Computers in Human Behavior: Artificial Humans 1 (2023) 100016

Contents lists available at ScienceDirect

Computers in Human Behavior: Artificial Humans

journal homepage: www.journals.elsevier.com/computers-in-human-behavior-artificial-humans

Virtual selfhood and consumer behavior: Exploring avatar attachment and

consumption patterns in Second Life’s metaverse
Maria Kalyvaki *, Heather McIntosh, Kyle Nash
Minnesota State University, Mankato, USA

A R T I C L E I N F O A B S T R A C T

Keywords: In the vast digital landscape of the Metaverse, users can create and personalize their avatars as virtual repre­
Avatar sentations of themselves. This study delves into the emotions users experience in relation to their avatars and
Attachment theory examines how this attachment influences their consumption behaviors within the virtual world. The research
Consumption
employed a sample of 214 active users participating in Second Life, a prominent virtual world platform. By
Metaverse
Virtual reality
analyzing survey data, we explore the dynamics of self-presentation and attachment between users and their
Second life virtual personas across this well-established platform. Our research offers valuable contributions to the existing
literature on the Metaverse, providing empirical evidence on how virtual reality platforms like Second Life foster
avatar customization and how this, in turn, affects consumer behavior. As the Metaverse gains prominence in the
business world, understanding the habits and preferences of virtual reality users is increasingly crucial. We aim
to enhance our understanding of consumer behavior by incorporating attachment theory into our research on
long-standing virtual environments like Second Life. Our study of Second Life provides valuable insights into the
dynamics of consumer behavior within a well-established virtual world, which can be applied to the emerging
Metaverse platforms. This knowledge helps businesses identify consumer profiles, address their needs, and
enhance their virtual presence and success.

1. Introduction are compelled to explore innovative, personalized communication ap­

proaches that provide more interactive and customized experiences
The Metaverse, which allows for user interaction, creation, and (Kohler et al., 2011). For consumers, the most captivating aspect of the
digital asset transactions, signifies a novel frontier in both human Metaverse lies in the virtual worlds and immersive 3D digital environ­
interaction and commerce (Kalyvaki, 2023). The concept of the Meta­ ments they can explore (Vasalou & Joinson, 2009). These ground­
verse, an all-encompassing digital universe that seamlessly integrates breaking social platforms allow users to create new personas through
physical and virtual reality for multi-user environments, has garnered avatars, offering opportunities for creative expression, learning, and
substantial attention from marketing scholars and industry professionals adventure (Vasalou & Joinson, 2009). As a result, the Metaverse has the
alike (Alcañiz et al., 2018; Wedel et al., 2020). As a powerful content potential to revolutionize various aspects of society, including com­
strategy, the Metaverse has fueled exponential growth for companies merce, entertainment, education, and social interaction (Wedel et al.,
across the globe (Mystakidis et al., 2021; Pellas et al., 2021). Major 2020).
corporations, including Meta (formerly Facebook), Disney, Microsoft, The Metaverse presents a range of possibilities for marketers, as well.
Nvidia, Apple, Nike, and IKEA, have recognized the potential of the By leveraging the unique capabilities of these virtual environments,
Metaverse and are investing heavily in the development of extended businesses can develop more effective and engaging marketing cam­
digital realities rooted in this concept (Gao, 2022; Kowalczuk et al., paigns, create immersive brand experiences, and foster deeper connec­
2021; Mystakidis et al., 2022). Virtual worlds, which offer a unique tions with their target audiences (Pellas et al., 2021). Furthermore, the
blend of personal entertainment and business applications, epitomize a Metaverse can also serve as a valuable tool for conducting market
new 3D content strategy that effectively addresses the intricate needs of research, as it enables organizations to gain insights into consumer
today’s consumers (Alcañiz et al., 2018). Consequently, organizations behavior and preferences within virtual spaces (Mystakidis et al., 2021).

* Corresponding author. MH 256, 56001, MN, USA.

E-mail address: Maria.Kalyvaki@mnsu.edu (M. Kalyvaki).

https://doi.org/10.1016/j.chbah.2023.100016
Received 25 April 2023; Received in revised form 6 September 2023; Accepted 16 September 2023
Available online 22 September 2023
2949-8821/© 2023 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
M. Kalyvaki et al. Computers in Human Behavior: Artificial Humans 1 (2023) 100016

