1 Introduction

Augmented reality (AR) is a technology that can integrate computer generated virtual images with the real world (Berryman 2012). In recent years, with the upgrading of smart device hardware and the popularity of mobile devices (e.g., iPad, iPhone, Android devices), AR application developers have begun to tend to design and develop augmented reality applications on different mobile devices, and consumers can use augmented reality functions on their smartphones (Dacko 2017; Wei 2019). Nowadays, AR has been used in different fields and may be used as a breakthrough technology to subvert the market and become popular among people (Rauschnabel 2021). According to the report of Corporation(IDC) (2020), affected by the COVID-19 pandemic, the expenditure of the global augmented reality market is expected to grow rapidly, from $12 billion in 2020 to $72.8 billion in 2024. This shows that AR has received extensive attention from investors. In the future, augmented reality technology will gradually enter people’s daily lives and has the potential to change the way we interact with the world.

Digital sports solutions have become particularly important in recent years as outdoor exercise has been significantly restricted due to the COVID-19 pandemic (Ruth et al. 2022). In this context, virtual exercise has gradually attracted people’s attention. In a survey of the top 20 most popular fitness trends in the world published by ACSM in 2021, virtual training and exercise applications ranked 6th and 12th, respectively, as digital exercise technologies (Thompson 2021). And AR technology is gradually being used in sports. For example, an AREA called Tiantian Rope Skipping in the Apple Store has been popular and ranks quite high among exercise apps (China.com.cn 2022). The existing AREAs create a ubiquitous gym for users at their homes through the virtual objects embedded in reality and allow users to interact with virtual content for performance feedback. Therefore, it has changed the way people exercise compared to traditional forms. Considering the growing trend of AREAs in the population, it is important to explore how it is being accepted and further used by users. However, little is known about how users evaluate these applications and their continuance intention. In fact, there is an AR industry report that more than 1/4 of applications will no longer be used after the first download (Nikhashemi et al. 2021). With the gradual rise and application of AR services, it is necessary for AR application providers to understand the user experience when using them and user’s continuance intention.

In addition, AR has proven its value to consumers in previous marketing areas (Caboni and Hagberg 2019; Rauschnabel et al. 2019). With the application of AR in different fields, there are studies that acknowledge a similar effect of AR in the sports field (Goebert and Greenhalgh 2020). However, the previous literature on AR sports mainly focused on physical performance (Chen et al. 2020; Jeon and Kim 2020; Ng et al. 2019) or the proposal of the system (Bielli and Harris 2015). Few studies focused on the application of AR technology to user experience in the context of exercise, and we do not know much about how AREAs affect users’ psychological perception and experiential value. In fact, this is quite a lack of research in the current AR research field, and there is still much to be explored in understanding the perception and cognition of users (Qin et al. 2021; Tredinnick 2018).

This study also conforms to the appeal of scholars (Ng et al. 2019; Rogers et al. 2017), they propose to promote the research on AR in the field of sports and focus on different usage types and contents due to the lack of research on AR in this area. Previous studies have pointed out that the characteristic aspects of mobile applications can arouse users’ perception of utilitarian and hedonic benefits and influence engagement intention (Fang et al. 2017). While research in exercise apps has considered this aspects (Higgins 2016; Middelweerd et al. 2014), with AREAs have incorporated the new characteristics of AR, there is little literature to discuss how these new characteristics will affect user behavior in a mature theoretical framework at present.

Based on the previous discussion, how do the characteristics of AREAs affect user perception and experience have lacked necessary exploration in the existing literature, which largely limits the understanding of relevant researchers and practitioners on how AR shapes user experience and enables users to further use AREAs to obtain the potential benefits they provide. Current research questions include: (1) whether and how specific characteristics of AREAs affect users’ psychological feeling and experiential value; (2) Whether and how the user’s psychological feeling and experiential value affect the continuance intention of AREAs. This study aims to answer the above questions. Specifically, we attempt to fill in the theoretical gaps by proposing the original proposition of AREAs characteristic classification based on previous AR literature and dividing their characteristics into two categories (i.e., Service and System characteristics) and carefully examining how these specific characteristics of AREAs affect users' continuance intentions based on the Stimulus-organism-response framework and validate the research model through an online survey. Since virtual sports integrated with AR are currently an emerging concept, it is of vital importance to understand users’ behavioral intentions in this context, as this helps improve AREAs and encourages users to continue using such services, and this study takes a leading step in this direction through empirical investigation.

The rest of the study is further organized with the literature review and theoretical background section that discusses the gaps in the current research field and introduces the basis of the theoretical framework of this paper. In the next section, the hypotheses and model development are explained. Further, it continues with the discussion of the research methodology and followed by the results presentation section. The final part of the study deals with the theoretical and practical implications and limitations and future research.

