Article

The Impact of Virtual Streamer Anthropomorphism on

Consumer Purchase Intention: Cognitive Trust as a Mediator
Chunyu Li 1,2 and Fei Huang 1, *

1 Seoul Business School, aSSIST University, Seoul 03767, Republic of Korea; lichunyu@stud.assist.ac.kr
2 School of Digital Commerce, Heilongjiang Polytechnic, Harbin 150070, China
* Correspondence: huangfei@assist.ac.kr

Abstract: As an important tool for brand promotion and marketing, the status of virtual streamers
is gradually improving, especially in the Chinese market with a huge Internet user base. Virtual
streamer anthropomorphism has gradually become an important research content in the field of con-
sumer behavior. However, the specific mechanism by which the multidimensional anthropomorphic
characteristics of virtual streamers affect consumer trust and purchase intention requires further
investigation. Therefore, based on the avatar theory, this research explores how the anthropomorphic
characteristics of virtual streamers affect consumer purchase intention through cognitive trust. The
analysis was performed using SPSS 27.0 and AMOS 24.0, establishing a structural equation model.
Through the analysis of questionnaire data from 503 Chinese consumers, it was found that behavioral
anthropomorphism, cognitive anthropomorphism, and emotional anthropomorphism all exert a
notable influence on cognitive trust. Appearance anthropomorphism and emotional anthropomor-
phism directly affect purchase intention, and cognitive trust has a significant impact on purchase
intention. Moreover, cognitive trust fully mediates the effects of behavioral anthropomorphism and
cognitive anthropomorphism on purchase intention and partially mediates the effects of emotional
anthropomorphism on purchase intention. This study enriches the application of avatar theory in vir-
tual streamers in live e-commerce and provides theoretical backing for virtual streamer development
and enterprise marketing strategies. It also offers practical insights to help brands optimize virtual
streamers and improve consumer participation and purchase conversion rates.

Keywords: virtual streamer; anthropomorphism; cognitive trust; purchase intention; live streaming
Citation: Li, C.; Huang, F. The Impact
e-commerce
of Virtual Streamer Anthropomorphism
on Consumer Purchase Intention:
Cognitive Trust as a Mediator. Behav.
Sci. 2024, 14, 1228. https://doi.org/
1. Introduction
10.3390/bs14121228
As digital technology has advanced rapidly in recent years, virtual streamers have
Academic Editor: Minseong Kim become widely utilized in e-commerce live streaming, establishing themselves as a vital
Received: 13 November 2024
tool for brand marketing. Virtual streamers can establish interactive relationships with
Revised: 12 December 2024 consumers, bridging the gap between brands and consumers and thus demonstrating sig-
Accepted: 18 December 2024 nificant market potential [1,2]. In China, by 2024, the Internet user population is estimated
Published: 20 December 2024 to approach 1.1 billion, with the penetration rate reaching 78.0%. The number of online
video users has grown to 1.068 billion, while e-commerce users have reached 905 million.
Additionally, the number of live streaming users has approached 800 million [3]. This
highlights China’s vast Internet user base and the promising prospects of its live streaming
Copyright: © 2024 by the authors. e-commerce market. The metaverse and virtual reality, as two prominent innovations,
Licensee MDPI, Basel, Switzerland.
are transforming the digital landscape by merging the real and virtual worlds to create
This article is an open access article
immersive experiences for audiences. Within the metaverse, the value of e-commerce is
distributed under the terms and
projected to reach an estimated USD 2.6 trillion [4].
conditions of the Creative Commons
As artificial intelligence and animation technologies have progressed, digital characters
Attribution (CC BY) license (https://
have been widely applied in scenarios such as social media and e-commerce live streaming.
creativecommons.org/licenses/by/
4.0/).
Research shows that the anthropomorphic features of virtual characters can enhance users’

Behav. Sci. 2024, 14, 1228. https://doi.org/10.3390/bs14121228 https://www.mdpi.com/journal/behavsci

Behav. Sci. 2024, 14, 1228 2 of 18

willingness to interact and emotional resonance, thereby promoting engagement between

users and digital personas [5,6]. Within live stream e-commerce settings, virtual streamers,
as a new form of interaction, are gradually becoming key factors in enhancing consumer
engagement and increasing purchase intention [7]. Anthropomorphic design can enhance
users’ sense of trust and willingness to participate by simulating social interactions, which
is particularly significant in the process of interaction between virtual streamers and
consumers [8]. Social media influencers, in contrast to traditional celebrity endorsements,
offer greater credibility due to their closer connection with users, providing a reference for
researching virtual streamers [9].
Although existing research has increasingly concentrated on applying streamer at-
tributes within live stream e-commerce settings, the majority has examined the features
of human streamers operating on live or social media platforms. There is still limited
exploration of how the multidimensional features of virtual streamers influence audience
trust and buying intention [10]. Furthermore, studies have indicated that characteristics
such as the likability and reactivity of virtual streamers can significantly enhance buying in-
tention by increasing users’ sense of presence. This highlights the significant role of virtual
streamers in e-commerce live streaming, specifically in promoting purchasing behavior by
enhancing user interaction experiences [11].
Therefore, this research seeks to explore how the four anthropomorphic dimensions of
virtual streamers, namely appearance anthropomorphism, behavioral anthropomorphism,
cognitive anthropomorphism, and emotional anthropomorphism, affect consumer purchase
intention through cognitive trust, thereby filling this research gap.
This research uses a questionnaire survey to gather data from Chinese consumers and
utilizes structural equation modeling (SEM) to examine the associations among different
anthropomorphic dimensions, cognitive trust, and purchase intention, along with testing
mediating effects. The study provides three key contributions. First, we extend avatar
theory into the realm of virtual streamers, detailing the distinct impact pathways of the
four anthropomorphic dimensions of appearance, behavior, cognition, and emotion on con-
sumer trust and purchase intention. Secondly, by introducing cognitive trust as a mediating
variable, we reveal how the various anthropomorphic characteristics of virtual streamers
indirectly or directly influence consumers’ purchasing decisions, providing practical guid-
ance for brands in design and communication approaches for virtual streamers. Finally,
this research establishes a theoretical foundation for future exploration of the application of
virtual streamers in different cultural contexts and product types, expanding the application
scenarios of anthropomorphism and trust theory in digital marketing.
The structure of this study is organized as follows: Section 2 presents a thorough
discussion of the theoretical foundation and research hypotheses, including an introduction
to avatar theory, a discussion on the anthropomorphism of virtual streamers, and a literature
review on cognitive trust and purchase intention. It concludes with the formulation of
research hypotheses and the development of the research model. Section 3 elaborates on
the research methodology, encompassing the study design, measurement of variables, and
data collection methods. Section 4 presents the empirical analysis, including reliability
and validity tests, common method bias examination, path analysis, and mediation effect
analysis. Section 5 focuses on discussion and conclusions. This section identifies the
limitations of the study and proposes directions for future research. It also provides a
detailed discussion of the results, highlighting their theoretical and practical implications.
Finally, the section concludes with a comprehensive summary of the entire study.

