Research on Perceived Value and Usage Intention of Tactile Interactive Advertising Among Consumers

Abstract

With the maturity of haptic technology and complex systems, tactile interaction has gradually become realized through specific hardware and software configurations in the e-commerce and business industries. As an innovative form depending on haptic systems, tactile interactive advertising could help both advertisers and consumers enhance the haptic experience of products through technology-mediated virtual environments and provide tactile information for purchase decision making that relies on restoring the real sense of touch. On the basis of the value-based adoption model (VAM) and the need for touch (NFT) from a preference for haptic information in a system, we conduct quantitative research and construct a partial least squares structural equation model, which aims to study the influencing factors that characterize the user preference of tactile interactive advertisements empowered by haptic systems among Chinese consumers. A total of 509 valid questionnaires were collected through online and offline channels. The study revealed that the perceived enjoyment (PE) and telepresence (TEL) of tactile interactive advertisements as benefit factors positively influence the perceived value (PV) and that the perceived fee (PF) as a sacrifice factor negatively influences PV, which further impacts the attitude and intention to use (IU). In addition, the study verified that a higher NFT positively affected PE, PU, and PF and IU for the perception of tactile interactive advertising. Through this study, we aim to provide insights from a consumer perspective to enhance the advertising effect and user experience through tactile interaction in further e-commerce, which transforms how we interact with digital systems and virtual environments.

1. Introduction

Touch is one of the most important and prominent senses for humans [1]. With the rapid development of haptic technology, the main types of haptic systems comprising robotic hardware and tactile feedback device for users are graspable, wearable, and touchable. The users are allowed to interact with virtual environments through mechanically simulated haptic [2] and could be provided a digital sensory experience that reflects the texture, stiffness, size, shape, temperature, weight, and orientation of an object in healthcare [3], manufacturing [4], advertising, and marketing [5], etc. The commercial value of haptic technology continues to be released [6]. Specifically, the global haptic technology market reached USD 3.05 billion in 2023 and would grow at a compound annual growth rate of 11.7% between 2024 and 2030 in the Asia–Pacific market [7]. According to IDTechEx, the haptic market will be worth close to USD 5 billion by 2025 [8] with considerable potential for future growth.

As a vital E-commerce application of haptic technology [9], tactile interaction can be used in the field of advertising. This form of advertising can be referred to as tactile interactive advertising. Relying on the communication and interaction between the haptic device and the control system, the encoding tactile information (such as surface material, temperature, and texture) is simultaneously transmitted with audiovisual material from the advertiser (or television group or digital platform) to the consumers. The remote haptic interaction with the sensation of the goods (e.g., vibration, force, and motion feedback) can be provided to consumers through haptic devices [10,11,12]. Like haptic feedback in the application of gaming or other industrial training, consumers can obtain the physical touch of virtual items through tactile interactive advertising in addition to the audiovisual presentation of a commodity (e.g., video and VR/AR advertising without tactile elements) which are contacted and appear to be genuine and real via tactile feedback. Haptic technology enhances the advertising experience and purchase intention. It is feasible to apply touch in mobile and digital marketing environments [13]. This can increase the engagement rate by providing tactile dimensions, and then enhance consumer purchase intentions with greater brand awareness. The immersion and IPG Media Lab revealed that the incorporation of haptic technology into advertisements enhances brand favorability by 50%, increases the awareness of advertising by 75%, and boosts purchase intention by 41% compared with the audiovisual advertisements alone [14].

Need for touch (NFT) describes consumers’ preference to obtain product information through tactile experiences [15], which is realized through tactile interaction. Consumers learn the relevant information about product attributes and enjoy themselves during the haptic experience for purposeful purchase decisions [16]. On this basis, the significance of tactile interactive advertising for consumers is shown in enhanced engagement, personalization, accessibility, and inclusiveness [17]. Some studies have demonstrated that tactile interaction techniques can enhance the user experience (UX) due to the increased sense of presence and enjoyment [18]. The simulation of the tactile sensation of the original item occurs when consumers use haptic interactive devices [19]. The emergence of tactile interactive advertising further aims to satisfy the haptic needs and user experience of consumers when viewing advertisements.

Therefore, the impact of haptic interaction technology on consumer experience and its effect on consumer attitudes, intentions, and behaviors is considered to be a salient topic in advertising research [20]. Current studies focus on whether haptic interaction technologies have an impact on consumer purchases or experiences, but how haptic interactions influence consumer judgments still needs to be further clarified [21]. Past studies have used the technology acceptance model (TAM) to address acceptance issues involving the application of new technologies. For haptic interaction technology, Oh et al. explored the adoption of haptic enabling technology via innovation and diffusion theory (IDT) and TAM [22]. Van et al. explored the specific features related to haptics that are worth using in online sales, including the ease of use, usefulness, and enjoyment [23]. Limerick et al. introduced the perceived characteristics (e.g., usability, aesthetics, and rewards) and the intention to use tactile interactive technologies from a user engagement perspective [24]. On the other hand, understanding how consumers with different needs for touch can interact with virtually touchable products via a haptic system is considered a research direction for future technological development [25]. In this way, we proposed the following research question (RQ):

RQ1: Do consumers’ need for touch affect their perception and intention to use tactile interactive advertising?

For users or consumers, what is given and what is received are the criteria by which they determine the value and adoption of technology [26]. Few existing studies have considered the impact of the negative aspects of technological features on consumers’ perceived value and even their attitudes and intentions to use, e.g., due to the deficiency of TAM and UTAUT framework. The value-based adoption model (VAM) is considered a balanced analytical model [27], since it not only describes the perceived benefits, i.e., perceived usefulness and perceived enjoyment, but also takes into account possible perceived sacrifices, i.e., monetary (perceived fee) and non-monetary (technicality) aspects of the costs involved in adopting the new application [28].

However, while NFT describes the need for consumers to access product information through haptic technology, past NFT research has focused more on consumers’ desire to touch and has not addressed the users’ access preference of haptic experiences through haptic technology [18]. Lee et al. previously attempted to demonstrate the negative impact of NFT on the perceived ease of use, perceived enjoyment, and loyalty intention during online shopping using TAM, but did not consider the digital sensory impact of haptic interaction techniques [29]. Another research question is presented as follows:

RQ2: What are the perceived characteristics of haptic technology that affect consumers’ perceived value and intention to use tactile interactive advertising?

In recent studies, on the one hand, researchers have found that the tactile interaction can affect consumers’ experience and willingness to purchase [13] and then develop pragmatic and hedonistic consumer motivations [30]. To determine how tactile interfaces affect consumers’ product evaluation and purchase intentions, Racat et al. revealed the effects on consumers’ understanding of product information and their pleasure when interacting with the product [31]. On the other hand, in terms of consumers’ technology acceptance, Pino et al. explored how haptic interaction technologies affect consumers’ expected ease of use of products, as well as their attitudes and intentions toward these products [32], whereas Lee et al. explored the impact of virtual touch on satisfaction in virtual reality (VR) stores [33]. Combined with touchscreen interactions, multimodal sensory interactions would enhance spatial presence, which can effectively alleviate consumer uncertainty and information overload about products [34]. Unlike existing studies, the contributions of this study are summarized as follows:

• Based on NFT and the value-based adoption model (VAM), we examine the consumer adoption of tactile interactive advertising in terms of both benefits and sacrifices by incorporating the characteristics of tactile interactive advertising, and further explored the intention to use.
• We found that different levels of need for touch have a direct impact on the perception of tactile interactive advertising features, as well as on the willingness to use them. For Benefit factors, perceived enjoyment and telepresence contribute to consumers’ perceived value; For Sacrifice factors, only a perceived fee negatively affects perceived value, which implies that developers need to develop or choose more affordable haptic interfaces/devices.
• These findings will provide practical recommendations for technology developers and advertisers to promote the popularity of tactile interactive advertising and empower product sales.

The structure of this paper is as follows. Section 2 introduces the review of haptic technology applications and advertising with new technologies, including tactile interactive advertising. Section 3 illustrates the research models and hypotheses. Section 4 describes the method and the demographic characteristic. Section 5 presents the data analysis process. Section 6 discusses the research result and points out its implications and limitations. Section 7 makes a final conclusion of the study.