2 Literature review and theoretical background AR in exercise

The current literature shows that AR technology has been applied in sports training and shows the benefits it brings to users at the physical fitness level. Applying AR technology to the health management of the elderly will help improve their quality of life. For example, Chen et al. (2020) has found that AR-assisted training systems can improve balance control and lower limb muscle strength in older adults. Accordingly, Jeon and Kim (2020) found that augmented reality-based exercise programs were effective in preventing muscle atrophy in older adults, while also improving their exercise self-efficacy. In addition, playing AR games (e.g., Pokemon Go) have also been found to be good for players' physical health, as it increases the time spent outdoors and improves players' physical activity levels (Bonus et al. 2017; Kaczmarek et al. 2017; Ruiz-Ariza et al. 2018). A related experimental study also suggests that AR-based fitness games help with exercise planning, and it may provide users with an effective form of exercise (Wiederhold et al. 2019). Affected by the epidemic and the rapid spread of mobile devices, the combination of AR and VR assisted sports training is gradually becoming an important research field in the future (Chang et al. 2019). More and more people will come into contact with AR technology in sports and fitness. Considering the growth trend of AREAs and its potential, it is important to understand how it is further used by users. However, to the best of our knowledge, there are few empirical studies in the AREAs that investigate users' willingness to continue using. Most of the research on AR motion focuses on the technical level, such as the presentation of systems (Bielli and Harris 2015). More research is needed to better understand users' continuance intention and motivate them to continue using such services. In practice, the lack of a satisfying experience increasingly seems to be a non-negligible barrier to the adoption of immersive technologies (Ibáñez‐Sánchez et al. 2022). Practitioners are therefore required to focus on how to build the functional characteristics of AREAs to provide the best experience for users (Hsu et al. 2021). To address these gaps, we propose the original proposition of AREAs characteristic classification based on previous AR literature and explore the impact of AREAs characteristics on user perception and continuance intention through empirical research. Understanding how users perceive the experience of using AREAs can be beneficial because it helps practitioners retain and improve the characteristics of AREAs and encourages continued use of such applications.

2.1 The S–O-R paradigm

The "stimulus-organism-response" (S–O–R) paradigm was originally originated from environmental psychology, which believed that external environmental factors ("stimuli") could affect the internal state of an organism ("organism"), thereby leading to the subsequent psychological or behavioral response ("response") of the organism (Mehrabian and Russell 1974). The internal state of the organism includes individual perception, feeling and emotional response (Bagozzi 1986). These internal states finally lead to the approaching or avoiding behavior of organisms (Bitner 1992; Donovan and Rossiter 1982).

The S–O–R paradigm is closely related to the context of this study. As previously pointed out in the literature, the S–O–R model is very effective in explaining user behavior in the initial stage of adoption (Davis 1989; Hsiao et al. 2016). Accordingly, Fang et al. (2017) pointed out that SOR model is applicable to explain the mechanism by which product attributes drive user ultimate behavioral engagement. Based on the above logic, in this study, we take SOR as a general framework to understand which characteristics of AREAs drive user continuance use and the influencing mechanisms therein.

2.2 AR service characteristics and AR system characteristics as stimuli (S)

Existing IS study has confirmed that product features can act as external environmental stimuli to influence users’ participation in technology (Venkatesh et al. 2012). Considering that product functional characteristics of mobile applications may have an impact on users (Lee et al. 2011), we propose the original proposition of AREAs characteristic classification based on the previous AR literature. Specifically, this study refines and tests two types of AR characteristics, namely AR service characteristics and AR system characteristics. Service characteristics include environmental embedding, ubiquitous, and gamification, whereas system characteristics include interactivity and responsiveness. We expect them to act as stimuli that will have an impact on the user’s subsequent mental state and behavior.

2.3 AR service characteristics

Environmental embedding is defined as "the visual integration of virtual content into a person’s real-world environment" (Hilken et al. 2017) and is provided as a service for users by AR (Hilken et al. 2017). AR cannot only combine physical objects and virtual scenes, but also embed virtual objects into physical scenes (Chung et al. 2015). In the context of AREAs, users can see their body movements and poses in any virtual motion scene through environmental embedding. Immersive AR technology with environmental embedding brings new services to people, which also requires researchers to conduct more research to understand users’ immersive experience (Butt et al. 2021; Wang et al. 2020). Given that few studies have explored the impact of environmental embeddings on user continuance intention, it is necessary to explore this issue.

Ubiquitous refers to the characteristic that AR applications provide users with anytime, anywhere services, defined in this study as AR allowing individuals to access and interact with their personal IT anytime, anywhere to maintain contact with IT artifacts (Hu et al. 2022). Recently, some scholars have also proposed that portability or mobility can be a major feature of AR (Javornik 2016). This shows that AR appears on smart mobile devices in the form of mobile applications, providing users with services and convenience anytime and anywhere. People are allowed to enjoy ubiquitous services to exercise in their own private environment as AREAs can create a real-time virtual motion scene for users.

Gamification is generally accepted as "the use of game design elements in non-game contexts" (Deterding et al. 2011) and is considered as a service for users (Hamari and Koivisto 2015). It uses some incentive mechanisms to improve the original products or services and make them more compelling, so as to influence users’ behaviors and attitudes (Yang et al. 2017) and promote users’ participation (Simões et al. 2013). Points, badges, levels, mission objectives, leaderboards and other elements are common features of gamified design (Hamari et al. 2014). In the AREAs concerned by this research, users can earn points based on their performance and allow them to participate in leaderboard rankings after completing different events. Because these features are consistent with the description of gamification, this study attempts to propose gamification as a service characteristic of AR applications.

2.4 AR system characteristics

Interactivity is defined as the degree to which AR applications allow users to access different content and interact with different user interfaces in the context of AR applications (Hsu et al. 2021). The services provided by AREAs allow users to easily interact with virtual content and participate in the exercise process in real time. At present, interactivity has been considered as part of the system quality by IS scholars (Zheng et al. 2013) and can be used to understand users’ technology uses experience (Cyr et al. 2009; Voorveld et al. 2011).

Responsiveness is defined as the degree of delay in the exchange of information (Song and Zinkhan 2008). While other scholars focus on the appropriate and relevant degree of response in communication (Johnson et al. 2006). In the recent AR literature, responsiveness and response time are considered as important components of the processing quality of AR systems (Kowalczuk et al. 2021). While system quality captures the ability of the system to perform reliably and accurately and to respond to requests at an appropriate speed (Kowalczuk 2018). Considering that when users exercising with AREAs, the system will respond to users based on their real-time actions. Therefore, this study focuses on the dimension of response speed to understand how responsiveness affects user experience and behavior.

