Abstract
Purpose
Esports game developers are incorporating interactive virtual shopping (IVS) experiences into their products to facilitate consumer–brand interactions. Despite the potential influence of IVS, existing research lacks an understanding of its impact on consumers’ overall affect– particularly game enjoyment, a critical factor in sustaining engagement with esports products – and on attitudinal change toward brands in the real world.
Design/methodology/approach
An experiment was designed to examine the impact of shopping in the virtually interactive esports marketplace on game enjoyment. Mediating effects of positive change in brand attitude, perceived realism and character attachment and the moderating effect of self-avatar gender congruity were investigated using SPSS PROCESS Macro. Additionally, the impact of brand popularity on the relationship between IVS and brand attitude was examined using analysis of covariance (ANCOVA).
Findings
There was a positive impact of IVS on game enjoyment, mediated by a positive change in brand attitude, perceived realism and character attachment. Self-avatar gender congruity moderated the mediation effect of perceived realism. Notably, less popular brands exhibited a greater change in brand attitude compared to more popular brands.
Originality/value
The findings add to the body of literature on esports, particularly in investigating the role of IVS within the esports virtual world and advancing research on cognitive appraisal theory. This study informs practitioners in the esports industry about the value of creating interactive virtual marketplaces, as they serve to benefit sponsoring brands and enhance game enjoyment for players.
Keywords
Citation
Hwang, Y., Gupta, K. and Ock, D. (2025), "Gotta take my avatar shopping: impacts of interactive virtual shopping in esports", Internet Research, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/INTR-03-2024-0346
Publisher
:Emerald Publishing Limited
Copyright © 2024, Yongjin Hwang, Keshav Gupta and Deokkyung Ock
License
Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode
Introduction
The emerging practice of integrating interactive virtual shopping (IVS) represents a groundbreaking fusion of entertainment and commerce, promising to reshape user engagement and satisfaction. Unlike traditional shopping platforms, IVS enables users to explore products in highly dynamic ways, fostering deeper connections with their purchases through immersive interactions (Hamari et al., 2017). For instance, Alibaba’s Buy + platform enables consumers to enter a virtual mall where they can browse, inspect products up close and view them on models before purchasing. This innovative approach not only enhances the interactive quality of the shopping experience but also helps reduce product returns, creating substantial opportunities for marketers to increase consumer engagement (Cappasity, 2024).
In this evolving landscape, esports particularly stands out as a uniquely effective platform where brands can interact with consumers on a continuous basis. Unlike other IVS platforms that may allow only brief and limited product interactions (e.g. rotating or enlarging a product), IVS in esports keeps players engaged with their purchased items throughout gameplay. This continuous interaction is a distinctive feature of esports IVS: it transforms the traditional, one-time shopping experience into an ongoing journey where players can see, use and feel the impact of their purchases as they navigate the esports virtual world (EVW). For example, NBA2k22’s virtual town enables players to customize their avatars by shopping for authentic brands like Nike and Converse and enjoy these purchases within the EVW, creating an uninterrupted immersive experience (Rai, 2024). Furthermore, esports games, with their inherent interactivity and immersion, create an ideal environment for IVS, allowing players to seamlessly incorporate purchases – such as branded apparel and accessories – into gameplay without disrupting the experience (Steinkuehler and Duncan, 2008). Accordingly, IVS benefits brands by transforming gaming spaces into dynamic marketplaces where players continuously interact with branded items, allowing for longer brand visibility, and thereby unlocking substantial commercial opportunities within and outside of the esports ecosystem. IVS, therefore, can present a poised environment to drive both gamer satisfaction and brand value.
Despite IVS’s growing prominence, limited research has examined its impact in the EVW context, particularly in terms of its impacts on consumers’ affect and brand perception. A fragmented understanding exists around game enjoyment – a crucial factor in fostering sustained engagement – yet research on this topic is dispersed across marketing (e.g. Hwang et al., 2017), entertainment (e.g. Smith et al., 2021) and communication studies (e.g. Lee and Williams, 2019), often lacking cohesion and depth in exploring how IVS influences player experiences. In esports, gameplay remains the primary service, with IVS functioning as an auxiliary feature. If IVS enhances enjoyment, it could strengthen game loyalty and long-term player engagement; however, if IVS detracts from enjoyment, it risks undermining these goals. Another notable gap is the limited exploration of IVS’s impact within the EVW on consumer attitudes, especially for brands with varying levels of popularity. Given esports players can engage with branded items like skins or apparel for their avatars, understanding how the brands’ popularity impacts players’ attitudes toward the brand is crucial, as brands seek to leverage interactive shopping opportunities to enhance brand equity (Hwang et al., 2023). Therefore, understanding these dynamics is essential for game developers and marketers to effectively integrate IVS in esports, with the aim of fostering lasting consumption habits and loyalty (Caroux and Pujol, 2023).
Accordingly, guided by cognitive appraisal theory (CAT; Lazarus, 1991), this research examines the impact of IVS in esports on consumers’ overall game enjoyment and its effect on brand attitude based on brand popularity. To comprehensively explore the multifaceted impacts of IVS, a carefully structured experiment was devised, in which the mediating roles of positive change in brand attitude, perceived realism and character attachment were examined, while also evaluating the moderating influence of self-avatar gender congruity on these complex relationships. Furthermore, the analysis extends to how brand popularity affects the interplay between IVS and brand attitude, thereby illuminating the intricate dynamics of brand perception.
The findings of this study contribute significantly to the existing literature by integrating CAT (Lazarus, 1991) within the realm of IVS in esports. This research explores how these immersive shopping experiences affect consumers’ overall enjoyment of games, thereby deepening our understanding of emotional engagement in gaming contexts. Additionally, it fills existing gaps in the literature concerning the dynamics between brand popularity, consumer behavior and IVS, thus enriching the theoretical discourse on consumer cognition and engagement in esports environments. From a practical perspective, the study provides a strategic framework for utilizing in-game purchasing to boost player enjoyment in the game. Moreover, the investigation into the nuanced effects of brand popularity on consumer attitudes equips practitioners with essential knowledge for tailoring effective marketing strategies in the esports landscape, ultimately promoting deeper connections with target audiences within and outside of the esports ecosystem.
Literature review
Cognitive appraisal theory
According to CAT, emotion is elicited based on an individual’s appraisal of a stimulus, where “appraisal” refers to an individual’s interpretation of the characteristics of a phenomenon (Lazarus, 1991). The same stimulus can be appraised distinctively by individuals, thereby eliciting a different type of emotion. For example, while some individuals may perceive an amusement ride at a carnival as entertaining and elicit a joyful emotion, others who perceive the ride to be dangerous would feel afraid. In the past, scholars have proposed diverse appraisals that cause emotions; however, the inconsistencies in the types and definitions of appraisals prompted Watson and Spence (2007) to categorize the appraisals that predicted a wide range of consumption emotions. The four key appraisals advanced by them were outcome desirability, agency, certainty and fairness.
Outcome desirability refers to the cognitive assessment of whether the situation is positive or negative relative to a benchmark (Watson and Spence, 2007). The assessment can be driven by goals or an evaluation of pleasantness. Specifically, if an event offers hedonic value independent of the event’s ability to further goal attainment, then the event is assessed as being positive and generates a positive valanced emotion (Smith and Ellsworth, 1985). The agency or the causal agent is the entity that has control over the stimulus. An individual may perceive the agency as oneself, someone else, or a circumstance (Smith and Ellsworth, 1985). Past research highlights its link to different coping strategies based on perceived control and responsibility (Watson and Spence, 2007). Certainty, which refers to the level of predictability of an event, influences an individual’s emotion (Smith and Ellsworth, 1985), such that a familiar event elicits favorable emotions, such as joy and excitement. Lastly, fairness refers to the moral appropriateness of an event (Smith and Ellsworth, 1985). Specifically, a positive emotion is experienced if the moral assessment of an event aligns with an individual’s standards. The four appraisals can act independently to influence an individual’s emotions (Lazarus, 1991; Lazarus and Folkman, 1984). In the context of IVS within the EVW, individuals should elicit a positive emotion toward the game if they favorably evaluate the shopping activity based on the four appraisals.