2. Literature Review

2.1. Haptic Technology and Tactile Concept

Haptic interaction technologies can usually be categorized into three different domains, i.e., the master domain, the network domain, and the slave domain [35]. The master domain usually consists of either human controllers or machine controllers, with the former also including human–machine interfaces (HMIs), motion control devices, and auditory–visual feedback devices. Currently, the human–machine interfaces of high interest include encounter tactile displays (ETHDs) [36], mid-air ultrasonic haptics [5], and electro-haptic feedback [37]. The slave domain consists of various entities that are directly controlled by the master domain, and the constituent entities of the slave domain vary from case to case. For example, in the case of teleoperation, the slave domain is mainly robotic devices that interact and perform work based on command signals from the operator; and, in the case of networked control systems, the slave domain is mainly sensor and actuator systems. The network domain serves as the link between the master and slave domains and is the medium for their interaction. As a brief description, haptic interaction consists of a haptic device with sensors and actuators, and a control computer capable of inputting relevant human operations to a haptic information display. With the development of haptic technology in tactile rendering and information transfer, the concept of Tactile Internet (TI) has been proposed by both academia and the industry. According to the definition of the IEEE 1918.1 Working Group on Tactile Internet standard, the so-called Tactile Internet is a network (or network of networks) for remotely accessing, sensing, manipulating, or controlling real or virtual objects or processes perceived by a human or a machine in real time [38]. To realize the key performance requirements of the haptic Internet, in addition to accurate localization and manipulation, ultra-high reliability, ultra-low latency, and stable network connectivity are required from the data transmission layer to provide efficient multimodal information. Currently, the application of haptic Internet technologies has been widely explored in various fields and has resulted in initial applications in areas such as healthcare [39], industrial practices [40], and advertising [41].

As an application of human–computer interaction, haptic technology generally encompasses haptic perception/sense and haptic feedback between human and machine, and the tactile interaction within the body of the human, between the machine and the machine [2,42]. In fact, tactile emphasizes a subjective human perception, addressing tactile perceptions from the skin, such as vibration and temperature [43]. On the other hand, haptic emphasizes a technological concept, which plays an important role in today’s human–computer interaction (HCI) [44]. Haptic technology has expanded from simple haptic feedback to provide haptic experiences including HCI, which allows users to realize more tactile interactions [45,46]. Therefore, the concept of “tactile interactive advertising” is used in this paper.

2.2. The Applications of Tactile Interaction

We can see that the tactile interaction has been used in the fields of medical and healthcare, industrial simulation, and marketing campaigns. For medical and healthcare, tactile interaction can help the remote operation of various surgical operations and provide convenience for the disabled. Lacki and Rossa constructed a 3-degree-of-freedom haptic devices using brake actuators, which fills the gap in the tactile system that ensures transparency and stability in remote surgery [47]. Turkseven et al. proposed the tactile impedance model of Endoscopic Submucosal Dissection (ESD), which will be applied to the treatment of colon surgery [48]. In addition, TI technology has been studied in orthopedic milling surgery and the tactile localization of uterine fibroids [49,50]. TI will also provide convenience for the blind. Barontini et al. proposed an indoor navigation system based on wearable tactile technology to help visually impaired people avoid obstacles [51]. Espinosa-casta and Medell proposed a virtual tactile perception method that supports blind teaching by combining VR and tactile technology [52].

For industrial practice, remote operation based on tactile interaction will prevent people from working in a dangerous environment. Abi-Farraj et al. explored a new tactile sharing control method to help operators classify and separate radioactive waste [53]. At the practical application level, when Volkswagen commercial vehicles assembled the doors of T6 and T7, they implemented and tested “SenseGlove Nova” for vehicle VR training to improve the authenticity of VR simulation and protect expensive vehicles from possible damage during the handle training [54]. Additionally, tactile interactivity has been used in interactive marketing campaigns [24]. Especially for the usage of marketing campaigns, it will trigger wide application in advertising through tactile interaction.

2.3. Tactile Interaction in Advertising

The application of haptic technology in advertising and marketing is gradually emerging. Early haptic interfaces were mobile phones or gamepads. TouchSense had previously developed a haptic involvement toolkit that could provide designers with a solution for adding haptic effects to video advertising content and could set the parameters that matched the advertising campaign [55]. Peugeot released a haptic advertisement that simulates the roar of the car and the vibration of an engine using the vibration of a mobile phone [56]. With the sophisticated haptic technology, the kinds of haptic interfaces are developed to wearable or non-wearable interfaces, which are capable of reflecting physical properties, and in part can be integrated with VR/AR. Porsche has teamed up with Zerolight and Ultrahaptic to combine high-quality visualization with mid-air haptics to create personalized public advertisements [57]. P&G Health used Senseglove to simulate the state of patients with nerve injury, and organized a business activity with the theme of “raising awareness of nerve injury” worldwide [58]. Nissan, in conjunction with the company known as HaptX, realized the interior driving experience of the concept car Nissan IMs using VR and haptic technology [59].

In academia, Mulcahy et al., in addition to showing that tactile interaction can improve the advertising experience and enhance the willingness to purchase a product, the haptic touch is now a sensation that can be used in mobile and digital retail environments [13]. Consumers can already touch, feel, and manipulate objects in virtual reality environments via tactile interaction devices [60]. Some studies also show that mid-air haptic feedback significantly improves usability and aesthetics compared to digital signage without haptic feedback [24]. Kim et al. introduced Refinity, a digital signage using holographic projection and mid-air haptics, and exemplified a usage scenario in which a user is free to browse a room and pick up or feel the furniture of interest in the room [61]. Racat et al. demonstrated through exploratory experiments that haptic rendering technology can mimic product texture and influence product purchase intention by mobilizing sensory memory to feel an experience similar to offline shopping [31].

2.4. The Usage of New Media Advertisement Technology

In the era of traditional advertising, even with the birth of mobile advertising, most of the advertisements use static pictures or videos to promote products by highlighting product information [62], The way to attract consumers’ senses is to improve the effect of advertising. Bolognesi et al. studied the role of “metaphors” of different sensory cues in print advertising. Existing research on digital sensory marketing mainly focuses on audio-visual interaction [63]. Because technologies and devices related to vision and hearing are more mature and widespread [64]. With the development of new media, three-dimensional (3D) technology has brought advertising into a new era, which is enabling consumers to experience the floating 3D visualization of products with real depth off-screen [65]. Compared with 2D advertising, the virtual experience brought by 3D advertising brings consumers a more positive perception and sense of presence [66,67]. Nowadays, advertising with 3D display, interactive features, and personalized recommendations is more popular among consumers [68].

As an emerging form of advertising, virtual reality (VR) enable consumers to watch 3D images with its advantages of vividity and interactivity [69]. The researchers mainly focused on the consumers’ acceptance of VR advertising and its effects [70,71,72]. Through empirical research, VanKerrebroeck et al. proved that VR produced higher vividness than ordinary 2D video and positively affected consumers’ attitudes towards advertising [72]. DeGauquier et al. also confirmed that people using VR advertising have a higher purchase intention [70]. In terms of interactivity, Wang et al. revealed that users valued interactive experience most when they experience VR advertising [73]. When users experience VR interactive advertisements, users paid the most attention to interactive experience such as interactive content, interactive systems, interactive functions, and interactive interfaces. For VR ad personalization distribution, Kilic et al. compares algorithms for personalized advertising distribution in VR platforms to maximize expected returns [74].

In contrast to VR technology, augmented reality (AR) technology organically combines the virtual and the real environment [75], which have great application opportunities [76]. Using computer graphics to generate visual output, advertisers can transform real print advertisements into 3D virtual models or videos that can be played through AR. For the purpose of enhancing consumers’ physiological responses and participation, AR advertisements stimulate consumers’ positive attitudes [77], which consumers had a higher acceptance of in the commercial field [78]. The different content characteristics of AR advertisements will affect consumers’ attitudes towards AR advertisements [77]. For example, the controllability and fun dimensions of the perceived interaction of AR advertisements will affect the mental imagery, thereby affecting consumers’ attitudes and behavioral intentions towards products [79]. The immersive experience will also significantly affect people’s perception of AR applications. In addition, information quality, sensory quality, interactivity, etc. also directly or indirectly affect users’ willingness to use and even continue to use [80,81].