The concept of the Metaverse represents a bold new frontier for mar­ provider specializing in virtual currencies (Metaverse Economy Payment
keting research and practice. As an immersive digital universe that Solutions | Tilia, n. d.; Second Life and Tilia Parent Company Adds Key
merges physical and virtual realities, the Metaverse is reshaping the way Leadership Roles to Support Rapid Growth, n. d.). The resurgence of
businesses engage with consumers and offering exciting new opportu­ Second Life amid the COVID-19 pandemic highlights the enduring ap­
nities for innovation and growth (Alcañiz et al., 2018; Wedel et al., peal and adaptability of virtual worlds in the face of unprecedented
2020). global challenges. Despite the persistence of SL as a virtual platform,
To thrive in the Metaverse, brands must grasp the underlying factors little work has been conducted which uses SL to understand the
that influence consumption within a 3D environment. While the novelty connection that consumer may have to their avatars in virtual settings.
of virtual environments may initially pique users’ interest, sustaining Given the aforementioned benefits and complexity of SL, this platform
their engagement requires a deeper understanding of additional ele­ serves as a useful tool for researchers to conduct such tests.
ments (Vasalou & Joinson, 2009). The current research will focus on the
role of avatars in promoting the connection users feel to the virtual 2.2. Attachment theory, consumption, and the self-avatar relationship
realm. The connection between users and their avatars is of paramount
importance in this context since avatars act as digital extensions of users, In order to understand the link between avatars and consumer
allowing them to interact with the virtual realm (Belk, 2013). Previous behavior, one must first understand relevant theories in the psycho­
studies have highlighted that users tend to form a strong attachment to logical literature which have been extended into virtual settings. One
their avatars, which, in turn, impacts their behavior within the virtual example is that of attachment theory (Ainsworth & Bell, 1970). This
environment (K. M. Lee & Nass, 2003). For example, research has sug­ theory originally concentrated on the relationships between mothers
gested that users tend to favor avatars that which are more “human-like” and infants, but has since been expanded to encompass various human
and resemble their own gender (Nowak & Biocca, 2003), as well as those relationships, such as friendships, romantic partnerships, and even
who resemble their own physical appearance (Messinger et al., 1970; connections with celebrities (Candel & Turliuc, 2019; Stfania et al.,
Park & Kim, 2022). Users who develop self-similar avatars exhibit 2023; VanMeter et al., 2015). More recently, attachment theory has
stronger identification (Trepte & Reinecke, 2010), enhanced enjoyment been extended to provide valuable insights into consumers’ purchasing
of gaming experiences (Waddell et al., 2015), increased self-awareness behavior by examining emotional connections in various contexts
(Vasalou & Joinson, 2009) and in the context of violent games, (Thorpe et al., 2019). For example, attachment theory has been
heightened aggression (J. A. Lee et al., 2021). While it appears clear that employed to investigate emotional bonds with brands (Boateng et al.,
avatars serve as important vessels for connection to the virtual world, 2020; Bidmon, 2017; Kim et al., 2021; Park et al., 2013; Whan Park
much work has been conducted in the gaming world, while less work has et al., 2010) and retail environments, both physical (Brocato et al., 2015;
been conducted on avatars and consumer behavior. The current research Johnson et al., 2015; Shaw & Sullivan, 2013) and virtual (Horáková
seeks to identify prior work on the avatar-consumer relationship, then et al., 2022). Kerschbaumer et al. (2023) note that attachment styles
detail a descriptive study conducted which explored relationships be­ lead to varying levels of anxiety about interpersonal relationships—this
tween avatar use and consumer behavior in the Second Life virtual re­ anxiety then extends into consumer behavior by modifying the types of
ality platform. Gaining insight into the factors that propel consumption relationships consumer build with brands and products. Attachment
in the Metaverse is essential for brands aiming to establish a presence in theory also helps to inform our comprehension of consumer behavior
this domain, as it can assist in pinpointing consumer profiles and needs within virtual spaces, such as online retail environments and social
(Kohler et al., 2011). networking sites. For example, VanMeter et al. (2018) operationalized
attachment to social media (ASM) using eight components, extending
2. Literature review attachment theory to marketing within social media contexts. Subse­
quent research has demonstrated that social media attachment in­
2.1. Second Life and the impact of the COVID-19 pandemic fluences the quality of brand interactions (R. VanMeter et al., 2018).
Horáková et al. (2022) would later demonstrate that consumers’
The COVID-19 pandemic has not only created opportunities for attachment to online and traditional brick-and-mortar retail spaces did
growth in the realm of virtual worlds but has also led to the renaissance not significantly differ from one another.
of certain platforms, such as Second Life, which witnessed a consider­ The importance of attachment theory for the current research is
able surge in user engagement during this period (“Second Life”, understanding that individuals also develop attachments to digital ob­
“Minecraft” … The “Confined” Find Themselves in Virtual Worlds, n. d.; jects (Koles & Nagy, 2021). One prominent digital object is the avatar, a
Sun et al., 2022). As the pandemic unfolded, both long-time virtual re­ digital representation of the user in virtual environments (Bente et al.,
ality enthusiasts and newcomers flocked to these platforms in search of 2008; Johnson et al., 2015). Users form relationships with their avatars,
social interaction and novel experiences (Dwivedi et al., 2022; which has been termed the avatar-self relationship (Koles & Nagy, 2021;
Krischke-Leitão & Gomes, 2021; Riar et al., 2022). Second Life (SL), Nagy & Koles, 2014). This relationship can foster the creation of a vir­
which was launched in 2003, has remained one of the pioneering and tual self that users aspire to become, either reflecting their actual self or
most prominent virtual reality platforms for almost two decades (Bonsu an idealized version (Fig. A1) (Vasalou & Joinson, 2009). Procter (2020)
& Darmody, 2008). suggests that more recent research departs from the notion of the user as
SL, a virtual world environment based in the United States and the sole agent controlling their avatar and instead views the user-avatar
operated by Linden Research, Inc., enables users or “residents” to create relationship as a dialectical process (Procter & Studies, 2020).
free accounts and customize their avatars to their liking (Bonsu & Dar­ Individuals participating in virtual reality environments, such as
mody, 2008). The platform offers a variety of communication tools, Second Life, develop strong connections to the digital spaces they
including voice chat and item sharing, and allows users to teleport their occupy, which subsequently results in an increased likelihood of
avatars to different virtual locations (Bonsu & Darmody, 2008). The engaging in consumer behaviors in the virtual setting (Ducheneaut et al.,
dynamic nature of SL’s virtual environment is a result of its co-creative 2009), such as acquiring virtual goods and services to augment their
foundation, where residents actively participate in generating fashion, virtual personas (Lehdonvirta, 2009). These connections concentrate on
objects, animations, and decorations (Bonsu & Darmody, 2008). virtual products as crucial elements of users’ digital identities (Guo &
The strength of SL’s economy is evidenced by the ability of residents Barnes, 2011). Avatar attachment theory combines attachment theory
to exchange virtual currency (Lindens) for real-world currency and vice with the digital representations that users generate for themselves in
versa. This robust virtual economy is further underscored by the plat­ online settings (Banks & Bowman, 2016; Lewis et al., 2008), suggesting
form’s recent partnership with Tilia, a dedicated payment service that users establish bonds with their virtual selves and that this has