2.5 Experiential value and presence as users’ internal states (O)

Experiential value is a kind of psychological perception, which comes from users’ actual use or indirect observation of products or services (Mathwick et al. 2001). In this study, experiential value is defined as the user’s assessment of the value of positive outcomes resulting from the use of AREAs. This study specified utilitarian and hedonic as user experiential values, which were classified into different types of personal online experiential values by researchers in the recent IS study (Hsu et al. 2021). Mobile AR applications were initially developed to serve marketing purposes and have significantly influenced consumers' shopping experience, including two important dimensions: utilitarian and hedonic (Huang and Liao 2015). For example, AR helps consumers evaluate products by overlaying virtual information (Scholz and Duffy 2018). At the same time, the novel way of shopping that integrates AR technology also brings a sense of enjoyment to customers (Park and Yoo 2020). In the case of AREAs, they allow users to participate in different categories of exercises to meet different needs of users. In other words, users may experience the utilitarian and hedonic value when their needs are met. In addition, with the increasing use of technology in the combination of virtual and real, the experience brought by technology has become a topic of discussion (Wang et al. 2021). One such perspective that has received much attention is “presence,” which has been described by other scholars as telepresence (Nah and DeWester 2011) or spatial presence (Schubert et al. 2001). According to Steuer (1992), telepresence is defined as “the extent to which one feels present in the mediated environment, rather than in the immediate physical environment.” In other words, it describes an individual’s subjective feeling of being in a place or environment, even if he is not actually there (Mollen and Wilson 2010). In virtual environments, the theoretical perspective of presence has been increasingly adopted, as it can be used as a key structure for understanding user experiences in virtual environments (Wei et al. 2019). In this study, presence refers to the psychological feeling that users are intoxicated in the mediation environment guided by AR technology when they exercise by interacting with virtual objects in the process of using AREAs.

Therefore, we combine the internal state of the individual and model the internal psychological perception (i.e., value assessment) and psychological feeling (i.e., presence) of the user as the components of the organism in S–O–R paradigm.

2.6 Satisfaction and continuance intention as response (R)

The response in the SOR framework encompasses psychological and behavioral responses as the result of a sequence of events (Mehrabian and Russell 1974). In this study, these two responses refer to satisfaction and continuance intention, respectively. In the context of information systems, some literature considers satisfaction as the emotional response of users to information system features (Bailey and Pearson 1983; Doll and Torkzadeh 1988), which usually comes from an individual’s positive evaluation after a product or service meets or exceeds his expectations (Bhattacherjee and Premkumar 2004). Continuance intention in this context refers to the extent to which users intend to use the application again, and previous studies have identified the important role of intention as it can positively predict the actual behavior of users (Venkatesh and Davis 2000).

2.7 Model development and hypotheses

2.7.1 AR characteristics and utilitarian and hedonic value

Following previous research, in this paper, utilitarian value is defined as the user’s perception of how effective the technology is in achieving their goals or solving problems (Venkatesh et al. 2003; Venkatesh and Xu 2012). Utilitarian value is often goal-oriented and emphasizes the outcome of task completion, and it reflects the usefulness of using AREAs in achieving the user’s exercise and fitness goals. Hedonic value is defined in this paper as the user’s subjective perception of the degree of pleasure and enjoyment brought by the use of technology (Rauschnabel et al. 2018; Venkatesh and Xu 2012). Compared with utilitarian value, hedonic value emphasizes not the aspects that AREAs can help to complete tasks, but the potential positive emotions or feelings it brings to users, such as entertainment, enjoyment and play (Hirschman and Holbrook 1982; Zhou et al. 2014).

In the early marketing literature, sensory and experiential attributes of products were thought to be associated with hedonism, while instrumental and functional attributes were associated with utilitarianism (Batra and Ahtola 1991). Recent research in the context of AR marketing has shown that AR's unique visual technology supplements additional product information and reduces the uncertainty of consumers when choosing products (Hoyer et al. 2020). The interactive function of AR also promotes consumer participation, allowing them to experience pleasure (Chopdar and Balakrishnan 2020). In addition, the attributes of AR itself may also help shape consumers' shopping experience (Park and Yoo 2020). Therefore, based on the above evidence, we anticipate that the different characteristics of AREAs will affect the user experiential value, including both utilitarian and hedonic aspects.

Environmental embedding facilitates specific actions by changing the immediate environment (Hilken et al. 2017). In other words, the realization of specific actions may depend on the relationship between an individual and his or her environment (Hilken et al. 2017). AREAs help the user present a scene that combines abstract virtual content with the real-time physical environment of the user, without the need to go through the user’s own imagination (Huang 2021). In this way, they can visually see how their movements and postures match up with real-time scenes in any virtual movement scene, thus conducting their exercise training in a more natural and effective way. Besides, according to cognitive load theory, people’s cognitive load may be in a high level when they need to process a large amount of information due to the limited cognitive resources and abilities (Sweller 1988). If that happens, users may have negative or passive emotions (Liu and Goodhue 2012). As discussed previously, the environmental embedding characteristic of AR applications can help users directly display information about themselves and their surroundings, which greatly reduces the mental effort and cognitive load of users, which may help them form a positive emotion. Therefore, we posit the following:

Hypothesis 1

Environmental embedding characteristic of AREAs will positively affect users’ perception of (a) utilitarian value and (b) hedonic value.