With the emerging 3D stereoscopic technology, also known as autostereoscopic technology (non-glasses type), viewers can see stereoscopic images without wearing special glasses [82,83] and holographic advertisements belong to this category. In advertisements, holographic technology can display 360-degree dynamic images, stereoscopic images, and music effects. Depending on its strong high emotion, sense of space, perspective, contrast and color reproduction, the audio-visual effects of advertisements are more realistic and immersive [84]. Kim et al. designed an interactive holographic signage called “Refinity” for a new retail shopping, which used an autostereoscopic 3D display and combined various interactive tools to enhance the interactivity of online shopping [61]. Naked eye 3D advertisements can also directly see the stereoscopic effect via a “pseudo-3D” imaging advertisement on a special LED screen aided by optical illusions, which has been widely spread all over the world, such as the Xiaomi advertisement in Taikoo Li, Chengdu, China, and the “DNA City Life” advertisement in Italy.

3. Research Model and Hypotheses

However, the original VAM also has weaknesses as it neglects the necessity of integrating with the technical characteristics of a particular technology. Most of the past studies have made adjustments in utilizing the VAM as a framework, e.g., Chloe et al. replaced the original perceived fee with captivating inputs when using the VAM to explore AR adoption [85]. Similarly, the shortcoming of VAM is also valid in haptic interaction advertisements, but so far, there is no very well-developed model that can characterize the technology of haptic interaction advertisements. Although previously, in order to better standardize the assessment of haptic feedback in haptic technology, some researchers developed the Haptic experience model (HX) based on user experience [86] and proposed four dimensions of Expressivity, Involvement, Realism, and Harmony after subsequent validation, it has yet to fully pass the reliability test [18].

In this study, a modified value-based adoption model (VAM) is proposed to investigate consumers’ value, attitude, and behavioral intention toward tactile interactive advertising. The constructs in the structural equation model (SEM) consist of the need for touch (NFT), benefit and sacrifice factors, perceived value (PV), attitude (ATT), intention to use (IU). The benefit factors include perceived enjoyment (PE), perceived usefulness (PU), telepresence (TEL), and sacrifice factors represented by physical risk (PR), complexity (COMP), perceived fee (PF), for which these terms are defined in

Table 1. Research model and hypotheses.

Construct	Definition	Reference
Need for touch	A predisposition for the extraction and utilization of advertisement information obtained through the haptic system.	[15]
Perceived value	The overall assessment of the utility of the tactile interaction advertising based on the perceptions of what is received and what is given.	[87]
Attitude	The degree to which consumers are willing to use tactile interaction advertising.	[88]
Intention to use	The subjective probability of using and recommending tactile interactive advertising.	[89]
Perceived enjoyment	The extent to which the activity of using tactile interactive advertising is perceived to be enjoyable in its own right.	[28]
Perceived usefulness	The degree to which the tactile interactive advertising would improve consumers’ performance.	[89]
Telepresence	A psychological state or subjective perception of presence into the environment created by tactile interactive advertising.	[90]
Physical risk	The possibility that may cause physical harm to consumers when using the tactile interactive advertising.	[91]
Complexity	The degree to which the tactile interactive advertising is perceived as relatively difficult to understand and use.	[89]
Perceived fee	The judgment of consumers on the cost of using tactile interactive advertising.	[28]

.

3.1. Attitude and Intention to Use

Attitude is defined as a psychological evaluation of an object, while the intention to use relates to the level of intensity in performing a particular behavior [89,92]. The Technology Acceptance Model (TAM) has determined that ATT and IU are positively correlated. Further studies exploring sensory interaction techniques based on TAM have validated this relationship. For instance, Zhang et al. revealed that the attitudes of construction workers towards using VR technology in safety training positively influenced their intention to use [88]. Pantano et al. found that users’ attitudes towards AR in online try-on positively and significantly affected their subsequent behavioral intentions to use the system [81]. Vo et al. discovered that the attitude of users towards mobile AR application had a positive effect on their usage intention [93]. Seong & Hong found that a positive attitude also influenced the intention to use VR movement [94]. Based on previous research regarding sensory technologies, we propose the following hypothesis:

H1: 

Attitude will positively influence the intention to use of tactile interactive advertising among consumers.

3.2. NFT and Its Formative Construction

NFT, as a preference for extracting and utilizing information from the tactile system, is based on differences in motivation and ability among individuals [95]. When evaluating product characteristics, people with a higher NFT will pay more attention to the texture of the product and continuously pay attention to tactile information. NFT is a multi-dimensional structure, including autotelic NFT (aNFT) and instrumental NFT (iNFT). The former is driven by seeking fun and enjoyment (not necessarily for purchase), while the latter is result-oriented and gathers haptic information to reduce uncertainty, so as to pick out high-quality items for purchasing [15].

According to the conceptual explanation of NFT, aNFT and iNFT act as two factors of NFT and there exists an essential difference in their nature, respectively. A change in any of the variables causes the structure to change. aNFT is a more spontaneous process and iNFT is a more postulated one [15]. In other words, the higher-order variable represents the variance shared by the two indicator constructs [96]. Thus, transforming aNFT and iNFT into higher-order NFT has two advantages; on the one hand, it does not ignore the difference between aNFT and iNFT; At the same time it restores the parsimony and intuition of the variables. Whereas aNFT and iNFT are first-order formative constructs for NFT. Similarly, Kuhn et al. operationalized the second-order construct NFT by using the repetition indicator method in a reflective–formative hierarchical component model, where iNFT and aNFT comprise the total NFT [97]. Based on the above findings, we make the following hypothesis:

H2-1: 

The second-order construct NFT holds, while aNFT (H2-1a) and iNFT (H2-1b) have a significant relationship with NFT.

Citrin et al. argued that the willingness to collect tactile cues is more pronounced among those with a higher need for tactile input, which is reflected in the online shopping environment, where those with a high need for tactile input would be deterred by the absence of tactile cues [98]. And, Rodrigues et al. showed that tactile information has a positive effect on consumers’ perception and purchase intention [99]. And, in the study of Zheng et al., it was shown that autotelic NFT and instrumental NFT influence consumers’ purchasing decisions [100]. The above literature illustrates a mechanism due to which people with high NFT have a higher willingness to collect tactile cues, and the collected tactile cues increase consumers’ purchase intention. However, in the past, online shopping has been characterized by a lack of tactile cues, causing the higher NFT group to have a lower willingness to consume. In this study, tactile interactive advertising precisely supplements the lack of tactile cues in online shopping. Based on the above assertion, this study hypothesizes that the group with a higher NFT has a stronger willingness to use tactile interactive advertising, and hypothesis is listed as follows:

H2-2: 

The NFT positively affects the intention to use of tactile interactive advertising among consumers.

3.3. Perceived Value

Perceived value is the consumer’s assessment after weighing the benefits and sacrifices involved, which has a significant impact on consumer decision making [101]. Regarding the purchase of green products, Xia et al. showed the positive influence of perceived value on purchase intention [102]. Kim et al. combined VAM with TAM in mobile Internet research and found that there exists a correlation between perceived value with adoption intention [28]. The attitude was also introduced in VAM and proved the positive effect of perceived value on attitude [103]. Therefore, the hypothesis is made as follows:

H3-1: 

Perceived value positively influences the attitude of tactile interactive advertising among consumers.

H3-2: 

Perceived value positively influence the intention to use of tactile interactive advertising among consumers.

3.4. Benefit

Kim proposed that usefulness and enjoyment are the benefit components of perceived value [28]. In the research field of emerging ICT scenarios such as IoT, VR/AR, and smart home, several factors of benefit component had been proposed in VAM, such as perceived usefulness, perceived enjoyment and telepresence in technology adoption studies. Vishwakarma et al. examined how perceived enjoyment, perceived usefulness, and perceived immersion as benefits positively affected the perceived value of VR adoption by travelers, and demonstrated the impact of all three on PV [91]. Similarly, Mathavan et al. investigated the impact of PE and PU on PV in fitness wearables usage, and also incorporated three additional variables in the benefit dimension, namely perceived design aesthetics [104]. Based on these studies, the hypotheses of this research are formulated:

H4-1: 

Perceived enjoyment (H4-1a) and perceived usefulness (H4-1b) will positively influence the perceived value of tactile interactive advertising among consumers.