Interactive virtual shopping in esports
The rise of e-commerce has made shopping convenient as its personalization and recommendation engines allow consumers to search and purchase products with minimal effort (Jobanputra, 2023). Regardless, a limited proportion of the consumers in the gaming industry engage in shopping behaviors, thereby restricting the in-game revenue generation for organizations (Gupta et al., 2022). To counter this challenge, esports game developers are constructing virtual shopping environments conducive to heightened consumer interaction, thereby offering consumers a sense of realism during their shopping experience (Rai, 2024). Such esports games go beyond traditional gaming as consumers, in addition to playing the game, can explore virtual shopping areas, purchase virtual goods and interact with the products in the game. It is imperative to understand whether IVS influences consumers’ affect, specifically enjoyment of the game, as an unpleasant experience in such auxiliary services can be detrimental to the overall experience consumers obtain in the game. However, existing research offers a limited understanding of the impact of IVS on consumers’ enjoyment. Table 1 offers an overview of the current research on online shopping.
Existing research predominantly examines non-immersive online shopping, such as in-app purchases (e.g. Ghazali et al., 2018; Hamari et al., 2019) and e-commerce 2-D shopping (e.g. Gulfraz et al., 2022). Such research offers a comprehensive understanding of factors impacting the adoption of online shopping devices (Ghazali et al., 2018), loyalty toward the shopping platform (Mofokeng, 2021) and online purchase behavior (Gupta et al., 2022; Hamari et al., 2019). Additionally, past research examines online shopping as a consequence of online addictive behaviors (e.g. Erhel et al., 2024; Starcevic et al., 2023) and consumers’ interaction with the game elements (e.g. Gupta et al., 2024). However, this segment of the literature falls short of examining immersive and interactive shopping environments that are seamlessly integrated into the game’s design and allow individuals to interact with the purchased product throughout the game. While non-immersive shopping may offer interactive elements, such as product customization, consumer engagement with the service typically concludes at the point of purchase. Contrarily, IVS within the EVW allows consumers to be immersed in an alternate world, where the experience of interactive shopping extends beyond mere transactions, influencing their overall gaming experience as they observe the meaningful effects of their purchases while navigating the EVW.
Current research on IVS also offers an understanding of the ability of immersive environments to mimic real-world scenarios and the determinants of purchase behavior in these environments. Specifically, Lombart and colleagues (2020) highlighted the ability of IVS to offer consumers similar levels of perceived product quality as in the real world, prompting more purchases in the virtual environment. Moreover, scholars highlight the impact of the immersive aspects of IVS on consumers’ cognition, resulting in favorable purchase decisions. For example, Hooker et al. (2019) indicated the positive impact of a sense of flow derived from an IVS on consumer’s attitudes and purchase intentions toward the displayed brands. Furthermore, such purchase decisions were impacted by the perceived informativeness and playfulness derived from IVS (Kang et al., 2020). Saleem et al. (2024) further advanced the understanding of IVS by shedding light on the positive impact of IVS on telepresence and sense of escape, resulting in post-purchase intentions, such as re-purchase and word of mouth. Additionally, these relationships were stronger when consumers had a higher level of attachment to their avatars.
While research on IVS provides a comprehensive understanding of the factors that impact a purchase decision, they are limited in the ability to explain the affective response of IVS toward the service. Although IVS may be financially beneficial for esports organizations, their impact on consumers’ emotions toward the game, especially game enjoyment, could create enduring consumption habits (Caroux and Pujol, 2023). Specifically, the overall experience of an activity is the result of multiple touchpoints that customers interact with during engagement (Funk, 2017). Within the esports context, these touchpoints can be the core gameplay itself and auxiliary services within the game, such as IVS. Therefore, it is necessary to understand the relationship between IVS and esports game enjoyment as it will contribute to the understanding of the consequence of shopping in the EVW on the overall enjoyment of the game.
Hypothesis development
Impact on game enjoyment
According to CAT, IVS within the EVW should result in a positive emotion toward the esports game if individuals appraise the shopping event positively. Specifically, if the four key appraisals of CAT are evaluated favorably with respect to shopping behavior, individuals should experience more enjoyment within the esports game. The shopping behavior includes activities in the virtual marketplace that mimic real-world scenarios, such as exploring the marketplace using the avatar, comparing products and purchasing the preferred product. Compared to merely being exposed to brands, shopping is a more cognitively involved activity that requires processing brand-related information and decision-making (Petty and Cacioppo, 1986), resulting in higher levels of enjoyment (c.f. Hall, 2009). Therefore, IVS, compared to exposure toward virtual products, in the EVW will offer individuals hedonic value. Furthermore, the cognitive assessment of shopping behavior will be favorable as consumers will be immersed in a realistic marketplace that would require them to have similar levels of cognitive involvement as in the real world, thereby resulting in positive appraisal via hedonic experience (Peukert et al., 2019).
In addition, past research highlights that individuals appraise the virtual environment positively in the presence of their decorated avatars (Moon et al., 2013). In the esports setting, shopping often aligns with players’ personal goals, such as improving the performance and appearance of their avatars (Neustaedter and Fedorovskaya, 2009). When shopping leads to outcomes players desire – such as a stronger avatar or a new skin – this appraisal creates feelings of joy and satisfaction. Such positive emotions, tied to a sense of accomplishment within the game, elevate the overall enjoyment of the esports game (c.f. Lazarus, 1991). Furthermore, in IVS within the EVW, players feel a sense of control and ownership over their avatars, making choices that directly impact their avatars, thereby positively appraising the sense of agency and fairness and reinforcing the enjoyment toward the game. Moreover, IVS within the EVW offers a sense of certainty as such environments mimic real-world scenarios that feel stable and predictable (Lombart et al., 2020), resulting in a favorable affective evaluation toward the game. Therefore, we hypothesize the following:
IVS within the EVW positively affects an individual’s game enjoyment.
Individuals’ evaluation of shopping in the esports marketplace depends on their subjective experience, which is influenced by the interaction of three elements, the brand and/or product (i.e. the sponsoring organization), the avatar (i.e. a virtual character analogous to the user) and the digital environment (i.e. the context; Funk, 2017). Specifically, individuals are likely to experience a greater enjoyment of the game when their shopping engagement leads to an improved attitude toward the brand, an increased sense of gameplay realism, and a stronger attachment to their avatar – stemming from positive assessments of fairness, certainty and agency appraisals, respectively. The following sections elaborate on these elements through the lens of CAT to explain the relationship between IVS within the EVW and game enjoyment.
Positive change in brand attitude
Like in the real world, individuals’ interaction with brands in digital space influences their attitude toward the brand (Hooker et al., 2019). Specifically, the presence of a brand in the virtual world and the emotional response toward the brand affect an individual’s brand attitude. Moreover, the appraisal of fairness is based on the assessment of the moral alignment of IVS with individual standards, which can shape brand attitude and elicit a favorable affective response toward the game (Watson and Spence, 2007). If brands within the EVW appear fair (i.e. reasonably priced with no overbearing in-game promotions) and satisfying, where individuals feel they are getting good value for their investment, the IVS experience will foster trust and value perceptions, creating a positive brand attitude that enhances consumers’ enjoyment of the game. Particularly, an enhanced brand attitude shaped by fairness appraisal would elicit a favorable view of the game developer, and this goodwill can make consumers more receptive to engaging with various game elements, thereby reinforcing their enjoyment of the game.
