The impact of trust building mechanisms on purchase intention towards metaverse shopping the moderating role of age

International Journal of Human–Computer Interaction

ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/hihc20

The Impact of Trust-Building Mechanisms on
Purchase Intention towards Metaverse Shopping:
The Moderating Role of Age

Lin Zhang, Muhammad Adeel Anjum & Yanqing Wang

To cite this article: Lin Zhang, Muhammad Adeel Anjum & Yanqing Wang (10 Mar 2023): The
Impact of Trust-Building Mechanisms on Purchase Intention towards Metaverse Shopping:
The Moderating Role of Age, International Journal of Human–Computer Interaction, DOI:
10.1080/10447318.2023.2184594
To link to this article: />
Published online: 10 Mar 2023.
Submit your article to this journal
Article views: 1488
View related articles
View Crossmark data
Citing articles: 4 View citing articles

Full Terms & Conditions of access and use can be found at
/>
INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION
/>
The Impact of Trust-Building Mechanisms on Purchase Intention towards
Metaverse Shopping: The Moderating Role of Age

Lin Zhanga, Muhammad Adeel Anjumb, and Yanqing Wanga

aSchool of Management, Harbin Institute of Technology, Harbin, China; bBalochistan University of Information Technology, Engineering and
Management Sciences (BUITEMS), Quetta, Pakistan

ABSTRACT

Given the uncertainty of online transactions in metaverse shopping, the digital economy encour-
ages building a trustworthy virtual environment. Based on media richness theory, this article
examines how the perceived media richness of the metaverse helps engender multidimensional
trust (i.e., cognitive trust and affective trust) and leads to purchase intention in the context of
metaverse shopping. The proposed model is tested based on survey data from 332 consumers on
an online scenario-based platform pertaining to metaverse initiatives. Structural equation model-
ing is used to examine the proposed research model. The empirical research findings show that
the perceived media richness of the metaverse builds cognitive trust and affective trust, which in
turn affects purchase intention towards metaverse shopping. Furthermore, we classify consumers
into digital natives (DNs) and digital immigrants (DIs) based on chronological age and examine
the different influences of the two dimensions of trust on purchase intention towards metaverse
shopping between the two groups. We identify and address several knowledge gaps in the extant
trust literature. We also discuss the theoretical and managerial implications and propose several
suggestions for future research.

1. Introduction respondents distrust virtual shopping activities in the meta-
verse (Berthiaume, 2022). Therefore, trust in the metaverse
Nowadays, the rapid technological development of artificial is identified as a salient factor for alleviating uncertainty
intelligence (AI), virtual reality (VR), augmented reality (AR), when consumers make purchase decisions in the metaverse
etc., has spawned a plethora of online shopping platforms shopping context. There is a call for more empirical studies
that have evolved from a static webpage into a more dynamic to uncover individuals’ trust formation process and the con-
three-dimensional (3D) space (Darbinyan, 2022). A new era, text-specific antecedents in this emerging research area.
termed “metaverse shopping,” has emerged in which consum-
ers experience immersive shopping with virtual avatars and The majority of the existing literature has concentrated
engage in interactive shopping activities by guiding their ava- on trust in different contexts, such as online commerce

tars through 3D stores. According to an Accenture report, (Cheng et al., 2017; N. Wang et al., 2013; W. Wang et al.,
94% of retail executives believe that the future digital econ- 2016) and the sharing economy (Shao, Zhang, Li, et al.,
omy needs to offer metaverse initiatives (Standish, 2022). For 2022; Shao & Yin, 2019). From a theoretical perspective,
instance, enterprises such as Facebook recognize the potential previous research has commonly emphasized two main
of the metaverse and are starting to activate metaverse- research foci regarding trust-related behaviors: the determi-
enabled social commerce on their platforms (Nix, 2022). A nants of trust (e.g., Cheng et al., 2021; Shao et al., 2019);
Chinese leading electronic commerce (e-commerce) platform, and the dimensions of trust (e.g., Chi et al., 2021; Shao,
Taobao, has also invested significant marketing efforts in Zhang, Brown, et al., 2022). Regarding the determinants of
metaverse shopping (Ryder, 2022). Virtual shopping in the trust, most studies have focused on exploring trust antece-
metaverse is expected to have an US$800 billion market dents that lead people to have different tendencies toward
opportunity by 2024 (Darbinyan, 2022). trust-related behaviors (Cheng et al., 2021). For example,
Cheng et al. (2017) investigated the joint knowledge-based,
Despite the fact that metaverse shopping has become a institution-based, calculative-based, cognition-based, and
significant trend, the interconnected nature of the metaverse personality-based trust antecedents in influencing social
heightens the related risks for security and privacy, which media communication behaviors. Shao and Yin (2019)
potentially leads to distrust issues among consumers regard- found that context-specific platform institutional mecha-
ing making purchase decisions in the metaverse shopping nisms have positive effects on trust in the ridesharing plat-
context (Di Pietro & Cresci, 2021). According to a report form, which in turn affects continuance intention. Regarding
from the consumer insights data platform, Zappi, 80% of

CONTACT Yanqing Wang School of Management, Harbin Institute of Technology, Harbin, China
ß 2023 Taylor & Francis Group, LLC

2 L. ZHANG ET AL.

the dimensions of trust, most studies have investigated multi- natives (DNs) and digital immigrants (DIs), which is applied
faceted trust concepts, such as competence, benevolence, and to explain individuals’ differentiated attitudes and technol-
integrity (McKnight et al., 2002). For instance, Pavlou (2002) ogy adoptions (Shao, Benitez, et al., 2022). DIs are used to
found that institution-based antecedents help engender two traditional communication mediums (e.g., email or social
specific trust dimensions (cognition-based credibility and ben- media) as the preferred social tools for the workplace or

evolence) and indirectly influence transaction success in daily life, whereas DNs prefer to interact with one another
online marketplaces. Some scholars have also defined trust as via the interactive virtual world, such as Second Life (Hong
a multidimensional construct comprising cognitive and affect- et al., 2013; L. Zhang et al., 2021). Moreover, DNs grew up
ive components (Cummings & Bromiley, 1996) and examined with digital technologies in a networked world and are more
the two dimensions of cognitive and affective trust in e-com- comfortable adopting innovative technologies than their DI
merce (Leong et al., 2021). counterparts (Kesharwani, 2020). Hence, attitudes toward
technology may vary largely depending on one’s age. While
We identify several knowledge gaps in the extant litera- the role of age (DNs vs. DIs) has been recognized in various
ture regarding trust literature. First, most studies have exam- contexts (Niehaves & Plattfaut, 2014; Ollier-Malaterre &
ined personality and institutional factors as the determinants Foucreault, 2021; Shao, Benitez, et al., 2022), there is scant
of trust (J. Liang et al., 2022; Shao et al., 2019), where the literature that theorizes age-difference issues regarding the
trustee is either a real human or an entity (e.g., a platform). relationships between multidimensional trust and purchase
More recent IS research has called for the characteristics of intention in the context of metaverse shopping.
human–system interactions to be examined where the
trustee is a technological environment, such as an AI chat To fill these research gaps, this study aims to develop a
channel or a blockchain-enabled mutual aid environment theoretical model to comprehensively understand how the
(Bao et al., 2021; Chi et al., 2021; Choung et al., 2022; Shao, perceived media richness of the metaverse helps engender
Zhang, Brown, et al., 2022). Specifically, many features of multidimensional trust (i.e., cognitive trust and affective
the metaverse work towards visualization and enhancing trust) and indirectly influences purchase intentions in the
natural communication through a virtual technological metaverse shopping context. Current literature has used
environment. Considering that the rich media-enabled vir- media richness theory (MRT) to explain how digital media’s
tual world is created to interact with products, brands, and ability to convey channel richness can affect purchase deci-
communities in metaverse shopping service delivery (Kim, sions and outcomes (Tseng & Wei, 2020; Zhu et al., 2010),
2021), consumers may have a high perception of media rich- which is considered an appropriate theoretical framework
ness, which is beneficial to triggering trust in the metaverse. for our study. Meanwhile, we further incorporate age as a
However, to the best of our knowledge, few studies have so salient contingency factor in the theoretical model to explore
far investigated the role of the media richness of the meta- the specific differences between DNs and DIs in terms of
verse in building trust beliefs and subsequently facilitating trust-related behaviors. To this end, this study aims to shed
purchase intention. light on the role and nature of multidimensional trust in the
metaverse shopping context by providing theoretical insights

Second, prior studies have mainly focused on the dimen- and empirical findings for the following research questions:
sion of cognitive trust related to competence and reliability
(Shao et al., 2019; Shao, Zhang, Li, et al., 2022), while affect- 1. How does the perceived media richness of the metaverse
ive trust, in the context of the new generation of technolo- affect consumers’ purchase intention through the medi-
gies, is yet to be advanced in the online shopping field. In ation effects of multidimensional trust (cognitive and
contrast to cognitive trust, affective trust is rooted in emo- affective trust)?
tional bonds and connections (Cummings & Bromiley,
1996). Considering the uncertainty and potential risks of the 2. How does age moderate the relationships between multi-
metaverse (Kim, 2021), consumers may make a rational dimensional trust and purchase intention in the meta-
assessment regarding the security and reliability of the meta- verse shopping context?
verse and form the cognitive trust belief when making pur-
chase decisions. On the other hand, consumers may also The remainder of this study is structured as follows.
generate the affective trust belief towards the metaverse, Section 2 reviews the related literature and presents the the-
resulting from the virtual world’s immersive, entertaining, oretical foundation. Section 3 develops the research model
and interactive shopping experience. As a result, it remains and formulates the hypotheses. Section 4 describes the
unclear whether or not the perceived media richness of the research method and data analysis. Section 5 concludes the
metaverse will exert different effects on cognition-based and article, presenting the key findings, implications, limitations,
affect-based trust in the metaverse shopping context. and future research directions.