In previous research, Lee et al. found the effect of NFT on TAM and demonstrated the positive effect of autotelic NFT on PE in e-commerce [29]. This suggests that NFT is an influential factor that affects perceptual dimensions in the market application of haptic technology. Gatter et al. demonstrated that people with a high iNFT usually touch objects in a goal-oriented, utility-oriented manner to evaluate product attributes, utility, and quality [25]. Therefore, the hypothesis is proposed in this study:

H4-2: 

Consumers’ level of NFT will be positively related to perceived enjoyment (H4-2a) and perceived usefulness (H4-2b) of tactile interactive advertising among consumers.

Presence is a dimension that can be utilized in various technological studies. For instance, telepresence (TEL) is a fundamental concept in virtual reality (VR) technology [105]. Additionally, Maier et al. state that the haptic Internet transmits both non-tactile controls and data while also imposing haptic control capabilities, making it impossible for end-users to distinguish whether they are operating the system locally or from a remote location [106]. Therefore, telepresence constitutes a crucial dimension of haptic technology. In the study of AR mobile applications under the VAM framework, Yoon et al. included presence in the Benefit category based on PE and PU, and demonstrated its favorable impact on PV [107]. As a result, in this research, TEL serves as a component of Benefit and is hypothesized as follows:

H4-3: 

Telepresence will positively influence the perceived value of tactile interactive advertising among consumers.

3.5. Sacrifice

Sacrifice consists of both non-monetary and monetary dimensions, and prior studies have used technicality and perceived fee as sacrifice dimensions of perceived value [28]. For instance, the studies confirmed that technicality negatively impacts the perceived value of self-customization services and AR, respectively, while perceived cost is negatively influential [107]. Ease of use is a commonly used indicator of technicality, and consumers tend to accept technologies that are uncomplicated and easy to use [108]. Perceived complexity is considered to be low perceived ease of use, which has a negative impact on the perceived value of VR [91]. Therefore, this study posits the following hypothesis:

H5-1: 

Perceived fee (H5-1a) and Complexity (H5-1b) will negatively influence the perceived value of tactile interactive advertising among consumers.

For NFT, past research illustrates that consumers with a high NFT will compensate for the lack of tactile information by focusing on price reductions when they are unable to touch the item or capture tactile information, and will also be more confident in their judgment of the product at the time of the promotion [109]. In addition, it has been shown that people with a higher aNFT have more positive attitudes toward saving money [110]. All of the above studies are suggesting that groups with a higher aNFT are more sensitive to the perception of money costing. On the other hand, according to Pino et al., even alternative touches can increase consumers’ expectations about the ease of use of product features [32]. In contrast, Lee et al. demonstrated that NFT also influences technology acceptance, as evidenced by the positive impact of aNFT on PEOU when consumers fully interact with the seller [29], and this is further corroborated by the fact that groups with a higher NFT have higher perceived ease-of-use, and correspondingly will have higher perceptions of technological complexity. Therefore, this study hypothesizes the following:

H5-2: 

Consumers’ level of NFT will be positively related to the perceived fee (H5-2a) and complexity (H5-2b) of tactile interactive advertising among consumers.

Current haptic technologies have the potential to enhance the sense of touch, but latency remains a prominent issue [39]. If the delay between the feedback and the interaction exceeds 1 ms, as is often the case with multiple sensory feedback like audio, video, and haptic, cybersickness can also occur [111]. Cybersickness presents a physical risk, and prior research on VR and 360-degree advertising has highlighted potential negative outcomes, such as eye strain, dizziness, and motion sickness [112,113]. Vishwakarma et al. defines physical risk and echoes the negative effects of this technology, and it also demonstrates that the perceived physical risk of using virtual reality technology has a negative impact on perceived value [91]. Hence, the hypothesis of this study is proposed as

H5-3: 

Perceived physical risk will negatively influence the perceived value of tactile interactive advertising among consumers.

Based on the proposed hypothesis, the conceptual model is illustrated in Figure 1.

4. Method

4.1. Sample and Data Collection

During the period from November 2022 to August 2023, we conducted purposive sampling and made an interview among consumers to investigate their perceptions and attitude toward tactile interaction and the form of tactile interactive advertising. The pre-test of the questionnaire was also delivered before large-scale distribution. The formal questionnaire survey was conducted from August 2023 to January 2024 via SurveyPlus and Credamo survey platforms and 576 questionnaires were collected totally.

In order to ensure that this study is not limited by geographical factors and to reach respondents more efficiency, especially those that are not active on the Internet, we conducted online and offline questionnaire distribution. Three channels were included: (1) offline on-site collection using SurveyPlus in Jinan, Shandong Province and Kunming, Yunnan Province, where 151 and 62 questionnaires were collected, respectively. (2) Online snowball sampling, where 213 questionnaires were collected by snowball sampling on the SurveyPlus platform; and (3) Online sample service, where 168 questionnaires were collected using Credamo platform in January 2024. After preliminary ad hoc analysis, there was no channel difference between online and offline.

In the screening session, this study removed invalid questionnaires based on the following criteria: (1) The questionnaire completion time was less than 5 min (300 s); (2) Setting a screening question “Please choose very dissatisfied” and choosing other options; (3) The answers to the questions were identical. Finally, 509 valid responses were retained after removing outliers and incomplete responses. Respondents’ characteristics are shown in

Table 2. Demographic characteristics of respondents (a).

Characteristics	N	%
Sex
Male	200	39.3
Female	309	60.7
Age
≤25	309	60.7
26–35	120	23.6
36–45	46	9.0
≥46	34	6.7
Education
Junior high school and below	6	1.2
High school/professional high school	26	5.1
Three-year college	36	7.1
Undergraduate	265	52.1
Postgraduate and above	176	34.6
Income/CNY
$\le 3000$	225	44.2
3001–5000	80	15.7
5001–7000	79	15.5
7001–9000	43	8.4
9001–10,000	28	5.5
≥ 10,001	54	10.6

and

Table 3. Demographic characteristics of respondents (b).

	M	SD
Consider purchasing the advertised product or service [114]	4.70	1.450
Openness to new technologies [115]	5.99	1.079

. The final sample consisted of 200 males and 309 females. A total of 429 people (94.3%) are less than 35 years old. Most respondents are well educated, with 441 people (86.7%) have at least completed an undergraduate degree if not more. The respondents are from 32 provinces (n = 507) in China and Australia (n = 2). Regarding the user’s habit of watching advertisements as shown in Figure 2, we measured the advertisement forms they have been exposed to, the frequency of watching advertisements and the time of each viewing, and the characteristics of the advertisement content, respectively. Among them, respondents experience more 2D screen ads (N = 440) and naked eye 3D ads (N = 325), which are more popular nowadays. In addition, 48.5% of users viewed ads 1–2 times per day, while 35.2% reached 3–10 times. Meanwhile, 42.2% of users accept advertising duration for 6–15 s each time, while 31.0% watched for 15–30 s each time. Regarding the characteristics of the ads, information (N = 269), entertainment (N = 206), and contextual overlap (N = 281) were more valued.

4.2. Instrument Procedure

Since haptic devices and tactile interactive advertisements have not yet been widely adopted, at the beginning of the questionnaire, an introduction video of the existing application related to haptic devices and tactile interactive advertising to stimulate the respondents’ awareness and subjective feelings about tactile interactive advertisements. The content of the video consists of two parts:

• The first part describes the haptic rendering principles of the haptic vest and haptic gloves, respectively. (1) The haptic vest video presents the haptic voice coil motors lined up on the haptic undershirt with fire ripples (presenting a sense of heat), wind ripples (presenting a sense of cold), and green waves (presenting a sense of impact). (2) The Haptic Glove introductory video tells viewers to use the Haptic Glove “if you can touch the objects you see in virtual reality” and shows the user switching between lifting with the gloves on and holding an apple in their bare hands. Later, it also introduced the haptic glove through the compressed air bit, using hundreds of tactile brakes to provide power, and this power directly touches the human skin to obtain realistic touch, but also introduced the haptic glove force feedback function. The images are presented with the user’s movements and the objects touched by the bare hand switching between the images.
• The second part describes the actual picture of the operator experiencing the Nissan concept car utilizing the Haptx gloves and VR equipment and the vision seen in the virtual space.

