1 Introduction

The Internet of Things (IoT) is a term that was introduced by Kevin Ashton during a presentation at the Massachusetts Institute of Technology (MIT) in 1999 [1]. Ashton visualized a fantasy world with objects connected via the Internet using sensors and actuators that are able to produce real time information and enhance the quality of humans’ daily life [2]. Within the context of IoT, wearables are introduced as the next-generation market demands after smartphones. In a report published by ABI Research [3, 4], it was predicted that 485 million wearable devices are going to be in use by 2018. According to Canhoto and Arp [5], wearable technologies have gained notoriety and a specific state of public awareness, in particular, devices that support healthy lifestyles.

Furthermore, Information Systems (IS) academics have emphasized that issues of morbid obesity, an unhealthy diet, and a dearth of physical activity may have significant consequences on the health of the younger generation which directly affects the healthcare system in both developed and developing countries [6, 7]. Therefore, devices that are able to track activities can play an important role in users’ lives by motivating them to adopt a healthier lifestyle. This can be achieved by reporting daily activities, such as step counters, sleep patterns, caloric intake, calories burned, heart rates, blood pressure, and body temperature, respectively [8, 9].

Although subjects related to smart technologies have been extensively considered in academic and practical contexts, a majority of previous researches have tended to focus on the concepts, general descriptions, challenges, business models, architecture, design, and implementation from the perspective of technology [10,11,12,13,14,15,16,17,18,19,20]. Only a small number of empirical researches have investigated the success factors and determinants of smart technology adoption, particularly smart wellness wearables, from the user’s perspective [21]. However, the adoption diffusion of smart wellness wearable technologies is relatively low [22]. Since users may not gain the promised benefits of smart wellness wearables in terms of health and fitness issues [5], there is a possibility that almost half of wearable users would abandon their devices during the first 6 months [5, 23]. Therefore, attracting and motivating users to continue using their smart wellness wearables is an important challenge for business managers. Hence, research studies that improve the understanding of the drivers of acceptance and continuous use of smart wellness wearable devices may have a positive effect on society and policy-making issues [5]. On the other hand, understanding the factors behind the adoption of new innovations is a critical issue for designers and developers when developing successful products so as to increase the speed of diffusion. Therefore, it is important to explore the particular requirements and preferences of users who own wearables for the purpose of wellness-tracking and to determine the priorities of the general public in various countries [21]. However, recent research studies on users’ acceptance of wearables have examined a limited number of critical factors from the technological perspective [24,25,26]. A unified and comprehensive framework is needed to explain the behavioral intention of users for using smart wellness wearable devices more clearly [27].

Therefore, the purpose of this study is to answer the following research questions about smart wellness wearables:

  • What are the significant factors and impediments that influence the intention to use smart wellness wearables?

  • What is the proper research model that could facilitate the intention of using smart wellness wearable technologies?

Consequently, this study has concentrated on previous IoT and smart wearable studies to identify the direct and indirect factors that have influenced the intention of the general public to use smart technologies, specifically smart wellness wearables. Thus, an integrated framework is proposed, and the validity and reliability of the measurement model are examined. The findings of this study will be evaluated through a large-scale survey as the next step in the near future.

The rest of this article is structured as follows. The second section reviews several previous studies; section three presents the justifications for integrating the “Unified Theory of Acceptance and Use of Technology 2 (UTAUT2)” and “Value-based Adoption Model (VAM)” and the hypotheses; section four presents the research methodology; section five discusses the results of the measurement model; and finally, section six presents the conclusions and future direction of the study.

2 Literature review

Wearable technology is not a new concept in the wellness and healthcare industry. Around two decades ago, the first smart shirt was created when the US Navy Defense invested in a research project at a technology institution in Georgia to track the physical condition of soldiers [28]. Since then, scholars began to extend the invention of wearables in the medical arena to track vital signs and forward biofeedback data to hospitals or physicians’ clinics [9]. Since their emergence, smart wearables have gradually improved from being inconvenient, heavy, and large technologies to more comfortable, portable, and weightless devices. Nevertheless, wearables have several disadvantages currently causing concern among researchers, developers, and users, such as privacy concerns and high prices [9, 29, 30].