The ubiquitous means that people can shed the restrictions on the business hours and locations of offline sports venues, which brings them great convenience, especially for those with busy schedules (Liu et al. 2020). With the help of the ubiquitous characteristic of AREAs, users can flexibly choose the time to exercise in their own family, thus helping them to better implement and complete the exercise plan according to their intention. Besides, previous studies have shown that the ubiquitous connectivity of smart technology allows people to get and keep in touch with family, friends and colleagues, which satisfies users’ needs and increases their satisfaction (Looney et al. 2004; Valenzuela et al. 2009). AREAs can meet the needs of users for exercising anytime and anywhere and may save a lot of trouble for people with busy schedules, which helps them to improve their life satisfaction, and then, they may have a happy psychological feeling about the use of this technology. Thus, we posit the following:

Hypothesis 2

Ubiquitous characteristics of AREAs will positively affect users’ perception of (a) utilitarian value and (b) hedonic value.

In previous literature, Hamari and Koivisto (2015) believe that the purpose of gamification is utilitarian; specifically, the ultimate goal of gamification is usually to accomplish some external utilitarian goals. Furthermore, several studies on exercise games have shown that the utilitarian benefit of playing games is to improve physical performance (Hamari and Koivisto 2015; Tu et al. 2019). Since AREAs rely entirely on the user’s physical activity during exercise, gamified design may be related to the ultimate utilitarian goal, such as improving physical performance. Besides, gamified design in smart phones has been found to bring users more fun than non-gamified design (Triantoro et al. 2019). Meanwhile, studies have found that gamification affects flow state (García-Jurado et al. 2019). Flow here refers to the state of being completely immersed in an event when a person is involved (Buchanan and Csikszentmihalyi 1991), and when people are in a state of flow, they tend to experience a higher level of enjoyment and fun (Buchanan and Csikszentmihalyi 1991). Therefore, we propose the following:

Hypothesis 3

Gamification characteristic of AREAs will positively affect users’ perception of (a) utilitarian value and (b) hedonic value.

Previous studies have shown that the interactive function of AR can help consumers to get more information from all aspects of the product, thus helping them to better evaluate the product (Scholz and Duffy 2018). Through the interactive technology of AR, users can seek to obtain more and higher quality information to reap utilitarian benefits (Zheng et al. 2019). In the context of AREAs, interactivity allows users to constantly receive information feedback during interaction with virtual objects, which helps them to understand their own motion and thus better evaluate their performance. Besides, the existence of interaction enables users to communicate between the virtual and the real, which helps them generate positive emotional responses (Fiore et al. 2005). Recent AR research shows that users have improved their participation and enjoyment when interacting with mobile AR applications (Chopdar and Balakrishnan 2020). When interacting with virtual objects on the screen, a highly interactive experience may give them fun, which helps to trigger pleasant emotions. Therefore, the following are proposed:

Hypothesis 4

Interactivity characteristics of AREAs will positively affect users’ perception of (a) utilitarian value and (b) hedonic value.

Responsiveness as an important component of IS quality has been generally accepted by researchers (Gorla et al. 2010; Lee et al. 2007). Some previous studies found the positive impact of quality on perceived usefulness by taking IS quality as the antecedent of usefulness (Ahn et al. 2007; Lin 2008). In fact, users want to simply and quickly received feedback from a system on their action recognition; a reasonable response time may help them to get the needed information, faster to increase their usefulness to the system of perception, thus users believe that with the help of the system can better achieve their goal. Besides, in the context of previous online websites, the impact of response speed on users as an attribution of the website has attracted the attention of researchers (Wixom and Todd 2005). Song and Zinkhan (2008) found that the length of website response time would affect users’ perception of website quality and their satisfaction. Objectively speaking, if the system cannot quickly and effectively recognize and respond to the user’s input actions during movement, it may bring bad experience to the user and cause their negative emotions. Therefore, we posit the following:

Hypothesis 5

Responsiveness characteristic of AREAs will positively affect users’ perception of (a) utilitarian value and (b) hedonic value.

2.7.2 AR characteristics and presence

Situated view of presence considers that when an individual is in the feeling of presence, he will feel that he can do something in the present situation (Schultze 2010). The environmental embedding characteristic of AR combines virtual elements (such as objects or scenes) with the user’s current body image, and the establishment of this relationship can cause people’s strong sense of presence and realism (Schultze 2010; Suh and Prophet 2018) and influences the perception of the possibility of action, which in turn may also help users feel presence. In addition, previous studies have found a positive relationship between perceived presence and the amount of sensory stimulation and the degree of control over the environment (Hilken et al. 2017; Steuer 1992). In other words, if virtual objects can be experienced from multiple dimensions and adjusted or manipulated in their own physical space, the gap between virtual objects and real objects will be narrowed (Suh and Chang 2006), and users will feel these experiences are real as if they have really gone to another real environment. Therefore, the interactivity of AREAs may make presence felt by users. Further, as discussed earlier, responsiveness is considered an aspect of IS quality. In previous studies on IS, researchers have found that IS quality can have a significant impact on user perception of telepresence (Kim and Hyun 2016). Accordingly, Wei et al. (2019) demonstrated the connection between the functional quality of VR and the sense of presence of tourists in the context of tourism. In a sense, it seems that the faster the response of the system in the virtual environment, the more likely the user will have a similar feeling to that in the real world. Therefore, based on the above discussion, we propose the following:

Hypothesis 6

(a) Environmental embedding, (b) interactivity and (c) responsiveness characteristic of AREAs will positively affect users’ feeling of presence.