4.3. Measurement

After watching the video, we conducted the survey among these participants through a scale. This procedure is similar with the prior studies of travelers’ VR adoption intention [91] and the diffusion of VR glasses [112], where a few participants had tried VR devices and made their experiences uniform. Despite the fact that a proportion of respondents had not previously encountered tactile interactive advertising, the video employed in this study proved to be a persuasive medium. A 7-point Likert scale ranging from “Totally Disagree = 1” to “Totally Agree = 7” was used for type measurement. Each item of the measure was adapted or used based on an existing validated scale, as shown in

Table A1. Measurement items and their sources.

Constructs	NO.	Item	Source
aNFT	aNFT_1	When walking through stores, I can’t help touching all kinds of products.	[15]
	aNFT_2	Touching products can be fun.
	aNFT_3	I like to touch products even if I have no intention of buying them.
	aNFT_4	When browsing in stores, I like to touch lots of products.
iNFT	iNFT_1	I place more trust in products that can be touched before purchase.	[15]
	iNFT_2	I feel more comfortable purchasing a product after physically examining it.
	iNFT_3	If I can’t touch a product in the store, I am reluctant to purchase the product.
	iNFT_4	I feel more confident making a purchase after touching a product.
PE	PE_1	Applying tactile interaction into advertisement will be fun.	[22]
	PE_2	Applying tactile interaction into advertisement will make life enjoyable.
	PE_3	Applying tactile interaction into advertisement will improve my happiness.
	PE_4	Applying tactile advertisement will be enjoyable and satisfying.	[27]
PU	PU_1	Tactile interactive advertisement makes it easier for me to access product information.	[89]
	PU_2	Using tactile interactive advertisement would enhance my understanding of product.
	PU_3	Using tactile interactive advertisement would enhance my effectiveness.	[135]
	PU_4	I find tactile advertisement is useful.
TEL	TEL_1	I like I was in the world the tactile interactive advertisement created.	[90]
	TEL_2	I will forget about my immediate surroundings when I watch the tactile interactive advertisement.
	TEL_3	After watching the tactile interactive advertisement, I felt like I came back to the real world after a journey.
	TEL_4	If I were actually using tactile interactive advertisement, I would easily visualize what the actual production is like	[136]
COMP	COMP_1	I feel the procedure of using tactile interaction devices may not be intuitive to me.	[81]
	COMP_2	Learning to use tactile interaction devices would be difficult for me.	[89]
	COMP_3	I would find tactile interactive advertisement difficult to use.
	COMP_4	It would be difficult for me to become skillful at using tactile interaction devices to experience tactile interactive advertisement.
PF	PF_1	The price threshold required to experience tactile interactive advertisement is too high for me.	[104]
	PF_2	Compared to other existing advertisements, tactile interactive advertisement can bring me financial burden.
	PF_3	I am not satisfied with the cost of using tactile interactive advertisement.
PR	PR_1	I am concerned that using tactile interactive advertisement may cause nausea.	[137]
	PR_2	I am concerned that using tactile interactive advertisement may cause dizzy.
	PR_3	I am concerned that using the tactile interactive advertisement may lead to other uncomfortable physical side effects.
	PR_4	I am concerned that the use of tactile interactive advertisement may bring long-term chronic health risks.
PV	PV_1	Compared to the fee I need to pay, the use of tactile interactive advertisement is worthwhile to me.	[28]
	PV_2	Compared to the time I need to spend, the use of tactile interactive advertisement is worthwhile to me.
	PV_3	Overall, the use of tactile advertisement is worthwhile to me.
ATT	ATT_1	It make sense to use tactile interaction into advertising.	[138]
	ATT_2	The thought of using tactile interaction is appealing to me.
	ATT_3	I think the use of tactile interaction for advertisement is a wise choice.	[88]
	ATT_4	I will be satisfied with the application of tactile interaction in advertising.
IU	IU_1	I plan to use tactile interactive advertisement in the future.	[28]
	IU_2	I intend to use tactile interactive advertisement in the future.
	IU_3	I predict I would use tactile interactive advertisement in the future.	[89]
	IU_4	I intend to recommend the enjoyable tactile interactive advertisements I think to people around me.

with 42 question items. Because common method bias (CMB) can negatively affect the validity of the research results, we evaluated the data using Harman’s one-factor test. The results of the principal component factor analysis showed that the first (largest) of the seven factors with eigenvalues greater than 1 accounted for 35.212% of the total, and no factor accounted for more than 50% of the total, indicating that CMB was not a serious problem for the collected dataset.

4.4. Model Estimation

As shown in

Table 4. Results of the normality test.

Components	Kolmogorov–Smirnov a			Shapiro–Wilk
	Statistic	df	Sig.	Statistic	df	Sig.
aNFT	0.130	509	0.000	0.931	509	0.000
iNFT	0.118	509	0.000	0.919	509	0.000
PE	0.138	509	0.000	0.921	509	0.000
PU	0.155	509	0.000	0.906	509	0.000
PU	0.106	509	0.000	0.936	509	0.000
COMP	0.049	509	0.005	0.987	509	0.000
PF	0.132	509	0.000	0.932	509	0.000
PR	0.108	509	0.000	0.960	509	0.000
PV	0.107	509	0.000	0.964	509	0.000
ATT	0.104	509	0.000	0.944	509	0.000
IU	0.129	509	0.000	0.928	509	0.000

Notes: Lilliefors significance correction.

, we conducted a normality test using the Kolmogorov–Smirnov and Shapiro–Wilk tests, which showed that these samples did not conform to a normal distribution. The partial least squares’ structural equation model (PLS-SEM) does not assume that the sample data is normally distributed, and is capable of exploring the relationships between latent variables and identifying critical paths [116]. Meanwhile, for the estimation of formative indicators, PLS-SEM is also considered a more recommended method compared to CB-SEM [117]. Therefore, we used SmartPLS 3 to evaluate our proposed model using the PLS algorithm to test the measurement model with a maximum number of 300 iterations and a stopping criterion of ${10}^{-7}$. We set 5000 sub-samples and performed PLS bootstrapping, bias correction, and acceleration, and a two-tailed hypothesis test.

The formative second-order structure NFT consists of two indicators aNFT and iNFT as reflective first-order structures according to [15,118]. In this way, our measurement model contains a Type II (reflective–formative) second-order structure [119]. There are three main approaches, i.e., repeated indicator, two-stage, and hybrid approach [120]. For endogenous conformations, it is more appropriate to use two-stage [121]. We adopt the two-stage approach and the steps measuring the first-order reflective structures to obtain the latent variables scores of aNFT and iNFT, and then use the aNFT and iNFT scores in the first stage as explicit variables in the NFT measurement model. The measurement results are detailed in the next section.

5. Results

5.1. Measurement Model

This section presents the measurement model results for the first-order reflective constructs (PE, PU, TEL, PR, COMP, PF, PV, ATT, IU) as well as for the components of the NFT (aNFT as well as iNFT). We also tested the assessment of the second-order formative constructs.

In

Table 5. Cronbach’s $\alpha$, rho_A, Composite Reliability, AVE, and Fornell–Larcker criterion.