2
M. Kalyvaki et al. Computers in Human Behavior: Artificial Humans 1 (2023) 100016

consequences for consumer behaviors in digital spaces (Waddell et al., effort in customizing my avatar”, “I care about my avatar’s appearance”,
2015). and “My avatar is an extension of myself”.
Prior research has identified ley factors which can influence avatar
The items listed above, although constructed by the current re­
attachment. These factors include identity expression (Ridden, 1997),
searchers, are consistent with prior work asking users about their ex­
social interactions with other users’ avatars (Yee & Bailenson, 2007),
periences with avatar construction and involvement (e.g., Takano &
immersion and realism (Slater & Wilbur, 1997), and having a wide range
Taka, 2022; Wang et al., 2019). By exploring these hypotheses, we can
of customization options (Bessière et al., 2007) Offering incentives and
more fully understand the relationship between users’ attachment to
rewards for users to customize and invest time in their avatars can also
their avatars and their consumption behaviors in virtual spaces such as
motivate them to continue using the platform and develop a deeper
SL. A better understanding of these connections will offer insights into
connection with their virtual selves (Lehdonvirta, 2011). In under­
the psychological aspects of virtual identity formation and how they
standing the self-avatar relationship more fully, prior work has also
may influence user engagement and spending habits in digital envi­
attempted to characterize what this relationship looks like. For example,
ronments (Ducheneaut et al., 2009; Lehdonvirta, 2009; Banks &
Nagy and Koles (2014) found that some users developed their avatar to
Bowman, 2016; Bowman et al., 2021).
represent functional aspects of themselves in the digital setting, while
others developed an avatar which represented their ideal selves in “the
3. Method
real world”. Kim et al. (2023) found that more positive user-avatar re­
lationships were found when avatars were more realistic and made users
3.1. Research design and procedures
feel psychologically closer to their avatar persona. Most recently, Eti­
enne, Leclercq, Remacle, Dessart, & Schyns, (2023) found that avatar
In this research study, a survey was conducted using Qualtrics, an
non-verbals are essential for creating a more immersive Metaverse
online survey platform that included Likert scale questions. The survey
experience for users. These findings are, perhaps, also explained by the
questionnaire was hosted on a third-party website, and no personally
extended-self theory (Belk, 2013), which cites that users find ways to
identifiable data was collected from any of the respondents. The purpose
become attached to their avatar in digital worlds. However, attachment
of this study was to collect both qualitative and quantitative data from
theory adds depth to this theory by suggesting that users differ in the
users of Second Life, a virtual world platform. To ensure high-quality
form and strength of their avatar relationship.
data and response rates, the focus of the study was explained to the
Despite these findings, preliminary relationships between the user-
respondents on the welcome page, emphasizing the importance of their
avatar relationship and actual consumer behavior are limited. For
participation and the confidentiality of their responses. This was done in
example, Arachchi and Samarasinghe (2023) found that aspects of the
line with the recommendations of Dillman (1978) for survey design (D.
avatar construction were related to the acceptance of fashion retail
Dillman, 1991; D. A. Dillman, 1978). Participation in the study was
services, although this study did not directly assess purchase/purchase
voluntary, and the participants were allowed to withdraw their partic­
intentions. Both Wang et al. (2019) and Zhang (2022) found relation­
ipation at any time during the process of filling out the surveys.
ships between avatar identification and intentions to purchase in-game
A total of 214 useable responses were received from Second Life
virtual items in a mobile game setting. Finally, in the most direct test of
users who were recruited through solicitation on social media platforms
the relationship of interest, a recent paper published in a conference
such as Facebook and Flickr, as well as in various in-world locations such
proceeding by Park and Kim (2022) found relationships between
as malls, beaches, and clubs. Purposive sampling was used to recruit
appearance similarity with an avatar and purchase intentions, and that
users who were at least 18 years of age and whose avatars were at least a
this link was explained by avatar identification.
month old. The latter criterion was to ensure that respondents had some
Overall, a few things are increasingly clear. First, prior work has
Second Life experience to tap on when responding to the survey.
suggested that online objects receive interactions which are similar to
The survey consisted of Likert scale questions that aimed to collect
“real world” objects, which includes the development of attachment and
data on users’ opinions, attitudes, and behaviors concerning their ex­
pseudo-relationships (Bente et al., 2008; Johnson et al., 2015; Koles &
periences on Second Life. Likert scales are widely used in social science
Nagy, 2021; Nagy & Koles, 2014). Also, preliminary work has shown
research as they provide a standardized format for collecting data on a
that aspects of the avatar-user relationship may be tied with purchasing
range of constructs (Beglar & Nemoto, 2014). The survey was conducted
behavior (Arachchi & Samarasinghe, 2023; Park & Kim, 2022; Wang
online, and the questionnaire was designed to be completed in approx­
et al., 2019; Zhang, 2022), although much of this work merely measured
imately 15 min. The use of an online survey platform and Likert scale
intentions and was assessed in a gaming setting. Therefore, the current
questions enabled the researchers to efficiently collect data from a large
work intends to address multiple gaps in the literature. Our first hy­
sample size.
pothesis is that users will tend to view avatars as an extension of
themselves, a conceptual replication of prior work which would verify
that avatars continue to serve a similar function as they have in prior 3.2. Participants
work.
The study’s participants ranged from 25 to 65+ years. The largest
H1. The proportion of users who agree with items such as “My avatar is age group was represented by individuals between 55 and 64 years,
an extension of myself”, “I care about my avatar’s appearance”, and “I constituting 36.02% of the total sample (refer to Appendix, Table 1).
invest considerable effort in customizing my avatar” will be significantly With regard to gender, the majority of participants were female, making
greater than 50%. up 59% of the sample. In terms of ethnicity, the predominant group
Next, we intend to address if there are relationships between user’s
attachment to their avatars and the amount of money they spend in SL. Table 1
Relationships will finally address if avatar relationships actually Summary (number, proportion, p-value) of users who responded with 4 or 5 on
contribute to the purchasing behavior in this specific online setting. the Likert scale on which users find their avatar to be extensions of themselves.
Question Number Proportion p-value (one-
H2. A positive correlation will exists between users’ attachment to
sided)
their avatars and purchasing behavior in Second Life (SL), manifested in
My avatar is the extension of myself 148 0.75
a significant association between the weekly amount of money spent in <.01
I do care how my avatar looks 176 0.90 <.01
SL/amount of purchases in SL and responses to following questions such I put a lot of work customizing my 166 0.84 <.01
as “I have a strong connection with my avatar”, “I invest considerable avatar