Moreover, individuals who purchase a brand’s product may develop stronger brand attitudes due to their perception of a long-term relationship with the brand. In the context of EVW, those who do not engage in shopping behaviors may become aware of brands through logos but will lack a sense of long-term commitment, as they do not own the products. In contrast, individuals who actively participate in shopping are more cognitively engaged with the brand; they assess and compare features across different offerings before making a purchase decision (Häubl and Trifts, 2000). As a result, these consumers are likely to find their shopping experiences within the EVW pleasurable, stemming from their anticipated long-term association with the brand they have purchased. Consequently, this positive appraisal of their shopping behavior will elicit favorable emotions toward the digital environment. Thus, we propose the following hypothesis:
The positive relationship between IVS within the EVW and an individual’s game enjoyment is mediated by a positive change in brand attitude.
Perceived realism
A factor that makes the EVW attractive and immersive to individuals is its ability to provide alternate realities for individuals to explore (Shafer et al., 2014). Perceived realism refers to the situation when a virtual environment appears real to an individual (Tavinor, 2019). According to Shafer et al. (2014), individuals perceive the virtual environment as more realistic when the environment is interactive and allows the individual to control their outcomes. Scholars have suggested that engaging in activities that resemble real-world behaviors, like shopping, in the virtual environment would add to the individual’s perception of realism (Moon et al., 2013). Specifically, when individuals engage in IVS within the EVW, they interact with the digital shopping environment that is designed to resemble a real-world shopping mall. Moreover, according to CAT, when individuals encounter situations that are predictable and consistent with their general behavior, they appraise the environment positively (Tiedens and Linton, 2001). Therefore, increased realism fosters a deeper enjoyment of the game as consumers feel fully absorbed in a lifelike environment. Accordingly, we hypothesize the following:
The positive relationship between IVS within the EVW and an individual’s game enjoyment is mediated by perceived realism.
Character attachment
Individuals create avatars that allow them to interact in the virtual environment. Past research highlights the significance of avatars in the EVW as they provide individuals with a sense of emotional intimacy, control over interactions in the virtual environments and self-character identification (e.g. Trepte and Reinecke, 2010; Trepte et al., 2009). Furthermore, scholars have found that individuals prefer to have control over the appearance of their avatar (Neustaedter and Fedorovskaya, 2009) and perceive the virtual environment better when the avatar is relatable and realistic (Latoschik et al., 2017). Specifically, in the esports context, consumers often see themselves as agents controlling their avatars’ success and appearance, thus deepening attachment to their characters. This appraisal of attachment with the avatar (i.e. agency) in IVS reinforces the consumer’s bond with their avatar, fostering a more personal experience. This attachment then enhances game enjoyment by creating a more immersive and personally relevant gaming environment. Therefore, we hypothesize the following:
The positive relationship between IVS within the EVW and an individual’s game enjoyment is mediated by character attachment.
Self-avatar gender congruity
One of the factors that is often considered in avatar-based research is the concept of gender congruity (Pitardi et al., 2023). Gender congruity refers to the match between an individual’s own gender and the gender of an entity (Trepte et al., 2009). According to the concept of self-congruity, individuals strive to protect their self-image by reducing the discrepancy between their self-image and the image of the entity with which they interact (Sirgy, 1982). In the context of EVWs, developers and creators often utilize avatars that resemble human features to offer a more realistic virtual environment (Latoschik et al., 2017). This highlights the importance of carefully considering gender representation in avatars, ensuring they align with users’ gender identities to enhance the authenticity and relatability of the experience.
Building on this, the concept of self-congruity suggests that individuals are more likely to form positive attitudes toward objects, brands or entities that mirror their self-image (Sirgy, 1982). In virtual environments, users who perceive gender congruity between themselves and their avatars tend to form stronger emotional connections with the virtual world and the brands within it (Oh et al., 2023). When an avatar’s gender aligns with the player’s self-perception, the experience can become more immersive, fostering a positive evaluation of the products and brands being interacted with. On the other hand, gender incongruence can weaken emotional engagement, diminishing the effectiveness of IVS on brand attitude. Therefore, we hypothesize the following:
Self-avatar gender congruity moderates the relationship between IVS within the EVW and a positive change in brand attitude.
Another important factor in virtual environments is perceived realism, which significantly influences the sense of immersion and engagement (Tavinor, 2019). When an avatar’s gender matches that of the user, the virtual experience can feel more authentic and personally relevant, thereby enhancing its realism (Schwind et al., 2017). Gender congruity can thus increase the user’s sense of immersion, making the virtual shopping experience appear more natural and credible. Conversely, a gender mismatch between the user and their avatar may create a disconnect, reducing the perceived realism of the virtual world and the shopping behaviors within it. This leads us to hypothesize:
Self-avatar gender congruity moderates the relationship between IVS within the EVW and perceived realism.
Finally, research has shown that avatar attachment is stronger when avatars align with users’ gender identities, as this congruence resonates more deeply with their self-concept (Suh et al., 2011). In the context of EVWs, such attachment can play a crucial role in how users interact with the virtual environment and engage in virtual shopping behaviors. A mismatch in gender may weaken users’ ability to form a strong emotional connection with their avatar, hindering their immersion in the shopping experience. Conversely, self-avatar gender congruity may foster a greater sense of identification with the avatar, leading to stronger attachment and more meaningful emotional investment in the IVS experience. Thus, we hypothesize the following:
Self-avatar gender congruity moderates the relationship between IVS within the EVW and character attachment.
The role of brand popularity
While IVS within the EVW can benefit the game by creating favorable affective responses of consumers toward the game, it can also offer a means for brands to improve their value. Specifically, esports offers an alternate environment for brands to become visible and connect with consumers in an interactive way (Hwang et al., 2023). Such digital interactions between brands and consumers can enhance brands’ value in the real world as consumers’ perception of the brands in the digital environment can spill over into the real world. However, such effects may not be consistent across all brands and may vary based on their popularity. Specifically, sponsoring brands in IVS within the EVW may have varying levels of popularity. Brand popularity is defined as the extent to which a brand is sought after and perceived by the general population (Kim and Chung, 1997). While the purchase of a branded product enhances consumers’ attitudes toward the brand, the magnitude of attitudinal change would vary based on the level of brand popularity. Attitude change toward a brand occurs when individuals are invested in appraising how well the brand satisfies dispositional needs (Petty and Cacioppo, 1986). From a CAT perspective, since consumers’ attitudes toward more popular brands are strong even prior to shopping behavior, their assessment of the outcome desirability of the brand in meeting their needs is adequate. Contrarily, consumers’ expectation of a less popular brand is marginal as such brands are not sought after. During the shopping behavior, consumers would rely on past experiences, compare the virtual products and choose one that best satisfies their needs. Consumers purchasing a more popular brand’s product would do so following their assessment of the brand’s ability to meet their expectations while those purchasing a less popular brand’s product would perceive the brand to exceed their prior expectation. Resultantly, the attitudinal change for less popular brands would be greater than for more popular brands.
IVS within the EVW will improve brand attitudes toward ‘less popular brands’ more than toward ‘more popular brands’ following a purchase.
Method
Participants
A total of 130 undergraduate students were recruited from a research university in the Southeastern part of the United States. Subjects were invited to study sports and entertainment management classes via an email solicitation. Those who agreed to participate in the experiment and successfully completed the required tasks received extra credit for the course they signed up for. A student sample was appropriate for this study because university students closely represent the demographic characteristics of the broader esports player base, who are typically young, tech-savvy and active on social media (Caroux et al., 2015).