Third, we consider individual characteristics contingent 2. Research background and theoretical foundations
on trust-related behaviors (Shao et al., 2019; Shao, Zhang,
Li, et al., 2022). Prior literature has indicated that human 2.1. Literature review in the field of the metaverse
perceptions and behavioral outcomes in technology imple-
mentations are contingent on age (e.g., Ghobadi & The term metaverse originated in the 1992 science fiction
Mathiassen, 2020; Hong et al., 2013). In particular, age is novel Snow Crash and gained global popularity in recent
one of the criteria used to differentiate between digital

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION 3

years (Oh et al., 2023). Based on the recent research work to walk and shop in the 3D shopping environment and to

(Hennig-Thurau et al., 2022; Oh et al., 2023), metaverse can complete specific interactive actions. Despite the more uni-
be conceptualized as a new computer-mediated environment fied experience (i.e., a blended virtual and physical experi-
that consists of virtual worlds in which people can act and ence) that metaverse shopping offers, to the best of our
communicate with each other via avatars. In past studies, knowledge, few studies have focused on the impact of media
the impacts of the metaverse have been examined from dif- richness created by the metaverse environment on consum-
ferent perspectives. For example, some studies have focused ers’ trust-building process and related purchase outcomes.
on the metaverse’s technological infrastructures, finding that In order to address this research gap, our study aims to
a heterogeneous crowd of technological capabilities and uncover trust-building mechanisms and subsequent purchase
tools can afford or constrain the potential action spaces decision-making behaviors towards metaverse shopping
available within the metaverse environments (Grupac et al., from the theoretical perspective of MRT, which will be
2022; Grupac & Lazaroiu, 2022; Hudson, 2022; Jenkins, described in the next section.
2022; Kliestik et al., 2022; Zvarikova et al., 2022). Another
study indicated that the implementation of 3D space com- 2.3. Media richness theory (MRT)
puter-generated simulations, data-driven artificial intelli-
gence, and text-to-image synthesis models is beneficial for Originating from computer science, MRT is defined as the
meeting customer dynamic demands in the metaverse envir- ability of a computer-mediated communication channel to
onment (Nica et al., 2022). Apart from examining the meta- deliver rich information and messages (Cummings &
verse’s technological infrastructures, attention has also been Bromiley, 1996). According to Daft and Lengel (1986), the
paid to the impacts of user perceptions and experiences (Oh theoretical concept of media richness can be conceptualized
et al., 2023; Shin, 2022; Wongkitrungrueng & Suprawan, as a set of objective features, including multiple information
2023; Xi et al., 2022). Further, Hennig-Thurau et al. (2022) cues, language variety, immediate feedback, and personal
found the metaverse-enabled environment to be positively focus. Specifically, multiple information cues mean that indi-
related to interaction performance outcomes through the viduals can gather more information in a variety of ways
mediation effects of psycho-physiological mechanisms. To (e.g., texts, images, videos, etc.) with the support of digital
extend the existing research, recent studies have emphasized technology. Language variety refers to the ability of digital
to comprehend how trust can be developed among users in technology to support individuals in communicating with
the metaverse context (Tan & Saraniemi, 2022) because the natural languages (e.g., language symbols, emoticons, etc.).
metaverse may expose user identifications to the service pro- Immediate feedback refers to the ability of digital technology
viders and cause privacy concerns (Dwivedi et al., 2022). to receive and send feedback quickly. Personal focus is the
However, scant attention has been paid to the antecedents degree to which individuals can customize messages or per-

of trust-building in facilitating users’ behavioral intention in sonal profiles according to their personal needs and the sur-
the metaverse shopping context. rounding environment.

2.2. Metaverse shopping As presented in Table 1, the earliest application of MRT
in enterprise organizations (Dennis & Kinney, 1998; Kishi,
The metaverse has become a new economic paradigm that 2008; Suh, 1999; Yoo & Alavi, 2001) explains how organiza-
builds on sharing an interactive and immersive virtual world tions can meet individual demands by reducing information
environment (Kim, 2021). Multifarious applications for the complexity and fuzziness. In recent years, MRT has been
metaverse have gained considerable attention in different introduced and widely applied in e-commerce research (D.
fields, such as improving work productivity (Xi et al., 2022), K. L. Lee & Borah, 2020; Li et al., 2022; Mirzaei &
social media value creation (Kraus et al., 2022), interactive Esmaeilzadeh, 2021; Shen et al., 2021). For example, Tseng
learning environments (Rospigliosi, 2022), and advertising and Wei (2020) showed the impact of mobile advertising
strategy (Kim, 2021; Taylor, 2022). Recently, the metaverse with various degrees of media richness on consumer deci-
has been introduced as a new business model to facilitate sion-making behavior. Zhu et al. (2010) found the positive
the immersive shopping experience among peers online influence of navigation support and communication support
(Grupac et al., 2022; Hudson, 2022; Jenkins, 2022; Zvarikova on collaborative online shopping behaviors. Despite great
et al., 2022), and many famous platforms have rapidly attention having been paid to online marketplaces, MRT
invested in and developed metaverse shopping initiatives was originally used only to examine traditional media tech-
(Nix, 2022; Ryder, 2022). nologies (e.g., telephone, electronic documents, email, etc.);
however, a few attempts have expanded its use to the meta-
Compared with traditional layouts of shopping platforms verse context. Boughzala et al. (2012) argued that the meta-
(e.g., plain text, images, or video), a key feature of the meta- verse is a special case of a socially interactive media channel;
verse is that it allows customers to immerse themselves sur- thus, MRT is suitable for gaining a rich understanding of
rounding a virtual shopping world, with the support of rich context-specific purchase behaviors in the emerging meta-
digital content (Kim, 2021). Specifically, in the metaverse verse realm.
shopping world, consumers are immersed in a media-rich-
ness-enabled virtual world where they can control their In particular, drawing upon MRT literature, there are two
movement through smartphones, guiding virtual characters key approaches for operationalizing media richness (see
Table 1). The first focuses on the category match approach

4 L. ZHANG ET AL.

Table 1. A literature review of MST adoption in studies.

Operationalization of Research data and context Main findings Research method Author
perceived media richness Experiment (Dennis & Kinney, 1998)
The use of richer media rather than Survey (Kishi, 2008)
Category match approach 132 Team members from leaner media did not lead to better Survey (D. K. L. Lee & Borah, 2020)
predesigned media- performance on the higher
enabled tasks equivocality task. Econometrics (Li et al., 2022)
Survey (Mirzaei & Esmaeilzadeh, 2021)
Perceptual approach 1062 Managers were studied Organizational interpretation of the Survey (Shen et al., 2021)
Perceptual approach from social media use in environment substantially affects Experiment (Suh, 1999)
the workplace. the use of rich media. Experiment (Otondo et al., 2008)
Experiment (Tseng & Wei, 2020)
671 Participants were Perceived media richness and self-
recruited using a social presentation can affect friendship Experiment (Yoo & Alavi, 2001)
media tool development through the mediating Experiment (Zhu et al., 2010)
effect of perceived functionality and
Category match approach A panel data of 87,540 posts the moderating effect of personality
Perceptual approach was collected from the trait.
Sina Weibo platform
Perceptual approach The relationship between information
Category match approach 348 Users from online health timeliness and public engagement is
Category match approach communities were moderated by information richness.
Category match approach surveyed.
Perceived channel richness affects
3309 Users from the blogging perceived social support, willingness
service were surveyed. to exchange information, and
engagement outcomes.

Data were gathered from 316
participants using a Immediate feedback and personal
computer-mediated mode. focus positively affect social identity,
which in turn, leads to we-intention.
688 Participants were
surveyed using a Communication media richness can
predesigned web interface. positively influence task
performance and satisfaction.
259 Consumers of mobile
advertising contexts Media richness features can influence
communication outcomes, i.e.
Category match approach 135 Participants from a effectiveness and satisfaction.
Category match approach decision-making task
The influence of media richness on
128 Participants were studied consumer decision-making behavior
using a Web collaboration is greater in the early stage than in
tool the later AS stage, and the
relationship is moderated by
product type.

Different media conditions can
influence task participation

Navigation and communication
support can positively influence
collaborative online shopping.

by designing different levels of communication media 2.4. Multidimensional trust: cognitive trust vs. affect
tools, such as using computer-mediated text, audio, video, trust
and face-to-face to present lowest to highest richness (e.g.,

Otondo et al., 2008; Suh, 1999; Tseng & Wei, 2020; Yoo Originating from social psychology science, trust is defined
& Alavi, 2001; Zhu et al., 2010). In contrast, using a sur- as a general belief in a person who will act in line with
vey approach, the second approach focuses on the percep- favorable expectations towards the trustee (Gefen, 2000).
tion of media richness (Kishi, 2008; D. K. L. Lee & Borah, Two distinct dimensions of trust (i.e., cognitive trust and
2020; Mirzaei & Esmaeilzadeh, 2021). The perceptual sur- affective trust) were identified by McAllister (1995), and
vey approach synthesizes the influence of perceived media they have been widely applied in recent studies (Chih et al.,
richness through general psychological multidimensional 2017; Goles et al., 2009; J. Lee et al., 2015; K. Z. K. Zhang
measurement (i.e., multiple information cues, language var- et al., 2014). In particular, cognitive trust (generated by
iety, immediate feedback, and personal focus). Given that rational assessment regarding the trustees’ ability and reli-
we are interested in consumers’ perceived media richness ability) is important in online exchange relations where
of the metaverse, we adopt the second approach and intro- uncertainty is present (Pavlou, 2002; Shao et al., 2019; Shao
duce four significant dimensions, namely the extent to & Yin, 2019). Affective trust (generated by perceived
which a user perceives: (1) the use of various cues, e.g., strength and the level of emotional attachment, caring, and
texts, images, graphical symbols, physical presence, ges- social reciprocity between a trustor and a trustee) also plays
tures, etc.; (2) communication with natural languages; (3) an important role in the context of online marketplaces
immediate interaction; and (4) personalized virtual images (Chih et al., 2017; Goles et al., 2009; J. Lee et al., 2015; K. Z.
(avatars). Accordingly, this study identifies the construct of K. Zhang et al., 2014).
the perceived media richness of the metaverse as a
significant trust-building antecedent in the research This study extends the two dimensions of trust from the
framework. traditional communication context (i.e., buyer–seller inter-
personal relationships) to the metaverse shopping environ-
ment. This extension can be justified by the fact that

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION 5

The metaverse-enabled shopping environment Age
Digital natives (DNs) vs. Digital immigrants (DIs)
Multiple
information cues Affective trust H6b (DNs > DIs)

Language Perceived media H5 (+) H6a (DIs > DNs) Purchase intention
variety richness of the towards the

Immediate metaverse metaverse shopping
feedback

Cognitive trust Control variables
Gender, Education, Shopping experience
Personal focus

Figure 1. The research model.

consumers may tend to rationally evaluate the security and interact with brands, stores, or other consumers to obtain
reliability of the virtual space created by the metaverse, and more realistic, unambiguous product information (Kishi,
to emotionally assess the enjoyable and interactive experi- 2008). Thus, the perceived media richness of the metaverse
ence created by the metaverse environment (Chih et al., has the potential to improve consumers’ cognitive trust. In
2017; W. Wang et al., 2016). Therefore, this study focuses addition, the metaverse (in a 3D virtual world) is more vivid
on trust in the metaverse (the trustee is a technological and interesting than traditional static web page-based shop-
environment) and divides it into two dimensions (i.e., cogni- ping (Kim, 2021), which is beneficial to facilitating affective
tive trust vs. affective trust) to examine their separate influ- trust. Therefore, we hypothesize the following:
ence on purchase intention.
H1: There is a positive relationship between the perceived
3. Research model and hypotheses media richness of the metaverse and cognitive trust.