3
M. Kalyvaki et al. Computers in Human Behavior: Artificial Humans 1 (2023) 100016

consisted of those who identified as Americans, accounting for 70.42% Table 2

of participants. Cramer’s V and p-values (in parentheses) for Fisher’s exact test/chi-square test
A closer look at the participants’ occupational backgrounds revealed on which users find their avatar to be extensions of themselves.
that 29.5% were retired. Educational attainment varied among partici­ Avatar. Care.avatar. Customizing.
pants, with nearly half (46%) having completed some college course­ extension looks avatar
work or obtained a 4-year college degree. Additionally, marital status Avatar.extension –
indicated that approximately 40.67% of participants were married. Care.avatar.looks .32 ** –
Expanding upon these findings, it is essential to consider the diverse Customizing. .44 ** .40 ** –
avatar
demographic characteristics when analyzing the study’s results. These
demographics may offer valuable insights into potential factors that Note: **p < .01.
could influence participants’ perceptions and experiences in the context
of the research topic. weekly in Second Life (SL). We computed Pearson r correlations and p-
values for relationships between avatar attachment and Lindens spent.
3.3. Statistical power and analytic strategy Table 3 shows the results, indicating that there were significant associ­
ations between avatar attachment and the amount of Lindens spent
The importance of having an appropriate sample size in statistical weekly in SL for each of the avatar outcomes. This provides support for
analysis cannot be overstated, as it is essential for maintaining an Hypothesis 2.
acceptable level of test power and for ensuring accurate p-value calcu­ Next, we analyzed relationships between the avatar items and how
lations that can effectively detect the necessary effect size or differences often users purchased items in SL. As shown in Table 4, significant
between groups (Serdar et al., 2021). Obtaining a sufficient number of correlations emerged between how often users purchased items and the
participants for survey research can pose significant challenges, and in avatar attachment questions. More specifically, purchased items were
this particular study, the required sample size was determined post hoc, related to avatar connection, customizing the avatar, changing the look
following the completion of the survey and the collection of data from of the avatar, caring about the avatars looks, and seeing the avatar as an
214 respondents. A significance level of 0.05 was established for this extension of oneself (all ps < .001).
analysis. Next, although this test was exploratory in nature, we conducted
To calculate statistical power, the GPower software developed by mediation analyses to examine if the relationship between sense of
Heinrich Heine University Düsseldorf (2019) was utilized (Heinrich connection to the avatar and Lindens spent was explained by the number
Heine University Düsseldorf, 2019). In the context of the Chi-square test, of purchases. Tests of indirect effects were conducted using the PRO­
a medium effect size of 0.3 for Cramer’s V was assumed, based on the CESS macro in SPSS (Hayes, 2017; Model 4). Indirect effects testing
recommendations of Gravetter et al. (2020). With these parameters, the showed a significant mediated effect (Mediated Effect—ME = 0.08, SE
achieved power was found to be 0.80 indicating that the study’s sample = 0.03, 95% CI [0.029, 0.145], p < .05). This model can be found in
size was adequate for this analysis. Fig. 1.
In order to test our hypotheses, we utilized statistical tests such as
Pearson’s r correlations, Cramer’s V, and mediation tests using boot­ 5. General discussion
strapped tests of 95% confidence intervals for indirect effects (Gravetter
& Wallnau, 2004). Furthermore, for the proportion z-test, a difference in The current research sought to explore associations between aspects
proportion of 0.1 was considered. The computed power for this test was of avatar engagement and actual purchase behaviors in SL. Findings
calculated to be 0.89, which is comfortably above the widely accepted suggested avatar engagement was linked with the amount of Linden’s
threshold of 0.80. This finding further supports the adequacy of the spent, as well as links to the total amount of purchases. Furthermore, the
sample size for the purposes of the study and suggests that the statistical link between avatar connection and Lindens spent and was explained by
analysis carried out was robust and reliable. number of purchases. It does appear that those who care more about
their avatars are reporting more purchases and Lindens spent in SL than
4. Results those who cared less about their avatars. These findings are consistent
with prior work suggesting that avatars may promote purchasing
4.1. An analysis of user perspectives and spending patterns in Second Life behavior in digital settings (Bessière et al., 2007). In terms of the
mediation model, it does suggest that those who care more about their
Our first hypothesis was that users will find their avatars to be an avatars are spending more Lindens, and this is because they are making a
extension of themselves, care about their avatar’s appearance, and put higher number of overall purchases in SL. This suggests that avatar
effort into customizing their avatars. To test this hypothesis, we engagement creates a pipeline toward buying more frequently and
analyzed data from the users Likert scale responses, considering users spending more money in-game int this particular setting.
who responded with 4 or 5 on the Likert scale as in agreement with the In virtual worlds such as social media platforms or video games,
item. Table 1 presents a summary of the responses where we found that users often create a persona (e.g., avatar) that represents their agency
75% of users see their avatar as an extension of themselves, 90% care within that space. The relationship between users and their avatars is
about their avatar’s appearance, and 84% put effort into customizing complex, as noted in psychological studies (Banks & Bowman, 2021).
their avatars. All three proportions are significantly higher than the While many researchers have explored the psychological merger of users
neutral point of 50% (p < .01), providing support for Hypothesis 1 that and their avatars, fewer have examined the relationship between avatars
users do indeed form an attachment to their avatars in virtual as an extension of oneself, which can drive customization efforts and
environments.
Table 2 presents the Cramer’s V effect sizes and associated p-values
from the Fisher’s exact test or Chi-square test for the avatar-related Table 3
Pearson r correlations p-values for association of Lindens spent weekly in SL.
questions. The results indicate significant associations between the
variables, further supporting the notion that users find their avatars to Pearson’s r p-value
be an extension of themselves. I have a strong connection with my avatar .29 <.001
Our second hypothesis is that users’ attachment to their avatars will My avatar is the extension of myself .24 <.001
correlate with their consumption behavior in virtual environments. I do care how my avatar looks .21 .004
I put a lot of work customizing my avatar .44
First, we tested for relationships with the amount of Lindens spent
<.001

4
M. Kalyvaki et al. Computers in Human Behavior: Artificial Humans 1 (2023) 100016

Table 4 are more invested in their avatars and the virtual environment not only
Fisher’s exact test p-values. tend to purchase more in SL, but they also engage in more adjacent
Variable 1. 2. 3. 4. 5. 6. 7. avatar-related activity. While this is not all that surprising, it does sug­
gest that avatar behaviors share heavy overlap, and that failed in­
1. Lindens – .38** .25** .20** .33** .22** .31**
2. Purchase – .29** .44** .47** .21** .24** terventions which target specific avatar behaviors may be easily
3. Connection – .38** .25** .15* .23** adjusted to target different, adjacent avatar-related behaviors instead.
4. Custom – .30** .09 .17* This should instill a sense of promise in creating avatar interventions by
5. Look – .57** .58** marketers.
6. Care – .70**
7. Extension –