The influence of spatial presence on individual behavior has attracted the attention of scholars, and it has been reported that it can predict user responses through positive emotions (Fiore et al. 2005; Tussyadiah et al. 2018). For example, studies have found that the degree of presence positively affects customers’ attitude toward products (Klein 2003). Shin (2018) also points out that spatial presence positively influences user mobility levels through enjoyment. Moreover, in the context of online games, studies have shown that telepresence positively affects perceived usefulness (Lin and Chiang 2013). Based on the discussion of these findings, we hypothesize that the perception of presence may positively predict the utilitarian and hedonic value of AREAs. Therefore:

Hypothesis 7

Users’ perception of presence is positively correlated with their perception of (a) utilitarian value and (b) hedonic value.

2.7.3 Utilitarian and hedonic value and response

Previous literature has shown that user satisfaction with products and services in the IT environment is similar to consumer satisfaction in marketing (Deng et al. 2010), this study tries to explain users’ satisfaction and behavioral intention in the context of AR by using findings in the literature related to consumer satisfaction. Previous research has shown that utilitarianism and hedonism are two important dimensions that consumers use to evaluate products and services (Mano and Oliver 1993). After experiencing products and services, consumers are more likely to be satisfied with products and services if they have positive feelings about them (Oliver and DeSarbo 1988). In addition, the literature on IS adoption also shows that positive perceptions of users influence satisfaction with IS (Thong et al. 2006). Based on the above discussion, this study proposes that users' perception of the hedonic and utilitarian value of AREAs is used as a predictor of satisfaction, and users will be more likely to feel satisfied when they get value from the experience. Therefore:

Hypothesis 8

Users’ perception of (a) utilitarian value and (b) hedonic value will positively affect their satisfaction with AREAs.

Previous research has confirmed that the users’ experience (e.g., usefulness and enjoyment) in the virtual environment positively affected their attitudes (White Baker et al. 2019). Recent studies on AR show that the services provided by AR technology help users to feel higher enjoyment, thereby increasing their behavioral intention (Kowalczuk et al. 2021; Yim et al. 2017). On the other hand, previous studies on IS adoption have shown that users are more likely to adopt a technology if they believe it can help improve performance (Davis 1989). At present, quite a few studies have demonstrated the influence of perceived usefulness on users’ behavioral intention (Fu et al. 2018; Liu et al. 2019). AREAs can provide functional support to help users complete specific exercise plans; meanwhile, it may also help bring pleasant emotions during exercise, which stimulates users’ continuance intention. Thus, we posit the following:

Hypothesis 9

Users’ perception of (a) utilitarian value and (b) hedonic value will positively affect their continuance intention.

According to the theory of rational action, satisfaction as a positive attitude can lead to positive behavioral intention. In addition, according to the expectation confirmation theory, researchers found evidence that consumer satisfaction can positively influence the intention to keep buying (Bhattacherjee 2001). Similarly, in the context of IS, users who are satisfied with technologies are more likely to continue using them (Thong et al. 2006). Therefore:

Hypothesis 10

User satisfaction with AREAs will positively affect the continuance intention.

Based on the above hypotheses, the research model is shown in Fig. 1.

Fig. 1
figure 1

Conceptual model

Full size image

3 Methodology

3.1 Data collection

Tiantian Rope Skipping (i.e., the Chinese name of a famous AREA) is a well-known augmented reality app in China, and the developer offers unique services to users by incorporating elements of AR technology into the apps (China.com.cn 2022). This app ranks highly in the relevant partitions and in different app markets (i.e., Android market and iOS Apple Store). When users enter a specific exercise program in the Tiantian Rope Skipping program, they will be provided with different virtual sports scenes. Figure 2 shows screenshots of the application's usage process and a complete real-world environment. When users move their body positions in real life, these virtual objects or images generated by AR technology will also change on the screen correspondingly with the user's actions.

Fig. 2
figure 2

Example screenshots of the AREA. Notes: The figure has been mosaicked for privacy purposes

Full size image

We targeted the AREA of Tiantian Rope Skipping and conducted an online experimental survey on the Credamo platform. Specifically, we have provided a download link for the target AREA on this platform. Online participants are randomly selected, and all participants will answer questions based on their own experience after experiencing the AREA. In this study, the survey data were collected in the form of self-reported questionnaires. Credamo has been used by researchers as a professional integrated data platform because it has been shown to provide a wide and diverse sample of the population while providing data as reliable as traditional methods. The survey process was designed to make sure respondents had downloaded Tiantian Rope Skipping and used the functions in the app. Therefore, before data collection, respondents were asked to answer filtering questions (i.e., respondents need to answer whether they had used this application and the kinds of the item used in the app). Afterward, those who successfully passed the filter were then asked to answer subsequent scale questions based on their experience with the AREA. Each respondent who successfully completes the questionnaire will be offered a financial reward. At the same time, to ensure the quality of the respondents, the researchers restricted the IP address of the respondents to prevent possible duplicate responses, and respondents were asked to participate in the survey only by computer to avoid sample bias due to inconsistencies in device types.

3.2 Measurement

The different structures in the research model were measured using items in the existing literature. At the same time, the wording of the scale will be modified in some cases to suit the research context of the specific AR application. The items measuring interactivity were adapted from Lin et al. (2012), and the environmental embedding scale was adapted from Huang (2021). The ubiquitous items were taken from Hu et al. (2022). The scale for measuring gamification was adapted from Pasca et al. (2020). The responsiveness items were adapted from Song and Zinkhan (2008). Further, measures for utilitarian and hedonic value and presence were adapted from Zhou et al. (2014)and Lee et al. (2018), respectively. The measures for satisfaction were adapted from Jung et al. (2015), whereas items for continuance intention were adapted from Ghazali et al. (2019). All items were measured using a 7-point Likert scale, ranging from “1 = strongly disagree” to “7 = strongly agree.”