Constructs	$\mathit{\alpha }$	rho_A	CR	AVE	VIF	ATT	COMP	IU	PE	PF	PR	PU	PV	TEL	aNFT	iNFT
ATT	0.897	0.897	0.928	0.763	2.371~2.924	0.874
COMP	0.870	0.897	0.884	0.661	1.977~2.776	−0.154	0.813
IU	0.869	0.871	0.911	0.718	1.948~2.626	0.843	−0.134	0.848
PE	0.859	0.866	0.905	0.704	1.699~2.362	0.739	−0.103	0.712	0.839
PF	0.857	0.868	0.913	0.777	1.732~2.730	−0.122	0.611	−0.102	−0.017	0.881
PR	0.900	0.937	0.930	0.768	3.136~2.300	−0.122	0.603	−0.096	−0.096	0.653	0.876
PU	0.874	0.876	0.914	0.727	1.966~2.444	0.653	−0.069	0.662	0.754	0.080	−0.028	0.852
PV	0.846	0.851	0.906	0.763	1.936~2.168	0.764	−0.139	0.730	0.606	−0.216	−0.134	0.497	0.874
TEL	0.877	0.879	0.915	0.730	1.854~2.859	0.667	−0.141	0.718	0.665	−0.136	−0.042	0.601	0.631	0.855
aNFT	0.929	0.931	0.949	0.823	1.534	0.500	−0.028	0.460	0.438	0.012	−0.059	0.435	0.419	0.415	0.907
iNFT	0.796	0.852	0.869	0.632	1.534	0.512	0.060	0.521	0.537	0.170	0.078	0.613	0.434	0.436	0.590	0.795

, for the first-order constructs, we first measured reliability and convergent validity. The measured Cronbach’s alpha values ranging between 0.796 and 0.929, which is higher than the required critical value of 0.7. composite reliability values ranging between 0.859 and 0.929, and rho_A values ranging from 0.851 to 0.937. For convergent validity, factor loadings, as shown in the bolded portion of

Table A2. Cross-factor loading.

	ATT	COMP	IU	PE	PF	PR	PU	PV	TEL	aNFT	iNFT
ATT_1	0.856	−0.094	0.740	0.649	−0.055	−0.057	0.635	0.645	0.576	0.438	0.505
ATT_2	0.868	−0.096	0.778	0.682	−0.039	−0.069	0.593	0.656	0.598	0.418	0.464
ATT_3	0.885	−0.169	0.700	0.617	−0.205	−0.171	0.490	0.689	0.553	0.446	0.390
ATT_4	0.886	−0.182	0.728	0.634	−0.130	−0.131	0.563	0.681	0.602	0.445	0.429
COMP_1	−0.093	0.893	−0.075	−0.066	0.531	0.505	−0.036	−0.112	−0.081	−0.006	0.095
COMP_2	−0.076	0.643	−0.034	−0.028	0.374	0.448	−0.081	0.017	0.060	0.047	0.049
COMP_3	−0.181	0.915	−0.165	−0.110	0.577	0.564	−0.075	−0.148	−0.195	−0.051	0.027
COMP_4	−0.127	0.771	−0.094	−0.099	0.454	0.527	−0.097	−0.065	−0.021	0.009	0.013
IU_1	0.743	−0.117	0.855	0.603	−0.174	−0.085	0.512	0.684	0.656	0.361	0.385
IU_2	0.734	−0.177	0.879	0.593	−0.126	−0.105	0.539	0.655	0.657	0.376	0.431
IU_3	0.686	−0.084	0.833	0.602	−0.041	−0.047	0.623	0.559	0.561	0.420	0.494
IU_4	0.694	−0.072	0.822	0.617	0.004	−0.088	0.579	0.570	0.555	0.405	0.461
PE_1	0.560	−0.071	0.538	0.764	0.088	−0.003	0.708	0.390	0.474	0.339	0.481
PE_2	0.581	−0.046	0.604	0.865	−0.013	−0.048	0.604	0.503	0.561	0.325	0.421
PE_3	0.660	−0.093	0.612	0.857	−0.087	−0.116	0.554	0.571	0.595	0.385	0.424
PE_4	0.670	−0.127	0.630	0.867	−0.027	−0.140	0.678	0.553	0.591	0.414	0.480
PF_1	−0.028	0.569	−0.006	0.067	0.885	0.569	0.151	−0.115	−0.085	0.034	0.183
PF_2	−0.066	0.531	−0.061	0.005	0.894	0.552	0.091	−0.172	−0.129	0.016	0.160
PF_3	−0.203	0.519	−0.179	−0.094	0.864	0.598	−0.007	−0.263	−0.137	−0.011	0.114
PR_1	−0.112	0.556	−0.071	−0.111	0.586	0.856	−0.038	−0.106	−0.032	−0.051	0.057
PR_2	−0.136	0.548	−0.090	−0.104	0.585	0.923	−0.020	−0.150	−0.061	−0.044	0.090
PR_3	−0.078	0.503	−0.093	−0.062	0.571	0.883	−0.009	−0.097	−0.024	−0.053	0.078
PR_4	−0.090	0.505	−0.084	−0.052	0.552	0.842	−0.031	−0.104	−0.020	−0.063	0.041
PU_1	0.592	−0.072	0.582	0.642	0.026	−0.053	0.867	0.457	0.550	0.389	0.542
PU_2	0.538	−0.067	0.575	0.654	0.058	−0.038	0.874	0.439	0.509	0.349	0.511
PU_3	0.564	−0.062	0.548	0.674	0.067	−0.011	0.842	0.408	0.528	0.386	0.523
PU_4	0.532	−0.034	0.553	0.601	0.128	0.010	0.825	0.388	0.457	0.359	0.511
PV_1	0.589	−0.088	0.580	0.455	−0.237	−0.113	0.330	0.860	0.535	0.301	0.307
PV_2	0.673	−0.146	0.628	0.524	−0.188	−0.129	0.424	0.883	0.525	0.368	0.404
PV_3	0.729	−0.127	0.697	0.598	−0.149	−0.110	0.531	0.878	0.591	0.419	0.417
TEL_1	0.636	−0.106	0.684	0.690	−0.030	−0.010	0.643	0.541	0.835	0.390	0.446
TEL_2	0.519	−0.088	0.575	0.519	−0.083	0.026	0.496	0.477	0.863	0.335	0.387
TEL_3	0.558	−0.140	0.597	0.543	−0.123	−0.021	0.479	0.540	0.889	0.338	0.343
TEL_4	0.559	−0.142	0.592	0.516	−0.215	−0.124	0.438	0.586	0.831	0.352	0.319
aNFT_1	0.439	−0.006	0.420	0.390	0.012	−0.062	0.372	0.402	0.400	0.909	0.528
aNFT_2	0.481	−0.032	0.447	0.431	0.041	−0.047	0.430	0.387	0.376	0.914	0.571
aNFT_3	0.440	0.012	0.387	0.366	0.010	−0.014	0.370	0.383	0.347	0.911	0.500
aNFT_4	0.451	−0.072	0.409	0.399	−0.020	−0.088	0.402	0.350	0.382	0.897	0.538
iNFT_1	0.451	0.004	0.447	0.453	0.125	0.007	0.533	0.399	0.376	0.547	0.850
iNFT_2	0.449	0.045	0.466	0.496	0.198	0.101	0.556	0.340	0.366	0.524	0.881
iNFT_3	0.295	0.116	0.266	0.239	0.028	0.099	0.228	0.288	0.317	0.346	0.515
iNFT_4	0.419	0.063	0.444	0.467	0.146	0.062	0.550	0.362	0.351	0.447	0.874

, was as low as 0.515 and as high as 0.923, which are both higher than the criterion value of 0.5, as mentioned in [116], and therefore, we retained all the question items. In addition, the AVE values ranged from 0.632 to 0.823, which is higher than 0.50.

For the measurement of discriminant validity, three tests were performed, including the Fornell–Larcker criterion, cross-loading, and Heterotrait–Monotrait (HTMT) ratio [122], and all three methods were used in this study. We can see that, as for the Fornell-Larcker criterion, the data in Table 5 demonstrate that the AVE square root exceeds the correlation coefficients for each latent variable. For cross-loading, as shown in Table A2, all indicator loading exceeded the cross loading. For the HTMT ratio in

Table 6. Heterotrait-monotrait (HTMT) ratio.

