Content uploaded by Guanxiong Huang
Author content
All content in this area was uploaded by Guanxiong Huang on Dec 02, 2019
Content may be subject to copyright.
1
Linking Technological Functions of Fitness Mobile Apps with Continuance Usage among
Chinese Users: Moderating Role of Exercise Self-Efficacy
Guanxiong Huang
Department of Media and Communication, City University of Hong Kong
Email: ghuang3@cityu.edu.hk
Yuchen Ren
School of Media and Communication, Shenzhen University
Email: renyuchen@szu.edu.cn
Citation:
Huang, G., & Ren, Y. (2020). Linking technological functions of fitness mobile apps with
continuous usage among Chinese users: Moderating role of exercise self-efficacy. Computers
in Human Behavior, 103, 151-160. https://doi.org/10.1016/j.chb.2019.09.013
Acknowledgement:
This work was supported by the Department of Media and Communication of City University
of Hong Kong under the Faculty Research Grant 9618013.
2
Abstract
Mobile apps are effective tools for administering health interventions and changing user
behaviors in key lifestyle areas, such as physical activity, but the attrition rates of fitness app
users are high. To understand how to increase user retention rates, the present study draws
from the Technology Acceptance Model (TAM) and investigates the role of exercise self-
efficacy, in addition to the original TAM constructs, namely, perceived usefulness, perceived
ease of use, and perceived enjoyment, in predicting current users’ intention to continue using
the apps. Moreover, this study extends TAM from a human-technology interaction
perspective by elucidating the antecedents of perceived usefulness in terms of specific
functions provided by fitness mobile apps. Samples were drawn from a large online Chinese
subject pool to test the hypotheses via a survey (N = 449). The results showed that four
technological functions—instruction provision, self-monitoring, self-regulation, and goal
attainment—had an indirect effect on continuance intention through perceived usefulness,
and this indirect effect was moderated by exercise self-efficacy such that the association
between perceived usefulness and continuance intention was stronger for those with low
exercise self-efficacy.
Keywords: fitness mobile apps, Technology Acceptance Model, exercise self-efficacy,
technological functions, human-technology interaction
3
1. Introduction
The prevalence of smartphones has fueled the rapid development and widespread
usage of mobile applications (apps) aimed at facilitating health behavior change. Mobile apps
specialized in physical activity and weight loss have become the most popular type of health
app product, accounting for over 70% of the total usage in the Health & Fitness App category
(eMarketer, 2017). Research has shown that mobile apps are effective tools for administering
health interventions and changing user behaviors in key lifestyle areas, such as physical
activity (Mateo, Granado-Font, Ferré-Grau, & Montaña-Carreras, 2015; Schoeppe et al.,
2016; Stiglbauer, Weber, & Batinic, 2019; Turner-McGrievy et al., 2013). Nevertheless, for
fitness mobile apps to produce their intended effects, users need to use the apps for a
continuous period of time, during which the desired behavior changes are incorporated into
their daily routines (Chin et al., 2016). In 2017 alone, there were 3.7 billion mobile health app
downloads worldwide (Statista, 2018). However, a national survey in the United States
revealed that the attrition rate of health app users was as high as 45.7% (Krebs & Duncan,
2015). Therefore, despite high download rates, understanding how to increase user retention
rates is an imperative issue for health practitioners and app developers (Molina & Sundar,
2018).
To tackle this issue, the present study explores which technological functions predict
current users’ continuance usage intention under the framework of the Technology
Acceptance Model (TAM; Davis, 1989). TAM is the most used theory in researching the
initial adoption of newly developed technologies and information systems (Venkatesh &
Bala, 2008), but it has been less frequently applied to investigating user post-adoption
behaviors. The predictors of user adoption behaviors proposed by TAM—perceived
usefulness, perceived ease of use, and perceived enjoyment—are subject to change based on
user interactions with the technologies. These perception-based constructs may change after
4
initial adoption, and the changes may affect continuance usage (Bhattacherjee, 2001; Cho,
2016). In this vein, TAM is applicable to examining post-adoption behaviors as well.
The current study is one of the few attempts at extending the scope of TAM to
unveiling the mechanisms behind current users’ continuance usage of mobile technologies
(e.g., Beldad & Hegner, 2018; Cho, 2016). More specifically, it enriches our understanding of
fitness app use and contributes to the TAM scholarship in two aspects. First, the TAM
predictors focus on information system benefits in terms of generic feelings, such as
usefulness, ease of use, and enjoyment, while providing limited insights for technology
system designers as to which specific functions or features predict technology adoption
(Bagozzi, 2007). To address this limitation, this study extends TAM from a human-
technology interaction perspective and investigates which technological functions would
influence user perceptions and feelings toward fitness mobile apps and, in turn, promote
future use. Second, this study elucidates the role of exercise self-efficacy in relation to TAM
when predicting continuance usage of fitness mobile apps. Self-efficacy, defined as one’s
belief in his/her personal ability to successfully organize and execute the actions needed to
achieve the desired outcome, is a significant predictor of the acquisition of new behaviors and
the maintenance of existing behaviors (Bandura, 1977). Given that the main function of
fitness mobile apps is to aid users’ physical activities, exercise self-efficacy is related to
people’s adoption and continued use of fitness mobile apps (Lim & Noh, 2017; Litman et al.,
2015; McAuley Lox, Rudolph, & Travis, 1994). However, the role of exercise self-efficacy
remains unclear, since little academic attention has been given to this particular scenario. The
present study seeks to fill this void by uncovering the direct and moderating effects of
exercise self-efficacy in relation to the TAM framework. Specifically, this study aims to
answer the following questions: What factors predict current fitness app users’ continuous
usage intention? What role does exercise self-efficacy play in relation to the other predictors?
5
This study was situated in the Chinese context. China has been experiencing a
breakthrough in the application of mobile technologies in the field of healthcare, with the
estimated value of China’s mHealth market exceeding 1.5 billion USD in 2017 (TechNode,
2017). With an enormous physically inactive population that accounts for two thirds of the
population (General Administration of Sport of China, 2015) and the prevalence of
smartphone ownership (CNNIC, 2018), there is no doubt that China has vast potential for
promoting physical activity via mobile technologies. This study will provide useful insights
into Chinese user behaviors and relevant theory-driven technological functions for health
practitioners and app developers. Given that the large body of TAM studies are in Western
contexts, this study will also contribute to the cross-cultural generalizability of the theory.
2. Literature Review
2.1. Technology Acceptance Model and fitness app use
Originating from the Theory of Reasoned Action (Fishbein & Ajzen, 1975), TAM
(Davis, 1989) postulates that perceived usefulness and perceived ease of use predict user
acceptance of a technology. Specifically, perceived usefulness refers to the extent to which
one perceives that a technology can improve his/her performance of a certain task, and
perceived ease of use refers to the extent to which one perceives that using a technology is
free of effort (Davis, 1989). A meta-analysis shows that perceived usefulness and perceived
ease of use explain approximately 50% of the variance in technology use intentions (King &
He, 2006). Additionally, a new construct, perceived enjoyment, is integrated into TAM and
has proven to be an important predictor of the use of various technologies that satisfy users’
hedonic desires, such as mobile games (Lee & Tsai, 2010) and social networking sites
(Hsiao, Chang, & Tang, 2016). Perceived enjoyment is defined as “the extent to which the
activity of using a specific system is perceived to be enjoyable in its own right, aside from
any performance consequences resulting from system use” (Venkatesh, 2000, p. 351). In
6
summary, perceived usefulness, perceived ease of use, and perceived enjoyment capture the
utilitarian motivation, the effort expectancy, and the hedonic motivation for using a
technology, respectively.
TAM is commonly used in research on the adoption of health care information
systems (for a review, see Holden & Karsh, 2010), while research on the use of mobile health
apps based on TAM is “still in its nascent phase” (Beldad & Hegner, 2018, p. 883). One line
of research has confirmed TAM’s validity in the mobile health context by showing that
perceived usefulness, perceived ease of use, and perceived enjoyment jointly contribute to
users’ continuance intentions to use healthcare apps with various specialties, such as weight
management, chronic disease information management, and medication information (Cho,
2016; Wang, Park, Chung, & Choi, 2014). In the context of fitness mobile apps, only a few
recent studies used TAM as a theoretical foundation to predict people’s intentions to use
these apps, most of which focused on initial adoption (Chen & Lin, 2018; Jeon & Park,
2015). To our knowledge, only one study used TAM to predict users’ continuance usage
(Beldad & Hegner, 2018); the researchers found that perceived usefulness and perceived ease
of use, along with injunctive social norms, predicted intentions to continue using a particular
fitness app among 476 German users.