Experimental procedure and manipulation
An experiment was designed to test the proposed hypotheses using an esports title, NBA2k22. While the chosen game title was primarily associated with basketball, the game offered a rich virtual world experience, including avatar customization, virtual town exploration and IVS. This study was designed to focus specifically on the IVS experience rather than traditional basketball gameplay, aiming to evaluate the impact of incorporating virtual shopping on overall esports game enjoyment. Participants created their own virtual selves by customizing their appearances such as their face, height, weight, skin color and hairstyle. While the creation of an avatar was highly customizable, only a male character was available due to the nature of the game, which created a dichotomous experimental condition of congruent or non-congruent gender identity with the character. Another experimental condition was IVS, which was also dichotomous of no-shopping vs. shopping. With respect to the gender identity condition, none of the participants reported non-matching biological sex and gender orientation, and thus female participants were automatically assigned to the self-avatar gender-incongruent group. Regarding the IVS condition, half of the participants were directed to purchase a pair of brand sneakers, and the other half explored the virtual shopping mall without purchasing any items. Purchase was limited to brand sneakers to control for the product category. Overall, the experimental design of the study was a two (IVS: no-shopping vs. shopping) by two (self-avatar gender congruity: incongruent vs. congruent) between-subjects structure (see Table 2 for experimental conditions).
Figure 1 presents the flow of the experimental design. Participants were first asked to fill out a pretest survey one week prior to their participation in the main experiment to assess any change in their attitude toward the brands after engaging in the IVS. The pretest survey included a list of brands featured in the game, including Adidas, Converse, Jordan, New Balance, Nike, Puma and Under Armour. To mitigate potential order bias, the sequence of brand names was randomized within the questionnaire. In addition to the target brands, dummy brands were also included in the pretest survey preventing participants’ attention to the target brands. Another purpose of the pretest survey was to control subjects’ previous experience by asking whether they had previously played the game title. Only those who had never played the NBA2k series were invited to the experiment.
After the pretest survey, a one-week period was employed before the main experiment to diminish any potential impacts of pretest experiences or judgments on the main experiment. Participants were then invited to visit a research laboratory located on the university’s campus that featured multiple Sony PlayStation 5 (PS5) stations. Each gaming station was separated with headphones and visual barriers. Once they arrived at the laboratory, participants were instructed on how to use the PS5 controller and the overall procedure of the experiment. Approximately 10 min of training time was assigned for the participants to get used to using the controller. Playing the NBA2k22 required elaborate control of in-game characters with complex skills. However, participants did not play an actual basketball game; instead, they created an avatar and traveled the virtual town in career mode, which did not demand complicated techniques of using the controller. Therefore, the 10 min of preparation time appeared to be appropriate. Once the practice session was over, all participants had 10 min to create their avatars without any limitations other than gender. After the creation of the avatar, they watched a short video clip of the introduction video embedded in the game by default. Then, in a third-person camera view, they controlled the avatar to visit a shopping center in the virtual town with a printed map. The virtual shopping center was designed to include a variety of brand stores, making the experience both engaging and realistic. At the shopping center, the shopping groups were allowed to freely visit the presented brand stores, try on items on display and purchase a pair of sneakers they liked the most within the budget of 5,000 virtual tokens, which was the average price of the available sneakers on sale. The purpose of the limited budget was to mimic a more realistic shopping experience. Participants in the no-shopping group were instructed to visit the virtual shopping center and were allowed to try on the displayed items, similar to the shopping group. However, they did not receive any virtual currency, and thus, no purchasing behavior was elicited from them. The key difference between the shopping and no-shopping conditions was the availability of virtual currency, which enabled participants in the shopping condition to make purchases. This design was specifically intended to separate the impact of the purchasing experience from the mere interaction with the items. By doing so, we aimed to isolate and evaluate the effects of making a purchase versus simply viewing and trying on items. After spending 15 min in the shopping center, participants were asked to explore the virtual city for an additional 15 min. During this time, those in the shopping group could enjoy their avatars adorned with the items they had purchased. This extended interaction reinforced the manipulation of the shopping condition by allowing participants to actively engage with and use their purchased items, enhancing the visual appeal of their avatars. Such a scenario, encompassing both the purchase and post-purchase usage, more accurately simulated the real-life experience of utilizing a newly acquired product, rather than merely focusing on the act of purchasing, thus further strengthening the manipulation of the shopping condition. In contrast, participants in the no-shopping group explored the town under the same conditions but without any enhancements to their avatars.
Upon completing their gameplay, each participant was provided with a QR code to access an online survey for data collection. In the posttest survey, participants were asked to answer questions regarding attitudes toward the brands, perceived realism of the gameplay, attachment to the character and game enjoyment. Their level of involvement in basketball was also captured.
Measures
Participants were randomly assigned to either the no-shopping (coded as “0”) or the shopping (coded as “1”) group without being given any information about the group assignment. In each group, participants were further allocated to self-avatar gender incongruent (coded as “0”) or congruent (coded as “1”) group based on their reported gender. Brand attitude was measured two times before and after the main experiment in both the pretest and posttest surveys. Each participant rated a 4-item, 7-point semantic differential scale. Based on the results of the pretest survey, the four brands were clustered together as “less popular brands” whose attitude scores were lower than 5 (i.e. Converse, M = 4.23, SD = 1.13; New Balance, M = 4.83, SD = 1.26; Puma, M = 4.23, SD = 1.05; Under Armour, M = 4.89, SD = 1.25) and the other three brands were classified as “more popular brands” (i.e. Adidas, M = 5.75, SD = 0.69; Jordan M = 5.52, SD = 0.97; Nike, M = 6.05, SD = 0.84). A paired samples t-test was further conducted to ensure that the distance between less and more popular brands was wide enough. Findings indicated that there was a significant difference between the average scores of less popular (M = 4.72, SD = 0.69) and more popular brands (M = 5.77, SD = 0.61), t(129) = 14.703, p < 0.001. In addition, positive change in brand attitude was also calculated by subtracting the pretest scores from the posttest scores. Other measures were assessed along 7-point numeric scales (1 = Strongly Disagree to 7 = Strongly Agree).
Table 3 provides a list of measures and their reliability scores. Notably, game enjoyment and basketball involvement were measured using single-item scales. Single-item measures have been shown to be reliable when the sample is not large and multiple items measuring the construct are homogeneous and semantically redundant (Diamantopoulos et al., 2012). Moreover, single-item measurements of enjoyment have been found to correlate with multi-item scales of enjoyment (Fu et al., 2009) and have been utilized in previous video game research, thus indicating the appropriateness of the game enjoyment measure applied in the current study (Rogers, 2017). Lastly, the multi-item scales were averaged to build a composite score.
Statistical analysis
We examined the relationships among IVS, positive change in brand attitude, perceived realism, self-avatar gender congruity and game enjoyment using SPSS PROCESS Macro (Hayes, 2017). IVS and game enjoyment were the independent and dependent variables, respectively. Positive change in brand attitude, perceived realism and character attachment were the three mediators, which were also moderated by self-avatar gender congruity. Following the guidelines outlined by Hayes (2017), we tested the direct effect of IVS on game enjoyment first and subsequently examined the indirect effects via the moderated mediating variables with 5,000 bootstrap samples. Additionally, we conducted a two-by-two mixed-method ANCOVA to compare brand attitudes toward “purchased items” before and after gameplay. While the moderated mediation analysis assessed the impact of IVS on the overall change in brand attitude across all in-game brands, this additional analysis was performed to gain a more detailed understanding of how the shopping experience influenced players’ attitudes specifically toward the brands they purchased. In the ANCOVA model, purchased brand popularity (less vs. more popular) and self-avatar gender congruity (incongruent vs. congruent) were between-subject factors and pre-post gameplay was a within-subject factor for comparing brand attitude before and after the gameplay. Finally, basketball involvement was included in both analyses as a covariate.