By integrating a trust-building framework with MRT, we H2: There is a positive relationship between the perceived
aim to explore the impact of the perceived media richness media richness of the metaverse and affective trust.
of the metaverse on purchase intention through the medi-
ation effects of cognitive trust and affective trust. 3.2. Cognitive trust, affective trust, and purchase
Meanwhile, we incorporate age as a moderator between intention towards metaverse shopping
multidimensional trust and purchase intention. Additionally,

gender, education, and shopping experience are controlled Trust is considered to be a key factor in determining behav-
in the structural model. Figure 1 shows the proposed ioral intention (Cheng et al., 2021; Shao, Zhang, Brown,
research model. et al., 2022). McAllister (1995) proposed that trust is a
multidimensional construct, including cognitive trust and
3.1. Perceived media richness of the metaverse, affective trust. On the one hand, cognitive trust depends on
cognitive trust, and affective trust the trustor’s rational evaluation, which can mitigate risk per-
ceptions and facilitate behavioral outcomes (Chih et al.,
Based on MRT, we propose that perceived media richness is 2017; Shao et al., 2019). On the other hand, affective trust
a second-order construct expressed by multiple information originates from the emotional attachment between the
cues, language variety, immediate feedback, and personal trustor and the trust target, which can generate comfortable
focus (Daft & Lengel, 1986; Shen et al., 2021). Prior research and positive attitudes (W. Wang et al., 2016). Following the
has argued that the level of intuitionistic and abundant theory of reasoned action, affective trust as a form of trust-
information content, vividness, and social cues provided by ing attitude will predict individuals’ behavioral intentions to
multimedia technologies is highly related to message persua- perform an action (Komiak & Benbasat, 2006). In the con-
sion (Kishi, 2008; D. K. L. Lee & Borah, 2020; Shen et al., text of online shopping, both cognitive trust and affective
2021), which can induce effective communication perform- trust play significant roles in affecting consumers’ purchase
ance (Dennis & Kinney, 1998; Zhu et al., 2010). Considering intention (Kimiagari & Malafe, 2021; Wu et al., 2023).
that effective interactions help build trust (Bao et al., 2021;
Pavlou, 2002; Shao et al., 2022b; Shao & Yin, 2019), it is In the metaverse shopping context, consumers with high
plausible to expect that trust will be affected by the modality cognitive trust in the metaverse will perceive a lower level of
type or level of perceived media richness. uncertainty and potential risks (Chih et al., 2017), which is
likely to enhance their willingness to buy recommended
In the metaverse shopping context, consumers use virtual goods in the metaverse. In addition, consumers who have
avatars to move freely in a 3D shopping world in a more affective trust in the metaverse will form emotional and
interactive way (Kim, 2021), which enables rich media to pleasant feelings (Kimiagari & Malafe, 2021; W. Wang et al.,

6 L. ZHANG ET AL.

2016) and may be motivated to buy the products recom- be more comfortable with digital technologies (Gurtner
mended by the metaverse. Therefore, we formulate the fol- et al., 2014). Specifically, DNs have more experience

lowing hypotheses: with the virtual world (such as Second Life or 3D gam-
ing) and may show positive and hedonic attitudes
H3: There is a positive relationship between cognitive trust towards emerging technologies (e.g., AR, VR, etc.) to
and purchase intention towards metaverse shopping. complete virtual shopping processes (L. Zhang et al.,
2021). As such, in the metaverse shopping context, DNs
H4: There is a positive relationship between affective trust are more likely to form comfortable and emotional per-
and purchase intention towards metaverse shopping. ceptions; thus, they may be more affected by affective
trustworthiness when making purchase decisions. Based
Furthermore, (McAllister, 1995) proposed that cognitive on this logic, we argue that affective trust plays a more
trust is a prerequisite for affective trust. When a consumer significant role in enabling DNs to purchase in meta-
makes a rational assessment of purchase behavior, he/she is verse shopping. The above analysis leads to the follow-
more likely to respond emotionally (Chen et al., 2019; Legood ing hypothesis:
et al., 2023). Therefore, in the metaverse shopping context,
cognitive trust may be necessary for affective trust to develop; H6b: Affective trust has a stronger influence on purchase
accordingly, we propose the following hypothesis: intention towards metaverse shopping for digital natives
compared with digital immigrants.
H5: There is a positive relationship between cognitive trust
and affective trust. 4. Research design and execution

3.3. The moderating role of age: digital natives (DNs) 4.1. Research context and data collection
vs. digital immigrants (DIs)
This study employed the scenario-based survey method to
Age has a significant functional meaning in understanding collect data since it is an effective way to promote partici-
individuals’ attitudes with regard to technology adoption pants’ contextualized understanding of metaverse shopping
(Hong et al., 2013; Shao, Benitez, et al., 2022; L. Zhang in a hypothetical situation (Chang et al., 2013; Kwak et al.,
et al., 2021). According to Hong et al. (2013), chronological 2021; X. Wang et al., 2020), especially when research on the
age plays an important role in distinguishing between DNs role of the metaverse is still in its infancy (Hua et al., 2018).
(who were born after the 1980s) and DIs (who were born Using insights from the past literature, vignettes were used
before the 1980s) (Shao, Benitez, et al., 2022). Previous lit- to present subjects with written descriptions of realistic sit-
erature has suggested significant age-related generational dif- uations (Trevino, 1992). The use of vignettes can provide

ferences between DNs and DIs in predicting consumption control by placing all subjects in the same scenario with the
orientation and purchasing behaviors. For example, Gurtner same fictitious metaverse shopping context. The vignette-
et al. (2014) found that DIs are more likely to adopt new based approach has been widely adopted in past studies due
technologies based on cognitive evaluation, while DNs value to numerous benefits such as the relevance of scenarios to
more the hedonic experience of using new technologies. L. realistic situations, lesser social desirability and memory
Zhang et al. (2021) explained that DNs’ purchase behaviors lapse biases, and ease of data collection from large samples
are driven by enjoyment perception, while DIs are more cir- (Tong et al., 2013; Vance et al., 2012; S. Zhang & Leidner,
cumspect or judicious and tend to perform purchase behav- 2018). In summary, respondents were required to make
iors via the cognitive evaluation process. behavioral decisions based on true-to-life vignettes
embedded in the hypothetical metaverse shopping scenario
From the perspective of age (DNs vs. DIs) (Hong et al., (see Appendix A).
2013; Tams et al., 2018), there will be high transaction risk
and vulnerability perceptions for DIs because they are not In the scenario-based investigation, we entrusted the
familiar with digital technologies. Therefore, DIs always third-party questionnaire website (www.sojump.com) to
require conscious calculation through a certain degree of randomly invite 500 consumers1 from its database to
information search, rational thinking, and risk assessment complete an online questionnaire from June 01 to June
when making decisions (Gurtner et al., 2014; L. Zhang et al., 10, 2022 (L. Zhang et al., 2021). The questionnaire
2021). As such, in the metaverse shopping context, DIs focus included three sections. The first section comprised a sim-
more on making rational assessments to eliminate uncertainty ple explanation of the concept of the metaverse and our
and perceived risk towards the surrounding purchase envir- hypothetical shopping scenario and contained one screen-
onment; thus, cognitive trust may exhibit a stronger influence ing item (Appendix A presents the scenario used in the
on DIs’ purchase intention towards metaverse shopping. The current study, and please see the endnote for more discus-
following hypothesis is therefore proposed: sion about the screening item).2 We stated that the ano-
nymity of the participants would be ensured, and that the
H6a: Cognitive trust has a stronger influence on purchase collected data would only be used for academic research.
intention towards metaverse shopping for digital immigrants The second section contained the measurements for each
compared with digital natives. variable included in the research model. The third section
comprised demographic information, including gender,
Compared with DIs, DNs are more exposed to new
technologies and digital media; therefore, they tend to

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION 7

age, etc. A monetary reward was given to any respondent cause the indicators, and some indicators can be inter-
completing the questionnaire online, and 479 responses changed without altering the meaning of the latent vari-
were received, with a response rate of 95.8%. After remov- ables (Benitez et al., 2020). In summary, the proposed
ing 103 respondents who chose items 1 or 2 for the research model comprises formative and reflective
screening item and 44 invalid responses (e.g., an extreme constructs.
answer to all questions), a total of 332 questionnaires
were used for data analysis. In order to assess the nonres- Perceived media richness of the metaverse
ponse bias, we compared the demographic variables Conistent with Shen et al. (2021), the construct was opera-
between the first 50 respondents and the last 50 respond- tionalized in terms of multiple information cues, language
ents using t-tests (Armstrong & Overton, 1977). The stat- variety, immediate feedback, and personal focus. That is, we
istical analysis results suggested that there were no operationalized perceived media richness of the metaverse as
significant differences between the two groups in terms of a broader concept having four dimensions: (1) the use of
gender, age, and shopping experience (p > 0.1); thus, non- various cues (e.g., texts, images, graphical symbols, physical
response bias does not exist in our study (H. Liang et al., presence, gestures, etc.); (2) communication with natural
2019). languages (e.g., language symbols, and emoticons); (3)
immediate interaction (e.g., receive and send feedback
The statistical characteristics of the sample are presented quickly); and (4) personalized virtual images (e.g., avatars).
in Table 2. Male and female consumers were almost evenly The dimensions were measured using reflective scales with
distributed. In terms of age, 6.0% were below 18 years old, three items each.
24.5% were 18–29 years old, 29.5% were 30–39 years old,
32.8% were 40–49 years old, and 7.2% were over 50 years Multidimensional trust
old. Most of the participants had a Bachelor’s degree and The scales by McAllister (1995) and W. Wang et al.
over three years’ online shopping experience. The results (2016) were used to measure cognitive trust and affective
suggest that no significant differences exist between our trust. Specifically, cognitive trust reflects consumers’
sample and the actual online consumers in China (CNNIC, rational expectations that the metaverse has the necessary
2021), demonstrating that our sample is representative of attributes to ensure a proficient and reliable virtual shop-
the actual online consumers in China. ping space while affective trust refers to consumers’ com-

fort and emotional bonds regarding the virtual shopping
4.2. Measurements space created by the metaverse. Both of the constructs
were operationalized as reflective variables with three
A seven-point Likert scale with the categories from items each.
“Strongly agree” (7) to “Strongly disagree” (1) was used
to measure constructs. Constructs can be operationalized Purchase intention
as reflective or formative (Benitez et al., 2020; L. Zhang Following K. Z. K. Zhang et al. (2014), purchase intention
et al., 2022). Multiple information cues, language variety, was conceptualized as consumers’ willingness to purchase
immediate feedback, personal focus, cognitive trust, goods or services in the metaverse shopping space. It was
affective trust, and purchase intention were specified as specified as a reflective construct and measured with five
reflective constructs, while perceived media richness of items.
the metaverse was specified as a second-order formative
construct, which emerges from the first-order constructs A few revisions were made to adapt our measurement to
of multiple information cues, language variety, immedi- the metaverse shopping context. A pilot study was con-
ate feedback, and personal focus (Benitez et al., 2020). ducted using 54 college students with online shopping
Other constructs were operationalized as reflective meas- experience. Several items with factor loadings lower than 0.4
urement models because they (i.e., latent variable) can were adjusted based on the respondents’ feedback (L. Zhang
et al., 2022). Table 3 describes the definition and measure-
Table 2. Descriptive statistics. ment items of the constructs.

Attributes Options Frequency Percentage (%) 4.3. The estimation strategy

Gender Male 164 49.4 We use partial least squares (PLS) path modeling to test the
Age 168 50.6 proposed research model, which is recognized as an appro-
Female 20 6.0 priate statistical tool for structural equation modeling (SEM)
Education <18 81 24.5 analysis (Benitez et al., 2020). Compared with the covari-
Shopping experience 18–29 98 29.5 ance-based approach, PLS is more robust for concurrently
30–39 109 32.8 handling formative and reflective constructs (Benitez et al.,
40–49 19 5.7 2020, 2022; Castillo et al., 2021). In our study, the perceived
50–59 1.5 media richness of the metaverse was estimated by a

>60 5 24.3
81 63.0
High school and below 209 12.7
Bachelor’s degree 42 1.2
4 13.6
Master’s degree and above 45 42.8
<1 year 142 18.4
1–2 years 61
3–4 years
>4 years

8 L. ZHANG ET AL.

Table 3. Constructs and items. Multiple information cues Definitions Items References
MIC1-MIC3 (Shen et al., 2021)
Constructs A consumer’s perception that the
Perceived media richness of metaverse supports multiple information LV1-LV3 (McAllister, 1995; W. Wang
the metaverse through a variety of ways (e.g., texts, et al., 2016)
images, and videoes) IF1-IF3
Language variety PF1-PF3 (K. Z. K. Zhang et al., 2014)
A consumer’s perception that the
Immediate feedback metaverse enables them to CT1-CT3
Personal focus communicate with natural languages
(e.g., language symbols, and emoticons). AT1-AT3
Cognitive trust PI1-PI3
A consumer’s perception that the
Affective trust metaverse helps them to receive and
Purchase intention send feedback quickly.