Numerous academic research studies have been conducted on smart wearables, including a research conducted by Kim and Shin [31]. The authors examined the intention of end users for continuous use of smart watches by extending the basic Technology Acceptance Model (TAM) with additional factors, such as cost and subcultural appeal. They also identified different predictors for perceived usefulness and ease of use. However, job-related measurement items were used in some parts of their survey, such as “this smart watch is useful in doing my job.” One of the flaws in their research is that they considered individuals that use activity trackers (such as Fitbit) in their survey, while this group of users usually does not use these devices for their job-related purposes [32]. In another study [33], patients’ recovery was monitored after cardiac surgery using the activity tracker, Fitbit. The results indicated that patients who used Fitbit to count their steps in the early stage of recovery, recuperated sooner than others.

Academic researchers believe that smart wellness devices have an impressive effect on users’ health principles. According to the various useful features on smart wellness devices, such as sleep monitoring patterns and caloric intake measurement, people may attain better understanding of their physical activity and become motivated to keep their body healthier. In this regard, Prayoga and Abraham [34] examined the indicators of users’ intention to use smart health devices based on the basic TAM. Their results demonstrated that perceived usefulness is the utmost determinant factor of behavioral intention for using smart health devices. Similarly, in another study, Holzinger et al. [35] evaluated the perceived usefulness of using wearable devices for tracking vital signs in the elderly. The researchers emphasized that wearables’ usability factors should be considered at the design and engineering level since the perceived usefulness of wearable devices has a direct effect on the acceptance rate of wearables.

Miller [36] examined the usefulness of applying a fitness wearable device in self-tracking behavioral changes in accordance with prescribed medical treatment. It was assumed that wearing self-tracking wearables while engaged in multifold medical settings may be helpful for behavioral adaptation for an overweight patient with Type 2 diabetes. Tsao et al. [21] inspected user requirements of activity trackers. The authors used a mixed method of questionnaire and semi-structured interviews for collecting data from Chinese and Germans users, and compared the findings between these two nationalities. The results of their study showed the influence of cultural differences on user requirements of wearables for monitoring daily activities. Based on their findings, the wellness wearables have shown different results in different countries [21].

Consequently, this study contributes to these streams of research by integrating UTAUT2 and VAM including perceived trust and perceived health increase. It aims to investigate the intention of using smart wellness wearables by the general public in Malaysia.

3 Asia Pacific and Malaysia market forecasts about smart wearables

According to one of the leading global market research companies, Data Bridge Market Research [37], the smart wearables market of the Asia Pacific region (APAC) is split up into nine areas based on geographical factors, namely: Malaysia, Singapore, South Korea, India, Taiwan, Japan, Australia, China, and the rest of the Asia Pacific region. It is expected that Japan will dominate the marketplace in this area, because of the strong awareness of smartwatches and pedometers in that country. Moreover, it is predicted that India and China will reach the main revenue area of this market, because of the growing trend of smart wearables and cheap devices in these countries. Figure 1 shows the total number of wearable devices around the world based on the region from 2015 to 2016 with projections for 2020 and 2021. As shown in this figure, the number of wearables in the APAC region reached 30.4 million units in 2015, while it is estimated that wearables will reach 258.2 million units in 2021 [38].

Fig. 1
figure 1

Worldwide forecasts concerning wearable unit sales based on the region [38]

Full size image

Statista market research [39], an online business intelligence and statistics portal, reports that the revenue of the smart wearables sector in Malaysia reached US $24 million in 2018 (Fig. 2). Moreover, it is predicted that the revenue of this sector will reach US $29 million in 2022, which represents an annual growth rate of 5.7% in Malaysia. In addition, reports indicate that the penetration of smart wearables’ users is at 3.5% in 2018, while it is predicted to reach more than 4.7% in 2022 in Malaysia. According to the reports of Statista [39], concerns about health and wellbeing (such as being overweight and smoking rates among males and females) are increasing in Malaysia. Based on the wearables’ projected revenue of 29.5 million dollars in 2022, it could be concluded that the overall tendency of the general public to pursue a healthier lifestyle is steadily rising.

Fig. 2
figure 2

Amounts of revenue in the wearables segment in Malaysia [39]

Full size image

As shown in Fig. 3, it is predicted that the number of smart wearables’ users in Malaysia will increase from 0.8 million in 2016 to 1.6 million in 2022 [39]. These statistics confirm that studying the intentions of Malaysians toward using smart wearables is a very beneficial research area and needs more investigation.