Perceived enjoyment, otherwise termed as hedonic motivation, is one of the most
important motivations that drive technology adoption and media use (Ferguson & Perse,
2000; Venkatesh, Thong, & Xu, 2012). If users derive entertainment value or obtain
emotional gratifications (e.g., fun and pleasure) from the experience of using a particular
media or information system, they are more likely to continue using it due to “an inner need
to keep an individual at an optimal, preferred state of comfort, congruent with external
stimulation” (Lu, Liu, & Wei, 2017, p. 4). This general argument has been supported by
empirical studies on mobile apps (Hew, Lee, Ooi, & Wei, 2015; Kang, 2014; Lee, Goh,
7
Chua, & Ang, 2010; Merikivi, Tuunainen, Nguyen, 2017). However, some other studies have
different findings. For example, Lee and Cho (2017) found that entertainment value was not a
significant predictor of continued use of diet/fitness mobile apps; Lu et al. (2017) revealed
that perceived enjoyment was positively related to users’ post-usage attitude, but had no
significant impact on their continuance intentions of mobile apps. Though no consensus has
been reached regarding the role of perceived enjoyment in predicting fitness app use from
prior studies, the increasing popularity of gamification, which refers to the application of
game elements in other activities, in fitness app design demonstrates that app designers are
placing more and more emphasis on the entertainment value of the apps in addition to the
instrumental value (Lister, West, Cannon, Sax, & Brodegard, 2014; Payne, Moxley, &
MacDonald, 2015). Therefore, we argue that perceived enjoyment is a positive predictor of
continuance intention of fitness apps among Chinese users. Taken together, the hypothesis
below is posited.
H1: (a) Perceived usefulness, (b) perceived ease of use, and (c) perceived enjoyment
are positively associated with continuance usage intention of fitness mobile apps.
2.2. Technological functions of fitness apps
Despite TAM’s popularity, it has been criticized for being a “black box” that focuses
on an information system’s benefits in terms of generic feelings, such as usefulness, ease of
use, and enjoyment. It provides limited insights for technology system designers as to which
specific functions or features predict technology adoption and continuous usage (Bagozzi,
2007). In response, the present study goes a step further to examine the key functions of
fitness mobile apps as antecedents of perceived usefulness.
One stream of research has focused on technological functions of fitness mobile apps
and related behavior change techniques (e.g., Conroy et al., 2014; Huang & Zhou, 2018;
Middelweerd et al., 2014; Payne et al., 2015). Through content analyses of fitness apps, these
8
studies identified several most prevalent technological functions, including instruction
provision, self-monitoring, self-regulation, and goal attainment. Almost all fitness apps on the
market in different countries are equipped with at least one of these four functions (Conroy et
al., 2014; Huang & Zhou, 2018; Middelweerd et al., 2014), indicating the app designer’s
perspective on the types of benefits that fitness apps can bring to their users. Instruction
provision involves models verbalizing instructions and/or performing desired actions so that
users can observe and learn from the models by mimicking their behaviors and acquiring
proven skills (Bandura, 2001). One example of this function is that some Yoga apps provide
demonstration/instruction videos featuring personal coaches giving verbal instructions and
demonstrating Yoga poses. The self-monitoring, self-regulation, and goal attainment
functions enable users to observe their own progress, evaluate their performance against pre-
set goals, and adjust their goals to a realistic level, which further facilitates self-regulation in
performing subsequent exercise behaviors (Bandura, 1991; 1997). For example, some fitness
apps allow users to set up specific goals such as work out for 30 minutes per day. Then the
apps would track and record, sometimes with the users’ manual inputs, the length and type of
physical activities that the users were engaged with. Based on the records, the apps are able to
provide feedbacks as to whether the users have achieved the goal. Then the users may adjust
their goals to balance their exercise capacity and the attainability of the goals.
Another stream of research from the user’s perspective has confirmed instruction
provision, self-monitoring, self-regulation, and goal attainment are the most desired functions
among fitness app users in various countries (Belmon, Middleweerd, teVelde, & Brug, 2015;
Rabin & Bock, 2011; Turner-McGrievy et al., 2013); however, the analyses of the prior
studies were mainly descriptive in nature, and did not systematically link those technological
functions with users’ post adoption behaviors or behavioral intentions. On the other hand, a
few other studies have investigated the effects of social networking features on user
9
behaviors. Lee and Cho (2017) identified networkability as one of the significant motivations
for continued use of diet/fitness apps; Zhu, Dailey, Kreitzberg, and Bernhardt (2017) found
that the social sharing and social competing features of wearable fitness devices significantly
predicted users’ exercise intentions via attitude, subjective norms and perceived behavioral
control. They indicated that certain functions or features may play a critical role in shaping
user behaviors via influencing users’ perceptions and attitudes toward the apps. Moreover,
the post-acceptance model (Bhattacherjee, 2001) posits that, when users confirm their initial
expectations of the main functions of an adopted mobile app, they begin to perceive the app
as useful for improving task performance, and thus continue to use it (Cho, 2016). Therefore,
the current study hypothesizes a positive relationship between the four technological
functions and perceived usefulness in the context of fitness mobile apps:
H2: Four main functions of fitness mobile apps (i.e., instruction provision, self-
monitoring, self-regulation, and goal attainment) are positively associated with perceived
usefulness.
2.3. Exercise self-efficacy
TAM has been used as a generic approach to technology adoption; the model does not
include any specific constructs to address the characteristics of the task that the technology is
involved with (Dishaw & Strong, 1999). For instance, one task characteristic is whether the
task is mandatory. For a mandatory task (e.g., a job), the adoption of information technology
(IT) may be heavily dependent on perceived usefulness and perceived ease of use; on the
other hand, for a free-choice task (e.g., seeking health information), the intrinsic motivation
to engage in the task itself is also important for predicting technology use (e.g., health beliefs
and concerns; Kim & Park, 2012). Physical activity typically falls into the second category
and, therefore, people’s interest, determination, and confidence levels for performing exercise
play a role in their adoption and continuous usage of fitness-related technologies. As such,
10
this study incorporates exercise self-efficacy, one of the most consistently identified
determinants of physical activity (Annesi, 2012; McAuley & Blissmer, 2000), into TAM to
predict fitness app use.
Self-efficacy refers to “people’s beliefs about their capabilities to exercise control
over their own level of functioning and over events that affect their lives” (Bandura, 1991, p.
257). As a key factor in social cognitive theory, self-efficacy influences all aspects of
behavior, including the acquisition of new behaviors and the maintenance of existing
behaviors (Bandura, 1977). Exercise self-efficacy is therefore defined as an individual’s belief
in his/her ability to engage in physical activity. Prior research has shown that, the greater self-
efficacy individuals have, the more motivation they have to perform exercise, thereby
resulting in their greater investment in physical activity (Annesi, 2012; McAuley & Blissmer,
2000). In light of the strong predictive power of self-efficacy in health-related behavior
change, health intervention and promotion programs have treated improvement in exercise
self-efficacy as a major outcome (Annesi, 2012; McAuley & Blissmer, 2000), which, in turn,
has been positively correlated with physical activity (Annesi & Marti, 2011), healthy eating
(Fleig, Küper, Lippke, Schwarzer, & Wiedemann, 2015), and weight loss (Faghri, Simon,
Huedo-Medina, & Gorin, 2017; Kim, Faw, & Michaelides, 2017; Teixeira et al., 2015). In
other words, exercise self-efficacy is not only a result of a supportive exercise environment,
but also a determinant of physical activity adherence (McAuley & Blissmer, 2000; Sheeran et
al., 2016).
In the context of fitness app use, to the best of our knowledge, one study found partial
evidence for the direct effect of exercise self-efficacy on fitness app use intentions (Lim &
Noh, 2017). Nevertheless, in the context of other types of mobile apps and services, empirical
support has been found regarding the positive relationship between self-efficacy and adoption
behavioral intention or actual behavior (Alalwan, Dwivedi, Rana, & Williams, 2016; Luarn &
11
Lin, 2005; Yang, 2010). Based on the literature above, we propose the following hypothesis
to examine the association between exercise self-efficacy and continuance usage intention of
fitness mobile apps.
H3: Exercise self-efficacy is positively associated with continuance usage intention of
fitness mobile apps.