Note: Ns range from 193 to 200 due to missingness. **p < .01; *p < .05. 5.1. Limitations and future research

Several limitations of the study should be acknowledged to interpret

increase consumption. These findings provide an important first step
its findings effectively. First, although various virtual communities exist,
into understanding the relationships between avatar usage and con­
our focus in this study was solely on Second Life (SL), limiting the
sumer behavior in online settings like SL. Although prior work has begun
generalizability of the results. Consequently, the findings may not apply
to explore this in the gaming world, the expansion of online markets
to other virtual worlds or communities. Second, the data for the study
such as the Metaverse require further research that establishes the
were collected exclusively using a web-based survey, raising the possi­
importance of avatars in promoting consumer purchases. The relation­
bility that the results might have been affected by common method
ships between the number of purchases/Lindens and various avatar
variance. Third, while the sample comprised a fair mix of males and
items also suggests that producers can target these specific components
females, the study did not specifically focus on the potential con­
to encourage their consumer to buy more products. For example, Chung
founding effects of gender. Finally, and perhaps most importantly, this is
(2005) notes that the ease of avatar use is an important factor for pro­
a single, preliminary step towards understanding the link between
moting sustained avatar involvement. If avatar use is linked with pur­
avatar use and purchasing behavior. While we feel this is critical for
chasing, as was found in this study, creating more effective avatar
building more complex models and frameworks moving forward, future
maintenance tools online will, in theory, help to ensure that the positive
work should consider replicating these findings and continuing to build
effects of avatar use will translate into consumption. In order to effec­
a strong body of evidence which demonstrates links between avatar use
tively cater to the needs of these VR residents, it is vital for companies to
and purchasing outcomes.
delve deeper into the factors that influence avatar attachment and cus­
Future research could address these limitations and explore new
tomization. By doing so, they can create personalized experiences that
directions. For instance, examining and comparing the perceptions of
resonate with users on a profound level. This, in turn, can lead to
males and females concerning identifiability in virtual spaces could offer
increased user engagement and retention rates within the metaverse.
valuable insights into potential gender differences. Moreover, repli­
By understanding these factors, businesses can develop targeted
cating the results of this study using experimental methodology would
strategies to engage and retain users in the metaverse, ultimately
provide more robust evidence and further insights into the research
building a strong and loyal consumer base within the virtual world. This
question. Findings from an experiment-based approach also help
could involve offering unique virtual goods, experiences, or services that
establish causal relationships and reduce concerns about common
cater to users’ preferences, providing opportunities for social in­
method variance.
teractions, and creating a seamless and immersive virtual environment.
Additionally, future research could explore the effects of time users
Recognizing the metaverse as an extension of users’ real lives is crucial
spend in customizing their avatars on their perceptions of identifiability.
for businesses and organizations to succeed in this growing digital
Investigating the relationship between avatar customization efforts and
landscape. By focusing on the factors that influence avatar attachment
users’ attachment to virtual representations may provide a deeper un­
and customization, companies can create compelling experiences that
derstanding of how users perceive virtual identities. Finally, extending
resonate with users, fostering a dedicated and loyal consumer base
the study to other virtual worlds or communities would increase the
within the virtual world. Our mediation model also suggests that the
generalizability of the findings, contributing to a more comprehensive
avatar is not just an engagement tool which increases likeability of a
understanding of avatar attachment and identifiability across different
platform—it can literally serve as a pipeline for purchases within the
online environments.
platform. It is likely that avatars promote attachment to the services
being provided by the game, encouraging users to make more purchases
and spend more money. 5.2. Conclusion
Lastly, the consideration of avatars as extensions of oneself exhibited
significant associations with an array of factors, including purchase To conclude, the Metaverse (an all-encompassing digital universe) is
frequency, avatar customization, avatar appearance, and concern for the connected to virtual reality platforms such as Second Life, which allow
avatar’s looks (Yee et al., 2009). These associations imply that users who users to create and customize avatars. The attachment that users form
with their avatars then influences their consumer behavior within the

Fig. 1. Images of Sl classic starter avatars and a female and male customized avatar (Linden Lab, 2015).

5
M. Kalyvaki et al. Computers in Human Behavior: Artificial Humans 1 (2023) 100016

virtual world, leading to virtual consumption of products and services. elsewhere for publication.
This, in turn, can impact the success of businesses operating within the
Metaverse. Finally, money spent based on income groups, gender, and Permission notes
other demographic factors may provide additional insights into
spending patterns. The limited data points and small sample size pre­ The authors confirm that all materials, figures, and data included in
sents an opportunity for future research. the manuscript are original content, and no permission is required from
third parties for their use in this submission.
Funding source declaration
Declarations of interest
The authors report that no external funding was received for the
research, authorship, or publication of this manuscript. None.

Author agreement/declaration Declaration of competing interest

All authors have contributed significantly to the research, writing, The authors declare that they have no known competing financial
and revision of this manuscript. They have seen and approved the final interests or personal relationships that could have appeared to influence
version of the manuscript and confirm that it has not been submitted the work reported in this paper.

Appendix
Table 1
Sociodemographic characteristics of participants

Sociodemographic Characteristic N %

Gender
Female 127 59.35
Male 79 36.92
Non-Binary 3 1.40
Prefer Not to Say 5 2.34
Age
25–34 years old 16 7.58
35–44 years old 25 11.85
45–54 years old 56 26.54
55–64 years old 76 36.02
65+ years old 38 18.01
Highest educational level
Less than high school degree 5 2.43
High school graduate (high school diploma or equivalent including GED) 35 16.99
Some college but no degree 53 25.73
Associate degree in college (2-year) 30 14.56
Bachelor’s degree in college (4-year) 43 20.87
Master’s degree 25 12.14
Doctoral degree 5 2.43
Professional degree (JD, MD) 10 4.85
Married Status
Divorced 66 31.58
Married 85 40.67
Never Married 35 16.75
Separated 8 3.83
Widowed 15 7.17
Income
Less than $10,000 21 11.86
$10,000 to $19,999 19 10.73
$20,000 to $29,999 25 14.12
$30,000 to $39,999 13 7.34
$40,000 to $49,999 10 5.66
$50,000 to $59,999 10 5.66
$60,000 to $69,999 12 6.78
$70,000 to $79,999 12 6.78
$80,000 to $89,999 13 7.34
$90,000 to $99,999 8 4.53
$100,000 to $149,999 25 14.12
$150,000 or more 9 5.08
Note: n = 214. Not all participants completed the entire questionnaire.