Fig. 3
figure 3

Number of wearables users in Malaysia [39]

Full size image

4 Conceptual research model

As reported by several IS academics in literature [40, 41], the dearth of user acceptance represents a salient challenge to the integration of new information technologies all over the world. As pointed out by Holzinger et al. [42], there is a strong dependency between previous exposure to technology and user acceptance. However, considerable attention has been focused on the TAM [40] and UTAUT [43] models to explain the acceptance and usage behaviors for Information Technologies (IT). TAM postulates that user acceptance can be described based on two dimensions, namely perceived ease of use and perceived usefulness [40], while UTAUT demonstrates that user acceptance can be defined by four indicators: performance expectancy, effort expectancy, social influence, and facilitating conditions. Many IS scholars have studied TAM and UTAUT and agreed that these two theories are reliable and valid for predicting an individual’s acceptance of new technologies [43,44,45,46]. However, academic researchers [11, 47, 48] believe that individuals may be resistant to the use of expensive technologies, even if it is beneficial for them. Since IT products and services are applied in both personal life and organizational settings, Venkatesh et al. [41] and Kim et al. [49] recommended that these theories may be more useful for investigating the adoption and intentional behavior of employees in organizations. This is because, in an organizational setting, companies (rather than their employees) are responsible for providing technologies and paying expenses. Consequently, the costs and structure of pricing may have a potential impact on users’ intentions to use new technologies [41, 49].

This matter may diminish the capabilities of TAM and UTAUT in describing and predicting individual acceptance of information technology services. TAM and the basic version of UTAUT only deal with perceived benefit constructs, while ignoring perceived monetary constructs (perceived fee) and non-monetary constructs (perceived privacy) [11, 48, 50]. Furthermore, UTAUT2 has disclosed a greater predictive power compared with other acceptance models applied in the literature [41], which indicates the exhaustive efficiency of the model [51].

Moreover, IT adoption may satisfy individual leisure and entertainment requirements as well as providing healthy lifestyle choices [52,53,54,55]. Accordingly, profits from both purposes should be taken into consideration when investigating user acceptance behavior. Thus, perceived value appears to be the appropriate variable for assessing IT acceptance behavior as it considers all related factors of benefits and sacrifices [11, 49, 56]. A vast number of research studies in the IS and marketing domain report that perceived value is a crucial variable in various contexts, such as: Mobile Internet (M-Internet) [49]; location-based services [57]; mobile data services [58, 59]; Internet of Things [11]; and electronic/mobile commerce [47, 48, 60,61,62,63,64,65]. Furthermore, many scholars assert that the main reasons behind the importance of perceived value are, namely: enhancing users’ loyalty and satisfaction [57, 66, 67]; raising users’ intention to adopt/use IT products [11, 48, 64, 68, 69]; and increasing users’ intention to purchase new technologies [48, 70].

In a study led by Kim et al. [49], the Value-based Adoption Model (VAM) was developed from the value maximization perspective to explain Mobile Internet adoption by users. The outcome of their study proved that users’ perception of value for using Mobile Internet is a principal indicator of the behavioral intention to use this technology. The validity of VAM in the context of behavioral intention scenarios was indicated in numerous studies [48, 68, 69, 71].

On the other hand, one of the limitations of well-known technology acceptance models (TAM, UTAUT, and UTAUT2) is the omission of a crucial trust-related factor in the context of wireless technologies, since researchers assume that there is nothing to stop individuals from adopting and using a technology if they choose to do so [72, 73]. In addition, many IS scholars claim that integrated models or extending original models by adding extra factors may increase the predictive power of technology acceptance and user behavior [43, 74,75,76,77].

Hence, the principal aim of this study is to fill the aforementioned gaps by combining the UTAUT2 and VAM models with two additional factors, namely perceived trust and perceived health increase. Figure 4 presents the proposed conceptual research model of this study.

Fig. 4
figure 4

Proposed research model

Full size image

Table 1 presents the potential variables that may increase the intention of end users to use smart wellness wearables based on a vast review of previous studies in the domain of smart and wireless technologies. Additionally, a brief description of each factor as well as their references and domains is summarized in the following table. In the remaining part of this section, a general justification is expressed relating to each construct and the hypothesis motivations based on the previous studies.

Table 1 Potential factors based on reviewed papers in the domain of smart and wireless technologies
Full size table

4.1 Perceived privacy (PP)

Researchers have considered privacy to be the principal obstacle for a full adoption of online services. Upon startup, almost every smart device, particularly in the context of wearables, begins by collecting user personal information [29, 87, 88]. Disclosure of wearables’ user data may expose personal characteristics and habits as well as the user’s location information. Unauthorized and illegal access to this sensitive information may adversely affect the privacy of the users [87, 89, 90]. In the case of smart wellness wearables, most activity trackers store the user’s location, heartbeat, and sleep patterns, which could be hacked and potentially exploited against the users [29].