2. Theoretical Background and Research Hypotheses

2.1. Avatar Theory and Virtual Streamers’ Anthropomorphism
Virtual figures have become increasingly vital in contemporary marketing strategies,
particularly as digital technology advances, as companies are gradually adopting virtual
avatars for brand endorsement to innovate marketing approaches. Avatar theory is widely
applied in advertising and brand promotion, positing that virtual avatars not only influence
Behav. Sci. 2024, 14, 1228 3 of 18

users’ emotional experiences and social interactions but also shape consumers’ perceptions
of brands [12,13].
Research has found that virtual self-discrepancy (the gap between users’ self-image
in virtual environments and their real self) significantly diminishes users’ self-presence
and immersion. Additionally, virtual other-discrepancy (differences in the appearance of
virtual others) indirectly impacts users’ trust and interaction smoothness by influencing
their social presence [14]. Virtual avatars also demonstrate potential for enhancing user
immersion and encouraging positive behaviors. For instance, virtual characters in games
can motivate users to adopt healthier lifestyles in real life [15], and personalized avatars
have been shown to help users alleviate anxiety [16].
The effectiveness of avatars depends on their realism in form and behavior, encom-
passing basic avatars, visually realistic but functionally limited avatars, intelligent avatars,
and fully human-like digital avatars in both appearance and intelligence. Different avatar
designs affect purchase intentions in various ways, offering companies guidance for appli-
cations across diverse contexts [17]. In metaverse environments, highly realistic avatars
strengthen users’ sense of self-presence and emotional connection, thereby increasing their
intention to use [5]. In summary, virtual avatars significantly influence consumer behav-
ior, brand perception, and emotional management, providing fresh strategic insights for
marketing in domains such as real-time e-commerce.
In the development of the Brand Anthropomorphism Scale (BASC), brand anthropo-
morphism was categorized into four dimensions: appearance, moral virtues, cognitive
perception, and emotional awareness. Validation studies indicated that these dimensions
positively influence brand trust and commitment [18]. Additionally, anthropomorphism
involves assigning human-like traits, motivations, or emotions to non-human agents,
making AI assistants more personable and prompting positive social and psychological
responses from users [19]. A virtual avatar, meanwhile, is a digital entity with an anthro-
pomorphic appearance, controlled by humans or software, and capable of interaction. Its
design elements include appearance realism and behavioral realism [17]. Similarly, virtual
humans are highly human-like characters, including virtual influencers, streamers, and
idols. Despite their rapid development, how users perceive these anthropomorphized
characters remains unclear [20]. Furthermore, virtual influencers’ anthropomorphic charac-
teristics are divided into four key aspects: appearance, moral virtues, cognitive awareness,
and emotional sensitivity, representing the human-like attributes of virtual entities [6].
Building on prior studies, this research classifies virtual streamer anthropomorphism into
four categories: appearance anthropomorphism, behavioral anthropomorphism, cognitive
anthropomorphism, and emotional anthropomorphism.

2.2. Virtual Streamers’ Anthropomorphism and Cognitive Trust

Cognitive trust represents a type of trust based on rational evaluation, relying on
objective judgments of another’s competence, reliability, and professionalism, with a focus
on the ability to fulfill responsibilities and behavioral consistency [21]. In e-commerce
contexts, cognitive trust stems from consumers’ assessments of the platform’s expertise,
goodwill, and integrity, helping them mitigate risks in an environment of information
asymmetry [22]. Without direct interpersonal interaction, users establish cognitive trust
through rational assessments of merchants’ reliability and honesty, which is crucial in their
purchasing decisions [23]. Consumers typically form cognitive trust in service providers
by evaluating their reputation and track record, thus enhancing the stability of service
relationships [24]. Within real-time e-commerce environments, cognitive trust depends on
viewers’ assessments of a streamer’s expertise and reliability; for instance, a streamer’s
timely response strategies can significantly strengthen this trust [25]. The essence of cog-
nitive trust lies in evaluating others’ professional quality and task performance, which
provides a stable trust foundation, especially in highly uncertain collaborative contexts [26].
Additionally, in live streaming, consumers’ cognitive trust in a CEO derives from perceived
expertise and reliability, fostering brand acceptance and engagement [27]. Research sug-
Behav. Sci. 2024, 14, 1228 4 of 18

gests that quality of interaction, empathy, and anthropomorphic psychological traits are
key factors in enhancing AI device acceptance in the service industry. When AI devices
demonstrate empathy and high-quality interactions, particularly by understanding and
responding to user needs, trust and acceptance levels increase significantly, facilitating
everyday use [28]. Anthropomorphic features such as a human-like voice, cognitive intelli-
gence, and interaction etiquette have also been shown to enhance social presence, thereby
increasing trust, interaction frequency, user reliance, and engagement [29]. Studies indicate
that for low-cost environmental behaviors, highly anthropomorphic virtual influencers
who appear similar to the audience increase trust and engagement intention; conversely,
for high-cost behaviors, appearance dissimilarity is more effective in enhancing trust and
engagement [30]. Accordingly, this study proposes the following hypotheses:

H1: Virtual streamer anthropomorphism positively affects cognitive trust.

H1a: Appearance anthropomorphism positively affects cognitive trust.

H1b: Behavioral anthropomorphism positively affects cognitive trust.

H1c: Cognitive anthropomorphism positively affects cognitive trust.

H1d: Emotional anthropomorphism positively affects cognitive trust.

2.3. Virtual Streamers’ Anthropomorphism and Purchase Intention

Virtual streamers’ characteristics impact consumers’ purchase intention by enhancing
social and transmission presence. Specifically, virtual streamers’ likability and responsive-
ness directly increase purchase intention and also indirectly enhance it by improving social
and transmission presence. Moreover, the vitality of virtual streamers further contributes
to purchase intention indirectly [11]. Additionally, virtual influencers’ anthropomorphic
features positively impact purchase intention by increasing trustworthiness and parasocial
bonds, particularly as ethical values and cognitive awareness significantly boost consumer
trust and purchase intentions [6]. Accordingly, the study proposes the following hypotheses:

H2: Virtual streamer anthropomorphism positively affects purchase intention.

H2a: Appearance anthropomorphism positively affects purchase intention.

H2b: Behavioral anthropomorphism positively affects purchase intention.

H2c: Cognitive anthropomorphism positively affects purchase intention.

H2d: Emotional anthropomorphism positively affects purchase intention.

2.4. Cognitive Trust and Purchase Intention

In e-commerce, cognitive trust is considered a core driver of purchase intention.
Without face-to-face interaction, users establish trust through a platform’s professionalism,
security, and merchant integrity, which increases their willingness to make purchases [23].
Cognitive trust is mainly influenced by utilitarian value, such as saving time, reducing
costs, and improving efficiency, which effectively promotes purchase intention [31]. In
metaverse shopping scenarios, higher cognitive trust allows consumers to make purchasing
decisions with greater confidence, particularly impacting older consumers, while younger
consumers are more driven by emotional trust [32]. Drawing from these insights, this
hypothesis is formulated:

H3: Cognitive trust positively affects purchase intention.

Behav. Sci. 2024, 14, 1228 5 of 18

2.5. Mediating Role of Cognitive Trust

Consumers’ intrinsic and extrinsic motivations for online shopping can influence
their purchase intentions through cognitive trust [31]. In addition, studies have shown
that the physical attractiveness and attitude homophily of social network influencers im-
pact their credibility, which, in turn, positively influences users’ intentions to purchase
niche products [10]. Within live stream e-commerce, cognitive trust serves as a media-
tor between a streamer’s response strategies and audience feedback. Positive response
strategies enhance cognitive trust, stimulating positive word-of-mouth, while avoidance
strategies decrease trust, leading to negative feedback [25]. Cognitive trust is also a crucial
mediator between streamer characteristics and impulsive buying behavior, increasing the
credibility of the streamer’s information, reducing decision uncertainty, and prompting
consumers to make quicker purchase decisions [33]. In virtual streamer broadcasts, an-
thropomorphic features make virtual streamers easier to understand and trust. Cognitive
trust converts consumers’ initial impressions into trust in the streamer’s recommendations,
thereby increasing purchase intention [7]. Building on the above, the study suggests the
following hypotheses:

H4: Cognitive trust mediates the effect of virtual streamer anthropomorphism on purchase intention.

H4a: Cognitive trust mediates the effect of appearance anthropomorphism on purchase intention.

H4b: Cognitive trust mediates the effect of behavioral anthropomorphism on purchase intention.

H4c: Cognitive trust mediates the effect of cognitive anthropomorphism on purchase intention.

H4d: Cognitive trust mediates the effect of emotional anthropomorphism on purchase intention.

2.6. Research Model

Overall, this study examines how virtual streamer anthropomorphism influences
purchase intention through cognitive trust. As shown in Figure 1, the research model cate-
gorizes virtual streamer anthropomorphism into four dimensions as independent variables:
appearance anthropomorphism, behavioral anthropomorphism, cognitive anthropomor-
Behav. Sci. 2025, 15, x FOR PEER REVIEW 6 of 20
phism, and emotional anthropomorphism. Cognitive trust in virtual streamers serves as
the mediating variable, and purchase intention as the dependent variable.