Results
Preliminary analysis
Most of the respondents were between 18 and 21 years of age (72%). Fifty-one percent of the sample were male, and the racial breakdown of study participants was 62% Caucasian, 16% African American, 12% Asian, 4% American Indian, 2% Native Hawaiian and 4% self-identified others. All constructs were reliably measured with the used scales (see Table 3). We also examined the correlations among the variables to check for multicollinearity. Since IVS and gender congruity were dichotomous variables, we performed a point-biserial correlation analysis (Tate, 1954). Overall, correlations showed expected directions of association, and no correlation coefficient was above 0.70 (see Table 4). In addition, the variance inflation factor scores of the predictor variables were all between 1 and 2. Therefore, the likelihood of multicollinearity was negligible (Tabachnick and Fidell, 2013). Finally, the result of an ordinary least square regression showed that IVS significantly predicted game enjoyment, F(1,129) = 3.89, p < 0.001, R2 = 0.11, which provided a preliminary result for the support of H1. This result also suggested that the conceptual model and the collected data were suitable for conducting additional data analysis of moderated mediation.
Testing of the moderated mediation model
We performed the bias-corrected bootstrap method with 5,000 samples to test the moderated mediating effects (Zhao et al., 2010) of positive change in brand attitude, perceived realism and character attachment on the relationship between IVS and game enjoyment. Basketball involvement was controlled in the model by including it as a covariate, whose effect was significant for positive change in brand attitude (β = 0.22, p = 0.03) and character attachment (β = 0.17, p = 0.03) but not for perceived realism (β = 0.07, p = 0.22) and game enjoyment (β = 0.05, p = 0.45). The summary of the PROCESS model showing all path coefficients is presented in Figure 2 and Table 5.
The results of the moderated mediation model identified that the direct paths from IVS to positive change in brand attitude (β = 0.98, p = 0.003), perceived realism (β = 1.10, p < 0.001) and character attachment (β = 0.80, p = 0.002) were all significant. However, the direct path from IVS to game enjoyment was insignificant (β = 0.17, p = 0.30), indicating that the direct impact of IVS on game enjoyment was diluted after the inclusion of the mediators. In addition, the paths from positive change in brand attitude (β = 0.13, p = 0.02), perceived realism (β = 0.30, p = 0.004) and character attachment (β = 0.25, p < 0.001) to game enjoyment were all significant, showing that the effect of IVS was fully mediated via positive change in brand attitude (supporting H2), perceived realism (supporting H3) and character attachment (supporting H4).
With regards to the moderation effects of self-avatar gender congruity, the results showed that the direct effects of self-avatar gender congruity were significant on perceived realism (β = 0.87, p < 0.001) and character attachment (β = 1.44, p < 0.001) but insignificant on positive change in brand attitude (β = 0.33, p = 0.31). Additionally, the interaction between IVS within the EVW and self-avatar gender congruity significantly and positively affected perceived realism (β = 0.99, p < 0.001), supporting H5b. However, the interaction between IVS and self-avatar gender congruity did not significantly influence positive change in brand attitude (β = 0.11, p = 0.80), rejecting H5a, and character attachment (β = 0.40, p = 0.27), rejecting H5c. We additionally investigated conditional mediating effects with the given two levels of self-avatar gender congruity. Intriguingly, the indirect effect of IVS on game enjoyment via positive change in brand attitude was significant for both gender incongruent (β = 0.14, SE = 0.08, 95% CI [0.01, 0.33]) and gender-congruent group (β = 0.13, SE = 0.08, 95% CI [0.01, 0.31]) and without containing zero in the 95% confidence interval (Hayes, 2017). The significant indirect effect of positive change in brand attitude for both groups suggests that IVS positively affected game enjoyment by first increasing players’ attitudes toward the brands in the game. This effect was present regardless of self-avatar gender congruity. Thus, both male and female participants, regardless of whether they controlled a gender-congruent or gender-incongruent avatar, experienced increased game enjoyment through improved brand attitudes resulting from IVS. However, the indirect effects via perceived realism (β = 0.33, SE = 0.15, 95% CI [0.07, 0.66]) and character attachment (β = 0.20, SE = 0.08, 95% CI [0.06, 0.37]) were significant only for the gender-congruent group.
Effects of IVS on brand attitude by brand popularity
We employed a two-by-two mixed-method ANCOVA to further analyze the effects of IVS within the EVW on brand attitude by brand popularity. Since it was necessary to examine whether the game setting itself engendered any unintended consequences, we first investigated the pre- and post-brand attitude scores of the no-shopping group. The results of paired samples t-tests revealed that, in the no-shopping groups, subjects’ brand attitude scores did not significantly change after the gameplay across the seven brands, which indicates the presentation of brands in the gaming context was not strong enough to change brand attitude. Once controlling for the impact of the game setting, a mixed-model ANCOVA was performed to test whether subjects’ shopping experience in the EVW environment affected their attitude toward the in-game brands they purchased.
As presented in Table 6, the ANCOVA model indicated significant main effects for both brand popularity, F(1, 60) = 9.45, p = 0.003 and pre-post gameplay, F(1, 60) = 9.84, p = 0.003. These results suggest that IVS within the EVW positively influenced their attitudes toward the items they purchased. Additionally, the overall attitude scores varied significantly between the two categories of brands – less and more popular. The main effect of self-avatar gender congruity was also significant, F(1, 60) = 4.68, p = 0.03, implying that male and female participants did not show similar levels of attitudes toward the in-game brands they purchased. Furthermore, the interaction between brand popularity and pre-post gameplay was significant, F(1, 60) = 8.64, p = 0.005, meaning that the increases in brand attitude scores were greater for less popular brands, supporting H6 (see Figure 3 for the individual brand attitude scores). No other significant interactions were identified in the model as presented in Table 6. Lastly, basketball involvement was not a significant covariate for brand attitude, F(1, 60) = 0.13, p = 0.72. Based on the overall results of the ANCOVA model, we concluded that IVS within the EVW significantly and positively changed brand attitudes. The changes in brand attitude for purchased brands, along with the interaction effect between brand popularity, pre-post gameplay and self-avatar gender congruity, are presented in Figure 3.
Discussion and theoretical implications
The results of the current study revealed that participants who shopped in-game brands felt a significantly higher level of enjoyment than those who did not purchase in-game brands (H1), supporting the belief that virtual shopping behavior favors esports consumers’ overall gaming experience. In the context of IVS, players engage in cognitive processing as they assess the value and relevance of virtual items. This cognitive assessment is a core component of the perception of stimuli, where the player’s enjoyment is derived not just from the purchase itself but from the evaluation and decision-making process associated with it (Petty and Cacioppo, 1986). The positively induced appraisal of the IVS experience, therefore, leads to positive emotions such as game enjoyment. Correspondingly, CAT was found to be a sound measure to explicate the EVW service consumption – players’ cognitive evaluations of IVS situations positively influenced their emotional responses, reinforcing the link between cognitive appraisals and emotional outcomes.
The findings of the study also extend the current understanding of IVS. Specifically, while past research on IVS highlights the ability of such environments to minim real-world conditions (Lombart et al., 2020) and the determinants of purchase behavior within IVS (Hooker et al., 2019; Kang et al., 2020), the current research sheds light on the impact of IVS on the enjoyment with the service. Moreover, the research examines a particularly unique IVS setting, i.e. EVW, thereby showcasing the ability of the esports landscape to foster meaningful customer–brand interactions that enhance consumers’ affective evaluation of the game and increased attitudes toward in-game brands within and outside of the esports ecosystem.