6
M. Kalyvaki et al. Computers in Human Behavior: Artificial Humans 1 (2023) 100016

Fig. A1. Mediation model in the study.

References Computing Systems - Proceedings, 1151–1160. https://doi.org/10.1145/

1518701.1518877
Dwivedi, Y. K., Hughes, L., Baabdullah, A. M., Ribeiro-Navarrete, S., Giannakis, M., Al-
Ainsworth, M. D. S., & Bell, S. M. (1970). Attachment, exploration, and separation:
Debei, M. M., Dennehy, D., Metri, B., Buhalis, D., Cheung, C. M. K., Conboy, K.,
Illustrated by the behavior of one-year-olds in a strange situation. Child Development,
Doyle, R., Dubey, R., Dutot, V., Felix, R., Goyal, D. P., Gustafsson, A., Hinsch, C.,
41(1), 49. https://doi.org/10.2307/1127388
Jebabli, I., … Wamba, S. F. (2022). Metaverse beyond the hype: Multidisciplinary
Alcañiz, M., Parra, E., & Giglioli, I. A. C. (2018). Virtual reality as an emerging
perspectives on emerging challenges, opportunities, and agenda for research,
methodology for leadership assessment and training. Frontiers in Psychology, 9(SEP).
practice and policy. International Journal of Information Management, 66. https://doi.
https://doi.org/10.3389/FPSYG.2018.01658
org/10.1016/J.IJINFOMGT.2022.102542
Arachchi, H. D. M., & Samarasinghe, G. D. (2023). Impact of artificial autonomy avatar
Etienne, E., Leclercq, A.-L., Remacle, A., Dessart, L., & Schyns, M. (2023). Perception of
on consumer acceptance of fashion retail services. In 2023 international research
avatars nonverbal behaviors in virtual reality. Psychology & Marketing, 1–18. htt
conference on smart computing and systems engineering (SCSE) (Vol. 6, pp. 1–8). IEEE.
ps://doi.org/10.1002/mar.2187.
Banks, J., & Bowman, N. D. (2016). Emotion, anthropomorphism, realism, control:
Gao, Y. (2022). Research on the disney’s IP brand-building process and shaping tactics.
Validation of a merged metric for player–avatar interaction (PAX). Computers in
Journal of Education, Humanities and Social Sciences, 5, 81–88. https://doi.org/
Human Behavior, 54, 215–223. https://doi.org/10.1016/j.chb.2015.07.030
10.54097/ehss.v5i.2886
Banks, J., & Bowman, N. D. (2021). Some assembly required: Player mental models of
Gravetter, F. J., & Wallnau, L. B. (2004). Statistics for the behavioral sciences. Thomson/
videogame avatars. Frontiers in Psychology, 12, Article 701965. https://doi.org/
Wadsworth.
10.3389/fpsyg.2021.701965
Guo, Y., & Barnes, S. (2011). Purchase behavior in virtual worlds: An empirical
Beglar, D., & Nemoto, T. (2014). Developing likert-scale questionnaires. JALT2013
investigation in Second Life. Information and Management, 48(7), 303–312. https://
Conference Proceedings, 1–8.
doi.org/10.1016/j.im.2011.07.004
Belk, R. W. (2013). Extended self in a digital world. Journal of Consumer Research, 40(3),
Heinrich Heine University Düsseldorf. (2019). Universität Düsseldorf: G*Power. Version
477–500. https://doi.org/10.1086/671052
3.1.9.4 Germany https://www.psychologie.hhu.de/arbeitsgruppen/allgemeine-ps
Bente, G., Rüggenberg, S., Krämer, N. C., & Eschenburg, F. (2008). Avatar-mediated
ychologie-und-arbeitspsychologie/gpower.
networking: Increasing social presence and interpersonal trust in net-based
Horáková, J., Uusitalo, O., Munnukka, J., & Jokinen, O. (2022). Does the digitalization of
collaborations. Human Communication Research, 34(2), 287–318. https://doi.org/
retailing disrupt consumers’ attachment to retail places? Journal of Retailing and
10.1111/J.1468-2958.2008.00322.X
Consumer Services, 67, Article 102958. https://doi.org/10.1016/J.
Bessière, K., Seay, A. F., & Kiesler, S. (2007). The ideal elf: Identity exploration in world
JRETCONSER.2022.102958
of warcraft. CyberPsychology and Behavior, 10(4), 530–535. https://doi.org/
Johnson, K. K. P., Kim, H. Y., Mun, J. M., & Lee, J. Y. (2015). Keeping customers
10.1089/CPB.2007.9994
shopping in stores: Interrelationships among store attributes, shopping enjoyment,
Bidmon, S. (2017). How does attachment style influence the brand attachment–brand
and place attachment. International Review of Retail Distribution & Consumer Research,
trust and brand loyalty chain in adolescents? International Journal of Advertising, 36
25(1), 20–34. https://doi.org/10.1080/09593969.2014.927785
(1), 164–189.
Kalyvaki, M. (2023). Navigating the metaverse business and legal challenges: Intellectual
Boateng, H., Kosiba, J. P., Adam, D. R., Ofori, K. S., & Okoe, A. F. (2020). Examining
property, privacy, and jurisdiction. Journal of Metaverse. https://doi.org/10.57019/
brand loyalty from an attachment theory perspective. Marketing Intelligence &
jmv.1238344
Planning, 38(4), 479–494. https://doi.org/10.1108/MIP-03-2019-0161/FULL/XML
Kerschbaumer, R. H., Kreimer, D., Foscht, T., & Eisingerich, A. B. (2023). Subscription
Bonsu, S. K., & Darmody, A. (2008). Co-creating second life: Market-Consumer
commerce: An attachment theory perspective. International Review of Retail
cooperation in contemporary economy. Journal of Macromarketing, 28(4), 355–368.
Distribution & Consumer Research, 33(1), 92–115.
https://doi.org/10.1177/0276146708325396
Kim, D. Y., Lee, H. K., & Chung, K. (2023). Avatar-mediated experience in the metaverse:
Brocato, E. D., Baker, J., & Voorhees, C. M. (2015). Creating consumer attachment to
The impact of avatar realism on user-avatar relationship. Journal of Retailing and
retail service firms through sense of place. Journal of the Academy of Marketing
Consumer Services, 73, Article 103382.
Science, 43(2), 200–220. https://doi.org/10.1007/S11747-014-0381-X
Kim, S., Sam, Kim, J., Badu-Baiden, F., Giroux, M., & Choi, Y. (2021). Preference for
Candel, O. S., & Turliuc, M. N. (2019). Insecure attachment and relationship satisfaction:
robot service or human service in hotels? Impacts of the COVID-19 pandemic.
A meta-analysis of actor and partner associations. Personality and Individual
International Journal of Hospitality Management, 93. https://doi.org/10.1016/j.
Differences, 147, 190–199.
ijhm.2020.102795
Chung, D. (2005). Something for nothing: Understanding purchasing behaviors in social
Kohler, T., Fueller, J., Matzler, K., & Stieger, D. (2011). CO-creation in virtual worlds:
virtual environments. CyberPsychology and Behavior, 8(6), 538–554.
The design of the user experience. MIS Quarterly: Management Information Systems, 35
Dillman, D. A. (1978). Mail and telephone surveys : The total design method. Wiley.
(3), 773–788. https://doi.org/10.2307/23042808
Dillman, D. (1991). The design and administration of mail surveys. Annual Review of
Koles, B., & Nagy, P. (2021). Digital object attachment. Current Opinion in Psychology, 39,
Sociology, 17(1), 225–249. https://doi.org/10.1146/annurev.soc.17.1.225
60–65. https://doi.org/10.1016/J.COPSYC.2020.07.017
Ducheneaut, N., Don Wen, M. H., Yee, N., & Wadley, G. (2009). Body and mind: A study
Kowalczuk, P., Siepmann (née Scheiben), C., & Adler, J. (2021). Cognitive, affective, and
of avatar personalization in three virtual worlds. Conference on Human Factors in
behavioral consumer responses to augmented reality in e-commerce: A comparative