Moreover, it is worth investigating how exercise self-efficacy interacts with the key
constructs of TAM in predicting continuance usage intention of fitness apps. Despite the
scarce literature on this subject, we can draw insights from the task-technology fit model
(TTF) (Dishaw & Strong, 1999), which indirectly specifies the moderating role of exercise
self-efficacy in the relationship between perceived usefulness and continuance usage
intention of health-related information systems. TTF posits that the match between user task
needs and the available functionality of an IT system predicts technology use (Goodhue &
Thompson, 1995), suggesting that users with different needs may weigh the perceptions of an
information system differently when determining whether to continue using it. In the current
study, people with different levels of exercise self-efficacy may have different needs to fulfill
when using fitness mobile apps. Those with high exercise self-efficacy may have already had
a strong intrinsic motivation (McAuley, Lox, Rudolph, & Travis, 1994) and a regular
exercise routine (McAuley et al., 1999) before downloading a fitness app, so they may weigh
its instrumental value (i.e., perceived usefulness) less when determining whether to continue
using it. Instead, they may place a higher value on entertainment features, such as
gamification and social networking features. In contrast, for those with low self-efficacy, they
may need an extra push to facilitate their self-regulation mechanisms to engage in physical
activity on a regular basis (Litman et al., 2015). In this sense, the instrumental value of fitness
apps carries a greater weight in users’ decision making processes regarding continuance
usage of fitness apps. Besides, empirical research also supported the moderating role of self-
12
efficacy in mobile-related behaviors, such as mobile commerce services (Islam, Khan,
Ramayah, & Hossain, 2011), and mobile-based health interventions (Clarke et al., 2014).
Taken together, Hypothesis 4 is therefore proposed to address the moderating effects of
exercise self-efficacy on continuance usage intention. Given that there has been no prior
research regarding the interplay between exercise self-efficacy and other TAM predictors in
predicting continuous usage, Research Question 1 is posed to explore the relationship.
H4: Exercise self-efficacy moderates the relationship between perceived usefulness
and continuance usage intention of fitness mobile apps. Specifically, the association between
perceived usefulness and continuance usage intention is stronger for users with low exercise
self-efficacy.
RQ1: What is the role of exercise self-efficacy in relation to a) perceived enjoyment
and b) perceived ease of use for predicting continuance usage intention of fitness mobile
apps?
Prior research on fitness mobile apps demonstrated that social networking functions
play a pivotal role in influencing user behaviors via shaping users’ perceptions and attitudes
toward the apps (Lee & Cho, 2017; Zhu et al., 2017). However, this mediating mechanism
has not been tested with other key functions or features, in particular, self-regulatory
functions, which are considered the most prevalent functions of fitness mobile apps (Conroy
et al., 2014; Huang & Zhou, 2018; Middelweerd et al., 2014). To fill this gap, the current
study extends TAM by incorporating the technological functions as antecedents of perceived
usefulness, which, in turn, influences continuous usage intention. Accordingly, Hypothesis 5
is posited to address this inquiry.
H5: The four technological functions (i.e., instruction provision, self-monitoring, self-
regulation, and goal attainment) have a positive indirect effect on continuance usage intention
via perceived usefulness.
13
[PLACE FIGURE 1 HERE]
3. Method
3.1. Participants
The data used to test the hypotheses were collected from a cross-sectional survey
conducted in China. The samples were drawn from an online subject pool operated by a
nationwide market research firm. The subject pool consisted of over 2.6 million participants
(female: 48%, male: 52%), distributed across all provinces, autonomous regions, and
municipalities of China. The age distribution of the subject pool was skewed toward those
under 30 years old, such that 54.37% of the subjects were between 21 and 30 years old. The
subject pool also displayed a diverse pattern in terms of occupation, including students,
business managers, professionals, blue collar workers, government officials and clerks, and
self-employers. The demographic pattern was consistent with the overall distribution of
China’s netizens, 98% of which are mobile Internet users (CNNIC, 2018).
Given that this study focused on the usage of fitness mobile apps, smart phone users
over 18 years old were eligible to take part in the survey. Invitations were sent to eligible
participants, who were randomly selected from the subject pool. The data were collected
between November 6-15, 2017, and a total of 536 valid responses were obtained. Among
them, 449 were current users of fitness mobile apps, and were, therefore, included in the
analysis. We compared the demographic characteristics of the sample (see Table 1) with
official statistics of China’s netizen population (CNNIC, 2018). While our sample is more
educated and has slightly more women, the distributions in terms of age, geographic
locations, and occupation are consistent with the netizen population as a whole.
[PLACE TABLE 1 HERE]
14
3.2. Measures
Perceived usefulness. Using a 7-point Likert-type scale, with responses ranging from
1 = strongly disagree to 7 = strongly agree, respondents were asked to indicate their level of
agreement to the following statements: “Using fitness mobile apps improves my performance
in physical activity,” “Using fitness mobile apps improves my engagement in physical
activity,” and “Using fitness mobile apps improves my effectiveness in physical activity”
(Hsiao et al., 2016). The items were averaged to create a composite measure (M = 5.79, SD
= .73, α = .74).
Perceived enjoyment. Also on a 7-point Likert-type scale, this variable was assessed
by three items: “Using fitness mobile apps is pleasurable,” “I have fun using fitness mobile
apps,” and “I find using fitness mobile apps to be interesting” (Hsiao et al., 2016). The items
were averaged to create a composite measure (M = 5.76, SD = .78, α = .79).
Perceived ease of use. This variable was measured by four items adopted from
Holden and Karsh (2010), including “It is easy to use fitness mobile apps,” “It is easy to
interact with fitness mobile apps,” “It is easy to learn how to operate fitness mobile apps,”
and “It is easy to get fitness mobile apps to do what you want them to do.” The items were
averaged to create a composite measure (M = 5.84, SD = .70, α = .75).
Exercise self-efficacy. Respondents were asked to indicate how confident they were
that they could perform exercise routines regularly (three or more times a week) under 18
circumstances, including “when I am feeling tired,” “when I am feeling under pressure from
work,” “during bad weather,” and “after recovering from an injury that caused me to stop
exercising” (Shin, Jang, & Pender, 2001). The response categories ranged from 1 = not
confident at all to 7 = very confident. The items were averaged to create a composite measure
(M = 4.72, SD = .82, α = .95).
15
Technological functions. Drawn from Huang and Zhou’s (2018) study on the most
prevalent theory-driven functions of fitness mobile apps, four statements were developed to
measure perceived benefits related to four app functions: instruction provision, self-
monitoring, self-regulation, and goal attainment. These statements were: “Fitness mobile
apps give me useful instructions on exercise” (M = 5.98, SD = .81), “Fitness mobile apps
help me monitor my physical activity progress” (M = 5.94, SD = .80), “Fitness mobile apps
improve my self-regulation” (M = 5.96, SD = .86), and “Fitness mobile apps help me achieve
my physical activity goals” (M = 5.92, SD = .90). Each item was rated on a 7-point scale,
ranging from 1 = strongly disagree to 7 = strongly agree.
Continuance usage intention. This variable was operationalized as the intention to
continue using certain mobile apps. Respondents were asked to indicate their level of
agreement on a 7-point scale, with responses ranging from 1 = strongly disagree to 7 =
strongly agree, to the following statements: “I intend to continue using fitness mobile apps in
the future,” “I always try to use fitness mobile apps in my daily life,” and “I will keep using
fitness mobile apps as regularly as I do now” (Hsiao et al., 2016). The items were averaged to
create a composite measure (M = 5.97, SD = .70, α = .76).
The heterotrait-monotrait (HTMT) ratio of the correlations approach was employed to
assess the discriminant validity of the variables (Henseler, Ringle, & Sarstedt, 2015). The
HTMT values ranged from .332 to .885. Henseler et al. (2015) suggested that for TAM and
its variations, due to the strong associations between the involved variables, a liberal HTMT
criterion—a threshold of no greater than .90—should be adopted. In light of this criterion, the
discriminant validity of the measures is acceptable.
In addition, a Harman’s one-factor test was conducted to address common method
biases incurred by data collected by one method from a single source (Podsakoff,
MacKenzie, Lee, & Podsakoff, 2003). All the items were constrained to load on one factor,
16
which accounted for 27.33% of the total variance, lower than the threshold of 50%, suggesting
that this single factor cannot explain the majority of the covariance among the variables.
Therefore, there were no concerns regarding the common method bias.