Before formally sending the questionnaire to the respondents, we invited professors of information systems to evaluate the questionnaire and improved the design of the scale and questionnaire based on their suggestions. Then, we conducted a 30-person pre-test on the Credamo platform to ensure that the reliability and validity of the scale were applicable to the formal survey. The final valid items used in the survey are listed in Table 1.

Table 1 Measurement model: descriptive statistics and items source
Full size table

3.3 Sample

The survey was maintained on the platform for a week, and 472 responses were received. To ensure the quality of the data, we performed a manual review of the responses we received to exclude invalid samples. The invalid samples included responses that gave the same response to all questions, had missing values, and more than 10 min or the minimum completion time based on the experimental test. Then, 74 samples that met at least one of the above conditions were deemed invalid and excluded. Therefore, 398 valid responses were finally obtained. Table 2 shows the demographic profile of the respondents in this study.

Table 2 Sample demographics. (N = 398)
Full size table

4 Results

4.1 Measurement model

In this study, SmartPLS2.0 was used to test the measurement model and the structural model. Before testing the structural model, we first evaluated the measurement model to check its reliability, convergent and discriminant validity. According to the requirements of reliability test, both composite reliability (CR) and Cronbach α value should be greater than 0.70 (Fornell and Larcker 1981). As shown in Table 3, the composite reliability and Cronbach α values of all constructs were above 0.70, indicating good reliability. To achieve convergent validity, the average variance extraction (AVE) value should be no less than 0.50 (Fornell and Larcker 1981). As shown in Table 3, the AVE values of all structures are all greater than 0.50, indicating that the scale has good convergent validity. Moreover, discriminant validity tests the degree of difference between different constructs, which compares the correlation between target constructs and other constructs by the square root of AVE (on the diagonal) (Fornell and Larcker 1981). As shown in Table 4, the square root of AVE is larger than the correlation coefficient between any pair of constructs, indicating good discriminant validity.

Table 3 Convergent validity and internal reliability
Full size table
Table 4 Discrimination validity
Full size table

4.2 Structural model

This study used SmartPLS 2.0 via bootstrapping and PLS algorithm to test the path coefficients and explanatory power of the structural model. The results of path analysis of the relationship between constructs are displayed in Fig. 3. The general results of the hypotheses with estimated coefficients and its significance level are listed in Table 5.

Fig. 3
figure 3

Overall model: path estimates by PLS analysis

Full size image
Table 5 Structural model estimates and tests of hypotheses
Full size table

With reference to Table 5, environmental embedding (β = 0.274, p < 0.001), ubiquitous (β = 0.171, p < 0.001), and responsiveness characteristics (β = 0.175, p < 0.01) have notable impacts on utilitarian value. Therefore, H1a, H2a, and H5a are supported. Nevertheless, gamification (β = 0.061, p > 0.05) and interactivity (β = 0.031, p > 0.05) characteristics were not significantly related to the utilitarian value. Thus, H3a and H4a are rejected. Moreover, hypothesized positive relationships of environmental embedding (H1b, β = 0.191, p < 0.01), ubiquitous (H2b, β = 0.110, p < 0.001), gamification (H3b, β = 0.351, p < 0.001), interactivity (H4b, β = 0.151, p < 0.05) and responsiveness (H5b, β = 0.119, p < 0.001) with hedonic value are all significant, indicating that all the characteristics of AREAs positively associated with hedonic value significantly. As hypothesized, three factors, interactivity (β = 0.154, p < 0.001), environmental embedding (β = 0.284, p < 0.01), and responsiveness (β = 0.233, p < 0.001) are all positively related to presence. Consequently, H6a, H6b, H6c are accepted. The findings further show that presence has a positive and significant impact both on utilitarian value (β = 0.224, p < 0.001) and hedonic value (β = 0.237, p < 0.001). As a result, H7a and H7b are supported.

Concerning the relations between organism and response, the results imply that both utilitarian value and hedonic value positively influence user’s satisfaction significantly (H8a, β = 0.214, p < 0.01 and H8b, β = 0.535, p < 0.001). Hedonic value has a significant positive impact on user’s continuance intention (β = 0.250, p < 0.001), supporting H9b. However, H9a is rejected as the impact of utilitarian value on continuance intention is not significant (β = 0.063, p > 0.05). Finally, a strong relationship between satisfaction and continuance intention is found (β = 0.468, p < 0.001), which means that satisfaction is a powerful predictor of the behavioral response of continuance intention.

On the whole, the explanatory power (R2) for continuance intention is 69.6%, and the satisfaction is 64.8%. Furthermore, the variance explained in utilitarian value, hedonic value and presence was 42.7%, 61.4%, and 39.3%, respectively. Overall, the variance explanation rates of the dependent variables highlight the model’s explanatory power.

4.3 Common method bias

In behavioral science research, common method bias is often considered as a source of measurement error in self-reported data (Esfandiar et al. 2019; Malhotra et al. 2016). Considering that the data in this study were collected in the same period and may be influenced by the same data source, measurement environment or rater, testing for common method bias is necessary. In this study, Harman’s one-factor test was used to assess common method bias (Podsakoff and Organ 1986). As shown in the results in Table 6, the Harman univariate test produced a total of 8 main factors, among which the maximum variance explanation rate of the single factor was 34.14%, which was lower than 40%. The test results showed that there was no significant common method bias in this study.