Figure 1. Research
Figure 1. Research model.

3. Research Methodology
3.1. Research Design
This study used SPSS 27.0 and AMOS 24.0 software to analyze the data. First, SPSS
Behav. Sci. 2024, 14, 1228 6 of 18

3. Research Methodology
3.1. Research Design
This study used SPSS 27.0 and AMOS 24.0 software to analyze the data. First, SPSS
27.0 was employed to conduct a reliability analysis and exploratory factor analysis (EFA)
on the survey data. The reliability analysis, using Cronbach’s Alpha, assessed the internal
consistency of the scales, while the EFA examined the relationships between items and
latent variables, providing a robust structural foundation for a subsequent confirmatory fac-
tor analysis (CFA). Next, AMOS 24.0 was used to perform the CFA, evaluating the model’s
fit and calculating the composite reliability (CR) and average variance extracted (AVE) to
verify convergent validity. Discriminant validity was confirmed by comparing the square
roots of the AVE with the correlations between constructs. To address potential common
method bias, a single-factor test was conducted using AMOS. The results indicated that
the single-factor model’s fit was significantly worse than that of the measurement model,
suggesting that common method bias did not pose a significant concern. Furthermore,
SPSS was employed to perform multicollinearity diagnostics. The variance inflation factor
(VIF) values for all variables were below 3, indicating the absence of multicollinearity
issues. Finally, using SEM in AMOS, the research analyzed the path relationships be-
tween the anthropomorphic characteristics of virtual streamers and consumers’ purchase
intentions. Additionally, the mediation effect of cognitive trust was tested through the
Bootstrap method to thoroughly validate the research hypotheses. The adoption of SEM
was particularly suitable for this study as it allowed for precise analysis of the complex
causal relationships among latent variables and quantified the mediating role of cogni-
tive trust in the relationship between virtual streamer anthropomorphism and consumers’
purchase intentions.
This study focuses on virtual streamers and selected Chinese consumers as the research
sample for the following reasons. First, by 2024, the live streaming user base in China will
surpass 700 million, representing 70.6% of all Internet users. This indicates that China’s
live streaming e-commerce market is vast. Moreover, virtual digital humans are widely
applied in China’s live streaming e-commerce sector. Many companies have launched
virtual live streaming hosts who introduce products and engage in live broadcasts [3].
Secondly, virtual streamers have gained notable recognition within the cultural context of
China. Numerous virtual streamers with significant fan bases have emerged on Chinese
live streaming and video platforms, suggesting that Chinese audiences are receptive to
technological innovation and virtual characters [34]. Based on the rapid growth of China’s
live streaming e-commerce industry and the rising popularity of virtual streamers, this
study utilizes data collected from Chinese consumers for analysis.

3.2. Variable Measurement

The development of the measurement scale in this study followed a systematic process
comprising several stages.
First, the initial measurement items were designed by referencing established scales
from relevant literature. All items were evaluated using a 5-point Likert scale ranging from
1 (“strongly disagree”) to 5 (“strongly agree”). During the initial questionnaire design phase,
feedback was solicited from several scholars in the field of management, and the question-
naire was continuously refined and optimized based on their input. For items adapted
from English-language scales, a rigorous process of translation and back-translation was
conducted multiple times to enhance linguistic accuracy and conceptual alignment.
Secondly, a pilot study was conducted to screen and optimize the items. Using SPSS
27.0, an exploratory factor analysis was performed on the pilot survey data. Items with
low factor loadings or insufficient alignment with the intended constructs were removed,
resulting in an optimized formal scale.
Finally, SPSS 27.0 was used to analyze the valid data from the formal survey to verify
the reliability and validity of the finalized scale. At this stage, no items were removed from
the formal scale. The finalized measurement items are presented in Table 1.
Behav. Sci. 2024, 14, 1228 7 of 18

Table 1. Measurement items.

Variables Items Reference

A1 The virtual streamer looks human-like.
Appearance Anthropomorphism A2 The virtual streamer resembles a real human. [7,18]
A3 The virtual streamer has a human-like appearance.
B1 The virtual streamer’s movements appear natural.
B2 The virtual streamer’s voice sounds natural.
Behavioral Anthropomorphism [7,28]
B3 The virtual streamer has freedom of action.
B4 The virtual streamer has decision-making ability.
C1 The virtual streamer has consciousness.
C2 The virtual streamer has a mind of its own.
Cognitive Anthropomorphism [19,28]
C3 The virtual streamer is creative and has imagination.
C4 The virtual streamer is capable of reasoning.
E1 The virtual streamer has its own emotions.
E2 The virtual streamer feels remorse for actions it deems shameful.
E3 The virtual streamer can empathize with people who feel down.
Emotional Anthropomorphism [7,18]
E4 The virtual streamer feels guilt when it hurts someone.
E5 The virtual streamer feels shame when people have negative views and
judgments about it.
CT1 The virtual streamer is trustworthy.
CT2 I believe what the virtual streamer says.
CT3 The virtual streamer is reliable.
Cognitive Trust [33,34]
CT4 There is no need to worry at all when dealing with the
virtual streamer.
CT5 I believe in the expertise and capabilities of the virtual streamer.
PIN1 I would purchase the products promoted by the virtual streamer
during the live streaming.
PIN2 I intend to purchase the products promoted by the virtual streamer
during the live streaming.
PIN3 I would make the virtual streamer’s live streaming my preferred [34,35]
Purchase Intention
shopping channel.
PIN4 I am willing to recommend the products promoted by the virtual
streamer to my friends and family.
PIN5 I plan to frequently use the virtual streamer’s live streaming for
shopping in the future.

3.3. Data Collection Process

This study distributed an electronic questionnaire to Chinese respondents via an online
survey platform (http://www.wjx.cn). A pilot survey was conducted first, targeting users
who were familiar with or had watched virtual streamers. A total of 79 valid responses
were collected. Based on the analysis of the pilot survey data, the researchers refined the
questionnaire items to develop the formal questionnaire.
The formal questionnaire began with an introduction and instructions for respondents.
To ensure the relevance of the sample, a screening question was included: “Are you familiar
with or have you watched live streams hosted by virtual streamers?” Respondents who
answered “No” were excluded. Additionally, attention-check questions were embedded
Behav. Sci. 2024, 14, 1228 8 of 18

in the questionnaire to further ensure data quality. Responses failing these checks were
excluded from the analysis.
The formal survey was launched in October 2024 and remained open for approxi-
mately one month. After applying the screening criteria, a total of 503 valid responses
were collected.
Table 2 presents the descriptive statistics of the sample. Among the 503 respondents,
46.5% were male, and 53.5% were female. The largest age group was 26–35 years old,
accounting for 41.7% of the total sample, while respondents aged 18–35 comprised 68.7%,
indicating that live streaming e-commerce users are predominantly young. In terms
of educational attainment, 58.8% of the respondents held a bachelor’s degree or higher.
Regarding monthly income, 46.9% reported an income between RMB 5001 and 10,000.

Table 2. Demographic information of the participants (n = 503).

Characteristic Category Frequency Percentage (%)

Male 234 46.5
Gender
Female 269 53.5
18–25 136 27
26–35 210 41.7
Age (years)
36–45 128 25.4
45 or above 29 5.8
High school or below 57 11.3
Associate degree 150 29.8
Education level
Bachelor’s degree 224 44.5
Master’s degree or above 72 14.3
RMB 3000 or below 39 7.8
RMB 3001–5000 112 22.3
Average monthly income
RMB 5001–10,000 236 46.9
Above RMB 10,000 116 23.1

Overall, the distribution of sample characteristics aligns closely with patterns observed
in prior research [36–38]. These findings suggest that the sample effectively represents
the core user group of China’s live streaming e-commerce market—young individuals
with higher education levels and disposable income. This alignment underscores the
representativeness of the sample within the target population, providing a reliable data
foundation for this study.