The dilution of the direct effect due to the mediators suggests that individuals’ evaluation of the EVW was dependent on their subjective experience with surrounding factors. As examined in this study, when esports players process IVS stimuli, their evaluations involve complex interpretations influenced by brand attitude, perceived realism and character attachment, leading to enhanced game enjoyment. These factors align with the argument of CAT that appraisals are multifaceted interpretations of stimuli that combine to generate a particular emotion (Watson and Spence, 2007). The following paragraphs further discuss each mediating variable of the study in detail.
Impacts of positive change in brand attitude and brand popularity
The results revealed that players’ overall brand attitude scores were significantly higher with the IVS condition, which led to higher game enjoyment (H2). This finding is in line with the outcomes of other related studies in the literature (e.g. Hwang et al., 2017). Furthermore, the present paper discovered the unexamined brand attitude change by testing the in-game brands that were part of the primary gaming task. According to Petty (2018), the process of attitude change occurs either on a central route via a purposeful evaluation or on a peripheral route, where the attitude change is not from inference but rather through mere association. Unlike conventional in-game brand placement that probably counts on the latter type of attitude change, IVS allows players to intentionally appraise the stimuli by perceiving the transaction that follows a purchase as being fair (Smith and Ellsworth, 1985), which causes the building of emotion, in this case, game enjoyment. Since IVS within the EVW was not a stationary brand message but a highly involved game activity, attitude change took place on a central route via cognitive appraisal (Watson and Spence, 2007). Notably, both less and more popular brand categories benefited by being part of shopping in the EVW. Less popular brands, in particular, exhibited a substantial increase in their attitude scores, while more popular brands witnessed a further enhancement in their already positive brand perception (H6). These findings evidenced that the effectiveness of IVS was substantial.
However, the current study uncovered intriguing inconsistencies within the findings while self-avatar gender congruity did not significantly influence positive change in brand attitude in the moderated mediation analysis (H5a), a notable main effect of self-avatar gender congruity on positive change in brand attitude emerged. One plausible explanation for these findings lies in the distinctive characteristics of the EVW context, which may foster a gender-neutral environment during the overall shopping experience. Within the immersive virtual world of esports, the significance of self-avatar gender congruity might be overshadowed by the broader context of the gaming environment. However, when it comes to specific decision-making processes, such as purchasing virtual goods within the game, players may inherently lean toward selecting a character that aligns with their gender identity. This observation resonates with prior research indicating that avatars resembling the players are more likely to enhance game enjoyment (Trepte and Reinecke, 2010). This phenomenon can be attributed to individuals’ inclination to preserve their self-image by minimizing the incongruity between their self-perception and the avatar’s representation (Latoschik et al., 2017).
Furthermore, this study highlights the enduring significance of brand perception in shaping players’ shopping behaviors within the esports environment. Despite the potential confounding effect of self-avatar gender congruity, the mediation effect of positive change in brand attitude remained evident across both gender-incongruent and congruent groups. This suggests that players’ cognitive appraisal of IVS, as reflected in the alteration of brand attitude, transcends the influence of self-avatar gender congruity. It underscores the enduring impact of brand perception as a key determinant of players’ shopping decisions, emphasizing the pivotal role of brand management within the dynamic landscape of esports.
Impacts of perceived realism
Perceived realism is crucial in explaining the cognitive processes of gameplay and potential media effects. Hence, the extent to which players evaluate games as realistic has been associated with game preferences (Cho et al., 2014). Consistent with the framework of cognitive appraisal, the results demonstrated that perceived realism was a significant mediator for the relationship between IVS within the EVW and game enjoyment (H3). In other words, a more realistic game is experienced as more authentic, resulting in a higher appraisal of certainty, which in turn leads to enhanced enjoyment (Tiedens and Linton, 2001).
However, perceived realism should not be confused with the real-world truth of the situation. Previous literature demonstrated that the truth of information has little effect on the extent to which individuals become engaged in it or enjoy it (Prentice et al., 1997). Similarly, female players could still perceive the gameplay as realistic with the given IVS stimuli even when playing with a male character. Nevertheless, the significant indirect impact of perceived realism diminished once accounting for self-avatar gender congruity (H5b): the mediating effect of perceived realism was significant in the gender-congruent group only. The reason might be that, as Bruner (1986) asserted, individuals treat given stories differently and hold different standards for evaluating truth. For instance, in video games that portray real life, players may judge self-avatar congruity as a critical element for the gameplay to be realistic or “truthful.” On the other hand, in a nonfictional game genre, players may feel less sensitive to gender incongruity, which is still acceptable for a realistic gaming experience. In sum, the current study empirically tested the established theoretical link between perceived realism and enjoyment (Caroux et al., 2015). Our results suggest that players may selectively evaluate the perceived realism of the gameplay via self-avatar gender congruity depending on the circumstances.
Impacts of character attachment
Our findings present that IVS within the EVW effectively creates a psychological tie between players and characters, leading to enhanced game enjoyment (H4). CAT explains this result: players’ shopping behavior granted more control over the appearance of their characters as they purchased in-game items, which led to a higher level of attachment to the avatars. Subsequently, this appraisal of the agency resulted in a positively evaluated shopping behavior in the EVW (Neustaedter and Fedorovskaya, 2009), eventually inducing a positive emotion. Intriguingly, customized game characters are not necessarily duplicated or similar images of the self. Instead, game characters are sometimes perceived as role models that empower players to try out different personalities or offset deficits in real life. Past research suggests that players either prefer self-avatar congruence and create avatars that are analogous to themselves in terms of gender identity and appearance or create dissimilar characters that mirror their ideal self and can assist in the mastery of the game (e.g. Trepte et al., 2009). Accordingly, players’ preferences in avatar creation are twofold: similar images seem to enhance enjoyment experiences, and dissimilar images can serve self-motivated goals such as identity-play and improved self-presentation. In turn, it can be assumed that both factors might foster media enjoyment.
Additionally, the current study examined the role of self-avatar gender congruity as a moderator (H5c) for the mediating effects of character attachment, revealing intriguing insights into how female players perceive and engage with their male characters. Contrary to expectations, self-avatar gender congruity did not emerge as a robust moderator in the findings, suggesting that female players can still develop a strong attachment to characters of a different gender. While prior research suggests that individuals have better attitudes toward their avatars in the virtual world in situations of self-avatar congruity (e.g. Suh et al., 2011), the present study discovered that players may intentionally create characters that diverge from their own gender as a form of identity exploration (Hussain and Griffiths, 2008) or as a means of enhancing their self-image within the game world (Kafai et al., 2010), potentially facilitating wishful identification with game characters. The gender disparity prevalent in video gaming has long been a subject of criticism due to the underrepresentation of female protagonists, which could marginalize female players and impede their ability to forge psychological connections with game characters (Schwind et al., 2017). However, the findings of this study offer a more optimistic outlook, suggesting that IVS within the EVW presents a promising avenue for fostering inclusivity and appealing to a diverse audience, irrespective of gender.