A consumer’s perception that the

metaverse enables them to customize
avatars or personal profiles according to
their personal needs and preferences.

Consumers’ rational expectations that the
metaverse has the necessary attributes
to ensure a proficient and reliable virtual
shopping space

Consumers’ comfort and emotional bonds
regarding the virtual shopping space
created by the metaverse

Consumers’ willingness to purchase goods
or services in the metaverse shopping
space

Table 4. Overall model fit of the saturated model analysis. Second-order level
First-order level

Discrepancy Value HI99 Conclusion Value HI95 Conclusion

SRMR 0.048 0.049 Supported 0.023 0.025 Supported
dULS 0.825 0.854 Supported 0.056 0.066 Supported
dG 0.567 0.602 Supported 0.086 0.094 Supported

Notes: Perceived media richness of the metaverse is a second-order formative construct, whilst its first-order dimensions are operationalized with reflective meas-
urement. Formative constructs were estimated with mode B, and the reflective construct was estimated with mode A consistent (PLSc) (for a discussion, see
Benitez et al., 2020).

formative construct, which used a regression weighting 4.3.2. Evaluation of the measurement model
scheme represented by arrows pointing from indicators to We assessed the reflective constructs through the indicators
their corresponding constructs (see details in Appendix B) of factor loadings, rho_A, and average variance extracted
(Benitez et al., 2020, 2022; Castillo et al., 2021). We (AVE), and we tested for multicollinearity, weights, and the
employed the statistical software package ADANCO 2.3 significance level of the formative construct (i.e., the per-
Professional for Windows (posite-modeling. ceived media richness of the metaverse). As noted in Table
com/) to analyze the structural model. 5, all indicators (i.e., factor loadings, rho_A, and AVE) for
the reflective constructs met the threshold criteria (Benitez
4.3.1. Evaluation of the overall fit of the saturated model et al., 2020). For the second-order formative construct (i.e.,
Following Benitez et al. (2020), we evaluated the overall the perceived media richness of the metaverse), the results
model fit of the saturated model to assess the validity of the showed that the variance inflation factor (VIF) values ranged
formative and reflective measurement models. The goodness from 1.529 to 2.175, which are below the threshold value of
of fit of the saturated model was evaluated by the discrep- 10, suggesting that multicollinearity is not a problem in our
ancy between the empirical correlation matrix and the study (Benitez et al., 2020). All indicator weights (from
model-implied correlation matrix through the indicators for 0.311 to 0.464) and dimension weights (from 0.081 to 0.409)
SRMR, dULS, and dG (Benitez et al., 2020, 2022; Benitez were significant except for the dimension weight of language
et al., 2022; Castillo et al., 2021). In order to guarantee an variety (0.081). Following the guidelines of Benitez et al.
adequate model fit, SRMR should be less than 0.080, and (2020) for formative measurement, although the dimension
SRMR, dULS, and dG should be below the 95% quantile weight of language variety was small, the factor loading for
(HI95) of the bootstrapping discrepancies or at least below language variety (0.755) was significant on an alpha level of
the 99% quantile (HI99). Table 4 summarizes the results of 0.001. Thus, we retained language variety in our model. The
the goodness of fit for the saturated model. Overall, the fit results suggested that our variables possess very good meas-
of the saturated model was satisfactory to proceed with the urement properties (Benitez et al., 2020; Benitez et al.,
evaluation of the measurement model (Cheng et al., 2022). 2022), and that we could proceed with the hypothesis testing
(see details in Appendix B).

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION 9

Table 5. Validity and reliability of the scales.

Constructs rho_A AVE Loading VIF Weights

Multiple information cues (MIC) 0.876 0.655 0.828ÃÃÃ 1.821 0.362ÃÃ
0.973ÃÃÃ 1.833 0.464ÃÃÃ
MIC1 The metaverse shopping virtual world transmits

a variety of different cues beyond the explicit

text-based product information 0.740ÃÃÃ 0.353ÃÃÃ

MIC2 The metaverse shopping virtual world conveys 2.501

multiple types of product information (verbal

and nonverbal) 0.685ÃÃÃ 0.326ÃÃÃ

MIC3 The metaverse shopping virtual world presents 2.141

vivid product information through facial

expressions and body language 0.755ÃÃÃ
0.855ÃÃÃ
Language variety (LV) 0.840 0.609 2.175 0.081
2.16 0.429ÃÃÃ
LV1 The metaverse shopping virtual world transmits

varied symbols (e.g., texts, photos, videos,

audios, links, and so on) 0.620ÃÃÃ 0.311ÃÃÃ

LV2 The metaverse shopping virtual world 1.725

communicates rich meanings about products

using a large pool of language symbols 0.844ÃÃÃ 0.423ÃÃÃ

LV3 The metaverse shopping virtual world uses rich 1.790

and varied language 0.799ÃÃÃ 0.366ÃÃ
0.669ÃÃÃ 0.331ÃÃÃ
Immediate feedback (IF) 0.847 0.619 1.529
1.955
IF1 The metaverse shopping virtual world has the

ability to give and receive timely feedback 0.751ÃÃÃ 0.372ÃÃÃ

IF2 The metaverse shopping virtual world can 2.560

provide immediate feedback 0.920ÃÃÃ 0.456ÃÃÃ

IF3 The metaverse shopping virtual world enables 1.755

consumers to send/receive information

quickly 0.848ÃÃÃ 0.409ÃÃ
0.929ÃÃÃ 0.402ÃÃÃ
Personal focus (PF) 0.902 0.731 2.039
2.032
PF1 The metaverse shopping virtual world enables

consumers to personalize their virtual avatars 0.727ÃÃÃ 0.315ÃÃÃ

PF2 The metaverse shopping virtual world enables 3.488

consumers to edit personal profiles or

decorate virtual avatars 0.896ÃÃÃ 0.388ÃÃÃ

PF3 The metaverse shopping virtual world enables 3.995

consumers to tailor personal avatars and

share their virtual images in the interactive

shopping community

Cognitive trust (CT) 0.891 0.731 0.864ÃÃÃ 0.372ÃÃÃ

CT1 When I browse products and participate in tasks 2.598

in the metaverse shopping scenario, I feel it

is capable and proficient 0.829ÃÃÃ 0.357ÃÃÃ

CT2 When I browse products and participate in tasks 2.979

in the metaverse shopping scenario, I feel it

is a reliable and secure virtual shopping

world 0.872ÃÃÃ 0.375ÃÃÃ

CT3 When I browse products and participate in tasks 2.442

in the metaverse shopping scenario, I feel it

is competent and effective in presenting

product contents

Affective trust (AT) 0.916 0.780 0.851ÃÃÃ 0.348ÃÃÃ

AT1 When I browse products and participate in tasks 2.549

in the metaverse shopping scenario, I feel it

responded caringly 0.928ÃÃÃ 0.379ÃÃÃ

AT2 When I browse products and participate in tasks 4.062

in the metaverse shopping scenario, I feel it

displays a warm and caring attitude

towards me 0.868ÃÃÃ 0.355ÃÃÃ

AT3 When I browse products and participate in tasks 3.663

in the metaverse shopping scenario, I feel

comfortable and enjoyable

Purchase intention towards the metaverse shopping (PI) 0.954 0.869 0.878ÃÃÃ 0.329ÃÃÃ

PI1 Given a chance, I predict that I should spend 4.592

money on the products recommended by the

metaverse shopping scenario in the future 0.945ÃÃÃ 0.354ÃÃÃ

PI2 Given a chance, I intend to buy products from 7.433

the metaverse shopping scenario 0.972ÃÃÃ 0.364ÃÃÃ

PI3 If I could, I am very likely to buy products from 5.282

the metaverse shopping scenario

Notes: ÃÃp < 0.01; ÃÃÃp < 0.001. Formative constructs were estimated with mode B and the reflective construct was estimated with mode A consistent (PLSc)

(for a discussion, see Benitez et al., 2020).

10 L. ZHANG ET AL.

Table 6. HTMT analysis.

1 2 3 4 5 6 7 8 9 10

1. MIC 0.717 0.589 0.585 0.464 0.778 0.638 0.227 0.094 0.208
2. LV 0.609 0.783 0.456 0.455 0.638 0.067 0.135 0.007

3. IF 0.627 0.447 0.434 0.397 0.076 0.162 0.095
4. PF 0.496 0.421 0.369 0.021 0.191 0.099
5. CT 0.421 0.366 0.081 0.033 0.115
6. AT 0.374 0.043 0.120 0.044
7. PI 0.088 0.136 0.031
8. Gender 0.150 0.083
9. Education 0.095
10. Experience

Discriminant validity was analyzed using the heterotrait- Table 7. Estimated model fit evaluation.
monotrait ratio of correlations (also called HTMT)
(Henseler et al., 2015). Based on the multitrait-multimethod Discrepancy Value HI95 Conclusion
matrix, HTMT demonstrates a superior performance by
means of a Monte Carlo simulation test compared to the SRMR 0.023 0.025 Supported
Fornell–Larcker criterion when conducting PLS-SEM dULS 0.056 0.066 Supported
(Benitez et al., 2020). Thus, this study chose HTMT to dG 0.086 0.094 Supported
evaluate discriminant validity. As noted in Table 6, HTMT
values were below 0.8, indicating adequate discriminant val- 4.3.5. Hypotheses testing
idity for our measurement model. The results showed that the perceived media richness of the
metaverse positively affected cognitive trust (b ¼ 0.338,
4.3.3. Common method bias (CMB) test p < 0.001) and affective trust (b ¼ 0.109, p < 0.05), thus sup-
Given that a self-reported survey was conducted, there is a porting H1 and H2, respectively. Both cognitive trust
potential for common method bias (CMB) resulting from (b ¼ 0.218, p < 0.05) and affective trust (b ¼ 0.255, p < 0.05)
the consistency motif, social desirability, and common scale had strong influences on purchase intention towards meta-
formats (Podsakoff et al., 2003). Therefore, we employed a verse shopping. Meanwhile, cognitive trust was positively
marker variable approach to statistically assess CMB associated with affective trust (b ¼ 0.569, p < 0.001), thus
(Ronkko & Ylitalo, 2011). Following (Miller & Simmering, supporting H3–H5. Regarding the demographic variables,
2022), we selected the blue color attitude with three-item only gender (b ¼ 0.190, p < 0.001) had a significant influence
scales (i.e., “I like the color blue”) as a marker variable, on purchase intention.
which is theoretically unrelated to our research model. As

shown in Appendix C, the results showed that the color 4.3.6. Multi-group analysis results: digital natives (DNs)
blue attitude (marker variable) had no significant influence vs. digital immigrants (DIs)
on cognitive trust (b ¼ À0.033, n.s.), affective trust (b ¼ To further explore the moderating effect of age, we con-
À0.022, n.s.), and purchase intention towards metaverse ducted a multi-group analysis (MGA) following Benitez
shopping (b ¼ À0.052, n.s.). Therefore, we conclude that et al.’s (2020) guidelines. Age was operationalized as a
CMB is not a major concern in this study. dummy variable to represent DNs and DIs with an age
threshold of 40 years old (Shao, Benitez, et al., 2022).
4.3.4. Hypotheses testing Following previous studies (Shao, Benitez, et al., 2022), we
We examined the structural model using the bootstrapping adopted the MGA method to compare the path coefficients
resampling procedure to estimate the significance of the of the two groups (DNs vs. DIs). As illustrated in Table 8,
path coefficients. For reference, we assessed the estimated the relationship between cognitive trust and purchase inten-
model fit by calculating the SRMR, dULS, and dG (Benitez tion towards metaverse shopping was higher for DIs
et al., 2020). As noted in Table 7, the SRMR was 0.023 for (b ¼ 0.442, p < 0.001) than for DNs (b ¼ 0.113, n.s.), sup-
the estimated model, which is lower than the threshold of porting H6a. In contrast, the relationship between affective
0.080. The SRMR, dULS, and dG were also below the 95% trust and purchase intention towards metaverse shopping
quantile of the bootstrap discrepancies, suggesting a high was stronger for the DN group (b ¼ 0.327, p < 0.001) than
level of model fit. Moreover, the explanatory power of the for the DI group (b ¼ 0.072, n.s.), supporting H6b.
structural model was assessed by calculating the amount of
variance (R2) explained in the endogenous variable. As illus- 4.3.7. Post-hoc mediation test
trated in Figure 2, our theoretical model explicated 54.4% In order to examine whether multidimensional trust in the
variance in cognitive trust, 64.2% variance in affective trust, metaverse mediated the relationship between the perceived
and 54.2% variance in purchase intention towards metaverse media richness of the metaverse and purchase intention, our
shopping, demonstrating the good explanatory power of the study followed Hair et al.’s (2021) criteria to conduct a
proposed research model. mediation test using the bootstrapping method. First, we
linked the direct relationship between the perceived media
richness of the metaverse and purchase intention and tested