3.3. Analytic procedure
Given that this research focused on the incremental predictive power of exercise self-
efficacy on top of the original TAM, hierarchical regression was employed to test H1 through
H4 for its advantage of revealing “the relative importance of each independent variable in the
prediction of the dependent variable” (Hair, Black, Babin, Anderson, & Tatham, 2010, p.
190). H5 was assessed by PROCESS using the bootstrapping approach, by which indirect
effects were estimated with 5,000 replicates (Hayes, 2013).
The variance inflation factor (VIF) for each variable was calculated for
multicollinearity diagnosis. The VIF values ranged from 1.171 to 1.990 and did not exceed
the cutoff value of 10 (Hair et al., 2010). Therefore, multicollinearity was not a point for
concern.
4. Results
H1 predicted that three key constructs under the TAM framework—perceived
usefulness, perceived enjoyment, and perceived ease of use—would be positively associated
with current users’ intentions to continue using fitness mobile apps. To test this hypothesis, a
hierarchical regression was performed, taking into account the influence of demographics and
the body mass index (BMI), with continuance intention as the outcome variable (Model 1).
As shown in Table 2, with the influence of demographics and BMI being controlled,
perceived usefulness (β = .353, p < .001), perceived enjoyment (β = .219, p < .001), and
perceived ease of use (β = .251, p < .001) were positively associated with the outcome
variable. The block composed of the three TAM variables accounted for 61.4% (incremental
R2) of the total variance in continuance intention. Thus, H1 was supported.
17
H2 predicted that perceived benefits related to four key functions of fitness mobile
apps would be positively associated with perceived usefulness. To test it, a hierarchical
regression was performed to take the influence of demographics and BMI into account, with
perceived usefulness as the outcome variable. As Table 2 shows, with the influence of
demographics and BMI being controlled, the four perceived benefits on instruction provision
(β = .248, p < .001), self-monitoring (β = .166, p < .01), self-regulation (β = .133, p < .01),
and goal attainment (β = .221, p < .001) were positive and significant predictors of perceived
usefulness. They explained 36.4% of the total variance in perceived usefulness. Thus, H2 was
supported.
[PLACE TABLE 2 HERE]
H3, H4, and RQ1 investigated the role of exercise self-efficacy in predicting
continuance intention. Exercise self-efficacy and its interaction terms with perceived
usefulness, perceived enjoyment, and perceived ease of use were included as a new block in
the hierarchical regression model in addition to demographics and BMI, as well as the TAM
predictors (Model 2). The results showed that the new block accounted for 3.2% of the total
variance of continuance intention. Specifically, in line with H3 and H4, exercise self-efficacy
was a positive and significant predictor of continuance intention (β = .165, p < .001), and a
significant interaction between exercise self-efficacy and perceived usefulness was present (β
= -.096, p < .05). Therefore, the positive association between perceived usefulness and
continuance intention would be greater for users who had lower exercise self-efficacy and
vice versa. However, no significant moderating effects of exercise self-efficacy on the
relationship between the other two TAM constructs and continuance intention were observed.
[PLACE TABLE 3 HERE]
H5 investigated the indirect effects of perceived benefits related to the four functions
of fitness mobile apps on continuance intention, which could be moderated by exercise self-
18
efficacy. Four separate conditional process analyses (Hayes, 2013) were performed using
PROCESS, an SPSS macro, to test the effects of the four functions on instruction provision,
self-monitoring, self-regulation, and goal attainment, respectively. When one of the four
items was entered into the model as the independent variable, the remaining three items were
included as covariates. Asymmetric bootstraps with 1,000 bootstrap replicates were employed
to assess the moderated mediation effects (Preacher, Rucker, & Hayes, 2007). Table 4
presents the results of the conditional process analyses. The results showed that all the four
functions of fitness mobile apps had significant impacts on current users’ continuance usage
intention; the impacts shared a pattern, such that the total effects could be decomposed into
direct effects (technological functions à continuance usage intention) and indirect effects
(technological functions à perceived usefulness à continuance usage intention).
Furthermore, the indirect effects were moderated by the level of exercise self-efficacy.
Specifically, for users who had higher exercise self-efficacy, the indirect effects of
technological functions on continuance usage intention through perceived usefulness became
smaller, as opposed to those who had lower exerciser self-efficacy, though the direct effects
remained the same. Therefore, H5 was supported.
[PLACE TABLE 4 HERE]
5. Discussion
5.1. Theoretical implications
Founded on a TAM approach, this study sought to link certain technological functions
of fitness mobile apps with current users’ continuance intentions and explored the role of
exercise self-efficacy in the relationship. The results showed that the three key TAM
constructs—perceived usefulness, perceived ease of use, and perceived enjoyment—
accounted for 61.4% of the variance in users’ intentions to continue using fitness apps.
Previous studies using TAM to predict technology adoption were able to explain
19
approximately 35%-62% of the variance in adoption behavior or behavioral intentions
involving various technologies, such as computer technology (Davis, Bagozzi, &
Warshaw,1989; Davis, Bagozzi, & Warshaw, 1992), hedonic information technology (Van
der Heijden, 2004), and online shopping sites (Koufaris, 2002). Hence, echoing Beldad and
Hegner (2018) and Lee and Cho (2017), the present study serves as a piece of empirical
evidence demonstrating the strong predictive power of TAM in the context of continued use
of fitness apps. Moreover, among the three key TAM constructs, perceived usefulness (β
= .404, p < .001) was found to have more predictive power than perceived ease of use (β
= .270, p < .001) and perceived enjoyment (β = .233, p < .001) in terms of standardized beta
coefficients. This finding is consistent with previous research on continuance intentions to
use health apps (Cho, Lee, & Quinlan, 2015) in that perceived usefulness was the major
determinant of behavioral intentions, suggesting a predominantly utilitarian approach to
decision making regarding fitness app use. In addition, this study also confirmed the
significant role of perceived enjoyment, contributing to resolving the conflicting results
emerged in prior research (Hew et al., 2015; Kang, 2014; Lee et al., 2010; Lee & Cho, 2017;
Lu et al., 2017). Given that the two studies that found that perceived enjoyment was not
significantly related to continued use used American samples (Lee & Cho, 2017; Lu et al.,
2017), we speculate that the divergence in the findings regarding perceived enjoyment may
be derived from different cultural contexts. Future studies are encouraged to replicate the
study in various cultural contexts other than China and the US to enhance the cross-cultural
generalizability and further explore the role of culture in predicting fitness app use under the
TAM framework.
Despite the fact that TAM is a powerful and parsimonious theoretical model, this
study extends TAM by opening the “black box” in terms of which particular app functions
positively predict perceived usefulness, and in turn result in higher continuance usage
20
intention. The present study found that the four key functions of fitness apps (i.e., instruction
provision, self-monitoring, self-regulation, and goal attainment) explained 36.4% of the
variance in perceived usefulness. Besides, direct and conditional indirect effects of the four
functions on continuance intention were observed; for those with low exercise self-efficacy,
the indirect effects of technological functions on continuance usage intention through
perceived usefulness were greater than that for those with high exercise self-efficacy.
Previous research on fitness apps focused on either the technological functions (Conroy et al.,
2014; Middelweerd et al., 2014) of fitness apps or user post-adoption attitudes and behaviors
(Beldad & Hegner, 2018; Lee & Cho, 2017). This study contributes to this line of research by
connecting the designer’s perspective with the user’s perspective and systematically testing
how particular technological functions influence user attitudes and behaviors (Chib & Lin,
2018). In this vein, this study also contributes to the TAM scholarship by elucidating the
antecedents of perceived usefulness from a human-technology interaction perspective. TAM
researchers have long noted the importance of identifying the exogenous variables in TAM
(Kim & Chang, 2007; Venkatesh & Davis, 1996). Nevertheless, prior research in this regard
mostly focused on user characteristics (e.g., user experience; Igbaria, Zinatelli, Cragg, &
Cavaye, 1997) or generic feelings towards the technology (e.g., perceived risk; Pavlou,
2003); whereas few studies have investigated the role of specific technological functions in
relation to TAM. The current study addresses this gap in the context of fitness mobile app use
and provides useful insights for app designers as to which specific functions are effective in
promoting continuous usage of fitness mobile apps.