4. Data Analysis and Results

4.1. Reliability and Validity Analysis
SPSS 27.0 was used to evaluate the internal consistency of the scales through reliability
analysis. The results indicated that the Cronbach’s Alpha values for all constructs were
above 0.7 (refer to Table 3), with appearance anthropomorphism at 0.825, behavioral anthro-
pomorphism at 0.870, cognitive anthropomorphism at 0.866, emotional anthropomorphism
at 0.894, cognitive trust at 0.892, and purchase intention at 0.871, indicating good reliability
of the scales [39,40].
Subsequently, the data were analyzed for validity using SPSS. The results are shown
in Table 3. Bartlett’s test of sphericity showed significant results (χ2 = 8520.214, df = 325,
p < 0.001), and the KMO value was 0.953, indicating that it is suitable for the next factor
analysis [41]. According to the EFA of the data using SPSS, the factor loadings of all
measurement items exceeded 0.6 (see Table 3). This shows that each measurement item has
a significant correlation with the construct to which it belongs and has strong explanatory
power for the construct.
Confirmatory factor analysis (CFA) was conducted in AMOS 24.0 to confirm the
structural validity of the measurement model. The model fit was good, with specific
indicators shown in Table 4: χ2 = 389.554, χ2 /df = 1.372, RMR = 0.032, RMSEA = 0.027,
Behav. Sci. 2024, 14, 1228 9 of 18

GFI = 0.943, AGFI = 0.930, NFI = 0.955, RFI = 0.949, IFI = 0.987, TLI = 0.986, and CFI = 0.987,
all meeting the recommended standards and indicating good model fit.

Table 3. Reliability and exploratory factor analyses.

Construct Items Factor Loading Cronbach’s Alpha

A1 0.755
Appearance Anthropomorphism A2 0.698 0.825
A3 0.788
B1 0.708
B2 0.752
Behavioral Anthropomorphism 0.870
B3 0.677
B4 0.654
C1 0.755
C2 0.74
Cognitive Anthropomorphism 0.866
C3 0.675
C4 0.743
E1 0.761
E2 0.748
Emotional Anthropomorphism E3 0.702 0.894
E4 0.703
E5 0.709
CT1 0.759
CT2 0.681
Cognitive Trust CT3 0.702 0.892
CT4 0.733
CT5 0.752
PIN1 0.786
PIN2 0.707
Purchase Intention PIN3 0.713 0.871
PIN4 0.669
PIN5 0.719
KMO 0.953
Approx.χ2 8520.214
Bartlett’s Test df 325
Sig. 0.000

Table 4. Model fit indices.

Fitting Index CMIN/DF RMR RMSEA GFI AGFI NFI RFI IFI TLI CFI
Criterion <3 <0.05 <0.08 >0.9 >0.9 >0.9 >0.9 >0.9 >0.9 >0.9
Actual value 1.372 0.032 0.027 0.943 0.930 0.955 0.949 0.987 0.986 0.987

In addition, this study tested the standardized coefficients in AMOS, and the re-
sults showed that the standardized coefficients of all measurement items were significant
(p < 0.001). The coefficients and significance levels are shown in Table 5. Convergent va-
lidity was evaluated through composite reliability (CR) and average variance extracted
(AVE). As shown in Table 5, all constructs had CR values exceeding 0.7, and the AVE values
were above 0.5, satisfying the criteria for convergent validity [41]. This indicates that the
measurement items exhibit good consistency at the construct level and sufficiently explain
the latent variables.
Behav. Sci. 2024, 14, 1228 10 of 18

Table 5. Test of convergent validity.

Path Estimate S.E. C.R. p Std. Estimate CR AVE

A3 ← 1 0.745
A2 ← AA 0.899 0.059 15.327 *** 0.685 0.839 0.639
A1 ← 1.268 0.064 19.730 *** 0.945
B4 ← 1 0.752
B3 ← 1.054 0.060 17.577 *** 0.770
BA 0.874 0.637
B2 ← 0.963 0.058 16.688 *** 0.735
B1 ← 1.225 0.058 21.138 *** 0.921
C4 ← 1 0.724
C3 ← 1.120 0.069 16.331 *** 0.756
CA 0.869 0.626
C2 ← 1.099 0.066 16.601 *** 0.768
C1 ← 1.306 0.068 19.207 *** 0.904
E5 ← 1 0.779
E4 ← 0.933 0.054 17.277 *** 0.728
E3 ← EA 1.035 0.054 19.128 *** 0.791 0.896 0.635
E2 ← 0.955 0.054 17.778 *** 0.745
E1 ← 1.209 0.052 23.114 *** 0.926
CT1 ← 1 0.922
CT2 ← 0.798 0.040 20.035 *** 0.721
CT3 ← CT 0.863 0.039 22.252 *** 0.767 0.894 0.629
CT4 ← 0.855 0.038 22.416 *** 0.770
CT5 ← 0.832 0.037 22.417 *** 0.770
PIN5 ← 1 0.721
PIN4 ← 1.041 0.064 16.155 *** 0.749
PIN3 ← PIN 1.004 0.064 15.739 *** 0.730 0.875 0.585
PIN2 ← 0.920 0.061 15.098 *** 0.701
PIN1 ← 1.218 0.063 19.252 *** 0.905
Note: *** means p < 0.001. AA—appearance anthropomorphism; BA—behavioral anthropomorphism;
CA—cognitive anthropomorphism; EA—emotional anthropomorphism; CT—cognitive trust; PIN—purchase
intention.

To assess discriminant validity, the square root of each construct’s AVE was compared
against correlations with other constructs. Table 6 demonstrates that each construct’s AVE
square root surpasses its correlations with other constructs, confirming strong discriminant
validity for the scales [42].

Table 6. Test of discriminant validity.

1 2 3 4 5 6
1. Emotional Anthropomorphism 0.797
2. Cognitive Anthropomorphism 0.687 0.791
3. Behavioral
0.687 0.693 0.798
Anthropomorphism
4. Appearance
0.593 0.616 0.714 0.799
Anthropomorphism
5. Cognitive Trust 0.683 0.648 0.694 0.580 0.793
6. Purchase Intention 0.656 0.605 0.632 0.625 0.650 0.765

4.2. Common Method Bias Test and Multicollinearity Analysis

As the data were collected through a survey, there was a possibility of common method
bias. To address this concern, we conducted a confirmatory factor analysis of a single-
factor model using AMOS [43]. Specifically, we constructed a model where all measured
items loaded onto a single common factor to assess the fit of this single-factor model. The
results indicated that the fit indices of the single-factor model (χ2 /df = 8.312, RMR = 0.086,
RMSEA = 0.121, GFI = 0.669, AGFI = 0.612, NFI = 0.714, RFI = 0.689, IFI = 0.739, TLI = 0.716,
and CFI = 0.738) were worse than those of the measurement model, suggesting that
common method bias was not a serious issue in this study. To further assess the presence
Behav. Sci. 2024, 14, 1228 11 of 18

of multicollinearity among the independent variables in the model, we analyzed the data
using SPSS and calculated the variance inflation factor (VIF) [44]. The results showed that
the VIF values for all independent variables ranged from 1.804 to 2.186, all of which were
less than 3, indicating that there was no serious multicollinearity problem in this study.

4.3. Path Analysis and Hypothesis Testing

SEM in AMOS 24.0 was used to analyze the relationships among virtual streamer
anthropomorphism, cognitive trust, and purchase intention, and to test the hypotheses [45].
Table 7 presents the results.

Table 7. Path coefficients analysis.

Hypothesis Estimate S.E. C.R. p β Result

AA → CT 0.075 0.067 1.131 0.258 0.062 H1a: Not supported
BA → CT 0.366 0.077 4.749 *** 0.314 H1b: Supported
CA → CT 0.234 0.076 3.086 0.002 ** 0.181 H1c: Supported
EA → CT 0.346 0.064 5.387 *** 0.306 H1d: Supported
AA → PIN 0.232 0.056 4.132 *** 0.241 H2a: Supported
BA → PIN 0.057 0.065 0.886 0.375 0.062 H2b: Not supported
CA → PIN 0.091 0.063 1.457 0.145 0.089 H2c: Not supported
EA → PIN 0.217 0.055 3.922 *** 0.243 H2d: Supported
CT → PIN 0.193 0.047 4.075 *** 0.243 H3: Supported
Note: **, and *** mean p < 0.01 and p < 0.001. AA—appearance anthropomorphism; BA—behavioral anthro-
pomorphism; CA—cognitive anthropomorphism; EA—emotional anthropomorphism; CT—cognitive trust;
PIN—purchase intention.