In other words, users may be concerned that wearable providers collect too much personal information without notice or that providers may be involved in illegal use of personal information for gaining profit [11]. These concerns could increase non-monetary costs and will negatively affect the value of perception toward wearables. Researchers have concluded that the type of collected and used data via service providers could impact upon the level of users’ privacy concerns [91,92,93]. Moreover, previous studies have empirically shown that privacy concerns have negative relationships with, specifically: continued usage [94]; willingness to transact [95]; trust [88, 93]; intention to disclose location information [96]; and perceived value [11]. Since most smart wellness wearable devices collect data related to personal habits and location information, the following hypotheses are proposed:

H1a

Perceived privacy while using smart wellness wearables will have a negative effect on the perceived value of using such devices;

H1b

Perceived privacy of smart wellness wearables has a positive effect on perceived trust in using such devices.

4.2 Perceived fee (PF)

Previous researchers cite the significant impact of a perceived fee on technology adoption and usage in various contexts, such as: electronic shop [97]; M-Internet [98]; Internet of Things [11, 84]; and smart wearables [30]. Venkatesh et al. [41] considered perceived fee as a predictive factor of the intention to adopt technologies. Since users bear the monetary costs of using technology, this may affect their perception of technology adoption behavior. Moreover, in the context of behavioral intention, users often compare the advantages of using a technology with the costs of using it. If the perceived fee exceeds the perceived benefits, the technology may be seen as an expensive device and users will be less likely to adopt it [98, 99].

Furthermore, from a user’s viewpoint, the adoption of smart technologies requires additional costs, such as possessing a smartphone with near-field communication (NFC) capability [11]. The price and maintenance costs of wearables were investigated as a critical barrier for using these modern technologies in an empirical study conducted by Yang et al. [30]. They illustrated that financial risk is a monetary sacrifice caused by purchasing or maintaining wearable devices, which could negatively impact upon users’ perceived value. Thus, the following hypotheses are proposed:

H2a

Perceived fee has a negative influence on perceived value of smart wellness wearables;

H2b

Perceived fee has a negative effect on the intention to use smart wellness wearables.

4.3 Perceived enjoyment (PEJ)

The significance of perceived enjoyment (equivalent to hedonic motivation) for investigating the intrinsic motivations of individuals in adopting and using consumer products has been determined by many scholars [11, 30, 41, 100,101,102,103,104]. Additionally, the proposed UTAUT2 [41] extends the previous model by adding a hedonic motivation to elucidate users’ intrinsic perceptions more clearly [41]. Individuals would have heightened intention toward adopting a technology or an innovation if they find pleasure in carrying out a particular behavior related to using that technology [49].

Similarly, Kim et al. [49] revealed that perceived enjoyment is a significant predictor of perceived value in the context of M-Internet. Hsu and Lin [11] have also determined perceived enjoyment as being a crucial factor in the context of IoT that could affect a user’s perceived value. In another study, Yang et al. [30] concluded that the effect of perceived enjoyment has the strongest impact on the perceived value of users of smart wearable devices. Therefore, this study proposed the following hypotheses:

H3a

The perceived enjoyment of using smart wellness wearables will positively affect users’ perceived value;

H3b

The perceived enjoyment of using smart wellness wearables will positively affect users’ intention to utilize wearables.

4.4 Performance expectancy (PE)

One of the principal reasons for the slow diffusion of smart technologies is that the usefulness of these technologies is not well understood by potential users [15]. Previous studies supported performance expectancy and perceived usefulness as critical determinants of behavioral intention to accept IT usage in various domains, such as: mobile data service [105]; short message services [106]; Mobile TV [78]; Radio-frequency identification (RFID) [51]; and IoT [15, 30, 79]. In addition, Park and Chen [107] realized that the usefulness of a user’s smartphone has a potential impact upon his/her intention to adopt such devices. Moreover, IS researchers have confirmed the significant correlation between perceived value and perceived usefulness [11, 30, 49]. However, Venkatesh and his co-authors [43] claimed that performance expectancy is the greatest predictor of the intention to use new technologies compared with perceived usefulness and similar constructs in other models [43]. Hence, they considered performance expectancy as one of the key factors in improving behavioral intention to use and in increasing the perceived value of smart wellness wearables. Therefore, the following hypotheses are assumed:

H4a

Performance expectancy of wearables will positively influence the user’s intention to use smart wellness wearables;

H4b

Performance expectancy of smart wellness wearables will positively influence users’ perceived values.