Besides, the present study also goes a step further to explore predictors that are unique
to the context of fitness app use to better address the interplay between this technology and its
users. Addressing the uniqueness of the behavior facilitated by the technology, the addition of
exercise self-efficacy into the final model further increased the explained variance in
21
continuance intentions, which is aligned with its positive role in health intervention and
promotion (Annesi, 2012; Faghri et al., 2017). The findings indicate that exercise self-
efficacy has unique contributions to predicting continued use of fitness apps and that adding
it to TAM can increase its explanatory power in this particular context. More importantly, a
significant interaction between exercise self-efficacy and perceived usefulness was observed,
which is in line with previous research on other mobile apps and services (Clarke et al., 2014;
Islam et al., 2011). Specifically, those with high exercise self-efficacy weigh perceived
usefulness less than those with low exercise self-efficacy when determining whether to
continue using fitness apps. A possible explanation may lie in the key functions of fitness
mobile apps as suggested by the indirect effects discovered. According to content analysis
studies of fitness apps (Conroy et al., 2014; Huang & Zhou, 2018; Middelweerd et al., 2014),
one key set of functions of these apps involves facilitating self-regulation with the aim of
incorporating regular exercise into app users’ daily routines. In this sense, these functions
may be of more use to those with low exercise self-efficacy, since they are typically less
confident in performing exercise on a regular basis under various circumstances. Therefore,
they are in more need of a technological aid that monitors their progress and provides timely
feedback against their preset goals. As a result, when they decide whether they will continue
using the apps, the level of perceived usefulness, in terms of how effective they perceive the
app functions to be in helping them form a regular exercise habit, may carry greater weight in
the decision making process. In contrast, those with high exercise self-efficacy are confident
that they are able to perform exercise on a regular basis, so they rely on the self-regulatory
functions of fitness apps to a lesser extent. For them, perceived usefulness may play a less
important role in their decision making processes regarding whether to continue using the
apps. Hence, the interaction between exercise self-efficacy and perceived usefulness
22
underlines the importance of considering the boundary conditions under which the TAM
predictors are more effective in predicting continuous usage.
Table 5 summarizes the major findings of the present study compared to previous
works on health mobile app adoption and use. In summary, the major theoretical
contributions of this study lie in three aspects. One is extending TAM into the context of
continued use of fitness mobile apps, echoing Beldad and Hegner (2018). The second
contribution is unpacking the technological functions that drive users’ post adoption
behaviors through affecting their perceptions of the usefulness of the apps. This study is the
first to link self-regulatory functions with app user behaviors, while prior studies focused on
social functions only (Lee & Cho, 2017; Zhu et al., 2017). The third contribution is
identifying a moderator—exercise self-efficacy—that explains the differential impacts of
those functions to different groups of people in relation to TAM. The addition of this context
specific variable into the TAM model increases the explanatory power of the model.
[PLACE TABLE 5 HERE]
5.2. Managerial implications
From a practical standpoint, app developers may consider incorporating instruction
provision, self-monitoring, self-regulation, and goal attainment into new fitness app design,
as well as optimizing existing app interfaces and functions to improve users’ exercise
performance in the direction of providing useful instructions and complementing users’ self-
regulation mechanisms. Additionally, the moderating role of exercise self-efficacy indicates
the importance of the inclusion of relevant features in a single app to address the diverse
needs of users with different levels of exercise self-efficacy. Specifically, to appeal to those
with low exercise self-efficacy, app developers can place more emphasis on developing and
optimizing the four key functions directly associated with perceived usefulness. Whereas, to
attract those with high exercise self-efficacy, app developers may develop features that make
23
usage more enjoyable, such as gamification features, and optimize existing interfaces and
features to make them more user friendly (Lister et al., 2014; Payne et al., 2015). This finding
can be applied to general health interventions using non-mobile methods as well: intervention
programs aimed for promoting physical activity should be tailored to participants based on
their exercise self-efficacy levels; special attention needs to be paid to physically inactive
individuals with low exercise self-efficacy as they need more external motivations to help
them form a regular exercise routine than those with high exercise self-efficacy.
5.3. Limitations and future suggestions
The findings of this study should be interpreted with the acknowledgement of the
following limitations. First, the sample was randomly drawn from an online subject pool
consisting of over 2.6 million participants in China. Although the composition of the subject
pool displayed a diverse pattern in terms of demographics and social economic status, the
sample was not a representative sample of China’s mobile users. If resources allow, future
studies are encouraged to adopt more rigorous sampling methods to increase the
generalizability. It is worth noting that the sample of this study consists of slightly more
women and is more educated compared to China’s netizen population. Based on this finding,
we speculate that women are more conscious about weight management and are therefore
more likely to use fitness mobile apps; use of fitness mobile apps demands certain levels of
mobile media literacy and people with higher levels of education are more likely to use
fitness mobile apps. Future research is recommended to explore the demographic patterns of
fitness mobile apps. Second, this study identified exercise self-efficacy as an antecedent of
fitness app use, in that the level of exercise self-efficacy was a positive predictor of fitness
app continuance usage, while the literature suggests that it can also act as an outcome
variable, in that continuance usage may enhance users’ exercise self-efficacy (Cheng & Chen,
2018; McAuley & Blissmer, 2000). Given that this study was cross-sectional, we were unable
24
to examine the reciprocal relationship between exercise self-efficacy and continuance usage.
Future studies are encouraged to use longitudinal research design or experimental methods to
further elucidate the role of exercise self-efficacy in fitness app use. Lastly, this study
investigated continuance usage intention as the focal outcome variable rather than actual
behaviors. According to the theory of reasoned action (Fishbein & Ajzen, 1975) and the
theory of planned behavior (Ajzen, 1991), “the most immediate and important predictor of a
person’s behavior is his/her intention to perform it” (Sheeran, 2002, p. 1). This argument has
found empirical support with a wide range of human behaviors including physical activity
(Sheeran & Orbell, 2000). Moreover, a meta-analysis on the intention-behavior relationship
revealed that the weighted average effect size was r = .53 (Sheeran, 2002). According to
Cohen’s (1992), correlation coefficients greater than .50 indicate large effect sizes. This thus
gives us confidence that in the context of fitness app use, continuous usage intention is a
reasonable and strong predictor of actual use. We note, however, that there is a discrepancy
between behavioral intention and actual behavior. Future studies are recommended to
examine the effectiveness of technological functions in terms of promoting actual physical
activity behaviors in the field, which would help address the intention-behavior gap and
further contribute to the scholarship.
5.4. Conclusion
To understand how to increase user retention rates, the present study draws from
TAM and investigates the role of exercise self-efficacy, in addition to the original TAM
constructs, in predicting current users’ intention to continue using the apps. Moreover, this
study extends TAM from a human-technology interaction perspective by elucidating the
antecedents of perceived usefulness in terms of specific functions provided by fitness mobile
apps. The findings of an online survey indicate that four technological functions—instruction
provision, self-monitoring, self-regulation, and goal attainment—have an indirect effect on
25
continuance intention through perceived usefulness, and this indirect effect is moderated by
exercise self-efficacy such that the association between perceived usefulness and continuance
intention is stronger for those with low exercise self-efficacy. This research contributes to the
TAM scholarship by shedding light on the technological functions that positively predict
perceived usefulness and identifying exercise self-efficacy as a contingent variable that
explains the differential impacts of those functions to different groups of people. The
findings provide actionable insights into developing effective health interventions using
mobile technologies.
Disclosure Statement
No potential conflict of interest was reported by the authors.
26
References
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human
Decision Processes, 50, 179-211.
Alalwan, A. A., Dwivedi, Y. K., Rana, N. P., Williams, M. D. (2016). Consumer adoption of
mobile banking in Jordan: Examining the role of usefulness, ease of use, perceived risk
and self-efficacy. Journal of Enterprise Information Management, 29(1), 118-139.
doi:10.1108/JEIM-04-2015-0035
Annesi, J. J. (2012). Supported exercise improves controlled eating and weight through its
effects on psychosocial factors: Extending a systematic research program toward
treatment development. The Permanente Journal, 16(1), 7–18.
Annesi, J. J., & Marti, C. N. (2011). Path analysis of exercise treatment-induced changes in
psychological factors leading to weight loss. Psychology and Health, 26(8), 1081–1098.
doi:10.1080/08870446.2010.534167
Bagozzi, R. (2007). The legacy of the technology acceptance model and a proposal for a
paradigm shift. Journal of the Association for Information Systems, 8(4), 244–254.
doi:10.17705/1jais.00122
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84(2), 191–215. doi:10.1037/0033-295X.84.2.191
Bandura, A. (1991). Social cognitive theory of self-regulation. Organizational Behavior and
Human Decision Processes, 50(2), 248–287. doi:10.1016/0749-5978(91)90022-L
Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: W. H. Freeman.
Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of
Psychology, 52, 1-26.
Beldad, A. D., & Hegner, S. M. (2018). Expanding the technology acceptance model with the
inclusion of trust, social influence, and health valuation to determine the predictors of
27
German users’ willingness to continue using a fitness app: A structural equation
modeling approach. International Journal of Human-Computer Interaction, 34(9), 882–
893. doi:10.1080/10447318.2017.1403220
Belmon, L. S., Middleweerd, A., te Velde, S. J., & Brug, J. (2015). Dutch young adults’
ratings of behaviour change techniques applied in mobile phone apps to promote
physical activity: A cross-sectional survey. JMIR mHealth uHealth, 3(4), e103. doi:
10.2196/mhealth.4383
Bender, M. S., Choi, J. W., Arai, S., Paul, S. M., Gonzalez, P., & Fukuoka, Y. (2014). Digital
technology ownership, usage, and factors predicting downloading health apps among
Caucasian, Filipino, Korean, and Latino Americans: The Digital Link to Health
Survey. JMIR mHealth uHealth, 2(4), e43. doi:10.2196/mhealth.3710
Bhattacherjee, A. (2001). Understanding information systems continuance: An expectation-
confirmation model. MIS Quarterly, 25(3), 351–370. doi:10.2307/3250951
Chen, M. F., & Lin, N. P. (2018). Incorporation of health consciousness into the technology
readiness and acceptance model to predict app download and usage intentions. Internet
Research, 28(2), 351–373. doi:10.1108/IntR-03-2017-0099
Cheng, C. H., & Chen, C. H. (2018). Developing a mobile app-supported learning system for
evaluating health-related physical fitness achievements of students. Mobile Information
Systems, 2018. doi:10.1155/2018/8960968
Chib, A., & Lin, S. H. (2018). Theoretical advancements in mHealth: A systematic review of
mobile apps. Journal of Health Communication, 23(10-11), 909-955. doi:
10.1080/10810730.2018.1544676
Chin, S. O., Keum, C., Woo, J., Park, J., Choi, H. J., Woo, J.-T., & Rhee, S. Y. (2016).
Successful weight reduction and maintenance by using a smartphone application in
those with overweight and obesity. Scientific Reports, 6, 34563. doi:10.1038/srep34563
28
Cho, J. (2016). The impact of post-adoption beliefs on the continued use of health apps.
International Journal of Medical Informatics, 87, 75–83.
doi:10.1016/j.ijmedinf.2015.12.016
Cho, J., Lee, H. E., & Quinlan, M. (2015). Complementary relationships between traditional
media and health apps among American college students. Journal of American College
Health, 63(4), 248–257. doi:10.1080/07448481.2015.1015025
Cho, J., Quinlan, M. M., Park, D., & Noh, G.-Y. (2014). Determinants of adoption of
smartphone health apps among college students. American Journal of Health Behavior,
6(11), 860-870. doi:10.5993/AJHB.38.6.8
Clarke, J., Proudfoot, J., Birch, M-R., Whitton, A. E., Parker, G., Manicavasagar, V.,
Harrison, V., Christensen, H., & Hadzi-Pavlovic. (2014). Effects of mental health self-
efficacy on outcomes of a mobile phone and web intervention for mild-to-moderate
depression, anxiety and stress: Secondary analysis of a randomized controlled trial.
BMC Psychiatry, 14, 272.
CNNIC. (2018). China Internet Development Report. Retrieved from
http://www.cac.gov.cn/files/pdf/cnnic/CNNIC41.pdf
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159.
Conroy, D. E., Yang, C. H., & Maher, J. P. (2014). Behavior change techniques in top-ranked
mobile apps for physical activity. American Journal of Preventive Medicine, 46(6), 649–
652. doi:10.1016/j.amepre.2014.01.010
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of
information technology. MIS Quarterly, 13(3), 319–340. doi:10.2307/249008
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. doi:10.1287/mnsc.35.8.982
29
Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1992). Extrinsic and intrinsic motivation to
use computers in the workplace. Journal of Applied Social Psychology, 22(14), 1111-
1132. doi:10.1111/j.1559-1816.1992.tb00945.x
Dishaw, M. T., & Strong, D. M. (1999). Extending the technology acceptance model with
task-technology fit constructs. Information and Management, 36(1), 9–21.
doi:10.1016/S0378-7206(98)00101-3
eMarketer (2017). Fitness app usage grows by leaps and bounds. Retrieved from
https://www.emarketer.com/Article/Fitness-App-Usage-Grows-by-Leaps-
Bounds/1016486
Faghri, P. D., Simon, J., Huedo-Medina, T., & Gorin, A. (2017). Perceived self-efficacy and
financial incentives: Factors affecting health behaviors and weight loss in a workplace
weight loss intervention. Journal of Occupational and Environmental Medicine, 59(5),
453–460. doi:10.1097/JOM.0000000000000987
Ferguson, D. A., & Perse, E. M. (2000). The World Wide Web as a functional alternative to
television. Journal of Broadcasting & Electronic Media, 44(2), 155-174. doi:
10.1207/s15506878jobem4402_1
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to
theory and research. Reading, MA: Addison-Wes l ey.
Fleig, L., Küper, C., Lippke, S., Schwarzer, R., & Wiedemann, A. U. (2015). Cross-behavior
associations and multiple health behavior change: A longitudinal study on physical
activity and fruit and vegetable intake. Journal of Health Psychology, 20(5), 525–534.
doi:10.1177/1359105315574951
General Administration of Sport of China. (2015). Survey results of physical activity among
Chinese citizens in 2014. Retrieved from
http://www.sport.gov.cn/n16/n1077/n1422/7300210.html
30
Goodhue, D. L., & Thompson, R. L. (1995). Task-technology fit and individual performance.
MIS Quarterly, 19(2), 213. doi:10.2307/249689
Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2010). Multivariate
data analysis (6th ed.). Upper Saddle River, NJ: Pearson.
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process
analysis: A regression-based approach. New York, NY: Guilford Press.
Henseler, J., Ringle, C. M., & Sarstedt, M. (2015). A new criterion for assessing discriminant
validity in variance-based structural equation modeling. Journal of the Academy of
Marketing Science, 43, 115-135. doi:10.1007/s11747-014-0403-8
Hew, J.-J., Lee, V.-H., Ooi, K.-B., & Wei, J. (2015). What catalyses mobile apps usage
intention: An empirical analysis. Industrial Management & Data Systems, 115(7), 1269-
1291. doi: 10.1108/IMDS-01-2015-0028
Holden, R. J., & Karsh, B. T. (2010). The technology acceptance model: Its past and its
future in health care. Journal of Biomedical Informatics, 43(1), 159–172.
doi:10.1016/j.jbi.2009.07.002
Hsiao, C. H., Chang, J. J., & Tang, K. Y. (2016). Exploring the influential factors in
continuance usage of mobile social Apps: Satisfaction, habit, and customer value
perspectives. Telematics and Informatics, 33(2), 342–355.
doi:10.1016/j.tele.2015.08.014
Huang, G., & Zhou, E. (2018). Time to work out! Examining the behavior change techniques
and relevant theoretical mechanisms that predict the popularity of fitness mobile apps
with Chinese-language user interfaces. Health Communication. Advance online
publication. doi:10.1080/10410236.2018.1500434
31
Igbaria, M., Zinatelli, N., Cragg, P., & Cavaye, A. L. M. (1997). Personal computing
acceptance factors in small firms: A structural equation model. MIS Quarterly, 21(3),
279–305. doi:10.2307/249498
Islam, M. A., Khan, M. A., Ramayah, T., Hossain, M. M. (2011). The adoption of mobile
commerce service among employed mobile phone users in Bangladesh: Self-efficacy as
a moderator. International Business Research, 4(2), 80-89.
Jeon, E., & Park, H. (2015). Factors affecting acceptance of smartphone application for
management of obesity. Healthcare Informatics Research, 21(2), 74–82.
doi:10.4258/hir.2015.21.2.74
Kang, S. (2014). Factors influencing intention of mobile application use. International
Journal of Mobile Communications, 12(4), 360-379. doi:10.1504/IJMC.2014.063653
Kim, D., & Chang, H. (2007). Key functional characteristics in designing and operating
health information websites for user satisfaction: An application of the extended
technology acceptance model. International Journal of Medical Informatics, 76, 790-
800. doi:10.1016/j.ijmedinf.2006.09.001
Kim, H., Faw, M., & Michaelides, A. (2017). Mobile but connected: Harnessing the power of
self-efficacy and group support for weight loss success through mHealth intervention.
