior
position.
Van de Ven et al.
(2017)
.898
EN_6
I was envious that others perform more steps than me.
EN_7
I was jealous of the others performing more steps than me.
GOAL INCREASE
GI_1
Do you want to increase your step count goal?
Whether one increases
or does not increase their
step count goal after
viewing the leaderboard.
Self-developed based
on Liu et al. (2019)
1
DISPOSITIONAL BENIGN ENVY
DP_1
When I envy others, I focus on how I can become equally
successful in the future.
The extent to which one
experiences a general
motivation directed at
achieving a standard of
excellence.
Lange & Crusius
(2015)
.902
DP_2
If I notice that another person is better than me, I try to
improve myself.
DP_3
Envying others motivates me to accomplish my goals.
DP_4
I strive to reach other people's superior achievements.
DP_5
If someone has superior qualities, achievements, or
possessions, I try to attain them for myself.
Table A-1. Construct Definitions and Measurement Scales

Social Comparison in mHealth
Forty-First International Conference on Information Systems, India 2020
17
ID
Item
Definition
References
α
DISPOSITIONAL MALICIOUS ENVY
DP_6
I wish that superior people lose their advantage.
The extent to
which one
experiences a
general motivation
to avoid falling
short of a standard
of excellence.
Lange &
Crusius
(2015)
.918
DP_7
If other people have something that I want for myself, I wish to take it away
from them.
DP_8
I feel ill will towards people I envy.
DP_9
Envious feelings cause me to dislike the other person.
DP_10
Seeing other people's achievements makes me resent them.
SOCIAL COMPARISON ORIENTATION
SO_1
I often compare how my loved ones (boy or girlfriend, family members etc.)
are doing with how others are doing
The extent to
which one
generally makes
comparisons
about opinions,
abilities, and
general aspects of
themselves.
Gibbons &
Buunk
(1999)
.905
SO_2
I always pay a lot of attention to how I do things compared with how others do
things
SO_3
If I want to find out how well I have done something, I compare what I have
done with how others have done
SO_4
I often compare how I am doing socially (e.g., social skills, popularity) with
other people
SO_5
I am not the type of person who compares often with others
SO_6
I often compare myself with others with respect to what I have accomplished
in life
SO_7
I often like to talk with others about mutual opinions and experiences
[excluded from analysis]
SO_8
I often try to find out what others think who face similar problems as I face
SO_9
I always like to know what others in a similar situation would do
SO_10
If I want to learn more about something, I try to find out what others think
about it
SO_11
I never consider my situation in life relative to that of other people [excluded
from analysis]
Table A-1. Construct Definitions and Measurement Scales (Continued)
Low similarity in performance
Low similarity in attributes
(1)
Low similarity in performance
High similarity in attributes
(2)
High similarity in performance
Low similarity in attributes
(3)
High similarity in performance
High similarity in attributes
(4)
Figure A-1. Exemplary Leaderboard Manipulations (Male; 40 years; 10,000 Daily Steps)