4.5 Effort expectancy (EE)

Previous studies stated that effort expectancy and perceived ease of use are substantial factors of behavioral intention toward using new technologies [15, 51, 79, 108, 109]. IS researchers believe that individuals may have the intention to accept smart technologies due to their simplicity and ease of use [15, 110]. In the retail context, Evanschitzky et al. [111] proved that perceived ease of using smart retail technologies has a positive relationship with an individual’s intention to adopt them in the future. In another study, Balaji and Roy [112] proved that when users perceive smart technologies as easy-to-use devices, they will have more incentive to find the benefits of these devices and interact with them.

However, Kim et al. [49] stated that perceived ease of use is a technicality factor of using M-Internet and could significantly affect users’ perceived value. Lai et al. [113] examined the relationship between perceived ease of use and perceived value of employees toward using online business-to-business (B2B) banking. Their empirical study showed that, as a technological factor, ease of use has the most significant effect on employees’ perceived value in using e-banking. Since effort expectancy is the most powerful predictor of behavioral intention to adopt new technologies compared with similar constructs in other models [43], this study considers effort expectancy as an influential factor of perceived value and intention to use smart wellness wearables. Thus, the following hypotheses are proposed:

H5a

Effort expectancy is estimated to have a positive influence on the perceived value of smart wellness wearables;

H5b

Effort expectancy is estimated to have a positive influence on the intention to use smart wellness wearables.

4.6 Social influence (SI)

The correlation between social influence and behavioral adoption of new technologies has been clarified in various models, namely: TAM2 [114], UTAUT [43], and UTAUT2 [41]. Several researchers have verified that the opinions of important people in a user’s life could positively influence her/his adoption behaviors in an information-sensitive context, such as: mobile health services [74]; wireless application protocol-enabled smartphones [115]; and location-based services [116]. This observation also includes domains closely related to the smart wellness wearable technologies, such as: wearable locating systems [117]; personal safety wearables [118]; smartwatches [101]; and healthcare wearables [27, 119].

In other words, most smart wearable users wish to make their own decisions for accepting these devices based on other people’s opinions since these types of technologies are entirely new for them [27]. Additionally, most studies have proven that a positive social acceptance of wearable technologies encourages users to adopt and use such devices [118, 120]. Hence, this study hypothesizes that:

H6

Social influence has a positive relationship with the intention to use smart wellness wearables.

4.7 Facilitating conditions (FC)

Venkatesh et al. [41] indicated that facilitating conditions are the consequence of external and internal conditions [119]. External conditions refer to users’ beliefs concerning the availability of necessary resources for performing a particular activity. Internal conditions refer to users’ views regarding the evaluation of their personal abilities for performing the activity [74, 121]. In UTAUT2 [41], a new relationship was developed between facilitating conditions and the intention toward new technology usage. Therefore, this framework was proposed in order to investigate new technology usage from the perspective of users in their daily lives [41]. Previous studies have demonstrated the importance of facilitating conditions toward an individual’s behavioral intention to adopt and use new technologies, such as: learning management system (LMS) [122]; personal safety wearable devices [118]; RFID [51, 76]; and Telemedicine [79]. On the other hand, most users are using different devices (e.g., different types of smartphones) that may affect their intention to use smart wellness wearable devices [41]. Therefore, the following hypothesis is established:

H7

Facilitating conditions could have a positive effect on the intention to use smart wellness wearables.

4.8 Perceived trust (PT)

The direct impact of trust on behavioral intention was demonstrated as a significant relationship in various domains, such as: online shopping [123]; mobile banking acceptance [72]; and E-commerce [124]. Moreover, in various studies on IT acceptance, perceived trust is cited as a crucial predictor for adopting technology and the behavioral intention to use technologies [118, 125,126,127]. In line with this claim, Li et al. proved that perceived trust is a critical factor that affects the individual’s intention to use health wearables [126]. In addition, Lunney et al. [128] suggested that the effect of perceived trust on smart wearable adoption should be investigated as a future work. Meanwhile, they found a negative relationship existing between perceived usefulness and inaccurate and unreliable data generated by the wearables. Based on the aforementioned discussion, this study postulates the following hypothesis:

H8

Perceived trust has a positive influence on user intention to use smart wellness wearables.