Journal of Health Communication, 22, 395-402. doi: 10.1080/10810730.2017.1296510
Kim, J., & Park, H. A. (2012). Development of a health information technology acceptance
model using consumers’ health behavior intention. Journal of Medical Internet
Research, 14(5), e133. doi:10.2196/jmir.2143
King, W. R., & He, J. (2006). A meta-analysis of the technology acceptance model.
Information and Management, 43(6), 740–755. doi:10.1016/j.im.2006.05.003
32
Koufaris, M. (2002). Applying the technology acceptance model and flow theory to online
consumer behavior. Information Systems Research, 13(2), 205-223.
doi:10.1287/isre.13.2.205.83
Krebs, P., & Duncan, D. T. (2015). Health app use among US mobile phone owners: A
national survey. JMIR mHealth and uHealth, 3(4), e101. doi:10.2196/mhealth.4924
Lee, H. E., & Cho, J. (2017). What motivates users to continue using diet and fitness apps?
Application of the uses and gratifications approach. Health Communication, 32(12),
1445–1453. doi:10.1080/10410236.2016.1167998
Lee, C. S., Goh, D. H., Chua, A. Y. K., & Ang, R. P. (2010). Indagator: Investigating
perceived gratifications of an application that blends mobile content sharing with game
play. Journal of the American Society for Information Science & Technology, 61(6),
1244-1257. doi:10.1002/asi.21305
Lee, M.-C. & Tsai, T.-R. (2010). What drives people to continue to play online games? An
extension of technology model and theory of planned behavior. International Journal of
Human–Computer Interaction, 26(6), 601-620. doi:10.1080/10447311003781318
Lim, J. S., & Noh, G.-Y. (2017). Effects of gain-versus loss-framed performance feedback on
the use of fitness apps: Mediating role of exercise self-efficacy and outcome
expectations of exercise. Computers in Human Behavior, 77, 249-257.
doi:10.1016/j.chb.2017.09.006
Lister, C., West, J. H., Cannon, B., Sax, T., & Brodegard, D. (2014). Just a fad? Gamification
in health and fitness apps. Journal of Medical Internet Research, 16(8), e9.
doi:10.2196/games.3413
Litman, L., Rosen, Z., Spierer, D., Weinberger-Litman, S., Goldschein, A., & Robinson, J.
(2015). Mobile exercise apps and increased leisure time exercise activity: A moderated
33
mediation analysis of the role of self-efficacy and barriers. Journal of Medical Internet
Research, 17(8), e195. doi:10.2196/jmir.4142
Lu, J., Liu, C., & Wei, J. (2017). How important are enjoyment and mobility for mobile
applications? Journal of Computer Information Systems, 57(1), 1-12.
doi:10.1080/08874417.2016.1181463
Luarn, P., & Lin, H-H. (2005). Toward an understanding of the behavioral intention to use
mobile banking. Computers in Human Behavior, 21, 873-891.
Mateo, G. F., Granado-Font, E., Ferré-Grau, C., & Montaña-Carreras, X. (2015). Mobile
phone apps to promote weight loss and increase physical activity: A systematic review
and meta-analysis. Journal of Medical Internet Research, 17(11), e253.
doi:10.2196/jmir.4836
McAuley, E., & Blissmer, B. (2000). Self-efficacy determinants and consequences of
physical activity. Exercise and Sport Sciences Reviews, 28(2), 85–88. doi:0091-
6631/2802/85-88
McAuley, E., Katula, J., Mihalko, S. L., Blissmer, B., Duncan, T. E., Pena, M., & Dunn, E.
(1999). Mode of physical activity and self-efficacy in older adults: A latent growth curve
analysis. Journal of Gerontology, 54(5), 283–292. doi:10.1093/geronb/54B.5.P283
McAuley, E., Lox, C., Rudolph, D., & Travis, A. (1994). Self-efficacy and intrinsic
motivation in exercising middle-aged adults. Journal of Applied Gerontology, 13(4),
355–370. doi:10.1177/073346489401300402
Merikivi, J., Tuunainen, V., & Nguyen, D. (2017). What makes continued mobile gaming
enjoyable? Computers in Human Behavior, 68, 411-421. doi:10.1016/j.chb.2016.11.070
Middelweerd, A., Mollee, J. S., van der Wal, C. N., Brug, J., & Te Velde, S. J. (2014). Apps
to promote physical activity among adults: A review and content analysis. International
34
Journal of Behavioral Nutrition and Physical Activity, 11, 97. doi:10.1186/s12966-014-
0097-9
Molina, M. D., & Sundar, S. S. (2018). Can mobile apps motivate fitness tracking? A study
of technological affordances and workout behaviors. Health Communication. Advanced
online publication. doi:10.1080/10410236.2018.1536961
Pavlou, P. A. (2003). Consumer acceptance of electronic commerce: Integrating trust and risk
with the technology acceptance model. International Journal of Electronic Commerce,
7(3), 101–134. doi:10.1080/10864415.2003.11044275
Payne, H. E., Moxley, V. B., & MacDonald, E. (2015). Health behavior theory in physical
activity game apps: A content analysis. JMIR Serious Games, 3(2), e4.
doi:10.2196/games.4187
Podsakoff, P. M., MacKenzie, S. B., Lee, J-Y., & 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. doi:10.1037/0021-
9010.88.5.879
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing moderated mediation
hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research,
42(1), 185–227. doi:10.1080/00273170701341316
Rabin, C., & Bock, B. (2011). Desired features of smartphone applications promoting
physical activity. Telemedicine and e-Health, 17(10), 801-803. doi:
10.1089/tmj.2011.0055
Schoeppe, S., Alley, S., Van Lippevelde, W., Bray, N. A., Williams, S. L., Duncan, M. J., &
Vandelanotte, C. (2016). Efficacy of interventions that use apps to improve diet,
physical activity and sedentary behavior: A systematic review. International Journal of
Behavioral Nutrition and Physical Activity, 13(1), 127. doi:10.1186/s12966-016-0454-y
35
Sheeran, P. (2002). Intention—behavior relations: A conceptual and empirical review.
European Review of Social Psychology, 12(1), 1-36. doi:10.1080/14792772143000003
Sheeran, P., Montanaro, E., Bryan, A., Miles, E., Maki, A., Avishai-Yitshak, A., . . .
Rothman, A. J. (2016). The impact of changing attitudes, norms, and self-efficacy on
health-related intentions and behavior: A meta-analysis. Health Psychology, 35(11),
1178–1188. doi:10.1037/hea0000387
Sheeran, P., & Orbell, S. (2000). Self-schemas and the theory of planned behavior. European
Journal of Social Psychology, 30, 533-550.
Shin, Y., Jang, H., & Pender, N. J. (2001). Psychometric evaluation of the exercise self-
efficacy scale among Korean adults with chronic diseases. Research in Nursing &
Health, 24(1), 68–76. doi:10.1002/1098-240X(200102)24:1<68::AID-
NUR1008>3.0.CO;2-C
Statista (2018). Global mobile health app downloads 2013-2017. Retrieved from
https://www.statista.com/statistics/625034/mobile-health-app-downloads/
Stiglbauer, B., Weber, S., & Batinic, B. (2019). Does your health really benefit from using a
self-tracking device? Evidence from a longitudinal randomized control trial. Computers
in Human Behavior, 94, 131-139. doi:10.1016/j.chb.2019.01.018
TechNode (2017). Mobile medical healthcare trends in China, part 2: How mobile health care
can help. Retrieved from https://technode.com/2017/02/04/mobile-medical-healthcare-
trends-in-china-part-2-how-mobile-health-care-can-help/
Teixeira, P. J., Carraça, E. V., Marques, M. M., Rutter, H., Oppert, J. M., De Bourdeaudhuij,
I., . . . Brug, J. (2015). Successful behavior change in obesity interventions in adults: A
systematic review of self-regulation mediators. BMC Medicine, 13(1), 1–16.
doi:10.1186/s12916-015-0323-6
36
Turner-McGrievy, G. M., Beets, M. W., Moore, J. B., Kaczynski, A. T., Barr-Anderson, D.