	ATT	COMP	IU	PE	PF	PR	PU	PV	TEL	aNFT	iNFT
ATT
COMP	0.159
IU	0.954	0.126
PE	0.839	0.114	0.823
PF	0.153	0.662	0.145	0.1
PR	0.132	0.682	0.108	0.108	0.741
PU	0.737	0.097	0.762	0.875	0.121	0.041
PV	0.872	0.123	0.845	0.7	0.247	0.149	0.569
TEL	0.749	0.131	0.819	0.762	0.155	0.061	0.685	0.727
aNFT	0.547	0.051	0.512	0.487	0.031	0.065	0.481	0.468	0.458
iNFT	0.606	0.102	0.621	0.636	0.194	0.102	0.709	0.534	0.536	0.686

, it proved that all the constructs were less than the critical value of 0.9, except for ATT and IU which had a value of 0.954. However, according to [122], even an HTMT greater than 1.0 does not quite indicate a lack of discriminant validity. Thus, the discriminant validity of our model is satisfied.

Thus, it can be said that our measurements of first-order structure are adequate. After obtaining the latent variables scores for aNFT and iNFT and as the explicit variables of NFT, we evaluate the second-order formative structure.

For the second-order formative structure in

Table 7. NFT measurement model assessment.

Second-Order Construct	First-Order Construct	Weight	T-Statistic	p-Values	VIF
NFT	aNFT	0.276	3.066	0.002	1.534
	iNFT	0.812	11.776	0.000	1.534

, the weight and T-statistic, as well as the VIF were measured, which shows that the weights of aNFT and iNFT were 0.276 (T-Statistic = 3.066, p < 0.01) and 0.812 (T-Statistic = 11.776, p < 0.001), respectively, which were significant. Also, the VIF of each item were less than 5 [123]. So, there is no significant multicollinearity. Combining the above analysis, the second-order structure was established, so hypotheses H2-1a and H2-1b were supported.

5.2. Structural Model

Table 8. Path coefficients and hypothesis testing.

Hypothesis	Relationship	Path Coefficients ($\mathit{\beta }$)	T-Statistic	p-Values	${\mathit{f}}^2$	Result
H1	ATT→IU	0.636 ***	12.646	0.000	0.564	Supported
H2-2	NFT→IU	0.107 **	2.840	0.005	0.030	Supported
H3-1	PV→ATT	0.764 ***	27.992	0.000	1.402	Supported
H3-2	PV→IU	0.194 ***	4.045	0.000	0.059	Supported
H4-1a	PE→PV	0.330 ***	4.955	0.000	0.076	Supported
H4-1b	PU→PV	0.055	0.939	0.348	0.002	Rejected
H4-2a	NFT→PE	0.557 ***	11.526	0.000	0.449	Supported
H4-2b	NFT→PU	0.618 ***	15.026	0.000	0.616	Supported
H4-3	TEL→PV	0.359 ***	6.338	0.000	0.125	Supported
H5-1a	PF→PV	−0.225 ***	3.778	0.000	0.046	Supported
H5-1b	COMP→ PV	0.081	1.480	0.139	0.007	Rejected
H5-2a	NFT→PF	0.140 *	2.199	0.028	0.020	Supported
H5-2b	NFT→ COMP	0.039	0.568	0.570	0.002	Rejected
H5-3	PR→PV	0.012	0.238	0.812	0.000	Rejected

Notes: * p < 0.05; ** p < 0.01; *** p < 0.001.

reports the results of the path coefficient ($\beta$), p-value, T-statistic, $f^2$, and hypothesis testing for the structural model. The hypothesis testing results are shown in Figure 3. Ten of the fourteen hypotheses were confirmed to be significant. The ATT → IU ($\beta$ = 0.636, T-Statistic = 12.646, p < 0.001), PV→IU ($\beta$ = 0.194, T-Statistic = 4.045, p < 0.001) were confirmed to be strongly positively correlated with each other and the results supported H1 and H3-1. NFT was similarly positively correlated with IU ($\beta$ = 0.107, T-Statistic = 2.840, p < 0.01) and H2-2 was confirmed. A strong positive correlation was also shown between PV and ATT ($\beta$ = 0.764, T-Statistic = 27.992, p < 0.001), and H3-2 was established.

For the hypotheses of the effect of NFT on the four variables, except for NFT → COMP which was not significant, NFT → PE ($\beta$ = 0.557, T-Statistic = 11.526, p < 0.001), NFT → PU ($\beta$ = 0.618, T-Statistic = 15.026, p < 0.001) and NFT→PF ($\beta$ = 0.140, T-Statistic = 2.199, p < 0.05) were found to be significant, thus the results support H4-2a, H4-2b, and H5-2a.

Among the six variables of Benefit and Sacrifice with PV, PE ($\beta$ = 0.330, T-Statistic = 4.955, p < 0.001), and TEL ($\beta$ = 0.359, T-Statistic = 6.338, p < 0.001) positively affected PV, which was supported for H4-1a and H4-3. While PF ($\beta$ = −0.225, T-Statistic = 3.778, p < 0.001) is negatively correlated with PV, COMP and PR in Sacrifice have an insignificant relationship with PV and H5-1a, H5-1b, and H5-3 were rejected.

We used $R^2$, $Q^2$ to evaluate the predictability as shown in

Table 9. Path coefficients and hypothesis testing.

Constructs	${\mathit{R}}^2$	${\mathit{R}}^2$ Adjusted	${\mathit{Q}}^2$
ATT	0.584	0.583	0.442
COMP	0.002	0.000	0.001
IU	0.737	0.735	0.524
PE	0.310	0.309	0.215
PF	0.019	0.018	0.012
PU	0.381	0.380	0.274
PV	0.491	0.485	0.364

. According to [124] and [125], predictive relevance is present if $R^2$ is greater than 0, whereas strong, moderate, and weak prediction are distinguished by three thresholds of 0.75, 0.50, and 0.25. In addition, $Q^2$ greater than 0 is indicative of predictive relevance. As shown in Table 9, ATT ($R^2$ = 0.583), PV ($R^2$ = 0.485), and IU ($R^2$ = 0.735) have a moderate level of predictive accuracy, while $Q^2$ is greater than 0, which represents that the model has good predictive ability.

6. Discussion

6.1. Generate Discussion

The primary objective of this study is to reveal the adoption intention towards tactile interactive advertising from the consumer perspective. In doing so, we utilized the VAM as a research model and provided explanations for RQ1 and RQ2 from the consumer side as well as the technological characteristics of tactile interaction advertising.

The findings confirm that NFT has a direct positive effect on PE and PU in the benefit dimension. In previous studies, Lee et al. have confirmed the positive effect of aNFT on PE, but iNFT negatively affected PE. Because pragmatic and goal-oriented iNFT can dissipate fun due to the inability to perceive tactile sensations in the network [29]. In contrast, Gatter et al. confirmed that aNFT has a positive effect on PU due to the fact that it does not provide a haptic sensation from the point of view of AR functioning, whereas people with higher iNFT tend to believe that only a real touch can help them assess usefulness [25]. In this study, the reason why NFT as a second-order variable has a positive effect on both PE and PU is because haptic technology could compensate for the real sense of touch and complement the tactile needs of people with high iNFT. In addition, NFT had a significant positive effect on PF in the sacrifice dimension, which is the same as [110]. However, for hypothesis H5-2b, the effect of NFT on COMP was not significant, contradicting the results of [29]. As Gatter stated, AR does not provide haptics, and for high NFT consumers, it is still only possible to facilitate imagination through realistic visualization. Therefore, the usage of tactile interactive advertising caters to people with high NFT, where people are immersed in what amounts to live touch, and then ignores the operational complexity of tactile application [25]. Rodrigues et al. reported that people with a higher NFT are more reluctant to shop online shopping due to the lack of tactile information [99]. On the other hand, Kuhn et al. demonstrated that direct touch interfaces can reduce the negative impact of NFT [97]. Tactile interactive advertising utilizes haptic technology as a direct contact interface, resulting in higher usage intention for consumers with a higher NFT. Overall, the above discussion explains RQ1.