4.9 Perceived value (PV)

One of the key barriers to accepting smart products is the lack of perceived value of such devices [129]. According to a study by the Acquity Group [130], the lack of perceived value among end users presents a considerable obstacle for mass adoption of smart technologies. Extensive research studies have proven that perceived value could influence the intention to use technology [11, 30, 49, 131]. Individuals may consider that using smart wellness wearables is a valuable experience when they observe that the benefits received have greater priority compared with the monetary and non-monetary costs expended [11, 49]. Kleijnen et al. [131] confirmed that perceived value had positively affected individuals’ behavior of adopting mobile service delivery. The potential relationship between perceived value of using blogs and the behavioral intention of users was revealed by Chen and Lin [132]. Yu et al. [133] also investigated the significant relationship existing between perceived value of location-based social networking services and the behavioral intention to use such services. Moreover, Hsu et al. [11] specified that an individual’s perceived value of smart technologies could influence the behavioral intention of using them in the future. However, if the perception of value toward using smart wellness wearables has greater benefits compared with monetary and non-monetary factors (e.g., perceived privacy and perceived fee), consumers may have more incentive to use them in the future [49]. Thus, the following hypothesis is proposed:

H9

The perceived value has a positive correlation with behavioral intention to use smart wellness wearables.

4.10 Perceived health increase (PHI)

According to Lamb et al. [134], performance of sports and physical activities may have positive effects on an individual’s overall self-rating of their health [86]. Smart wellness wearables provide users with the opportunity to collect and monitor physical activities and wellness-related data on these devices and on their connected technologies, such as smart phones [135]. In other words, if users consider a particular behavior as beneficial to enhance their current health status, they would have more incentive to engage in that specific behavior. Consequently, it is expected that individuals would use smart wellness wearables if they believe these devices will improve their general health [86].

According to the Theory of Reasoned Action (TRA) [136], a person’s decision to become involved in a specific behavior is based on the consequences that he/she assumes will happen [136]. In this regard, the Health Belief Model (HBM) [137] asserted that individuals are likely to become involved in a particular behavior if they presume that such behavior will improve their current health status. By considering these explanations, it could be concluded that individuals may be encouraged to use smart wellness wearables if they presume that these devices will improve their current health status [86].

In a study conducted by Lunney et al. [128], the authors recommended that user perception of fitness wearables related to general health status should be examined through a casual structure as a future area of research. They found that there was a positive and direct relationship between using fitness wearables and perceived general health. However, the perceived general health construct was not considered in their proposed research model since there was a lack of literature to support this claim. Nonetheless, Ernst et al. [86] have determined that perceived health increase has a direct and positive impact on the behavioral intention to use fitness trackers. Based on this line of thought, this study proposes the following hypothesis:

H10

Perceived health increase can positively affect the behavioral intention to use smart wellness wearables.

5 Research methodology

5.1 Scale development

In order to test the proposed model, a web-based questionnaire was developed. This included the items for all constructs engaged in the conceptual research model. Almost all measurement items (indicators) were extracted based on a comprehensive search of previous studies, which were validated and widely used in IoT and wireless technology adoption researches. This study proposed two new measurement items, namely an item related to the “perceived trust” construct and an item related to the “intention to use” construct (see Table 3). All measurement items were improved with minor modifications to fit with the smart wearables domain. Indicators were evaluated according to the five-point Likert scale ranging from “Strongly Disagree” to “Strongly Agree.” To guarantee that the responses were not selected inattentively or randomly, the following attention-check item was embedded in the middle of the questionnaire: “You should answer ‘strongly disagree with this statement’ to ensure you have read all questions carefully” [138]. It is worth mentioning that this study is a small part of a PhD research and the main target population of the research will be Malaysian people. Therefore, a Malay version of the questionnaire was also developed, with the assistance of two senior PhD students from Malaysia. Then, to ensure that the Malay version of the questionnaire accurately reflects the English version, both versions were carefully perused by two senior lecturers at the Faculty of Computing, Universiti Teknologi Malaysia (UTM). Based on their feedback, some of the items were amended for purposes of clarity and understandability. In addition, any differences between the two versions were minimized by using the back-translation process [108]. To accomplish this, the Malay version of the questionnaire was translated back into English and then compared with the original English version, whereby no significant discrepancy was found. Both versions of the measurement items are presented in Table 3.