Practical implications
Consistent with previous findings, the results demonstrate that brand placement in an esports game product positively affects brand attitude, regardless of the level of brand popularity. Consequently, for well-established brands, the utilization of IVS emerges as a viable strategy to sustain their position as market leaders, while comparatively less popular brands can leverage in-game brand placements to enhance their competitiveness and align themselves with market frontrunners. It is noteworthy that IVS stands as a cost-effective approach, demanding only a fraction of the significant financial outlays associated with multimillion-dollar investments in traditional advertising or sponsorship activation. Furthermore, IVS within the EVW creates numerous opportunities for an additional revenue source. For instance, players need to spend real money to purchase in-game shoes, which infers that physical goods brands may expand their product line from tangible commodities to digital content service so that in-game brand placement is no longer superficial advertainment; rather, the intangible service can be another line of prosperous revenue stream.
Practitioners may move one step further by allowing players to create content. For example, NBA2k22 now includes a “sneaker designing” feature through which players build customized shoes and post their designs on social media. The current stage of content creation with in-game brands is, however, just sharing the gaming experience, not providing developers and business organizations with additional profit. The EVW could become a platform where players create, share and eventually buy or sell their digital intellectual property (e.g. in-game clothing). This could be a seamless way for organizations in the esports industry to step into a new service market, such as the non-fungible token (NFT), without disturbing the consumer experience.
Regarding the perceived realism of the gameplay, our results afford guidelines to software developers for better designing EVW content via IVS. The major challenge esports game developers are now facing is how to construct more gratifying EVW experiences. Our results demonstrate that just placing brand information on what they can see in real life is not compelling enough, which signifies that the EVW should go beyond simulating the image of the real world. To craft a genuine EVW experience, an esports product must enable the users to apply lifelike experiences such as IVS on top of the technological attraction with quality graphics and sound. Additionally, game developers must allow players to be part of the gameplay with communication and interaction to develop a sense of ownership of the experience (e.g. avatar creation and decoration). Such an engagement in the gameplay would empower players to feel they are creating a new life in the EVW transcending the physical verse resulting in a holistic user experience.
Limitations and future research
This study includes several limitations to address. Firstly, while moderated mediation models offer valuable insights into the relationship between a stimulus and a response, substantive researchers face the ongoing challenge of determining suitable sample sizes for accurately estimating moderated mediation effects. Recent findings by Sim et al. (2022) suggest that employing bootstrapping methods with a sample size of at least 110 can effectively detect effects in a complex path model. Nevertheless, larger sample sizes are advisable to minimize the risk of overlooking significant effects. Thus, researchers are encouraged to consider replicating the study with more extensive samples to ensure comprehensive analysis and robust findings. Secondly, participants spent roughly 30 min playing the esports game, which might not reveal the longitudinal impacts of IVS within the EVW. The temporal component of players’ playing experience could be an element that reinforces or dilutes the positive outcomes of this study, especially for brand attitude and character attachment. Therefore, it is necessary to investigate further how consumers’ IVS behavior affects their EVW experience over time. Thirdly, we were unable to thoroughly examine the self-avatar gender congruity effect, specifically the relationship between female players and female characters. With the given experimental setup, our investigation was limited to only a male character. Considering the significance of the self-avatar gender congruity impacts, it would be worthwhile to test whether the female–female relationship would result in consistent outcomes. Besides, another congruity factor, such as ethnicity beyond the gender effect, can also be examined with a more advanced experimental setup. Fourthly, even though the esports game was appropriate for testing the EVW context, the specific game genre (i.e. sports simulation) was not representative of the variety in the EVW. Accordingly, the generalizability of the findings might be limited to the gaming context of the real-life simulation type. Our results show that perceived realism was influential, but in other genres, such as science fiction, perceived realism might not be the critical determinant for gamers’ enjoyment. Future research should replicate the current study across multiple genres, particularly the surreal types. Lastly, the present study provides a milestone for the incipient stage of shopping behavior in the EVW. However, we measured participants’ game enjoyment via a single item. Future studies should comprehensively measure game enjoyment to fully capture the psychometric property of the construct and test which variable stimulates what aspect of game enjoyment.
Additionally, in considering the future direction of research on IVS within the EVW, an intriguing alternative explanation for the observed increase in enjoyment among participants also emerges. This notion suggests that the act of spending money itself may significantly contribute to heightened enjoyment (Hamari et al., 2017). While the current study focused on investigating three primary drivers of enjoyment – brand attitude, perceived realism and character attachment – it is imperative to explore the role of the transaction itself. Previous research in consumer psychology has highlighted the psychological mechanisms underlying the pleasure associated with spending (e.g. Norton et al., 2012). Therefore, delving deeper into these mechanisms could provide valuable insights into consumer behavior within the EVW context and its implications for gaming experiences. This will contribute to a deeper understanding of the complex interplay between spending, enjoyment and consumer behavior within the EVW landscape, thus informing future studies in this burgeoning field. Moreover, it would also contribute to the understanding of CAT as this line of investigation would allow scholars to explore the impact of appraisals, such as fairness, certainty and agency, that would play a role in eliciting emotions during such transactional activities.
Figures
Categorization of online shopping research
Context | Interactive shopping | Shopping behavior as a predictor | Shopping behavior as a consequence | Sources |
---|---|---|---|---|
In-app purchases in mobile game | No | No | Yes | Ghazali et al. (2018), Hamari et al. (2019), Gupta et al. (2024) |
Addictive shopping behaviors | No | No | Yes | Erhel et al. (2024), Starcevic et al. (2023), Tian et al. (2021), Kircaburun and Griffiths (2018) |
E-commerce | No | Yes | No | Mofokeng (2021), Gulfraz et al. (2022), Shah et al. (2024) |
Immersive shopping | Yes | No | Yes | Hooker et al. (2019), Kang et al. (2020), Lombart et al. (2020), Saleem et al. (2024) |
Experimental conditions
Group | Interactive virtual shopping | Self-avatar gender congruity | N |