For RQ2, the results of the study demonstrate that the perceived value (PV) has a direct and significant effect on the intention to use (IU), i.e., the tactile interactive advertising, thereby validating the Kim’s prediction of value structure on adoption intention [28]. After introducing attitude (ATT), it shows that PV has a positive effect on ATT and ATT also has a positive effect on IU, which is consistent with [103]. Perceived enjoyment (PE) and telepresence (TEL), as part of the benefit factor, have a positive impact on PV, while perceived usefulness (PU) has no significant effect on PV (H4-1b reject). In previous research on VAM, there was a significant positive correlation between PU and PV, except for Yoon’s research which concluded that PU was not significant for PV in the study of AR-based mobile applications [107]. This study reached the same results as [107]. The reason is that tactile interactive advertising, such as AR applications, will be more commonly applied to hedonic usage scenarios. In the descriptive analysis of this study, entertainment is one of the advertising features that people pay more attention to, which also confirms the above view.

As a part of the sacrifice factor, the perceived fee (PF) has a significant negative effect on PV, which is consistent with the conclusions in [28]. Geng et al. proposed that the price barriers of VR devices hinder the acceptance of VR advertising compared with smartphones [126]. Bian et al. revealed that the product price would have a negative effect on the value cognition of high-tech products. At this stage, haptic technology has yet to be popularized and has reached more consumers due to the demand for hardware equipment. The costs and fees for those haptic devices would be attracted, thus reducing the perceived value for this advertising form [127]. We also found that complexity (COMP) had no significant impact on PV, which led to the rejection of the hypothesis H5-1b, while Vishwakarma noted that COMP can have a significant negative effect on PV in VR technology use [91]. This is inconsistent with previous research results. The reasons include the fact that tactile interactive advertising can provide people with an immersive experience. In this process, consumers mainly focus on the enjoyment rather than what effort is put into the experience, which is confirmed by Yersuren et al. in the research into VR [128]. Meanwhile, physical risk (PR) has a nonsignificant effect on PV (H5-3 reject), similarly contradicting the results of [91]. For the perception of PR, it is influenced by other factors such as enjoyment [129], i.e., if the experience of tactile interaction is related to enjoyment, the user is more willing to experience a certain level of discomfort symptoms, in addition to the symptoms of physical discomfort that may be caused by the technology itself. Another reason for COMP and PR being insignificant to PV is that the age of the respondents collected for this study is generally concentrated around 25 years old, an age group that happens to belong to Generation Z. The age of the respondents is not significant to PV, but rather to COMP and PR. Since Generation Z has been exposed to technology from an early age [130], the perception of the complexity and risk of using new technologies is correspondingly low. This finding is also reflected in the study by Rafdinal et al. [131].

Empowered by new technologies, there are two major trends: shifting from sensory integration to experience integration, and one-way communication to interactive communication [132]. In the past, traditional advertising used words, images, videos, symbols, etc., to realize brand value shaping and value symbol abstraction. New technology advertisements (e.g., holographic and VR advertising) will pursue the scene and experience, and bring more diversification through immersive and interactive scenes [84,133]. The usage of new technology advertisements enables users to actively engage with products and services, which enhances their willingness to purchase [133].

6.2. Implications

This study explores consumer intentions to use tactile interactive advertising and fills a research gap in the context of user perceptions of tactile interactive technologies. The theoretical and practical implications are summarized as follows:

For the theoretical implications, we applied modified VAM to explore the usage intention of tactile interactive advertising among consumers. The effects of benefit and sacrifice factors on perceived value are based on the perceived technological characteristics, and their relationships are verified. There is no further theoretical validation of how haptic technology affects the user experience, although a preliminary study developed the haptic experience (HX) model and reported on the various elements of measuring haptic experience [86,134]. In our research, the effects between perceived value and attitude or intention to use are further investigated, and the user acceptance of tactile interactive technology as a new advertising medium is measured. This study also extends the original VAM based on haptic technology features to include two new structures, TEL and PR, and confirms the effectiveness of TEL, which can be considered another extension of the VAM in the application domain of tactile interactive advertising. These findings provide a basis understanding of consumers’ perceptions and acceptance for this emerging advertising form. For the NFT aspect, we construct the second-order formative structure including the first-order reflective structures of aNFT and iNFT. On the one hand, NFT was analyzed as a higher-order variable, which was ignored by past studies that analyzed aNFT and iNFT separately. It has been confirmed that NFT promotes the perception of fun and usefulness and can meanwhile enhance consumers’ willingness to use tactile interactive advertising. The role of NFT on the perceived cost in tactile interactive advertising was also found, i.e., consumers with a higher NFT are more dissatisfied with the cost of tactile interactive advertising. On the other hand, the role of NFT on VAM was further confirmed, in addition to the effect of NFT on TAM [29].

In terms of practical implications, our findings make a strong case for the need to apply and promote tactile interactive advertising in the future, echoing the recommendation that advertisers should implement tactile experiences [13]. Because consumers with a higher NFT are more willing to use tactile interactive advertising, a sense of enjoyment and presence promotes the perceived value even more. In addition, the negative impact of PF on PV suggests that the development of tactile interactive advertising needs to reduce the cost paying for extra devices and interfaces. As an evolution of the haptic interaction technology, there are a variety of information interfaces including encountered-type haptic displays [36], Mid-Air Ultrasound Haptics [5], and Electrotactile Feedback [37], it would be better for advertisers to combine mid-air haptics and display techniques to increase telepresence in the upcoming tactile interactive advertising. The specific types of tactile experiences need to be generated for commodities, which will reduce the aesthetic fatigue among consumers. In terms of advertising content, advertisers need to combine the characteristics of the product and consumer demand to create appropriate advertising scenes. For example, for fabrics (clothing, sofas, etc.), the texture of different fabrics needs to be emphasized, whereas for products with complex functions (e.g., automobiles, home appliances, etc.), the regulation of force feedback should be emphasized and combined with the perception of temperature as well as size and shape. Our findings also provide a direction for future efforts in the development of tactile interactive advertising. For the technology company, it is necessary to enhance the render effect of the sense of touch, while emphasizing the consistency between the haptic sense and the other senses, which will bring better and more realistic haptic experiences for consumers and enhance their telepresence. In terms of the perceived fee, a more feasible leasing model of haptic systems should be considered to reduce the device costs. In addition, to make it accessible to more consumers, the e-commerce companies need to develop suitable tactile interfaces to show the features of products and accelerate the popularization of tactile devices. Undoubtedly, the recovery of haptics restored in the virtual consumption scenario should also be realistic, i.e., close to the touch of real products, which may be the key to make sure that people accept tactile interactive advertising.

7. Conclusions

In this paper, we investigated the perceived value and acceptance to use tactile interactive advertising among consumers with different needs for touch (NFT). On the one hand, we can see that NFT has direct effects on perceived enjoyment (PE), perceived usefulness (PU), and perceived fee (PF) in the benefit and sacrifice dimension as well as the intention to use (IU). The results show that consumers with higher NFT positively impact the PE, PU, and PF of tactile interactive advertising and have a stronger usage intention, which confirms that the consumers’ NFT actually affects their perception and intention to use tactile interactive advertising. Research question 1 can be answered. On the other hand, this study has revealed that the influence of PE, telepresence (TEL), and PF on perceived value (PV), which has no significant relationship with perceived usefulness, physical risk, and complexity. The perception of enjoyment and telepresence can drive the perceived value of tactile interactive advertising, but the high price can dissipate the adoption. PV and attitudes towards tactile interactive advertising significantly impact IU. In this way, the perceived characteristics of tactile interactive advertising that affect consumers’ PV and IU have been revealed. Research question 2 can be answered.

The limitations of this research can be summarized as follows: First, the participants in this survey were mainly from China and the results are not certain to be generalizable to other countries. Second, this study was based on past literature and did not correlate NFT with TEL and PR, and subsequent studies may consider examining the relationships of TEL and PR. Third, the majority of participants were young.

In future work, first, a follow-up study could be conducted to make a side-by-side comparison among different countries, which will reveal the consumers’ adoption intention of tactile interactive advertising from a cross-cultural perspective. Concrete marketing planning will be designed by multinational brands or their advertisers in marketing campaigns. Second, considering the operational complexity of this emerging form of advertising, the middle-aged and elderly groups should be further explored. Compared with the younger generation, the digital divide affects their perceived value and acceptance of tactile interactive advertising. Third, the concrete brand or advertising of products should be focused on. Then, we can design inter-group or intra-group experiments to evaluate the advertising effect. In this way, some practical advice for the niche market will be obtained through the advertising cases.