7
M. Kalyvaki et al. Computers in Human Behavior: Artificial Humans 1 (2023) 100016

study. Journal of Business Research, 124, 357–373. https://doi.org/10.1016/J. Second life and Tilia parent Company adds Key leadership roles to support Rapid growth
JBUSRES.2020.10.050 (n.d.). Retrieved May 5, 2022, from https://www.prnewswire.com/news-releases/se
Krischke-Leitão, D., & Gomes, L. G. (2021). Second Life: Comme espace de sociabilité cond-life-and-tilia-parent-company-adds-key-leadership-roles-to-support-rapid-gr
pendant la pandémie de COVID-19. In Anthropologica (Vol. 63)University of Toronto. owth-301537083.html.
https://doi.org/10.18357/anthropologica6312021352. Issue 1. Serdar, C. C., Cihan, M., Yücel, D., & Serdar, M. A. (2021). Sample size, power and effect
Lab, L. (2015). New classic avatars are now. Second Life! Second Life Community. Available size revisited: Simplified and practical approachin pre-clinical, clinical and
in https://community.secondlife.com/blogs/entry/1803-new-classic-avatars-are-n laboratory studies. Biochemia Medica, 31(1), 1–27. https://doi.org/10.11613/
ow-available-in-second-life/. BM.2021.010502
Lee, K. M., & Nass, C. (2003). Designing social presence of social actors in human Shaw, K., & Sullivan, P. (2013). Retail place attachment: A qualitative study of apparel
computer interaction. Conference on Human Factors in Computing Systems - shoppers. Journal of Global Fashion Marketing, 4(4), 284–298. https://doi.org/
Proceedings, 289–296. https://doi.org/10.1145/642611.642662 10.1080/20932685.2013.822682
Lee, J. A., Sudarshan, S., Sussman, K. L., Bright, L. F., & Eastin, M. S. (2021). Why are Slater, M., & Wilbur, S. (1997). A framework for immersive virtual environments (FIVE):
consumers following social media influencers on Instagram? Exploration of Speculations on the role of presence in virtual environments. Presence: Teleoperators
consumers’ motives for following influencers and the role of materialism, 41(1), 78–100. and Virtual Environments, 6(6), 603–616. https://doi.org/10.1162/
https://doi.org/10.1080/02650487.2021.1964226 PRES.1997.6.6.603
Lehdonvirta, V. (2009). Virtual item sales as a revenue model: Identifying attributes that Stefania, C., Rogier, G., Beomonte Zobel, S., & Velotti, P. (2023). The relation of anxiety
drive purchase decisions. Electronic Commerce Research, 9(1–2), 97–113. https://doi. and avoidance dimensions of attachment to intimate partner violence: A meta-
org/10.1007/s10660-009-9028-2 analysis about victims. Trauma, Violence, & Abuse, 24(2), 1047–1062.
Lehdonvirta, V. (2011). Real-money trade of virtual assets: Ten different user Sun, J., Gan, W., Chao, H.-C., & Yu, P. S. (2022). Metaverse: Survey, applications,
perceptions. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.1351772 security, and opportunities. ACM Computing Surveys, 1(1), 35. https://doi.org/
Lewis, M. L., Weber, R., & Bowman, N. D. (2008). “They may be pixels, but they’re MY 10.48550/arxiv.2210.07990
pixels:” Developing a metric of character attachment in role-playing video games. Takano, M., & Taka, F. (2022). Fancy avatar identification and behaviors in the virtual
CyberPsychology and Behavior, 11(4), 515–518. https://doi.org/10.1089/ world: Preceding avatar customization and succeeding communication. Computers in
cpb.2007.0137 Human Behavior Reports, 6, Article 100176.
Messinger, P. R., Ge, X., Stroulia, E., Lyons, K., Smirnov, K., & Bone, M. (1970). On the Thorpe, S., Bolster, A., & Neave, N. (2019). Exploring aspects of the cognitive
relationship between my avatar and myself. Journal For Virtual Worlds Research, 1(2). behavioural model of physical hoarding in relation to digital hoarding behaviours.
https://doi.org/10.4101/JVWR.V1I2.352 Digital health, 5, Article 2055207619882172.
Metaverse economy payment Solutions | Tilia (n.d.). Retrieved December 13, 2022, from Trepte, S., & Reinecke, L. (2010). Avatar creation and video game enjoyment: Effects of
https://www.tilia.io/. life-satisfaction, game competitiveness, and identification with the avatar. Journal of
Mystakidis, S., Berki, E., & Valtanen, J. P. (2021). Deep and meaningful e-learning with Media Psychology: Theories, Methods, and Applications, 22(4), 171. https://doi.org/
social virtual reality environments in higher education: A systematic literature 10.1027/1864-1105/A000022
review. Applied Sciences, 11(5). https://doi.org/10.3390/APP11052412 VanMeter, R. A., Grisaffe, D. B., & Chonko, L. B. (2015). Of “likes” and “pins”: The effects