This study hypothesized that the appearance, behavioral, cognitive, and emotional
anthropomorphism of virtual streamers would positively influence consumers’ cognitive
trust (H1a, H1b, H1c, and H1d). The analysis results indicate that behavioral anthropo-
morphism (β = 0.314, p < 0.001), cognitive anthropomorphism (β = 0.181, p < 0.01), and
emotional anthropomorphism (β = 0.306, p < 0.001) significantly affect cognitive trust,
supporting H1b, H1c, and H1d. However, appearance anthropomorphism does not signifi-
cantly affect cognitive trust (β = 0.062, p = 0.258), thus not supporting H1a. Additionally,
the anthropomorphic characteristics of virtual streamers were hypothesized to positively
influence purchase intention (H2a, H2b, H2c, and H2d). The results show that appearance
anthropomorphism (β = 0.241, p < 0.001) and emotional anthropomorphism (β = 0.243,
p < 0.001) have a significant influence on purchase intention, supporting H2a and H2d.
Conversely, the influence of behavioral anthropomorphism (β = 0.062, p = 0.375) and cogni-
tive anthropomorphism (β = 0.089, p = 0.145) on purchase intention is not significant, thus
not supporting H2b and H2c.
Finally, the positive effect of cognitive trust on purchase intention (H3) is supported
(β = 0.243, p < 0.001).

4.4. Mediation Effect Test

This research employed SEM in AMOS 24.0 and used the Bootstrapping method to
test cognitive trust as a mediator between appearance, behavioral, cognitive, and emotional
anthropomorphism and purchase intention [46]. Detailed analysis results are shown in
Table 8.
The indirect effect of appearance anthropomorphism on cognitive trust is not signif-
icant (Est. = 0.015, p = 0.264, BC 95% CI = [−0.013, 0.054]), indicating that appearance
anthropomorphism does not enhance purchase intention through cognitive trust, and
therefore, H4a is not supported. The indirect effect of behavioral anthropomorphism is
significant (Est. = 0.076, p < 0.001, BC 95% CI = [0.031, 0.149]), while its direct effect on
purchase intention is not significant (p = 0.416), indicating that cognitive trust fully mediates
the link between behavioral anthropomorphism and purchase intention, supporting H4b.
The indirect effect of cognitive anthropomorphism is significant (Est. = 0.044, p = 0.004, BC
Behav. Sci. 2024, 14, 1228 12 of 18

95% CI = [0.014, 0.094]), and its direct effect is not significant (p = 0.191), indicating that cog-
nitive trust also fully mediates the relationship between cognitive anthropomorphism and
purchase intention, supporting H4c. The indirect effect of emotional anthropomorphism
is significant (Est. = 0.074, p = 0.001, BC 95% CI = [0.032, 0.132]), and there is a significant
direct effect on purchase intention (Est. = 0.243, p = 0.002), indicating that cognitive trust
partially mediates the link between emotional anthropomorphism and purchase intention,
supporting H4d.

Table 8. Mediation test.

Bootstrapping BC 95% CI
Est. Std. Error Lower Bound Upper Bound p-Value
Indirect effect AA 0.015 0.016 −0.013 0.054 0.264
BA 0.076 0.028 0.031 0.149 0.000
CA 0.044 0.020 0.014 0.094 0.004
EA 0.074 0.024 0.032 0.132 0.001
Direct effect AA 0.241 0.062 0.117 0.360 0.001
BA 0.062 0.079 −0.088 0.219 0.416
CA 0.089 0.070 −0.049 0.226 0.191
EA 0.243 0.067 0.097 0.363 0.002
Total effect AA 0.256 0.064 0.128 0.384 0.001
BA 0.139 0.078 −0.017 0.292 0.073
CA 0.133 0.069 −0.004 0.266 0.056
EA 0.317 0.066 0.176 0.435 0.002
Note: AA—appearance anthropomorphism; BA—behavioral anthropomorphism; CA—cognitive anthropomor-
phism; EA—emotional anthropomorphism; mediator—cognitive trust; dependent variable—purchase intention.

In summary, behavioral and cognitive anthropomorphism have a full mediating

effect on purchase intention through cognitive trust, while emotional anthropomorphism
has a partial mediating effect, and appearance anthropomorphism has no significant
mediating effect [47]. This suggests that appearance anthropomorphism alone is insufficient
to enhance cognitive trust; rather, behavioral and cognitive characteristics are crucial, with
emotional anthropomorphism also moderately increasing purchase intention.

5. Discussion and Conclusions

5.1. Limitations and Future Research Directions
This research has certain limitations that merit exploration in future studies. First,
the participants in this study were predominantly individuals familiar with or who had
watched virtual influencers. While this sampling approach ensured the relevance of the
sample, China is a vast country with diverse regions, economic backgrounds, and consumer
habits, which may lead to varying levels of acceptance of virtual streamers. Future research
could expand the sample coverage by considering consumers from different regions, age
groups, and cultural backgrounds. Furthermore, this study concentrated on examining
cognitive trust as a mediating factor and did not delve deeper into the potential impact
of other variables such as affective trust. Future studies could look into the influence of
other mediating variables or consider the moderating effects of individual characteristics
such as gender or age to conduct a more comprehensive study of the trust mechanism
between virtual steamers and consumers. Finally, the data for this study were collected
through a survey. Future research could consider conducting case studies or experiments to
investigate related issues or utilize machine learning to analyze the content of live streams.

5.2. Discussion on Results

This study empirically investigated the underlying mechanism through which the
anthropomorphic characteristics of virtual streamers influence consumer purchase intention
via cognitive trust. Our findings revealed that different dimensions of anthropomorphism
have varying impacts on cognitive trust and purchase intention. Notably, cognitive trust
Behav. Sci. 2024, 14, 1228 13 of 18

plays a significant mediating role in some paths. These results not only validate existing
theories but also provide new insights into the field of virtual streamer research.
Firstly, appearance anthropomorphism did not significantly influence cognitive trust.
As research has shown, not all virtual character appearances can evoke consumer trust and
reliance [48]. This finding suggests that a basic human-like appearance alone is insufficient
to directly stimulate consumers’ trust in virtual streamers. Humanoid entities need to
influence the social relationships that users can perceive, such as social presence, in order
to increase users’ trust in them [8]. In previous studies, the effect of anthropomorphism
on trust was not a direct relationship, which is consistent with the conclusion of this
paper [7]. This study further found that appearance anthropomorphism can directly have a
significant impact on consumers’ purchase intention. This finding reveals that the action
path of appearance anthropomorphism may bypass the trust mechanism and directly affect
purchase motivation. This may be because the appearance design conforms to consumers’
aesthetics or preferences, so it can directly increase purchase intention, which is consistent
with the results of prior research [49].
Secondly, behavioral anthropomorphism has a significant positive impact on cognitive
trust. This shows that users are more likely to have a certain degree of trust in virtual
streamers with humanized interactive behaviors. Virtual streamers are a form of artificial
intelligence, which is an intelligent technology that can interact with the surrounding
environment and simulate human behavior [50]. This anthropomorphism of artificial
intelligence has an impact on trust. Designers often consider the human-like qualities of
robots when designing robots to influence trust [51]. However, the behavioral anthropo-
morphism of virtual streamers has no direct effect on consumers’ purchasing intention,
which may indicate that virtual streamers mainly meet users’ entertainment or social needs,
and it is difficult for simple behavioral anthropomorphism to directly stimulate consumers’
purchasing intention [52]. Therefore, this reveals that the human-like behavior of virtual
streamers can obviously narrow the distance between them and consumers and make
consumers trust them, but this is not a direct factor in promoting purchases.
Thirdly, cognitive anthropomorphism has a significant positive impact on cognitive
trust, which shows that when virtual streamers have certain logical characteristics or knowl-
edge characteristics, they can effectively gain consumer trust [53]. This is consistent with
previous research results. Artificial intelligence technology is applied in many fields. In the
study of intelligent agents, the level of anthropomorphism can indeed significantly improve
users’ cognitive trust [54]. However, cognitive anthropomorphism does not show a signifi-
cant impact on purchase intention. This may be because consumers regard virtual streamers
as an information provider or tool. Therefore, the cognitive anthropomorphism of virtual
streamers does not constitute a factor in direct inducement of purchasing behavior [55].
Fourthly, emotional anthropomorphism has a significant positive impact on both
cognitive trust and purchase intention, which shows that if virtual streamers are given
the ability to express emotions, it will not only enhance users’ trust in the streamer but
also directly affect purchasing behavior through emotional resonance. In previous stud-
ies, researchers studying smart voice assistants found that users’ trust in voice assistants
was highly correlated with the intimacy between the user and the device, and that the
level of anthropomorphism was also moderately correlated with trust [56]. The anthro-
pomorphism of chatbots can affect perceived trust and emotional evaluation, which in
turn affects perceived enthusiasm and purchase intention [49]. Therefore, we extend this
mechanism of correlation between artificial intelligence and user emotions to the character-
istics of virtual streamers’ emotional anthropomorphism, showing that virtual streamers’
emotional anthropomorphism affects trust and can play an important role in consumer
purchasing decisions.
Furthermore, the findings revealed that cognitive trust positively influences purchase
intention, consistent with prior studies [31,57]. This underscores the pivotal role of trust in
virtual streamer live broadcasts. When verifying the mediating role of cognitive trust, it
was found that cognitive trust plays a complete mediating role in the impact of behavioral
Behav. Sci. 2024, 14, 1228 14 of 18