5.2 Survey administration

Generally, a questionnaire must be examined through a small group of actual users from the target population to investigate its weaknesses and potential problems. Therefore, a survey was designed by including all constructs involved in the conceptual research model. Previous studies on the adoption and usage of new technologies stated that students are heavy users of new technologies [120, 139, 140]. Since the main target population of this study comprises smart wellness wearables users, university students would be the best representative sample for this population. Therefore, this study was conducted at the Faculty of Computing, Universiti Teknologi Malaysia (UTM). After developing a web-based questionnaire using Google Forms, an invitation e-mail was sent to all students including a link to the online survey. To present the students with a clear understanding, a brief description of the research objectives and smart wellness wearables were prepared at the beginning of the survey. After eliminating inaccurate responses, a total of one hundred (100) usable responses remained. Table 2 presents a summary of the demographic information of the respondents.

Table 2 Demographic information of respondents
Full size table

6 Confirmatory factor analysis

The quality of the measurement model was evaluated by following the guidelines of [141, 142] using content validity, construct validity, and construct reliability, respectively. Face and content validation were applied in order to determine whether the items related to a construct had adequately covered all aspects of the construct, or whether the instrument had measured the right content [142,143,144]. In other words, content validity measures how much the questionnaire items can represent the relevant construct [91]. To conduct face validity analysis, researchers should consult with several lay persons to investigate whether the questionnaire is sound or relevant [143]. Therefore, seven PhD candidates from the IS department were selected to participate in the face validity analysis. Participants made several suggestions for revising the items to render them clearer and more understandable. After applying their comments, a panel of seven experts, having knowledge of new technology adoption from the IS and Computer Science departments, were invited to participate in the content validity process. This study applied the Content Validity Index (CVI) according to experts’ rankings for calculating the relevancy of items against the constructs [144]. Subsequently, two types of CVI were calculated, namely Item-level Content Validity Index (I-CVI) and Scale-level Content Validity Index (S-CVI) [144]. I-CVI includes the content validity of individual items, while S-CVI consists of the content validity of the entire questionnaire [144]. Based on the results of these two CVIs, almost all of the items as well as the entire questionnaire were found to have met the adequate criteria defined by Lynn [143]. Only one item from the Social Influence construct (SI5) was deleted from the final version of the questionnaire, because its I-CVI was lower than the required criteria. At the end of the content validity process, the indicators were modified according to the experts’ recommendations. The revised items are presented in Table 3.

Table 3 Measurement items and factor loadings
Full size table

According to Straub et al. [142], construct validity is referred to as a subject of measurement between constructs that can be evaluated through convergent validity and discriminant validity [91]. Generally, convergent validation is referred to as the degree to which an indicator correlates positively with other indicators of the same construct [141, 142]. Most academic researchers assume the factor loadings (outer loadings) of the indicators and Average Variance Extracted (AVE) to be convergent validity assessment [141]. To assess the measurement model, SmartPLS, version 3.2.7, was used. Figure 5 shows the measurement model assessment produced by SmartPLS software. To achieve an admissible convergent validity, factor loadings should be higher than 0.6 [145,146,147,148,149]. As presented in Table 3, all indicators (except for two items) met the desired criteria of factor loadings. An item from the effort expectancy section and an item from the perceived enjoyment section were removed due to low factor loading values. According to Hair et al. [141], AVE is defined as “the grand mean value of the squared loadings of the indicators associated with the construct.” To achieve a satisfactory convergent validity, all values of AVE should not be less than 0.5. As shown in Table 5, all AVE values are higher than the recommended value.

Fig. 5
figure 5

The measurement model assessment obtained from SmartPLS software

Full size image

Discriminant validity is referred to as the degree to which indicators of different constructs are truly distinct and each construct does not reflect other constructs [141, 158, 159]. To evaluate the discriminant validation of the constructs involved in the proposed model, two methods were applied [141]. The first method aimed to examine the cross-loadings of the measurement items. According to Hair et al. [141], an item’s factor loadings related to a construct must be greater than all outer loadings on the other constructs. As shown in Table 4, all indicators’ outer loadings meet the required criteria of discriminant validity. The second and more conventional approach is the Fornell-Larcker criterion. This test compares the square root of AVE values with the correlations of the latent variables. Hair and his co-authors [141] claimed that the square root of the intended construct’s AVE must be higher than its maximum association with the remaining constructs. As shown in Table 5, all the square roots of AVE values have satisfactory quantities for the discriminant validation.