Table 6 Harman’s one-factor test for common method bias
Full size table

5 Discussion

5.1 Research findings

Based on the SOR (stimulus-organism-response) paradigm, we examined how characteristics of AR applications affect users’ continuance intention. In previous studies, researchers have examined how utilitarian and hedonic benefits as independent variables influence users’ behavioral intention (Rauschnabel et al. 2019). This study further found that the characteristics of AREAs will positively influence users’ utilitarian and hedonic values, which further influence subsequent continuance intention. Specifically, all characteristics hypothesized in the model were found to be positively and significantly associated with hedonic value, with environmental embedding, ubiquitous, and responsiveness positively associated with utilitarian value. This suggests that while AREAs creating a virtual exercise environment for people, it also allows them to exercise anytime and anywhere without being limited by the place, which may meet their needs for fitness, thus making them feel utilitarian value, and in the process of movement, users can also experience the pleasure of AREAs. However, the correlation between interaction characteristics and utilitarian value has not been found, which seems to contradict previous research results. Specifically, Qin et al. (2021) found a positive relationship between interactivity and utilitarian gratification. However, the results presented here are consistent with Nikhashemi et al. (2021), and one possible explanation for our conflicting results is that users' perception of utilitarian value is related to the practical attributes of the product (Picot-Coupey et al. 2021). In other words, when the use of products helps users solve problems or achieve their goals, users are likely to have a perception of utilitarian value. Due to technological limitations in the interaction between users and virtual objects in reality, limited interaction may not be sufficient to enable them to better understand and evaluate their motor performance, which leads to a lack of utilitarian value. In addition, the relationship between gamification and utilitarian value was also found to be insignificant, which may stem from users’ perception of gamification in terms of its utility (Perry and Ballou 1997; Venkatesh 1999). Specifically, while there is evidence that many players play games to improve physical performance (Hamari and Koivisto 2015; Tu et al. 2019), users form their own judgments about gamified services and frameworks. If gamified design is perceived to be more for entertainment purposes, users' perception of utility may become weak. So, as things stand, users of the Tiantian Rope Skipping may see gamification as bringing them more pleasure than athletic performance. In addition, this study revealed the mediating effect of presence in the context of this study. Specifically, interactivity, environmental embedding, and responsiveness can positively contribute to user perception of presence and indirectly influence utilitarian and hedonic value through presence. This means that when AREAs integrate virtual content into the user’s realistic picture and allow them to interact with the virtual content, the user will feel a sense of presence and feel part of the virtual exercise scene. These findings of this research also confirm and extend previous studies on the mediating role of presence (Kim 2015; Peukert et al. 2019).

The results further show that there is a significant difference between utilitarian value and hedonic value in promoting users’ continuance intention. Only hedonic value is found to have a significant relationship with continuance intention. This finding is consistent with the results of Hsu et al. (2021), which was built on the background of AR cosmetics. As Javornik (2016) points out, AR brings users more hedonic experience than utilitarian experience. These similar findings seem to collectively convey the message that AR application features can bring both utilitarian and hedonic value to users, while hedonic value is a better predictor of user continuance intention than utilitarian value. However, it is worth noting that both utilitarian and hedonic values of AR are positively correlated with satisfaction, which extends and complements the study of Hsu et al. (2021), showing that users are more likely to be satisfied with products after experiencing positive value, and satisfaction thus positively predict the continuance intention AR applications.

5.2 Theoretical implications

The current research has the following theoretical contributions. First, this study builds a new model based on the SOR framework of psychology to illustrate the effects of two types of characteristics of AREAs on user's continuance intention. Specifically, this study conceptualized the "stimulus" in the SOR paradigm in a specific AREAs environment. Based on existing AR literature and specific research environments, we identified two original propositions as characteristics of AREAs (i.e., system characteristics and service characteristics) and considered them as external stimuli. We found that these characteristics are associated with the users’ inner states, and they can positively affect the users’ value assessment (i.e., utilitarian value and hedonic value) and psychological feeling (i.e., presence). And these positive inner states, which in return promote satisfaction and continuance intention.

Second, the basic findings of this study facilitate further research and understanding of AR technology and provide interesting insights from the perspective of technology acceptance. In previous AR literature, researchers have examined how utilitarian and hedonic benefits as independent variables have an impact on users' behavioral intentions (Rauschnabel et al. 2019). Our results further suggest that specific application characteristics will influence users' perceptions of utilitarian and hedonic value. This finding is important for researchers and practitioners because it helps them understand how different characteristics bring different experience values to users, and what types of characteristics drive users to reuse information systems based on user experience. The results of this study also reveal that the characteristics of AREAs can provide both utilitarian and hedonic benefits, and both utilitarian and hedonic values indirectly or directly predict continuance intention, which is consistent with the view of technology acceptance (Akel and Armağan 2020; Chang et al. 2014; Davis 1989; van der 2004). These findings confirm the original a priori hypothesis of this study that AREAs may contain both utilitarian and hedonic aspects. However, unlike the technology acceptance literature, the past literature seems to have tended to strictly define the categories of information systems and regard information systems as utilitarian (Davis 1989; van der 2004) or hedonic (van der 2004). Therefore, from this point of view, it is important to note that it is difficult to determine whether AREAs are hedonistic or utilitarian designs, rather, it is more like a hybrid system that combines utilitarianism and hedonism (Hamari and Koivisto 2015), and for certain application characteristics, it may also provide both utilitarian and hedonic values.

Third, this study further extends the theoretical perspective of presence to the background of AR and contextualizes the antecedents and consequences of presence in the environment of AREAs. AR can integrate virtual information into the real environment and has a unique characteristic of environmental embedding (Hilken et al. 2017). This characteristic of AR is very similar to the concept of augmentation in the study of human–computer interaction (Billinghurst and Kato 2002) and is therefore distinguished from virtual reality. Previous studies in the virtual reality context believe that interactivity and vividness are the antecedents of presence at the technical level (Steuer 1992). This study found a positive relationship between environmental embedding, interactivity and responsiveness and presence in the context of AREAs. This result is consistent with the previous VR literature on the one hand (Steuer 1992) and also indicates the important role of environmental embedding in the sense of presence, which supplements the pre-causes that promote presence in the context of AREAs and expands the previous literature in the context of AR. Considering that environmental embedding may be a unique feature of AR that differentiates it from other technologies and that the extant literature pays little attention to it, further research is needed to explore this feature of AR and understand its impact on user experience.