anthropomorphism and cognitive anthropomorphism on purchase intention, while it plays

a partial mediating role in the impact of emotional anthropomorphism. This is consistent
with existing research, which agrees that with the rapid growth of the number of Internet
users, trust is particularly important among the many factors that affect online purchase
intentions. In social media marketing, brand trust can have an impact on consumer pur-
chasing decisions, and trust can also regulate perceived anthropomorphism [38,52,58,59].
However, on the path of appearance anthropomorphism, the mediating effect of cogni-
tive trust is not significant, which shows that although appearance anthropomorphism
enhances the attractiveness of virtual streamers, it is difficult for consumers to develop a
sense of trust based on the appearance characteristics of virtual streamers alone.

5.3. Theoretical Implications

First, this study addresses a research gap concerning virtual streamers’ anthropomor-
phism’s effect on purchase intention. The existing literature primarily examines how virtual
imagery influences user experience, with limited systematic analysis of how anthropomor-
phic features influence purchasing decisions. This study explores the mechanism by which
anthropomorphic features affect purchase intention through cognitive trust, providing a
new perspective for academia. Secondly, this research broadens the application and mea-
surement of avatar theory. By categorizing virtual streamers’ anthropomorphism into four
dimensions of appearance, behavior, cognition, and emotion and validating their varying
effects on purchase intention, this classification offers a more detailed framework for future
measurement and reveals the complex mechanisms of anthropomorphism across various
contexts. This research also presents cognitive trust as a mediator, emphasizing its key
role in how virtual streamer anthropomorphic traits affect purchase intention. The results
indicate that cognitive trust fully mediates behavioral and cognitive anthropomorphism
and partially mediates emotional anthropomorphism, enhancing trust theory applications.
Overall, this study broadens the application boundaries of anthropomorphism and trust
theories, providing theoretical support for research on virtual streamers and practical
guidance for their design.

5.4. Practical Implications

Firstly, this research offers actionable guidance for designing virtual streamers. The re-
sults show that the multidimensional anthropomorphic characteristics of virtual streamers
have different effects on consumer trust and purchase intention, with emotional, cognitive,
and behavioral traits significantly enhancing cognitive trust. Companies should prioritize
emotional expression and cognitive–behavioral traits when optimizing virtual streamers to
strengthen connections with consumers and improve the credibility and effectiveness of
product recommendations. Secondly, this study offers insights for brand marketing strate-
gies. Cognitive trust serves as a mediator between virtual streamer anthropomorphism
and purchase intention, suggesting that companies can indirectly boost consumer purchase
intention by enhancing the professionalism and reliability of virtual streamers. Therefore,
companies should focus on creating a consistent and professional image for virtual stream-
ers to ensure reliable information delivery, thereby strengthening brand trust and loyalty.
Additionally, this study provides a new perspective for customer relationship manage-
ment. Cognitive and behavioral anthropomorphic traits promote purchase intention by
enhancing cognitive trust. Companies can leverage the anthropomorphic characteristics of
virtual streamers to reinforce customer relationships, for example, by optimizing interactive
design to increase emotional intelligence and responsiveness, thereby boosting consumer
engagement and trust. Finally, this study offers empirical support for AI-driven virtual
marketing technologies. The findings show that virtual streamers’ anthropomorphic traits
positively influence consumer behavior, with cognitive trust playing a mediating role. This
provides a scientific basis for further innovation in virtual marketing technologies and
encourages companies to focus on anthropomorphic traits and building trust to enhance
user experience and explore new business models.
Behav. Sci. 2024, 14, 1228 15 of 18

5.5. Conclusions
With the booming development of the digital economy, live streaming e-commerce has
become a new frontier for brand marketing. In this field, virtual streamers, as an emerging
interactive tool, can connect with consumers and enhance the shopping experience by
simulating human characteristics or behaviors. China, as one of the world’s largest Internet
markets, has seen a rapid development of live streaming e-commerce, providing a rich
practical scenario for the study of virtual streamers. However, although virtual streamers
have shown great potential in attracting consumer attention and promoting purchasing
behavior, systematic research on how their anthropomorphic characteristics affect consumer
psychology and behavior is still limited.
The existing literature mostly focuses on the technical implementation and initial user
reactions of virtual streamers and lacks in-depth discussions on how the anthropomorphic
characteristics of virtual streamers specifically affect the process of building consumer
trust and forming purchase intentions. Especially under the mediating role of cognitive
trust, the mechanism of how the anthropomorphic characteristics of virtual streamers
affect consumer purchasing decisions has not yet been clarified. This research gap limits
our comprehensive understanding of the role of virtual streamers in e-commerce live
broadcasts and also restricts brands from effectively using virtual streamers to formulate
marketing strategies. Therefore, based on the avatar theory, this study empirically analyzed
the impact of the personification of virtual streamers on cognitive trust and consumer
purchase intention. Through structural equation modeling analysis of questionnaire data
from 503 participants in China, we found that the behavioral anthropomorphism, cognitive
anthropomorphism, and emotional anthropomorphism of virtual streamers significantly
improved cognitive trust. Appearance anthropomorphism and emotional anthropomor-
phism can directly enhance purchase intention, while cognitive trust has a significant
impact on purchase intention.
The innovation of this paper is that it is the first to subdivide the anthropomorphic
characteristics of virtual streamers into the above four dimensions and link these four
dimensions with the mediating role of cognitive trust, providing a new perspective for
understanding how consumers build trust with virtual characters and how these characters
promote consumer purchase intention. This study also has some limitations, such as the
geographical limitations of the sample and the specificity of the research design. Future
research can expand the sample range, explore the impact of virtual streamer anthropomor-
phism in different cultural backgrounds, and consider other possible mediating variables
or control variables. In addition, the use of a longitudinal research design or experimental
methods may further verify and deepen the findings of this study.
The research results are of great significance for understanding and designing virtual
streamers suitable for live e-commerce activities. Theoretically, this research expands the
application of avatar theory in the field of digital marketing, particularly concerning how
virtual streamers influence consumer trust and purchasing behavior. On a practical level,
this study provides empirical support for how brands can design virtual streamers to
optimize consumer engagement, promote consumer purchases, and improve the com-
petitiveness of virtual streamers. In terms of specific practical suggestions, companies
or brands can consider developing advanced interactive functions for virtual streamers,
such as implementing responses to consumers’ personalized questions and interacting
with consumers in a vivid and real way during live broadcasts to increase consumers’
immersive experience. In addition, the image, actions, and expressions of virtual streamers
can be tailored to align with the preferences of their target audience to enhance appeal.
Virtual streamers can also be used to convey brand values and product information, en-
sure the consistency and transparency of information, gain consumers’ favor, and build
consumer trust.
Behav. Sci. 2024, 14, 1228 16 of 18