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION 11

The metaverse-enabled shopping environment

Multiple 0.3 62 ** * 0.109* Affective trust 0.255* Purchase intention 0.1 90 ** * Gender
information 0.081 R2=64.2% 0.218* 0.033
0.3 66 ** * Perceived media towards the metaverse 0.008 Education
cues richness of the 0.5 69 ** *
0.4 09 ** * shopping Shopping
Language metaverse R2=54.2% experience
variety
0.3 38 ** *
Immediate
feedback

Personal Cognitive trust
focus R2=54.4%

Figure 2. Structural model results. Note: Ãp < 0.05, ÃÃÃp < 0.001 (two-tailed test).

Table 8. Results of PLS-MGA.

Path coefficient

Paths DNs DIs Coefficient diff (p-value) Hypotheses
À0.329Ã (0.018)
H6a: Cognitive trust in the metaverse ! Purchase intention towards the metaverse shopping 0.113 ns 0.442ÃÃÃ 0.255Ã (0.045) Supported
H6b: Affective trust in the metaverse ! Purchase intention towards the metaverse shopping 0.327ÃÃÃ 0.072 ns Supported

Note: Ãp < 0.05, ÃÃÃp < 0.001, ns: Non-significant, sample size for DNs ¼ 199, sample size for Dis ¼ 133.

Table 9. Mediation test results. Indirect effect [95% CI] Conclusion
0.054ÃÃ [0.018, 0.087] Full mediation

Relationship Full mediation
0.032Ã [0.007, 0.063] Full mediation
Perceived media richness of the metaverse!Cognitive trust!
Purchase intention towards the metaverse shopping 0.032ÃÃ [0.011, 0.057]

Perceived media richness of the metaverse!Affective trust!
Purchase intention towards the metaverse shopping

Perceived media richness of the metaverse!Cognitive trust!
Affective trust!Purchase intention towards the metaverse
shopping

Note: Ãp < 0.05; ÃÃp < 0.01.

the significance level of the indirect relationships. Second, 5.1. Discussion of results
partial or full mediation effects were assessed depending on
whether the independent variable directly impacted the 5.1.1. Reconsidering the role of trust in metaverse
dependent variable after the mediator variable was included. shopping
Specifically, the direct relationship between the perceived The literature has indicated that the metaverse can shape
media richness of the metaverse and purchase intention was consumers’ sentiment and purchase behaviors through the
insignificant (b ¼ 0.107, p > 0.05). Thus, the bootstrapping synergetic integration of technological features in the infra-
test results demonstrate full mediation effects for our structure and application layers such as the implementations
research model (see Table 9). of data mining techniques, simulation modeling tools, and
cognitive enhancement technologies (Grupac et al., 2022;
5. Discussion and core conclusions Hudson, 2022; Jenkins, 2022; Zvarikova et al., 2022), which
can be categorized within the general discussion of bounda-
In recent years, trust issues in the metaverse have gained ries and scope of potential technology in actualized opera-
much attention across disciplines. However, little has been tions. However, there has been a dearth of research
done to explore trust-building mechanisms in the meta- exploring the underlying psycho-physiological mechanisms
verse context. This study proposes and empirically vali- that may affect users’ purchase behaviors towards metaverse

dates a theoretical model that explicates how the shopping (Hennig-Thurau et al., 2022). In particular, con-
metaverse-enabled shopping environment affects users’ sumers’ trust issues have been a major challenge in the
trusting beliefs and purchase intention toward the meta- metaverse environment (Tan & Saraniemi, 2022). It is also
verse shopping. The critical assessment and discussion of challenging to convince users that the online shopping is
the major findings and their implications are summarized safe within the metaverse (Dwivedi et al., 2022). This study
in the following subsections. demonstrates the importance of trust in the metaverse con-
text, i.e., the study highlights that trust plays an important

12 L. ZHANG ET AL.

role in shaping users’ purchase intention towards metaverse 5.1.3. Uncovering contingency effect of trust belief—Trust
shopping. This finding provides empirical support to the outcome relationship labeled by the “digital divide” in
theoretical work of McAllister (1995) and Chen et al. (2019) metaverse shopping
in proposing, operationalizing, and validating the multidi- It has been argued that consumers’ perceptions and pur-
mensional trust variables, and extends previous research chase behaviors are contingent upon the generation gap
(Chih et al., 2017; W. Wang et al., 2016; Wu et al., 2023) labeled by the digital age divide (Chan et al., 2017; L. Zhang
from traditional shopping to the metaverse shopping con- et al., 2021). However, little effort has been made to evaluate
text. More specifically, our study found that multidimen- age-related generational differences in the context of meta-
sional trust in the metaverse significantly influences verse shopping. To address this research gap, we incorporate
purchase intention towards metaverse shopping, highlighting age as a salient contingency factor in the theoretical model
the predictive power of cognitive and affective trust in trig- to explore the differences between DNs and DIs in the trust
gering the willingness to purchase in the metaverse shopping belief—trust outcome relationship regarding metaverse shop-
setting. With this construction of multidimensional trust- ping. Specifically, our research findings support the moder-
purchase intention relationships, a better proportion of the ating effect of age on the relationship between
variance in purchase intention towards metaverse shopping multidimensional trust and purchase intention towards
was captured compared to existing works (e.g., Hwang & metaverse shopping, implying that cognitive trust as a
Lee, 2022; Patil & Pramod, 2022). Moreover, previous rational assessment process plays a more salient role in fos-
research has shown that positive cognitive trust is strongly tering DIs’ purchase intention, while DNs’ purchase inten-
associated with positive affective trust (Chen et al., 2019; tion is more likely to be triggered when they perceive higher
Chih et al., 2017; W. Wang et al., 2016). Our research find- levels of affective trust. Consistent with the age-based stereo-

ings support the notion that cognitive trust directly leads to type’s predictions (Hong et al., 2013; Shao, Benitez, et al.,
affective trust and indirectly affects trust outcomes (i.e., pur- 2022; Tams et al., 2018; L. Zhang et al., 2021), age affects
chase intention), similar to previous empirical findings. individual attitudes, technology acceptance, and usage
Therefore, we believe that McAllister’s (1995) classification behaviors. We introduced age as a salient digital-divide fac-
of trust remains important; however, future studies should tor and showed the contingency influences of multidimen-
justify the interplay between cognitive and affective trust in sional trust on consumers’ purchase intention towards
a metaverse shopping context. metaverse shopping across DNs and DIs. Overall, the find-
ings suggest that DNs and DIs exhibit distinct cognitive and
5.1.2. Developing trust-building antecedents in metaverse behavioral patterns, i.e., the impacts of multidimensional
shopping trust on purchase behaviors in the metaverse shopping
Prior literature has conceptualized metaverse as a new com- context.
puter-mediated environment and shown that metaverse
shopping consists of virtual worlds in which users can com- 5.2. Key contributions to theoretical research
municate with each other and make shopping experiences
via avatars (Hennig-Thurau et al., 2022; Oh et al., 2023). Our study makes three major contributions. First, our study
However, the context-specific computer-mediated media brings MRT into the context of metaverse shopping and
richness that triggers users’ trust formation and purchase identifies the perceived media richness of the metaverse as a
intention towards metaverse shopping remains relatively significant trust-building determinant. Previous literature
underexplored. To address this research gap, this study vali- has mostly focused on personality and institutional factors
dated the relationships between the perceived media richness (such as structural assurances) as trust-building antecedents
of the metaverse and multidimensional trust, suggesting that in e-commerce contexts (Shao et al., 2019; Shao, Zhang,
consumers’ cognitive trust and affective trust formation Brown, et al., 2022; Shao, Zhang, Li, et al., 2022), whereas
depend on rich information gathered from the metaverse the influence of technological virtual space on trust-building
environment. Although the relationship between media rich- behaviors has been largely overlooked in the context of
ness and multidimensional trust (cognitive trust vs. affective metaverse shopping (Dwivedi et al., 2022; Tan & Saraniemi,
trust) has not been studied before, a positive association 2022). To fill this gap, grounded in the media richness the-
between these constructs is consistent with previous empir- ory, we unearth perceived media richness as a broader con-
ical evidence that indicates utilitarian and hedonic values of cept constituting four dimensions: multiple information
rich media features (e.g., D. K. L. Lee & Borah, 2020; cues, language variety, immediate feedback, and personal
Mirzaei & Esmaeilzadeh, 2021; Shen et al., 2021). This find- focus, which would strengthen consumers’ trustworthy

ing is also in line with previous literature in that virtual ava- beliefs in the metaverse environment and possibly contribute
tar-oriented interface design can increase cognitive trust and to purchase intentions. The research findings can enrich our
facilitate affective trust in AI agent contexts (Bao et al., knowledge of trust-building antecedents and related out-
2021; W. Wang et al., 2016). Therefore, it would be mean- comes in the context of metaverse shopping through the
ingful to develop trust-building mechanisms by focusing on lens of MRT.
rich media features of the metaverse that would send out
perceptual cues regarding multiple information cues, lan- Second, our study integrated McAllister’s (1995) multidi-
guage variety, immediate feedback, and personal focus. mensional trust framework with MRT, and uncovered the
mediation mechanism of cognitive vs. affective trust in the