Mystakidis, S., Christopoulos, A., & Pellas, N. (2022). A systematic mapping review of of consumers’ attachment to social media. Journal of Interactive Marketing, 32, 70–88.
augmented reality applications to support STEM learning in higher education. https://doi.org/10.1016/j.intmar.2015.09.001
Education and Information Technologies, 27(2), 1883–1927. https://doi.org/10.1007/ VanMeter, R., Syrdal, H. A., Powell-Mantel, S., Grisaffe, D. B., & Nesson, E. T. (2018).
S10639-021-10682-1 Don’t just “like” me, promote me: How attachment and attitude influence brand
Nagy, P., & Koles, B. (2014). My avatar and her beloved possession": Characteristics of related behaviors on social media. Journal of Interactive Marketing, 43, 83–97.
attachment to virtual objects. Psychology and Marketing, 31(12), 1122–1135. https:// https://doi.org/10.1016/J.INTMAR.2018.03.003
doi.org/10.1002/MAR.20759 Vasalou, A., & Joinson, A. N. (2009). Me, myself and I: The role of interactional context
Nowak, K. L., & Biocca, F. (2003). The effect of the agency and anthropomorphism on on self-presentation through avatars. Computers in Human Behavior, 25(2), 510–520.
users’ sense of telepresence, copresence, and social presence in virtual environments. https://doi.org/10.1016/j.chb.2008.11.007
Presence: Teleoperators and Virtual Environments, 12(5), 481–494. https://doi.org/ Waddell, T. F., Sundar, S. S., & Auriemma, J. (2015). Can Customizing an Avatar Motivate
10.1162/105474603322761289 Exercise Intentions and Health Behaviors Among Those with Low Health Ideals?, 18(11),
Park, C. W., Eisingerich, A. B., & Park, J. W. (2013). Attachment–aversion (AA) model of 687–690. https://doi.org/10.1089/CYBER.2014.0356. https://Home.Liebertpub.
customer–brand relationships. Journal of Consumer Psychology, 23(2), 229–248. Com/Cyber
https://doi.org/10.1016/J.JCPS.2013.01.002 Wang, H., Ruan, Y. C., Hsu, S. Y., & Sun, C. T. (2019). Effects of game design features on
Park, J., & Kim, N. L. (2022). From metaverse to the real world: The role of avatar player-avatar relationships and motivation for buying decorative virtual items. In
identification in consumer’s virtual purchasing behavior. In International textile and DiGRA conference (pp. 1–22).
apparel association annual conference proceedings (Vol. 79)Iowa State University Wedel, M., Bigné, E., & Zhang, J. (2020). Virtual and augmented reality: Advancing
Digital Press. No. 1. research in consumer marketing. International Journal of Research in Marketing, 37(3),
Pellas, N., Mystakidis, S., & Kazanidis, I. (2021). Immersive virtual reality in K-12 and 443–465. https://doi.org/10.1016/j.ijresmar.2020.04.004
higher education: A systematic review of the last decade scientific literature. Virtual Whan Park, C., Maclnnis, D. J., Priester, J., Eisingerich, A. B., & Lacobucci, D. (2010).
Reality, 25(3), 835–861. https://doi.org/10.1007/S10055-020-00489-9 Brand attachment and brand attitude strength: Conceptual and empirical
Procter, L., & Studies, G. (2020). I Am/We Are: Exploring the Online Self-Avatar differentiation of two critical brand equity drivers. Journal of Marketing, 74(6), 1–17.
Relationship, 45(1), 45–64. https://doi.org/10.1177/0196859920961041. https:// https://doi.org/10.1509/JMKG.74.6.1
doi-Org.Ezproxy.Mnsu.Edu/10.1177/0196859920961041 Yee, N., & Bailenson, J. (2007). The proteus effect: The effect of transformed self-
Riar, M., Xi, N., Korbel, J. J., Zarnekow, R., & Hamari, J. (2022). Using augmented reality representation on behavior. Human Communication Research, 33(3), 271–290.
for shopping: A framework for AR induced consumer behavior, literature review and https://doi.org/10.1111/J.1468-2958.2007.00299.X
future agenda. Internet Research. https://doi.org/10.1108/intr-08-2021-0611 Yee, N., Bailenson, J. N., & Ducheneaut, N. (2009). The proteus effect: Implications of
Ridden, G. M. (1997). Review: Life on the screen: Identity in the age of the internet. transformed digital self-representation on online and offline behavior.
Literature and History, 6(2), 117–118. https://doi.org/10.1177/ Communication Research, 36(2), 285–312. https://doi.org/10.1177/
030619739700600223 0093650208330254
“Second Life”, “Minecraft”... The “confined” find themselves in virtual worlds (n.d.). Zhang, K. (2022). How interactive narrative and avatar identification influence player’s in-
Retrieved April 8, 2022, from https://www.lemonde.fr/pixels/article/2020/04/0 game purchase intention in Otome mobile game.
1/second-life-minecraft-les-confines-se-retrouvent-dans-les-mondes-virtuels_60
35182_4408996.html.