Author Contributions: Conceptualization, F.H.; methodology, C.L.; validation, C.L.; formal analysis,
C.L.; investigation, C.L.; writing—original draft preparation, C.L.; writing—review and editing, F.H.;
supervision, F.H. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: This study was conducted in accordance with the Declaration
of Helsinki. Ethical review and approval were waived for this study according to Article 13 of the
Enforcement Rule of the Bioethics and Safety Act in the Republic of Korea.
Informed Consent Statement: Informed consent was obtained from all subjects involved in this study.
Data Availability Statement: Data are contained within the article. The original contributions presented
in this study are included in the article. Further inquiries can be directed to the corresponding author.
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. Appel, G.; Grewal, L.; Hadi, R.; Stephen, A.T. The Future of Social Media in Marketing. J. Acad. Mark. Sci. 2020, 48, 79–95.
[CrossRef] [PubMed]
2. Huaman-Ramirez, R.; Lunardo, R.; Vasquez-Parraga, A. How Brand Self-Disclosure Helps Brands Create Intimacy with Cus-
tomers: The Role of Information Valence and Anthropomorphism. Psychol. Market. 2022, 39, 460–477. [CrossRef]
3. China Internet Network Information Center (2024). The 54th statistical report on China’s Internet development. China Inter-
net Network Information Center. Available online: https://www.cnnic.net.cn/n4/2024/0829/c88-11065.html (accessed on 10
December 2024).
4. Vadhri, S. Digital Twins and Virtual Worlds: Influencing Technology’s Future. Available online: https://www.forbes.com/
councils/forbesbusinessdevelopmentcouncil/2024/11/27/digital-twins-and-virtual-worlds-influencing-technologys-future/
(accessed on 2 December 2024).
5. Kim, D.Y.; Lee, H.K.; Chung, K. Avatar-Mediated Experience in the Metaverse: The Impact of Avatar Realism on User-Avatar
Relationship. J. Retail. Consum. Serv. 2023, 73, 103382. [CrossRef]
6. Dabiran, E.; Farivar, S.; Wang, F.; Grant, G. Virtually Human: Anthropomorphism in Virtual Influencer Marketing. J. Retail.
Consum. Serv. 2024, 79, 103797. [CrossRef]
7. Chen, H.; Shao, B.; Yang, X.; Kang, W.; Fan, W. Avatars in Live Streaming Commerce: The Influence of Anthropomorphism on
Consumers’ Willingness to Accept Virtual Live Streamers. Comput. Hum. Behav. 2024, 156, 108216. [CrossRef]
8. Qiu, L.; Benbasat, I. Evaluating Anthropomorphic Product Recommendation Agents: A Social Relationship Perspective to
Designing Information Systems. J. Manag. Inform. Syst. 2009, 25, 145–182. [CrossRef]
9. Schouten, A.P.; Janssen, L.; Verspaget, M. Celebrity vs. Influencer Endorsements in Advertising: The Role of Identification,
Credibility, and Product-Endorser Fit. In Leveraged Marketing Communications; Routledge: London, UK, 2021; pp. 208–231, ISBN
978-1-003-15524-9.
10. Sokolova, K.; Kefi, H. Instagram and YouTube Bloggers Promote It, Why Should I Buy? How Credibility and Parasocial Interaction
Influence Purchase Intentions. J. Retail. Consum. Serv. 2020, 53, 101742. [CrossRef]
11. Gao, W.; Jiang, N.; Guo, Q. How do Virtual Streamers Affect Purchase Intention in the Live Streaming Context? A Presence
Perspective. J. Retail. Consum. Serv. 2023, 73, 103356. [CrossRef]
12. Low, G.S.; Mohr, J.J. Advertising vs Sales Promotion: A Brand Management Perspective. J. Prod. Brand Manag. 2000, 9, 389–414.
[CrossRef]
13. Jin, S.-A.A.; Sung, Y. The Roles of Spokes-Avatars’ Personalities in Brand Communication in 3D Virtual Environments. J. Brand
Manag. 2010, 17, 317–327. [CrossRef]
14. Jin, S.-A.A. The Virtual Malleable Self and the Virtual Identity Discrepancy Model: Investigative Frameworks for Virtual Possible
Selves and Others in Avatar-Based Identity Construction and Social Interaction. Comput. Hum. Behav. 2012, 28, 2160–2168.
[CrossRef]
15. Wang, X.; Butt, A.H.; Zhang, Q.; Shafique, M.N.; Ahmad, H.; Nawaz, Z. Gaming Avatar Can Influence Sustainable Healthy
Lifestyle: Be Like an Avatar. Sustainability 2020, 12, 1998. [CrossRef]
16. Pimentel, D.; Kalyanaraman, S. Customizing Your Demons: Anxiety Reduction via Anthropomorphizing and Destroying an
“Anxiety Avatar”. Front. Psychol. 2020, 11, 566682. [CrossRef] [PubMed]
17. Miao, F.; Kozlenkova, I.V.; Wang, H.; Xie, T.; Palmatier, R.W. An Emerging Theory of Avatar Marketing. J. Mark. 2022, 86, 67–90.
[CrossRef]
18. Golossenko, A.; Pillai, K.G.; Aroean, L. Seeing Brands as Humans: Development and Validation of a Brand Anthropomorphism
Scale. Int. J. Res. Mark. 2020, 37, 737–755. [CrossRef]
19. Li, X.; Sung, Y. Anthropomorphism Brings Us Closer: The Mediating Role of Psychological Distance in User–AI Assistant
Interactions. Comput. Hum. Behav. 2021, 118, 106680. [CrossRef]
20. Li, S.; Chen, J. Virtual Human on Social Media: Text Mining and Sentiment Analysis. Technol. Soc. 2024, 78, 102666. [CrossRef]
Behav. Sci. 2024, 14, 1228 17 of 18