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION 13

relationship between the perceived media richness of the Despite the potential benefits of the metaverse, difficulties
metaverse and purchase intention. Prior literature has arise from consumers’ lack of faith in taking the potential
mostly focused on the role of multidimensional trust in risks. Our study shows that visualization and natural com-
affecting consumers’ purchase intention in traditional e- munication features are beneficial to facilitating trusting
commerce contexts (Chih et al., 2017; Kimiagari & Malafe, beliefs in the metaverse. Therefore, online retailers and web-
2021; W. Wang et al., 2016; Wu et al., 2023), whereas less site designers should utilize the findings of this study to
attention has been paid to the metaverse shopping context. improve the design of their media richness mechanisms to
In particular, this study makes a theoretical and empirical institute a trustworthy virtual shopping environment.
contribution to the interpersonal or institutional trust litera- Specifically, providing multiple information cues, language
ture by specifying the trust target as the virtual metaverse variety, immediate feedback, and personal focus in the meta-
environment. Interestingly, our results suggest that cognitive verse shopping environment will increase consumers’ trust
trust and affective trust significantly mediate the relationship formation. This suggests that the management of a meta-
between the perceived media richness of the metaverse and verse marketplace needs to take steps to increase the effect-
purchase intention. The empirical research findings can iveness of media communication mechanisms to enhance
enrich our understanding of consumers’ cognitive beliefs trust in the metaverse.
and emotional reactions aroused by the rich media of the
metaverse when making purchase decisions in the virtual Second, our study proposes that multidimensional trust
3D shopping world (Hennig-Thurau et al., 2022). Moreover, signals to consumers that it is safe and reliable both to act

the bootstrapping analysis results showed that the indirect and to make purchases in the virtual shopping environment
effect of the perceived media richness of the metaverse on provided by the metaverse. Thus, online retailers should pay
purchase intention is fully and serially mediated by cognitive attention both to cognitive and affective trust. Furthermore,
trust and affective trust. The empirical findings contribute to online retailers and website designers should recognize that
the extant trust literature by uncovering the nomological the effects of cognitive trust vs. affective trust on purchase
network among cognitive vs. affective trust and their antece- intention towards metaverse shopping exhibit significant dif-
dents in the emerging context of the digital economy. ferences between DNs and DIs. Given that DNs are more
affected by affective trust when making purchase decisions
Third, we explicated the boundary condition of trust- in the metaverse shopping context, online retailers and web-
related purchase outcomes in the metaverse shopping con- site designers must invest more in the development of the
text by incorporating the moderating role of age. pleasurable and entertaining experience and a sense of
Specifically, we divided the overall sample into two groups belonging. On the contrary, DIs focus more on the quality,
(DNs vs. DIs) based on their age, categorized by the digital authenticity, and credibility of the content presented in the
divide (L. Zhang et al., 2021), and uncovered the path rela- virtual space and are more cautious in making purchase
tionship differences between the two dimensions of trust decisions in the metaverse shopping context. Therefore,
and purchase intention towards metaverse shopping. online retailers and website designers should attend to build-
Although previous studies have examined the individual atti- ing DIs’ cognitive trust by providing a safe and reliable
tudes and behavioral differences between DNs and DIs environment through the metaverse purchase process.
(Hong et al., 2013; Kesharwani, 2020; Shao, Benitez, et al., Overall, such efforts to understand the differentiated pur-
2022; L. Zhang et al., 2021), few studies have incorporated chase behaviors of DNs vs. DIs are important when design-
age as a key boundary condition of the multidimensional ing marketing strategies to bring greater efficiency and
trust–purchase intention relationships. To the best of our impact to product campaigns.
knowledge, this is one of the first studies to reveal the con-
tingency influence of multidimensional trust on purchase 5.4. Conclusions, limitations, and future research
intention in the emerging context of metaverse shopping. avenues
Specifically, we found that cognitive trust exhibits a stronger
influence on purchase intention towards metaverse shopping Our study has delineated the significance of multidimen-
for DIs than DNs, while affective trust demonstrates a stron- sional trust in affecting consumers’ purchase intention in
ger influence on purchase intention towards metaverse shop- the domain of metaverse shopping. Drawing on MRT, this
ping for DNs than DIs. Overall, this research’s findings study has developed a research model to examine the ante-

advance research on IS behavioral phenomena by highlight- cedents and impact outcomes of multidimensional trust in
ing the digital-divide difference in explaining trust-related the metaverse. A scenario-based survey was conducted in
behaviors in the metaverse field (Dwivedi et al., 2022). China in which 332 valid responses were collected from
users, and empirical results suggested that the perceived
5.3. Implications for practice media richness of the metaverse positively influences con-
sumers’ purchase intention through the joint mediating
Our study can be used to provide guidelines for online effects of cognitive and affective trust. Furthermore, we per-
retailers and website designers to enhance profitability in the formed a MGA and found that cognitive trust exhibits a
digital economy. First, online retailers need to recognize that stronger influence on purchase intention towards metaverse
building trust is critical to induce consumers to make risky shopping for DIs, while affective trust has a greater effect on
purchase decisions in the metaverse shopping context. purchase intention towards metaverse shopping for DNs.

14 L. ZHANG ET AL.

Our research provides fresh insights into the effective use of Disclosure statement
media richness characteristics, which reinterpret the trust-
building mechanisms in the emerging metaverse shopping No potential conflict of interest was reported by the author(s).
context and opens a promising avenue for future research in
considering the generation gap labeled by the digital divide. Funding

This study has several limitations that highlight future This research study was funded by the Humanities and Social Sciences
research directions. First, this study measured the media project, Ministry of Education in China (21YJA880065).
richness of the metaverse using a survey approach, which
may limit the generalizability of the results. Future studies References
can focus on objective media features of the metaverse using
experiments. Additionally, metaverse-enabled technology Armstrong, J. S., & Overton, T. S. (1977). Estimating nonresponse bias
affordances and multisensory customer experience during in mail surveys. Journal of Marketing Research, 14(3), 396–402.
the purchase process may also influence the trust-building />process. While this limitation was inevitable given the
study’s exploratory nature, further research should include Bao, Y., Cheng, X., De Vreede, T., & De Vreede, G.-J. (2021).
these excluded factors. Second, this study focused on con- Investigating the relationship between AI and trust in human-AI col-

sumers’ trust in the metaverse where the trustee is a techno- laboration [Paper presentation]. Proceedings of the Hawaii
logical environment, while we did not control the International Conference on System Sciences, 54th (pp. 607–616).
interpersonal or institutional trust. This constrains our />opportunity to observe the differential influences of trust
antecedents on different trust targets. Future studies can Benitez, J., Arenas, A., Castillo, A., & Esteves, J. (2022). Impact of
consider other trust targets (e.g., trust in the platform, trust digital leadership capability on innovation performance: The role of
in the product provider) and examine the trust transfer platform digitization capability. Information & Management, 59(2),
mechanism. Third, we explored the contingency influences 103590. />of cognitive trust vs. affective trust on purchase intention
towards the metaverse across DNs and DIs. The moderating Benitez, J., Henseler, J., Castillo, A., & Schuberth, F. (2020). How to
effects of other individual characteristics on purchase inten- perform and report an impactful analysis using partial least squares:
tion towards the metaverse can be examined in future Guidelines for confirmatory and explanatory IS research.
research. For instance, future studies can incorporate gender Information & Management, 57(2), 103168. />as a contingency factor in the research model to examine its im.2019.05.003
influence on the trust belief—trust outcome relationship.
Fourth, metaverse shopping platforms may be able to per- Benitez, J., Ruiz, L., & Popovic, A. (2022). Impact of mobile technol-
form better in terms of both instrumental outcomes and ogy-enabled HR gamification on employee performance: An empir-
experiential outcomes, which may limit the explanatory ical investigation. Information & Management, 59(4), 103647.
power of our research model to a certain extent. />Accordingly, future studies can test many other trust-related
outcomes (i.e., satisfaction) in the metaverse shopping Berthiaume, D. (2022). Survey: Are consumers ready for metaverse shop-
context. ping? Chain Store Age. /> sumers-ready-metaverse-shopping
Notes
Boughzala, I., de Vreede, G.-J., & Limayem, M. (2012). Team collabor-
1. We performed a statistical power analysis to determine the ation in virtual worlds: Editorial to the special issue. Journal of the
minimum sample size required to estimate the proposed Association for Information Systems, 13(10), 714–734. /> model. Assuming an anticipated effect size of 0.150, a 10.17705/1jais.00313
desired statistical power level of 0.95, a constructed number
of 11, and a confidence level of 0.99, the required minimum Braojos, J., Benitez, J., & Llorens, J. (2019). How do social commerce-
sample size to estimate our model is 232 (Faul et al., 2009). IT capabilities influence firm performance? Theory and empirical
evidence. Information & Management, 56(2), 155–171. https://doi.
2. To determine whether participants were focused on the org/10.1016/j.im.2018.04.006
metaverse shopping hypothetical scenario, the screening
question was adapted as follows: “Please choose which of Castillo, A., Benitez, J., Llorens, J., & Luo, X. (. (2021). Social media-
the following statements is in line with your current driven customer engagement and movie performance: Theory and

perception of the aforementioned hypothetical scenario.” A empirical evidence. Decision Support Systems, 145(June), 113516.
semantic differential scale was used, i.e. (1) “After reading /> the corresponding scenario, I still have no idea about
metaverse shopping;” (2) “After reading the corresponding Chan, T. K. H., Cheung, C. M. K., & Lee, Z. W. Y. (2017). The state of
scenario, I have some difficulty imagining metaverse online impulse-buying research: A literature analysis. Information &
shopping;” (3) “After reading the corresponding scenario, I Management, 54(2), 204–217. /> can imagine the metaverse shopping environment to some
extent;” (4) “After reading the corresponding scenario, I can Chang, H. H., Rizal, H., & Amin, H. (2013). The determinants of con-
easily imagine the metaverse shopping environment.” sumer behavior towards email advertisement. Internet Research,
23(3), 316–337. />
Chen, Y., Lu, Y., Wang, B., & Pan, Z. (2019). How do product recom-
mendations affect impulse buying? An empirical study on WeChat
social commerce. Information & Management, 56(2), 236–248.
/>
Cheng, X., Fu, S., & de Vreede, G.-J. (2017). Understanding trust influ-
encing factors in social media communication: A qualitative study.
International Journal of Information Management, 37(2), 25–35.
/>
Cheng, X., Fu, S., & de Vreede, G.-J. (2021). Determinants of trust in
computer-mediated offshore software-outsourcing collaboration.
International Journal of Information Management, 57(April), 102301.
/>
Cheng, X., Su, L., Luo, X., Benitez, J., & Cai, S. (2022). The good, the
bad, and the ugly: Impact of analytics and artificial intelligence-
enabled personal information collection on privacy and participation

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION 15

in ridesharing. European Journal of Information Systems, 31(3), 339– Gurtner, S., Reinhardt, R., & Soyez, K. (2014). Designing mobile busi-
363. ness applications for different age groups. Technological Forecasting
Chi, O. H., Jia, S., Li, Y., & Gursoy, D. (2021). Developing a formative and Social Change, 88(October), 177–188. /> scale to measure consumers’ trust toward interaction with artificially techfore.2014.06.020
intelligent (AI) social robots in service delivery. Computers in