Finally, this study expands the literature on AR field as well as sports digitization at a broader level. Specifically, this study expands the sports digitization framework proposed by Xiao et al. (2017) in practical experience. The issue of physical activity has come under the spotlight during the COVID-19 pandemic. But in fact, specific sports can be migrated into virtual worlds (Jenny et al. 2016). This study complied with the call of (Ng et al. 2019; Rogers et al. 2017) and investigated the scenario of AR integrated into exercise applications which were applied at the individual level. The context of this study shows that users can compete in sports through the virtual world of AR, which adds to the literature on the unique connections people have through technology due to “lockdowns” during the pandemic (Hacker et al. 2020). In addition, based on the topic that digital practices may persist for a long time after the COVID-19 (Hacker et al. 2020), it is proposed that virtual exercise through AR technology may complement traditional physical activity during the pandemic.

5.3 Practical implications

This study also provides some practical implications. Firstly, it is found that different characteristics of applications will affect users’ perception of different experiential values. However, utilitarian value was found to have no significant relationship with continuance intention. There are indications in the previous literature that user perception of utilitarian value may be related to the utility attributes of a product (Picot-Coupey et al. 2021). These suggest that when designing and developing AR applications, practitioners should think about the purpose for which the product is being used, and design hedonic and utilitarian functions for different uses. And at the same time, it is urgent to pay attention to and identify whether different AR application characteristics will bring utilitarian or hedonic value to users. The application development focus should be selectively adjusted according to the influence of different characteristics on user experiential value. It is recommended that practitioners can recruit experienced users for small-scale testing when designing AR applications to understand the actual differences between functional characteristics in utilitarian and hedonic situations. In addition, it is found that both hedonic and utilitarian value can affect users’ satisfaction with AREAs and further affect continuance intention through satisfaction. Application vendors should focus on and improve user satisfaction while providing services. This suggests that application providers can improve user satisfaction by regularly collecting users' feedback and constantly improving and upgrading applications while providing services.

Second, the results show that only hedonic value can promote continuance intention, further supporting the previous literature that AR provides hedonic rather than utilitarian (Javornik 2016). This highlights the need to focus more on the hedonic value an app can bring to users, such as the introduction of gamification. In this study, gamification was found to positively affect users’ perception of hedonic value, suggesting that gamification elements may be attractive. Previous IS literature has called for gamification to be more explored (Liu et al. 2013, 2017); however, gamification has rarely been mentioned in AR research. Since fitness apps can provide gamification related elements (e.g., badges, points, and leaderboards) (James et al. 2019), it is recommended that practitioners add relevant designs to their AR apps. In addition, interactivity has also been found to have hedonic value, so adding more interactive elements to gamification would be a potential choice as it could greatly enhance the attractiveness of the AREAs.

Third, many AR apps are location-specific, and virtual content is only added when the camera captures a suitable target, which means location-specific apps are limited by where they are used to some extent. Whether AR applications that can provide mobility and portability independently of specific locations will enable users to adopt a positive attitude has also attracted the attention of researchers (Javornik 2016). The AREAs in the context of this study allow users to use it in their home, and the results show that this ubiquitous characteristic is positively associated with users’ utilitarian and hedonic values. Practitioners can focus on the potential benefits of this characteristic for AR products and provide features that allow users to be able to use it when and where they define it, such as indoor activities and games that are less restrictive to the venue and require less space, which may contribute to a positive user response.

Finally, the study shows the mediating effect of presence, which suggests that AR practitioners should pay attention to how users feel about presence. Specifically, this study found that interactivity, environmental embedding and responsiveness have a positive impact on presence. Practitioners should be concerned about the environmental embedding of AR and pay attention to the effect of visual presentation and provide high-quality embeddings for users. For example, users can achieve a more immersive feeling by designing more vivid and realistic 3D images and even endowing sound. In terms of responsiveness, the previous AR literature has pointed out that responsiveness is an important part of system processing quality (Kowalczuk et al. 2021). AR engineers should focus on testing and improving the quality of the system and reducing the time it takes for the system to respond, as this affects how the user feels immersed in the AR environment.

5.4 Limitations and future research

While this study provides many important insights in theory and practice, the results of the research are inevitably bound by some limitations. First, it should be noted that in the context of AREAs, this study investigates one product category, so the results may be limited to specific AR environments. Although this product ranks high in the Apple Store and Android market and has received extensive media attention, the research findings still need to be cautiously inferred. Future research should examine the application of AR in different sports product categories and expand to other more fields to test the findings of this study. In addition, as a cross-sectional study, this paper only examined the situation at a single time point. As time goes by, the development iteration of the technology itself and the change of users’ familiarity and attitude toward the technology were not tested. This limits our additional understanding of research results due to technological developments or changes in user perceptions. In future research, users’ familiarity with AR (Bonnin 2020) and previous experience (Song et al. 2019) can be considered to examine the impact of changes in personal factors on the continuance intention of AR. At the same time, factors such as participant samples may also limit our deeper understanding of the results. As this study was conducted online, it means that there may be a lack of people in the sample who do not use the internet, which may include a considerable portion of the elderly. Future research could be performed offline and with a broader sample of the population. Finally, this study was conducted in the context of China. Ethnic groups with different cultural backgrounds may have different views on AR technology and services. Therefore, future research may promote the findings of this study among countries to examine the impact of cultural differences on continuance intention.