21. McAllister, D.J. Affect- and Cognition-Based Trust as Foundations for Interpersonal Cooperation in Organizations. Acad. Manag.
J. 1995, 38, 24–59. [CrossRef]
22. McKnight, D.H.; Choudhury, V.; Kacmar, C. Developing and Validating Trust Measures for E-Commerce: An Integrative Typology.
Inform. Syst. Res. 2002, 13, 334–359. [CrossRef]
23. Gefen, D.; Karahanna, E.; Straub, D.W. Trust and TAM in Online Shopping: An Integrated Model. Mis Quart. 2003, 27, 51–90.
[CrossRef]
24. Johnson, D.; Grayson, K. Cognitive and Affective Trust in Service Relationships. J. Bus. Res. 2005, 58, 500–507. [CrossRef]
25. Chen, J.; Gong, X.; Ren, R. Active or Avoidance Coping? Influencing Mechanisms of Streamers’ Coping Strategies on Viewers’
Word of Mouth after Livestreaming e-Commerce Failures. J. Retail. Consum. Serv. 2023, 72, 103278. [CrossRef]
26. Legood, A.; van der Werff, L.; Lee, A.; den Hartog, D.; van Knippenberg, D. A Critical Review of the Conceptualization,
Operationalization, and Empirical Literature on Cognition-Based and Affect-Based Trust. J. Manag. Stud. 2023, 60, 495–537.
[CrossRef]
27. Wei, K.; Xi, W. CEO vs. Celebrity: The Effect of Streamer Types on Consumer Engagement in Brands’ Self-Built Live-Streaming. J.
Res. Interact. Mark. 2024, 18, 631–647. [CrossRef]
28. Pelau, C.; Dabija, D.-C.; Ene, I. What Makes an AI Device Human-like? The Role of Interaction Quality, Empathy and Perceived
Psychological Anthropomorphic Characteristics in the Acceptance of Artificial Intelligence in the Service Industry. Comput. Hum.
Behav. 2021, 122, 106855. [CrossRef]
29. Fakhimi, A.; Garry, T.; Biggemann, S. The Effects of Anthropomorphised Virtual Conversational Assistants on Consumer
Engagement and Trust during Service Encounters. Australas. Mark. J. 2023, 31, 314–324. [CrossRef]
30. Wan, C.; Lee, D.; Ng, P. The Role of Anthropomorphism and Racial Homophily of Virtual Influencers in Encouraging Low- versus
High-cost Pro-environmental Behaviors. Psychol. Mark. 2024, 41, 1833–1853. [CrossRef]
31. Chang, S.-H.; Chih, W.-H.; Liou, D.-K.; Yang, Y.-T. The Mediation of Cognitive Attitude for Online Shopping. Inform. Technol.
Peopl. 2016, 29, 618–646. [CrossRef]
32. Zhang, L.; Anjum, M.A.; Wang, Y. The Impact of Trust-Building Mechanisms on Purchase Intention towards Metaverse Shopping:
The Moderating Role of Age. Int. J. Hum.–Comput. Interact. 2024, 40, 3185–3203. [CrossRef]
33. Li, X.; Huang, D.; Dong, G.; Wang, B. Why Consumers Have Impulsive Purchase Behavior in Live Streaming: The Role of the
Streamer. BMC Psychol. 2024, 12, 129. [CrossRef]
34. Wang, K.; Wu, J.; Sun, Y.; Chen, J.; Pu, Y.; Qi, Y. Trust in Human and Virtual Live Streamers: The Role of Integrity and Social
Presence. Int. J. Hum.–Comput. Interact. 2023, 40, 8274–8294. [CrossRef]
35. Guo, Y.; Zhang, K.; Wang, C. Way to Success: Understanding Top Streamer’s Popularity and Influence from the Perspective of
Source Characteristics. J. Retail. Consum. Serv. 2022, 64, 102786. [CrossRef]
36. Li, Y.; Li, X.; Cai, J. How Attachment Affects User Stickiness on Live Streaming Platforms: A Socio-Technical Approach Perspective.
J. Retail. Consum. Serv. 2021, 60, 102478. [CrossRef]
37. Lu, B.; Chen, Z. Live Streaming Commerce and Consumers’ Purchase Intention: An Uncertainty Reduction Perspective. Inf.
Manag. 2021, 58, 103509. [CrossRef]
38. Zhang, X.; Shi, Y.; Li, T.; Guan, Y.; Cui, X. How Do Virtual AI Streamers Influence Viewers’ Livestream Shopping Behavior? The
Effects of Persuasive Factors and the Mediating Role of Arousal. Inf. Syst. Front. 2023, 26, 1803–1834. [CrossRef]
39. Cronbach, L.J. Coefficient Alpha and the Internal Structure of Tests. Psychometrika 1951, 16, 297–334. [CrossRef]
40. Nunnally, J.C. Psychometric Theory—25 Years Ago and Now. Educ. Res. 1975, 4, 7–21. [CrossRef]
41. Hair, J.F. Multivariate Data Analysis, 7th ed.; Prentice Hall: Upper Saddle River, NJ, USA, 2010; ISBN 978-0-13-813263-7.
42. Fornell, C.; Larcker, D.F. Evaluating Structural Equation Models with Unobservable Variables and Measurement Error. J. Mark.
Res. 1981, 18, 39–50. [CrossRef]
43. Podsakoff, P.M.; MacKenzie, S.B.; Lee, J.-Y.; Podsakoff, N.P. Common Method Biases in Behavioral Research: A Critical Review of
the Literature and Recommended Remedies. J. Appl. Psychol. 2003, 88, 879–903. [CrossRef]
44. Katrutsa, A.; Strijov, V. Comprehensive Study of Feature Selection Methods to Solve Multicollinearity Problem According to
Evaluation Criteria. Expert Syst. Appl. 2017, 76, 1–11. [CrossRef]
45. Byrne, B.M. Structural Equation Modeling with AMOS: Basic Concepts, Applications, and Programming, 3rd ed.; Multivariate
Applications Series; Routledge: New York, NY, USA, 2016; Volume 396, ISBN 978-1-315-75742-1.
46. Preacher, K.J.; Hayes, A.F. Asymptotic and Resampling Strategies for Assessing and Comparing Indirect Effects in Multiple
Mediator Models. Behav. Res. Methods 2008, 40, 879–891. [CrossRef] [PubMed]
47. Baron, R.M.; Kenny, D.A. The Moderator–Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic,
and Statistical Considerations. J. Personal. Soc. Psychol. 1986, 51, 1173. [CrossRef] [PubMed]
48. Peña, J.; Yoo, S.-C. Under Pressure: Avatar Appearance and Cognitive Load Effects on Attitudes, Trustworthiness, Bidding, and
Interpersonal Distance in a Virtual Store. Presence-Teleop. Virt. 2014, 23, 18–32. [CrossRef]
49. Zhang, R.; Wang, N. Unravelling the Impact of Chatbot Anthropomorphism on Purchase Behavior: A Perspective of Halo Effect
and Task-Technology Fit. In Proceedings of the PACIS 2023 Proceedings, Nanchang, China, 9–12 July 2023.
50. Glikson, E.; Woolley, A.W. Human Trust in Artificial Intelligence: Review of Empirical Research. Acad. Manag. Ann. 2020, 14,
627–660. [CrossRef]
Behav. Sci. 2024, 14, 1228 18 of 18

51. Cohen, M.C.; Peel, M.A.; Scalia, M.J.; Willett, M.M.; Chiou, E.K.; Gorman, J.C.; Cooke, N.J. Anthropomorphism Moderates the
Relationships of Dispositional, Perceptual, and Behavioral Trust in a Robot Teammate. Proc. Hum. Factors Ergon. Soc. Annu. Meet.
2023, 67, 529–536. [CrossRef]
52. Malhotra, G.; Ramalingam, M. Perceived Anthropomorphism and Purchase Intention Using Artificial Intelligence Technology:
Examining the Moderated Effect of Trust. J. Enterp. Inf. Manag. 2023. ahead-of-print. [CrossRef]
53. De Visser, E.J.; Monfort, S.S.; McKendrick, R.; Smith, M.A.; McKnight, P.E.; Krueger, F.; Parasuraman, R. Almost Human:
Anthropomorphism Increases Trust Resilience in Cognitive Agents. J. Exp. Psychol. Appl. 2016, 22, 331. [CrossRef]
54. Cheng, P.; Meng, F.; Yao, J.; Wang, Y. Driving with Agents: Investigating the Influences of Anthropomorphism Level and
Physicality of Agents on Drivers’ Perceived Control, Trust, and Driving Performance. Front. Psychol. 2022, 13, 883417. [CrossRef]
55. Chung, S.I.; Han, K.-H. Consumer Perception of Chatbots and Purchase Intentions: Anthropomorphism and Conversational
Relevance. Int. J. Adv. Cult. Technol. 2022, 10, 211–229. [CrossRef]
56. Seymour, W.; Van Kleek, M. Exploring Interactions between Trust, Anthropomorphism, and Relationship Development in Voice
Assistants. Proc. ACM Hum.-Comput. Interact. 2021, 5, 1–16. [CrossRef]
57. Yu, T.; Teoh, A.P.; Bian, Q.; Liao, J.; Wang, C. Can Virtual Influencers Affect purchase Intentions in Tourism and Hospitality
E-Commerce Live Streaming? An Empirical Study in China. Int. J. Contemp. Hosp. M. 2024. ahead-of-print. [CrossRef]
58. Alfina, I.; Ero, J.; Hidayanto, A.N.; Shihab, M.R. The Impact of Cognitive Trust and E-Wom on Purchase Intention in C2c
E-Commerce Site. J. Comput. Sci. 2014, 10, 2518–2524. [CrossRef]
59. Hanaysha, J.R. Impact of Social Media Marketing Features on Consumer’s Purchase Decision in the Fast-Food Industry: Brand
Trust as a Mediator. Int. J. Inf. Manag. Data Insights 2022, 2, 100102. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.