Human Behavior, 118(May), 106700. Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2021). A pri-
2021.106700 mer on partial least squares structural equation modeling (PLS-SEM).
Chih, W.-H., Hsu, L.-C., & Liou, D.-K. (2017). Understanding virtual Sage Publications.
community members’ relationships from individual, group, and
social influence perspectives. Industrial Management & Data Hennig-Thurau, T., Aliman, D. N., Herting, A. M., Cziehso, G. P.,
Systems, 117(6), 990–1010. Linder, M., & Ku€bler, R. V. (2022). Social interactions in the meta-
0119 verse: Framework, initial evidence, and research roadmap. Journal of
Choung, H., David, P., & Ross, A. (2022). Trust in AI and its role in the Academy of Marketing Science. Advance online publication.
the acceptance of AI technologies. International Journal of Human– /> Computer Interaction. Advance online publication. /> 10.1080/10447318.2022.2050543 Henseler, J., Ringle, C. M., & Sarstedt, M. (2015). A new criterion for
CNNIC. (2021). The 42nd China statistical report on internet develop- assessing discriminant validity in variance-based structural equation
ment. modeling. Journal of the Academy of Marketing Science, 43(1), 115–135.
Cummings, L. L., & Bromiley, P. (1996). The organizational trust /> inventory (OTI). In R. M. Kramer & T. R. Tyler (Eds.), Trust in
organizations: Frontiers of theory and research (pp. 302–330). Sage Hong, S.-J., Lui, C. S. M., Hahn, J., Moon, J. Y., & Kim, T. G. (2013).
Publications, Inc. How old are you really? Cognitive age in technology acceptance.
Daft, R. L., & Lengel, R. H. (1986). Organizational information require- Decision Support Systems, 56(December), 122–130. /> ments, media richness and structural design. Management Science, 1016/j.dss.2013.05.008
32(5), 554–571. />Darbinyan, R. (2022). Virtual shopping in the metaverse: What is it and Hua, J., Chen, Y., & Luo, X. R. (2018). Are we ready for cyberterrorist
how will AI make it work. Forbes. attacks?—Examining the role of individual resilience. Information &
bestechcouncil/2022/03/16/virtual-shopping-in-the-metaverse-what- Management, 55(7), 928–938. /> is-it-and-how-will-ai-make-it-work/?sh=661defa85f27
Dennis, A. R., & Kinney, S. T. (1998). Testing media richness theory in Hudson, J. (2022). Virtual immersive shopping experiences in meta-
the new media: The effects of cues, feedback, and task equivocality. verse environments: Predictive customer analytics, data visualization
Information Systems Research, 9(3), 256–274. algorithms, and smart retailing technologies. Linguistic and
isre.9.3.256 Philosophical Investigations, 2022(21), 236–251. />Di Pietro, R., & Cresci, S. (2021). Metaverse: Security and Privacy Issues 22381/lpi21202215
[Paper presentation]. 2021 Third IEEE International Conference on
Trust, Privacy and Security in Intelligent Systems and Applications Hwang, R., & Lee, M. (2022). The influence of music content market-
(TPS-ISA) (pp. 281–288). ing on user satisfaction and intention to use in the metaverse: A
00032 focus on the SPICE model. Businesses, 2(2), 141–155. />Dwivedi, Y. K., Hughes, L., Baabdullah, A. M., Ribeiro-Navarrete, S., 10.3390/businesses2020010
Giannakis, M., Al-Debei, M. M., Dennehy, D., Metri, B., Buhalis, D.,
Cheung, C. M., Conboy, K., Doyle, R., Dubey, R., Dutot, V., Felix, Jenkins, T. (2022). Immersive virtual shopping experiences in the retail
R., Goyal, D. P., Gustafsson, A., Hinsch, C., Jebabli, I., … Wamba, metaverse: Consumer-driven E-commerce, blockchain-based digital
S. F. (2022). Metaverse beyond the hype: Multidisciplinary perspec- assets, and data visualization tools. Linguistic and Philosophical

tives on emerging challenges, opportunities, and agenda for research, Investigations, 2022(21), 154–169. /> practice and policy. International Journal of Information
Management, 66(October), 102542. Kesharwani, A. (2020). Do (how) digital natives adopt a new technol-
fomgt.2022.102542 ogy differently than digital immigrants? A longitudinal study.
Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical Information & Management, 57(2), 103170. /> power analyses using GÃ Power 3.1: Tests for correlation and regres- im.2019.103170
sion analyses. Behavior Research Methods, 41(4), 1149–1160. https://
doi.org/10.3758/BRM.41.4.1149 Kim, J. (2021). Advertising in the metaverse: Research agenda. Journal
Gefen, D. (2000). E-commerce: The role of familiarity and trust. Omega, of Interactive Advertising, 21(3), 141–144. /> 28(6), 725–737. 15252019.2021.2001273
Ghobadi, S., & Mathiassen, L. (2020). A generational perspective on
the software workforce: Precocious users of social networking in Kimiagari, S., & Malafe, N. S. A. (2021). The role of cognitive and
software development. Journal of Management Information Systems, affective responses in the relationship between internal and external
37(1), 96–128. stimuli on online impulse buying behavior. Journal of Retailing and
Goles, T., Rao, S. V., Lee, S., & Warren, J. (2009). Trust violation in Consumer Services, 61(July), 102567. /> electronic commerce: Customer concerns and reactions. Journal of conser.2021.102567
Computer Information Systems, 49(4), 1–9. /> 08874417.2009.11645335 Kishi, M. (2008). Perceptions and use of electronic media: Testing the
Grupac, M., Husakova, K., & Balica, R.-S¸. (2022). Virtual navigation and relationship between organizational interpretation differences and
augmented reality shopping tools, immersive and cognitive technolo- media richness. Information & Management, 45(5), 281–287. https://
gies, and image processing computational and object tracking algo- doi.org/10.1016/j.im.2008.02.008
rithms in the metaverse commerce. Analysis and Metaphysics,
2022(21), 210–226. Kliestik, T., Novak, A., & Lazaroiu, G. (2022). Live shopping in the
Grupac, M., & Lazaroiu, G. (2022). Image processing computational algo- metaverse: Visual and spatial analytics, cognitive artificial intelli-
rithms, sensory data mining techniques, and predictive customer ana- gence techniques and algorithms, and immersive digital simulations.
lytics in the metaverse economy. Review of Contemporary Philosophy, Linguistic and Philosophical Investigations, 2022(21), 187–202.
2022(21), 205–222. />
Komiak, S. Y. X., & Benbasat, I. (2006). The effects of personalization
and familiarity on trust and adoption of recommendation agents.
MIS Quarterly, 30(4), 941–960. />
Kraus, S., Kanbach, D. K., Krysta, P. M., Steinhoff, M. M., & Tomini,
N. (2022). Facebook and the creation of the metaverse: Radical busi-
ness model innovation or incremental transformation? International
Journal of Entrepreneurial Behavior & Research, 28(9), 52–77.
/>

Kwak, D.-H., Lee, S., Ma, X., Lee, J., Lara, K., & Brandyberry, A.
(2021). Announcement of formal controls as phase-shifting percep-
tions: Their determinants and moderating role in the context of
mobile loafing. Internet Research, 31(5), 1874–1898. /> 10.1108/INTR-10-2020-0581

16 L. ZHANG ET AL.

Lee, D. K. L., & Borah, P. (2020). Self-presentation on Instagram and outcomes. Information & Management, 45(1), 21–30. /> friendship development among young adults: A moderated medi- 10.1016/j.im.2007.09.003
ation model of media richness, perceived functionality, and open- Patil, K., & Pramod, D. (2022). Can Metaverse Retail lead to purchase
ness. Computers in Human Behavior, 103(February), 57–66. https:// intentions among the youth? A Stimulus-Organism-Response theory
doi.org/10.1016/j.chb.2019.09.017 perspective [Paper presentation]. 2022 ASU International Conference
in Emerging Technologies for Sustainability and Intelligent Systems
Lee, J., Lee, J.-N., & Tan, B. C. Y. (2015). Antecedents of cognitive (ICETSIS) (pp. 314–320). /> trust and affective distrust and their mediating roles in building cus- 2022.9888929
tomer loyalty. Information Systems Frontiers, 17(1), 159–175. https:// Pavlou, P. A. (2002). Institution-based trust in interorganizational
doi.org/10.1007/s10796-012-9392-7 exchange relationships: The role of online B2B marketplaces on trust
formation. The Journal of Strategic Information Systems, 11(3–4),
Legood, A., van der Werff, L., Lee, A., den Hartog, D., & van 215–243. /> Knippenberg, D. (2023). A critical review of the conceptualization, Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P.
operationalization, and empirical literature on cognition-based and (2003). Common method biases in behavioral research: A critical
affect-based trust. Journal of Management Studies, 60(2), 495–537. review of the literature and recommended remedies. The Journal of
Applied Psychology, 88(5), 879–903. /> 9010.88.5.879
Leong, L.-Y., Hew, T.-S., Ooi, K.-B., Chong, A. Y. L., & Lee, V.-H. Ronkko, M., Ylitalo, J. (2011). PLS marker variable approach to diag-
(2021). Understanding trust in ms-commerce: The roles of reported nosing and controlling for method variance [Paper presentation].
experience, linguistic style, profile photo, emotional, and cognitive Proceedings of the 32nd International Conference on Information
trust. Information & Management, 58(2), 103416. Systems. /> 1016/j.im.2020.103416 thods/8.
Rospigliosi, P. (2022). Metaverse or Simulacra? Roblox, Minecraft,
Li, K., Zhou, C., Luo, X. R., Benitez, J., & Liao, Q. (2022). Impact of Meta and the turn to virtual reality for education, socialisation and
information timeliness and richness on public engagement on social work. Interactive Learning Environments, 30(1), 1–3. /> media during COVID-19 pandemic: An empirical investigation 10.1080/10494820.2022.2022899
based on NLP and machine learning. Decision Support Systems, Ryder, B. (2022). Taobao launches ‘Metaverse Mall’ in time for China’s
162(November), 113752. 618 Shopping Festival. Jing Daily. /> launches-metaverse-mall-in-time-for-chinas-618-shopping-festival/
Liang, J., Gao, Q., Li, W., Shi, Y., Shen, M., & Gao, Z. (2022). Personality Shao, Z., Benitez, J., Zhang, J., Zheng, H., & Ajamieh, A. (2022).

Affects Dispositional Trust and History-Based Trust in Different Antecedents and performance outcomes of employees’ data analytics
Ways. International Journal of Human–Computer Interaction, 39(4), skills: An adaptation structuration theory-based empirical investigation.
949–960. European Journal of Information Systems. Advance online publication.
/>Liang, H., Xue, Y., Pinsonneault, A., & Wu, Y. A. (2019). What users do Shao, Z., & Yin, H. (2019). Building customers’ trust in the ridesharing
besides problem-focused coping when facing IT security threats: An platform with institutional mechanisms: An empirical study in
emotion-focused coping perspective. MIS Quarterly, 43(2), 373–394. China. Internet Research, 29(5), 1040–1063. /> INTR-02-2018-0086
Shao, Z., Zhang, L., Brown, S. A., & Zhao, T. (2022). Understanding
McAllister, D. J. (1995). Affect-and cognition-based trust as founda- users’ trust transfer mechanism in a blockchain-enabled platform: A
tions for interpersonal cooperation in organizations. Academy of mixed methods study. Decision Support Systems, 155(April), 113716.
Management Journal, 38(1), 24–59. /> Shao, Z., Zhang, L., Li, X., & Guo, Y. (2019). Antecedents of trust and
McKnight, D. H., Choudhury, V., & Kacmar, C. (2002). The impact of continuance intention in mobile payment platforms: The moderating
initial consumer trust on intentions to transact with a web site: A effect of gender. Electronic Commerce Research and Applications,
trust building model. The Journal of Strategic Information Systems, 33(January–February), 100823. /> 11(3–4), 297–323. 100823
Shao, Z., Zhang, L., Li, X., & Zhang, R. (2022). Understanding the role
Miller, B. K., & Simmering, M. J. (2022). Attitude toward the color blue: of justice perceptions in promoting trust and behavioral intention
An ideal marker variable. Organizational Research Methods. Advance towards ride-sharing. Electronic Commerce Research and
online publication. Applications, 51(January–February), 101119. /> j.elerap.2022.101119
Mirzaei, T., & Esmaeilzadeh, P. (2021). Engagement in online health Shen, X.-L., Li, Y.-J., Sun, Y., & Wang, F. (2021). Good for use, but
communities: Channel expansion and social exchanges. Information better for choice: A relative model of competing social networking
& Management, 58(1), 103404. services. Information & Management, 58(3), 103448. /> 103404 10.1016/j.im.2021.103448
Shin, D. (2022). The actualization of meta affordances: Conceptualizing
Nica, E., Poliak, M., Popescu, G. H., & P^arvu, I.-A. (2022). Decision affordance actualization in the metaverse games. Computers in
intelligence and modeling, multisensory customer experiences, and Human Behavior, 133(August), 107292. /> socially interconnected virtual services across the metaverse ecosys- 2022.107292
tem. Linguistic and Philosophical Investigations, 2022(21), 137–153. Standish, J. (2022). Retailers, meet me in the metaverse. https://www.
accenture.com/us-en/insights/retail/tech-vision
Suh, K. S. (1999). Impact of communication medium on task perform-
Niehaves, B., & Plattfaut, R. (2014). Internet adoption by the elderly: ance and satisfaction: An examination of media-richness theory.
Employing IS technology acceptance theories for understanding the Information & Management, 35(5), 295–312. /> age-related digital divide. European Journal of Information Systems, S0378-7206(98)00097-4
23(6), 708–726. Tams, S., Thatcher, J. B., & Grover, V. (2018). Concentration, compe-
tence, confidence, and capture: An experimental study of age, inter-