Table 4 Factor loadings and cross-loadings of measurement items
Full size table
Table 5 Internal consistency and Fornell-Larcker
Full size table

Generally, construct reliability is determined by two criteria, namely Cronbach’s Alpha (CA) and Composite Reliability (CR) [142]. CA is defined as a measure of internal consistency among all indicators of a construct that estimates whether all indicators are equally reliable [141]. CR refers to a similar concept; however, it is considered as a more precise reliability measure in Structural Equation Modeling (SEM) [141, 160]. To reach a desirable reliability, both CA and CR should be equal to or greater than 0.7 [141]. Table 5 shows the reliability indicators of each construct. All values for CA and CR met the recommended criteria of higher than 0.7 [141].

7 Conclusion, limitations and future directions

Smart wellness wearables are introduced as the next generation of digital revolution devices that could be accepted by a majority of the general public. Although the general public awareness toward the acceptance of smart wellness wearable technologies is rising, at the same time smart wearable abandonment cases are gradually increasing. Today, most individuals tend to monitor their physical activities with the aim of achieving healthy life styles and preventing chronic diseases. Thus, it is critical that designers and developers understand the specific requirements and priorities that users need when using such devices. Therefore, based on previous studies, this research has determined the potential factors that could affect user intention in the context of using smart wellness wearables. Subsequently, a measurement instrument was developed and a web-based survey was conducted among university students. The measurement model was then assessed and validated using the SmartPLS software.

The findings confirmed that all indicators and constructs met the adequate validity and reliability criteria. Moreover, it is envisaged that the proposed instrument of this study (both the English and Malay versions) could help academics and researchers to examine the same factors in other categories of smart wearables. This theory could also be applied to smart technologies, such as smart medical devices. In addition, this study has aimed to contribute to the collective IS researches by providing better understanding of the potential factors that could affect user willingness to use smart wellness wearables. To the best of the authors’ knowledge, this is the first effort to integrate two models (namely UTAUT2 and VAM), for the purpose of examining the behavioral intention of users toward using smart wellness wearables, particularly in Malaysia.

Due to the novelty of research, this study is not immune from limitations, one of which is the small sample size (100 participants). Since smart wellness wearables are essentially designed for general public usage, another potential limitation is the composition of the sample population of this study (university students). The population of students cannot encompass all segments of society. It should be noted that this paper is a small part of a larger empirical study. Consequently, in order to fulfill the aforementioned limitations, an extended survey is planned to repeat the similar analysis of this study on a larger scale. This survey will aim to corroborate the outcomes of the CFA, as well as examine and validate the proposed research model among the general public in Malaysia. Nevertheless, the authors of this research assume that the proposed model of the present study may complement the original models in explaining users’ behavioral intention to use smart wellness wearables.

However, considering the complex challenges in the evaluation of a research model, it is not applicable to involve all potential factors in a unified model. Thus, this research may not take into account all the potential factors related to smart wellness wearables. These can include cultural differences and technical characteristics of smart wearables which play significant roles in the behavioral intention of consumers toward using such devices [5, 161]. Accordingly, consideration of technical characteristics (such as device weight, portability, resilience, and data accuracy) of smart wellness wearables on the intention of consumers toward using these devices has been suggested for future research.

On the other hand, new technology adoption and usage may be affected by cultural dissimilarities in different countries [126, 150, 162,163,164]. Keikhosrokiani et al. [126] examined the moderating effect of nationality on the usage of a patient-centric healthcare system via smart wearables. The study revealed the effect of cultural differences in two different countries, specifically, Iran and Malaysia. Generally, Malaysia is considered as a multicultural country made up of three prominent ethnic groups, namely Malays, Chinese and Indians [165]. Hence, it is recommended to consider the moderating effect of ethnicity on the usage of smart wellness wearables so as to examine the effect of cultural differences among the general public in Malaysia.

By the same token, Hofstede and Bond [166] developed a cross-cultural framework and cultural dimensions to explain the influences of the culture of a society on its members’ value as well as evaluating the effect of these values on the members’ behavior. In this regard, it is recommended to investigate the impact of Hofstede’s cultural dimensions (such as individualism and collectivism, uncertainty avoidance, masculinity and femininity, and time perception) on the social influence construct of the presented research model so as to compare the cross-cultural differences of using smart wellness wearables in Malaysia and other countries.