J., & Tate, D. F. (2013). Comparison of traditional versus mobile app self-monitoring of
physical activity and dietary. Journal of the American Medical Informatics Association,
20(3), 513–518. doi:10.1136/amiajnl-2012-001510
Van der Heijden, H. (2004). User acceptance of hedonic information systems. MIS Quarterly,
28(4), 695–704. doi:10.2307/25148660
Venkatesh, V. (2000). Determinants of perceived ease of use: Integrating control, intrinsic
motivation, and emotion into the technology acceptance model. Information Systems
Research, 11(4), 342–365. doi:10.1287/isre.11.4.342.11872
Venkatesh, V., & Bala, H. (2008). Technology acceptance model 3 and a research agenda on
interventions. Decision Sciences, 39(2), 273–315. doi:10.1111/j.1540-
5915.2008.00192.x
Venkatesh, A., & Davis, F. D. (1996). A model of the antecedents of perceived ease of use:
Development and test. Decision Sciences, 27, 451–481.
Venketesh, V., Thong, J. Y. L., & Xu., X. (2012). Consumer acceptance and use of
information technology: Extending the unified theory of acceptance and use of
technology. MIS Quarterly, 36(1), 157-178. doi:10.2307/41410412
Wang, B. R., Park, J. Y., Chung, K., & Choi, I. Y. (2014). Influential factors of smart health
users according to usage experience and intention to use. Wireless Personal
Communications, 79(4), 2671–2683. doi:10.1007/s11277-014-1769-0
Yang, K. (2010). The effects of technology self-efficacy and innovativeness on consumer
mobile data service adoption between American and Korean consumers. Journal of
International Consumer Marketing, 22(2), 117-127. doi:10.1080/08961530903476147
37
Zhu, Y., Dailey, S. L., Kreitzberg, D., & Bernhardt, J. (2017). “Social networkout”:
Connecting social features of wearable fitness trackers with physical exercise. Journal of
Health Communication, 22(12), 974-980. doi:10.1080/10810730.2017.1382617
38
Table 1
Main Sociodemographic Characteristics of Study Participants (N = 449)
Study Participants (N = 449)
M/Percentage
SD
Sex
Female 57%
-
Age
31.85
6.91
Education
High School or Below 4.9%
Bachelor’s Degree 84.2%
Master’s Degree or Above 10.4%
Others 0.4%
-
Monthly Family Income
8000 CNY or Below 13.1%
8001~10000 CNY 14.5%
10001~20000 CNY 47.2%
20001~40000 CNY 16.9%
Above 40000 CNY 8.2%
-
Body Mass Index (BMI)
21.77
3.34
Note. 1 CNY 0.14 USD
39
Table 2
Hierarchical Regression Analysis Predicting Perceived Usefulness (N = 449)
Independent Variables
Perceived Usefulness
Beta
t
Block 1: Demographics and BMI
Sex
.021
.504
Age
.040
.995
Education
-.042
-1.077
Monthly Family Income
-.016
-.419
BMI
.021
.493
R2
.007
F
.664
Block 2: Technological Functions
Fitness mobile apps give me useful instructions on exercise.
.248
5.235***
Fitness mobile apps help me monitor my physical activity progress.
.166
3.441**
Fitness mobile apps improve my self-regulation.
.133
2.746**
Fitness mobile apps help me achieve my physical activity goals.
.221
4.569***
R2
.372
Incremental R2
.364
F Change
63.431***
Note. Entries are standardized regression coefficients from a final regression equation with all
blocks of variables in the model. *p < .05. **p < .01. ***p < .001.
40
Table 3
Hierarchical Regression Analysis Predicting Continuance Usage Intention (N = 449)
Independent Variables
Model 1
Model 2
Beta
t
Beta
t
Block 1: Demographics and BMI
Sex
-.043
-1.370
-.033
-1.088
Age
.011
.345
.004
.148
Education
-.046
-1.536
-.031
-1.082
Monthly Family Income
.009
.295
.010
.336
BMI
-.079
-2.505*
-.053 -1.712#
R2
.023
.023
F
2.099#
2.099#
Block 2: TAM predictors
Perceived Usefulness (PU)
.404
10.055***
.353
8.888***
Perceived Enjoyment (PE)
.233
5.869***
.219
5.710***
Perceived Ease of Use (PEOU)
.270
6.750***
.251
6.506***
R2
.637
.637
Incremental R2
.614
.614
F Change
247.669***
247.669***
Block 3: Exercise Self-Efficacy
Exercise Self-Efficacy (ESE)
.165
5.469***
ESE PU
-.096
-2.226*
ESE PE
-.031
-.698
ESE PEOU
-.004
-.080
R2
.669
Incremental R2
.032
F Change
10.543***
Note. Entries are standardized regression coefficients from a final regression equation with all
blocks of variables in the model. Variables were mean centered before the product terms
were created. #p < .10. *p < .05. **p < .01. ***p < .001.
41
Table 4
Results of Conditional Process Analysis of the Effects of Technological Functions on
Continuance Usage Intention
Technological
Functions
Effect Type
Exercise Self-
Efficacy Value
Effect (SE)
95% Confidence Interval
Instruction
Provision
Direct
-
.2065 (.0294)
[.1487, .2644]
Indirect
3.8333
.2713 (.0298)
[.2123, .3280]
4.7778
.2201 (.0257)
[.1706, .2714]
5.5000
.1809 (.0262)
[.1317, .2345]
Self-
Monitoring
Direct
-
.2273 (.0291)
[.1700, .2845]
Indirect
3.8333
.2594 (.0303)
[.2011, .3205]
4.7778
.2092 (.0260)
[.1601, .2622]
5.5000
.1708 (.0260)
[.1227, .2248]
Self-
Regulation
Direct
-
.2281 (.0263)
[.1764, .2799]
Indirect
3.8333
.2250 (.0278)
[.1686, .2787]
4.7778
.1871 (.0243)
[.1379, .2343]
5.5000
.1580 (.0248)
[.1096, .2069]
Goal
Attainment
Direct
-
.2203 (.0259)
[.1695, .2711]
Indirect
3.8333
.2276 (.0272)
[.1741, .2817]
4.7778
.1898 (.0238)
[.1438, .2373]
5.5000
.1608 (.0248)
[.1143, .2112]
42
Table 5
Summary of Representative Works on Health App Adoption and Continued Use in Comparison with the Present Study
Article
Focal
Behavior
Major Findings
Beldad & Hegner
(2018)
Continued
use
Perceived usefulness, perceived ease of use and injunctive social norms, predicted intentions to
continue using a particular fitness app.
Bender, Choi, Arai,
Paul, Gonzalez, &
Fukuoka (2014)
Initial
adoption
Education and family history of heart attack were positively associated with actual behaviors of
downloading health apps. Compared to Caucasians, Latino and Korean users were less likely to use
health apps. Age was a negative predictor of health app usage.
Chen & Lin (2018)
Initial
adoption
Optimism influenced perceived ease of use and perceived usefulness of dietary and fitness apps, which
in turn affected the intention to download and use the apps.
Cho, Quinlan, Park, &
Noh (2014)
Initial
adoption
Health consciousness and subjective norms affected intentions to adopt health apps through perceived
usefulness; eHealth literacy had an indirect impact on adoption intentions through perceived ease of
use.
Cho (2016)
Continued
use
Perceived usefulness, perceived ease of use, confirmation, and satisfaction were significant predictors
of intentions to continue using health mobile apps.
Jeon & Park (2015)
Initial
adoption
Compatibility, perceived usefulness, and perceived ease of use were significant predictors of the
intention to use a particular mobile obesity-management app.
Krebs & Duncan
(2015)
Continued
use
Lack of interest, hidden costs, high data entry burden, and concern about privacy intrusion were major
factors leading to discontinued use of health apps.
Lee & Cho (2017)
Continued
use
Recordability, networkability, credibility, comprehensibility, and trendiness were significant predictors
of users’ intentions to continue using diet/fitness apps.
Zhu et al. (2017)
Exercise
Two social features—social sharing and competing—influenced users’ exercise intentions through
attitudes, subjective norms, and perceived behavioral control.
The Present Study
Continued
use
Four technological functions—instruction provision, self-monitoring, self-regulation, and goal
attainment—had an indirect effect on continuance intention through perceived usefulness, and this
indirect effect was moderated by exercise self-efficacy. Perceived ease of use and perceived enjoyment
were also predictors of continuance intention.
43
Figure 1. The effects of technological functions on continuance usage intention.
H1a
Technological Functions
• Instruction
provision
• Self-monitoring
• Self-regulation
• Goal attainment
Perceived
Usefulness
Continuance
Usage Intention
Exercise Self-
Efficacy
H2
H4
H5
H3
Perceived Ease
of Use
Perceived
Enjoyment
H1b
H1c