Nix, N. (2022). Facebook is opening a physical store to show off its vir- ruption-based technostress, and task performance. Journal of the
tual gadgets. The Washington Post. hingtonpost.
com/technology/2022/04/25/facebook-meta-store/

Oh, H. J., Kim, J., Chang, J. J. C., Park, N., & Lee, S. (2023). Social
benefits of living in the metaverse: The relationships among social
presence, supportive interaction, social self-efficacy, and feelings of
loneliness. Computers in Human Behavior, 139(February), 107498.
/>
Ollier-Malaterre, A., & Foucreault, A. (2021). When are social network
sites connections with coworkers beneficial? The roles of age differ-
ence and preferences for segmentation between work and life.
Journal of the Association for Information Systems, 22(5), 1454–1471.
/>
Otondo, R. F., Van Scotter, J. R., Allen, D. G., & Palvia, P. (2008). The
complexity of richness: Media, message, and communication

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION 17

Association for Information Systems, 19(9), 857–908. Yoo, Y., & Alavi, M. (2001). Media and group cohesion: Relative influ-
10.17705/1jais.00511 ences on social presence, task participation, and group consensus.
Tan, T. M., & Saraniemi, S. (2022). Trust in blockchain-enabled MIS Quarterly, 25(3), 371–390. /> exchanges: Future directions in blockchain marketing. Journal of the
Academy of Marketing Science. Advance online publication. https:// Zhang, K. Z. K., Cheung, C. M. K., & Lee, M. K. O. (2014). Examining
doi.org/10.1007/s11747-022-00889-0 the moderating effect of inconsistent reviews and its gender differ-
Taylor, C. R. (2022). Research on advertising in the metaverse: A call ences on consumers’ online shopping decision. International Journal
to action. International Journal of Advertising, 41(3), 383–384. of Information Management, 34(2), 89–98. /> ijinfomgt.2013.12.001
Tong, Y., Wang, X., Tan, C.-H., & Teo, H.-H. (2013). An empirical
study of information contribution to online feedback systems: A Zhang, S., & Leidner, D. (2018). From improper to acceptable: How
motivation perspective. Information & Management, 50(7), 562–570. perpetrators neutralize workplace bullying behaviors in the cyber
world. Information & Management, 55(7), 850–865. />Trevino, L. K. (1992). Experimental approaches to studying ethical- 10.1016/j.im.2018.03.012

unethical behavior in organizations. Business Ethics Quarterly, 2(2),
121–136. Zhang, L., Shao, Z., Li, X., & Feng, Y. (2021). Gamification and online
Tseng, C.-H., & Wei, L.-F. (2020). The efficiency of mobile media rich- impulse buying: The moderating effect of gender and age.
ness across different stages of online consumer behavior. International Journal of Information Management, 61(December),
International Journal of Information Management, 50(February), 102267. /> 353–364. />Vance, A., Siponen, M., & Pahnila, S. (2012). Motivating IS security Zhang, L., Wang, Y., Anjum, M. A., & Mu, J. (2022). The impacts of
compliance: Insights from habit and protection motivation theory. point rewarding and exchanging on users’ loyalty toward mobile
Information & Management, 49(3–4), 190–198. payment applications: A dual channeling perspective. Internet
1016/j.im.2012.04.002 Research, 32(6), 1832–1861. />Wang, W., Qiu, L., Kim, D., & Benbasat, I. (2016). Effects of rational 0414
and social appeals of online recommendation agents on cognition-
and affect-based trust. Decision Support Systems, 86(June), 48–60. Zhu, L., Benbasat, I., & Jiang, Z. (2010). Let’s shop online together: An
empirical investigation of collaborative online shopping support.
Wang, N., Shen, X.-L., & Sun, Y. (2013). Transition of electronic word- Information Systems Research, 21(4), 872–891. /> of-mouth services from web to mobile context: A trust transfer per- 1287/isre.1080.0218
spective. Decision Support Systems, 54(3), 1394–1403. /> 10.1016/j.dss.2012.12.015 Zvarikova, K., Machova, V., & Nica, E. (2022). Cognitive artificial intel-
Wang, X., Xu, Y. C., Lu, T., & Zhang, C. (2020). Why do borrowers ligence algorithms, movement and behavior tracking tools, and cus-
default on online loans? An inquiry of their psychology mechanism. tomer identification technology in the metaverse commerce. Review
Internet Research, 30(4), 1203–1228. of Contemporary Philosophy, 2022(21), 171–187. /> 05-2019-0183 22381/RCP21202211
Wongkitrungrueng, A., & Suprawan, L. (2023). Metaverse Meets
Branding: Examining Consumer Responses to Immersive Brand About the authors
Experiences. International Journal of Human–Computer Interaction.
Advance online publication. Lin Zhang is a PhD Candidate at the School of Management, Harbin
2175162 Institute of Technology, and a visiting student at EDHEC Business
Wu, W., Wang, S., Ding, G., & Mo, J. (2023). Elucidating trust-build- School. His research focuses on trust, and IT adoption, and published
ing sources in social shopping: A consumer cognitive and emotional articles in journals such as Decision Support Systems, Internet
trust perspective. Journal of Retailing and Consumer Services, Research, International Journal of Information Management, etc.
71(March), 103217. />Xi, N., Chen, J., Gama, F., Riar, M., & Hamari, J. (2022). The chal- Muhammad Adeel Anjum is an Assistant Professor of Management
lenges of entering the metaverse: An experiment on the effect of Sciences at Balochistan University of Information Technology,
extended reality on workload. Information Systems Frontiers. Enineering and Management Sciences (BUITEMS), Quetta, Pakistan.
Advance online publication. His area of interest is Organizational Behavior and HRM. His research
10244-x has been published in journals including Internet Research, Journal of
Management and Organization, etc.

Yanqing Wang is an Associate Professor of Management Science and
Engineering at the Harbin Institute of Technology. His research pri-
marily focuses on E-learning, computer-supported communication,
implementation, adoption, and diffusion of information technology.
His work has been published in academic journals, including
Computers and Education, Computers in Human Behavior, etc.

18 L. ZHANG ET AL.

Appendix A. Hypothetical scenario

The corresponding scenario

The term “Metaverse” originated in the 1992 science fiction novel Snow Crash. I n the novel, i t
reflects a virtual world that can be linked with the physical world, which creates a digital virtual
space with a new social system. With the development of Web 3.0 Technology, many enterprises
try to launch their own Metaverse solutions as a preemptive digital economy initiative, such as
Metaverse shopping.

Assume that you are using a Metaverse shopping application called “Metastore”. In the Metaverse
environment, you will keep a virtual avatar-centered perspective and act naturally and intuitively in the
3D environment through navigation. Meanwhile, you will be able to interact with other shoppers, real
store attendants, and their brand in a unique immersive environment that tells a brand’s story, the
dimensions of your desired products, and multiple rich product information. Clicking through a specific
store icon, you can turn into a more detailed product store and purchase immediately.

Section one: Based on the screenshot, please look at the described Metaverse shopping scenario
and imagine it in mind, and please choose which of the following statement is in line with your current
perception of the aforementioned hypothetical scenario:

Key:
1 = After reading the corresponding scenario, I still have no idea about Metaverse shopping
2 = After reading the corresponding scenario, I feel a little difficult to imagine Metaverse shopping
3 = After reading the corresponding scenario, I can imagine the Metaverse shopping environment to some extent
4 = After reading the corresponding scenario, I can easily imagine the Metaverse shopping environment.

Section two: Based on the screenshot, please look at the described Metaverse shopping scenario
and imagine it in mind, and indicate the extent to which you agree or disagree with each of the
following statements. Use the key below to determine your response:
Key:
1 = Strongly disagree
2 = Disagree
3 = Somewhat disagree
4 = Neutral
5 = Somewhat agree
6 = Agree
7 = Strongly agree

Measurements of each variable included in our study 1234567

Section three: demographic information, including gender, age, etc.

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION 19

Appendix B. PLS analysis process based on Benitez et al. (2020)

Two-step estimation strategy Focus Operationalization procedural Function description

Step 1 First-order constructs 1. Create first-order constructs (including 1. Collect the standardized latent
Step 2 Second-order constructs all reflective constructs) variable scores (LVS) of multiple

information cues, language variety,
2. Freely link all the constructs, keep the immediate feedback, and personal
algorithm settings by default and run focus, respectively.
it on the ADANCO 2.2 application
2. Assess the first-order-goodness of
1. Create the second-order constructs as model fit (saturated model).
follows: (a) collect the LVS from Step
1; (b) copy the LVS and paste it into 1. Build the second-order construct of
the original dataset; (c) import the perceived media richness of the
new dataset (with the LVS) metaverse

2. Create the second-order constructs 2. Assess the second-order-goodness
using the LVS of model fit (saturated model)

3. Freely link the second-order construct 3. Assess the second-order-goodness
with all reflective constructs (saturated of model fit (estimated model)
model)
4. Analyze the structural model to
4. Create our conceptual model, examine the path relationship and
including the key variables, and link explanatory power of the research
them in the proposed way (estimated model (estimated model)
model)

Notes: Perceived media richness of the metaverse was specified as a second-order formative construct determined by multiple information cues, language variety,
immediate feedback, and personal focus. These four constructs were specified as reflective at the first-order level (Benitez, Ruiz, et al., 2022; Braojos et al.,
2019). We followed a two-step estimation strategy to describe the operationalization process: (1) the first step was to collect the standardized latent variable
scores (LVS) of the first-order constructs (i.e., multiple information cues, language variety, immediate feedback, and personal focus), and (2) the second step is
to create the second-order construct (i.e., perceived media richness of the metaverse) using LVS of the first-order constructs (Benitez et al., 2020).

Appendix C. Common method bias analysis

Relationship Baseline model without the marker CMB test model with the marker

Perceived media richness!Cognitive trust 0.338ÃÃÃ 0.337ÃÃÃ
Perceived media richness!Affective trust 0.109Ã 0.109Ã
Cognitive trust!Affective trust 0.569ÃÃÃ 0.573ÃÃÃ
Cognitive trust!Purchase intention 0.218Ã 0.216Ã
Affective trust!Purchase intention 0.255Ã 0.252Ã
Gender!Purchase intention 0.190ÃÃÃ 0.195ÃÃÃ
Education!Purchase intention 0.033 ns 0.0366 ns
Online shopping experience!Purchase intention 0.008 ns 0.006 ns
Marker variable!Cognitive trust –0.033 ns
Marker variable !Affective trust N/A –0.022 ns
Marker variable!Purchase intention –0.052 ns
N/A
Notes: ns not significant; Ãp < 0.05; ÃÃÃp < 0.001.
N/A