Keywords

1 Introduction

The Internet of Things (IoT) has contributed significantly to the formation of people’s everyday life by facilitating it. Wearable technology is based on IoT applications and refers to smart hearable devices, smart watches, smart clothing smart patches, and smart implantables. Smart devices use biosensors to collect medical data such as heart rate and abnormal heart rhythms, blood pressure, sleeping patterns, glucose meters, and fitness trackers. By 2024 the forecast for wearable devices will reach 109 billion dollars [1, 2]. Sports wearables account for 50% of the unit sales in the global market. The study aims to identify and explain the mechanism leading to consumers’ acceptance of sports wearables. The extant literature has yet to cover this field [3, 4]. This study examines the factors inhibiting and the factors contributing to the adoption of smartwatches. Research on wearable devices is focused mainly on smartwatches adoption, but few studies explore the adoption of smartwatches by amateur athletes who are also interested in their health [4].

2 Theoretical Foundation and Literature Review

Wearable devices are autonomous and they track physiological indicators in order to understand individuals’ habits and improve their physical activity levels. Such devices are the smartwatches, biosensors, activity trackers or wristbands. A smartwatch serves mostly as a satellite device which will gather useful data through a wireless Bluetooth connection. The access to information is faster, more accurate and also convenient when the use of the smartphone is not practical [2]. Smartwatch technology allows ubiquity, mobility, time and location independence, and also it incorporates context awareness [3]. Sport wearables indications allows users to improve their strategy and their performance, to observe their physical condition and preventing them from getting injured. Users benefit from the usage on fitness from continuous usage for long periods of time and that affects the quality of their life [4]. To provide a theoretical framework for our research we used a well-established model of Technology Acceptance Theory (TAM) [5]. Davis theorized TAM as a mechanism to explain and predict users’ acceptance of information and communication technologies and applications. TAM has proven to be an extremely robust model in its application to a wide variety of technologies such as sports wearables, sensors, smartwatch, smart health devices, wrist worn wearables [1, 4, 6,7,8,9,10].

2.1 Technology Acceptance Model

In the digital era, new technologies are developing constantly, and various theoretical models are proposed to explain and predict the adoption process. One of the most extensively applied is the theoretical model “Technology Acceptance Model” [11]. TAM is a popular model used to explain and predict the adoption of technologies in many organizations and cultural contexts; it is based on the Theory of Reasoned Action and Planned Behavior [12], but Davis [5] indulged in innovation adoption at the working environment. He posits that perceived ease of use and perceived usefulness are the main psychological constructs of predicting the intention to use a technology [2, 6, 13, 14]. Venkatesh and Davis in 2000, proposed an extension of TAM and new determinants of perceived usefulness were proposed: subjective norm, image, job relevance, output quality, result demonstrability. They also proposed new mediators of ease of use: experience and voluntariness [15, 16]. Perceived ease of use is defined as “the degree to which the prospective user expects the target system to be free of effort” [17]. Perceived usefulness is defined as the user's “subjective probability that using a specific application system will increase his or her job performance within an organizational context” [15]. TAM model is appropriate to explore users’ perceptions on smartwatch adoption.

2.2 Using Wearable Technology Devices in Running

Wearable devices promote health and safety in physical activities, positive lifestyles, and preventing injuries while training [3]. Running has become a prevalent outdoor activity with a high participation rate by amateurs. Smart devices data can help runners to adjust their rate of training, take different routes or change their exercising behavior. In that way a user becomes aware of his overall performance almost instantly. Runners’ performance benefits from the visualization of smartwatch indications, and they feel more motivated and supported in their training [5] indicating if their goals have been achieved. They share their digitally codified results with friends or coaches. Users’ virtual interactions and information sharing on social media is an extra motivation to accomplish training goals [9]. The acceptance of wearable devices differs among sociodemographic characteristics of users, such as gender, age, income, and culture [16, 18]. Kim and Chiu [4] explored gender differences among sports wearable users and identified adoption differences regarding their usage. They resulted that gender moderated the relationship between PU and intention to use [4]. The costs of acquiring a smart device affects the decision to adopt [19]. The non-monetary costs [13] includes time consumption, and psychological costs, including inner conflict, discomfort, anxiety, or mental fatigue [13, 20]. We define smartwatch anxiety as the degree of discomfort a user suffers when using a smartwatch. A user with high anxiety levels is less likely to adopt a smart health device as easy to use [21].

2.3 Participation at a Virtual Community

Information and communication technologies and broad participation in social media allow people to become co-creators in the communication process [22]. The interactions and the social ties between the members characterize a community. These communities consist of people, a common purpose, and similar policies, and they are using computers or smart devices. According to Koh et al. [22], a virtual community comprises members sharing information and knowledge for mutual learning and providing solutions. A virtual community uses a platform to share common interests to support and exchange information considered unbiased [23, 24]. Online communities develop relationships towards value co-creation, and they build trust and commitment between members. They interact in cyberspace and establish social relationships through repeated contact. They share feelings of belonging to the group, sense need-fulfillment, express an emotional connection with other members, and feel that they may influence other members. Members tend to exhibit addictive behavior via online daily communication. Thus, “the construct of immersion is adopted as an emergent property of the virtual community” [22]. Being a member of a running athletic community influence members’ perceptions [24]. Thus, we examine whether athletes’ participation at a virtual community of runners influences their perceptions of adopting a smart device.

3 Methodology

A survey was performed through an online sports magazine specializing in running addressed to amateur runners. The magazine posted a questionnaire at its web page. The questionnaire was based on previously validated TAM studies [25, 26] and received 128 responses in 15 days. Prior to the announcement of the questionnaire, a pilot study was performed on 20 athletes. Participants chose the degree of satisfaction or importance of each question in the questionnaire according to their feelings and cognition on a five-point Likert scale anchored by 1: strongly disagree, and 5: strongly agree. Descriptive statistical analysis was conducted on survey data through SPSS22.0. The reliability test results with Cronbach’s α showed that the measurement had strong internal consistency (Cronbach’s α > 0.7). In order to compare samples across demographic groups or perceptions we performed inferential tests such as ANOVA and t-test analyses. Multivariate analysis (factor analysis and regression analysis) to create variable composites from the original attributes and obtain a small number of variables that explain most of the variances along attributes. Then we applied the derived factor scores in multiple regression analysis. Regression analysis indicated a model to the adoption of the device. Based on literature we propose the following hypotheses: H1: PEOU, PU, anxiety, intention to use, and community participation in using smartwatches differ by the training frequency of the amateur runner [27]. H2: PEOU, PU, anxiety, intention to use, and community participation in using smartwatches differ between genders [4]. H3: PEOU, PU, anxiety, intention to use, and community participation in using smartwatches differ between different age groups of amateur runners [16].

4 Results

The sample consisted of 26.6% females and 73.4% males, of which 42% declare to be aged between 35 and 50 years old and 29% are male married with children while the 14% of women are singles. A large percentage of the male participants had a university degree as their highest level of education (78%). A percentage of 35% of male participants and 44% of female participants had family income between 10.000 and 20.000 euros. Amateur athletes used their smartwatch to check health indicators: Heart beats (71%), monitoring sleep habits (13.3%) level of organization exhaustion (6.2%). The 8.8% of female athletes use to observe their menstrual cycle using their smart watch. 50% of male athletes used to train four to five times per week while 32% of female athletes used to training once a week.

To assess the measurement model, we examine the loadings, the average variance extracted (AVE), Cronbach’s α, the composite reliability (CR), following the guidelines of Hair et al. [28]. The Kaiser–Meyer–Olkin Measure of Sampling Adequacy (KMO) is 0.854 higher than the threshold of 0.60 indicating the appropriateness of factor analysis and the Barlett’s Test of Sphericity, is χ2 = 1961; p < 0.001 supported the decision. Principal Component Analysis and orthogonal Rotation with Varimax method resulted in a five-factor solution representing 63.96% of total variance. Factor loadings for items range between 0.530 and 0.910 which are considered high (Table 1). All measures present internal consistency reliability as Cronbach’s α coefficients are surpassing 0.7 and the values of CR exceeded the criterion of 0.60 [29, 30] the five factors were labeled based on the core variables: Perceived Ease of Use, Perceived Usefulness, Intention to Use, Participation in Online Community, and Anxiety.

Table 1 Internal variability and convergent validity
Full size table

The impact of personal characteristics upon the five factors influencing the adoption of smart devices is examined, by using inferential statistics (t-tests and ANOVA tests). According to these tests: H1: PEOU, PU, anxiety, intention to use and community participation using smartwatches differ by the frequency of training. Equal variances assumed the hypothesis of relationship between PU and frequency of workout is supported [F(4, 4782) = 5.453, p = 0.00 < 0.01] and indicates differences among the means of groups of athletes and the frequency of their training. Athletes who used a smartwatch and trained four or five times per week (mean = 0.3495366) indicated a higher perception of the perceived usefulness of the smartwatch relative to users who trained once a week (mean =  − 0.3566247). H2: PEOU, PU, anxiety, intention to use and community participation using smartwatches differ between genders. Independent t-tests were employed to seek the differences between male (n = 94) and female (n = 4) users. Compared to female consumers, male users reported lower anxiety towards using a smartwatch. Equal variances assumed the hypothesis of differences between genders and anxiety using the smartwatches is supported [F(126, 1614) = 0.958 p = 0.001 < 0.00], women are more anxious than men (mean female = 0.46 and mean male = 0.166). H3: PEOU, PU, anxiety, intention to use and community participation using smartwatches differ between different age groups. Equal variances assumed the hypothesis of relationship between PU and age groups is supported [F(4, 6.124) = 7349, p = 0.00 < 0.01] and indicates differences among the means of age groups of athletes. Athletes at the age group of (18–24) have lower perception (mean = − 1.2723) of the perceived usefulness of a smartwatch relative to athletes at the age group of (35–50) years old (mean = 0.3666). According to the regression model: adjusted R2 = 0.694 (F = 96.905, p < 0.01). The model explains that almost 70% of the independent variables affect the intention to use a smartwatch. Specifically, Intention = 4.227 + 0.868 PU + 0.151PEOU + 0.90 HEALTH INDICATORS. The intention to use a smartwatch is positively related to the perceptions of easiness and usefulness but it is also related to the health indicators the runner checks.

5 Discussion and Conclusion

This research examines the factors contributing to user acceptance of smartwatches by amateur athletes who use smart devices to monitor their physical performance during training sessions. Results did not support anxiety or technology discomfort as a significant factor influencing amateur runners’ behavioral intentions. The participation in a virtual community did not influence their perceptions on smartwatch usage. Findings showed that perceived ease of use and perceived usefulness are important factors in predicting the usage of smartwatches. Athletes are also interested in the health indicators presented which is a significant factor in their intention. A study on amateur runners and the possibility of suffering cardiac problems resulted that intense and enduring exercise like running a Marathon race may result in medical disfunctions [27]. Intense and frequent training affects the users’ perceptions of smartwatches. They use smart device data to measure overall fitness rehabilitation of injuries, avoiding overexertion, and assigning training proposals based on individual performance. Findings need to be confirmed by further evidence from other sports given the differences in values and sports. Further research should include the collection and analysis of longitudinal data and other variables that affect smart device adoption such as compatibility, innovativeness, aesthetic image etc. to better explain the mechanisms behind the adoption/usage of a new wearable technology.