---|---|---|---|
1 | No-shopping | Incongruent | 32 |
2 | No-shopping | Congruent | 33 |
3 | Shopping | Incongruent | 32 |
4 | Shopping | Congruent | 33 |
Source(s): Authors’ own work
Measures and reliability scores of multi-item scales
Construct | Items | Cronbach’s α |
---|---|---|
Brand attitude (pre) Mitchell and Olson (2000) | How appealing is the brand to you? | α = 0.80 (Adidas) α = 0.90 (Converse) |
How likable is the brand to you? | α = 0.86 (Jordan) α = 0.93 (New Balance) | |
How valuable is the brand to you? | α = 0.87 (Nike) α = 0.89 (Puma) | |
How pleasant is the brand to you? | α = 0.93 (Under Armour) | |
Brand attitude (post) Mitchell and Olson (2000) | How appealing is the brand to you? | α = 0.80 (Adidas) α = 0.91 (Converse) |
How likable is the brand to you? | α = 0.90 (Jordan) α = 0.94 (New Balance) | |
How valuable is the brand to you? | α = 0.87 (Nike) α = 0.92 (Puma) | |
How pleasant is the brand to you? | α = 0.94 (Under Armour) | |
Perceived realism Cho et al. (2014) | This game showed something that could possibly happen in real life | α = 0.83 |
This game portrayed possible real-life situations | ||
The story in this game could actually happen in real life | ||
Never in real life would what was shown in this game happen | ||
Real people would not do the things shown in this game | ||
Character attachment Lewis et al. (2008) | While playing this game, I sometimes forgot my own feelings and took on those of my character | α = 0.86 |
While playing this game, I enjoyed pretending my character was a real person | ||
While playing this game, I considered my character a friend of mine | ||
While playing this game, I enjoyed pretending I was my character | ||
While playing this game, I could see myself being attracted to my character | ||
Game enjoyment Klimmt et al. (2009) | I found playing this game to be enjoyable | __ |
Basketball involvement Kowert et al. (2014) | I consider myself to be a fan of basketball | __ |
Descriptive statistics and correlations
Variable | M | SD | 1 | 2 | 3 | 4 | 5 | 6 |
---|---|---|---|---|---|---|---|---|
1. Interactive virtual shopping | – | – | – | |||||
2. Positive change in brand attitude | 0.54 | 1.41 | 0.37*** | – | ||||
3. Perceived realism | 5.45 | 0.82 | 0.36*** | 0.13 | – | |||
4. Character attachment | 4.33 | 1.21 | 0.25** | 0.13 | 0.40*** | – | ||
5. Self-avatar gender congruity | – | – | 0.01 | −0.12 | 0.24** | 0.49** | – | |
6. Game enjoyment | 5.78 | 0.98 | 0.33*** | 0.28** | 0.44*** | 0.45*** | 0.23** | – |
Note(s): Results of a point-biserial correlation analysis. Interactive virtual shopping and self-avatar gender congruity were dichotomous variables. **p < 0.01. ***p < 0.001
Source(s): Authors’ own work
Moderated mediation analysis
Path | Estimate | SE | p | 95% CI |
---|---|---|---|---|
Direct effects | ||||
IVS → PCBA | 0.98 | 0.32 | 0.003 | 0.34, 1.62 |
IVS → PR | 1.10 | 0.18 | <0.001 | 0.52, 1.22 |
IVS → CA | 0.80 | 0.25 | 0.002 | 0.31, 1.30 |
IVS → GE | 0.17 | 0.16 | 0.30 | −0.16, 0.50 |
PCBA → GE | 0.13 | 0.06 | 0.02 | 0.02, 0.24 |
PR → GE | 0.30 | 0.10 | 0.004 | 0.10, 0.50 |
CA → GE | 0.25 | 0.07 | <0.001 | 0.12, 0.38 |
SAGC → PCBA | 0.33 | 0.33 | 0.31 | −0.31, 0.98 |
SAGC → PR | 0.87 | 0.18 | <0.001 | 0.52, 1.22 |
SAGC → CA | 1.44 | 0.25 | <0.001 | 0.94, 1.94 |
SAGC × IVS → PCBA | 0.11 | 0.46 | 0.80 | −0.79, 1.01 |
SAGC × IVS → PR | 0.99 | 0.25 | <0.001 | 0.50, 1.48 |
SAGC × IVS → CA | 0.40 | 0.35 | 0.27 | −0.30, 1.10 |
Path | Estimate | SE | 95% CI | |
---|---|---|---|---|
Indirect effects | ||||
IVS → PCBA → GE | Gender incongruent | 0.14 | 0.08 | 0.01, 0.33 |
Gender congruent | 0.13 | 0.08 | 0.01, 0.31 | |
IVS → PR → GE | Gender incongruent | 0.03 | 0.06 | −0.11, 0.15 |
Gender congruent | 0.33 | 0.15 | 0.07, 0.66 | |
IVS → CA → GE | Gender incongruent | 0.10 | 0.06 | −0.01, 0.23 |
Gender congruent | 0.20 | 0.08 | 0.06, 0.37 |
Note(s): CA = character attachment, GE = game enjoyment, IVS = interactive virtual shopping, PCBA = positive change in brand attitude, PR = perceived realism, SAGC = self-avatar gender congruity, SE = standard error
Source(s): Authors’ own work
ANCOVA for in-game brand attitudes
ANCOVA results | |||
---|---|---|---|
Effect | F | p | η2 |
Main effects | |||
Brand popularity (aless popular vs. bmore popular) | 9.45 | 0.003 | 0.14 |
Pre-post gameplay | 9.84 | 0.003 | 0.14 |
Self-avatar gender congruity (gender incongruent vs. gender congruent) | 4.68 | 0.03 | 0.07 |
Brand popularity × Pre-post gameplay | 8.64 | 0.005 | 0.13 |
Brand popularity × Self-avatar gender congruity | 0.07 | 0.40 | 0.01 |
Pre-post gameplay × Self-avatar gender congruity | 0.50 | 0.49 | 0.01 |
Covariate | |||
Basketball involvement | 0.17 | 0.68 | 0.01 |
Estimated marginal means | ||||
---|---|---|---|---|
Condition | M | SE | ||
aLess popular brands | Gender incongruent | Pre | 4.60 | 0.24 |
Post | 5.89 | 0.18 | ||
Gender congruent | Pre | 5.36 | 0.35 | |
Post | 6.35 | 0.24 | ||
bMore popular brands | Gender incongruent | Pre | 5.80 | 0.26 |
Post | 6.27 | 0.18 | ||
Gender congruent | Pre | 6.01 | 0.20 | |
Post | 6.51 | 0.14 |
Note(s): ANCOVA = analysis of covariance. aLess popular brands = Converse, New Balance, Puma and Under Armour. bMore popular brands = Adidas, Jordan and Nike
Source(s): Authors’ own work
References
Bruner, J.S. (1986), Actual Minds, Possible Worlds, Harvard University Press, Cambridge, MA.
Cappasity (2024), “How retailers use immersive shopping experiences in stores”, available at: https://medium.com/cappasity-blog/how-retailers-use-immersive-shopping-experiences-in-stores-6e0f3e4bb156 (accessed 10 October 2024).
Caroux, L. and Pujol, M. (2023), “Player enjoyment in video games: a systematic review and meta-analysis of the effects of game design choices”, International Journal of Human-Computer Interaction, Vol. 40 No. 16, pp. 4227-4238, doi: 10.1080/10447318.2023.2210880.
Caroux, L., Isbister, K., Le Bigot, L. and Vibert, N. (2015), “Player–video game interaction: a systematic review of current concepts”, Computers in Human Behavior, Vol. 48, pp. 366-381, doi: 10.1016/j.chb.2015.01.066.
Cho, H., Shen, L. and Wilson, K. (2014), “Perceived realism: dimensions and roles in narrative persuasion”, Communication Research, Vol. 41 No. 6, pp. 828-851, doi: 10.1177/0093650212450585.
Diamantopoulos, A., Sarstedt, M., Fuchs, C., Wilczynski, P. and Kaiser, S. (2012), “Guidelines for choosing between multi-item and single-item scales for construct measurement: a predictive validity perspective”, Journal of the Academy of Marketing Science, Vol. 40 No. 3, pp. 434-449, doi: 10.1007/s11747-011-0300-3.
Erhel, S., Drouard, J., Jacob, F., Lumeau, M., Suire, R. and Gonthier, C. (2024), “Predictors of problematic internet use in the everyday internet activities of a French representative sample: the importance of psychological traits”, Computers in Human Behavior, Vol. 153, 108099, doi: 10.1016/j.chb.2023.108099.
Fu, F.L., Su, R.C. and Yu, S.C. (2009), “EGameFlow: a scale to measure learners' enjoyment of e-learning games”, Computers and Education, Vol. 52 No. 1, pp. 101-112, doi: 10.1016/j.compedu.2008.07.004.
Funk, D.C. (2017), “Introducing a Sport Experience Design (SX) framework for sport consumer behaviour research”, Sport Management Review, Vol. 20 No. 2, pp. 145-158, doi: 10.1016/j.smr.2016.11.006.
Ghazali, E.M., Mutum, D.S., Chong, J.H. and Nguyen, B. (2018), “Do consumers want mobile commerce? A closer look at M-shopping and technology adoption in Malaysia”, Asia Pacific Journal of Marketing and Logistics, Vol. 30 No. 4, pp. 1064-1086, doi: 10.1108/apjml-05-2017-0093.
Gulfraz, M.B., Sufyan, M., Mustak, M., Salminen, J. and Srivastava, D.K. (2022), “Understanding the impact of online customers' shopping experience on online impulsive buying: a study on two leading E-commerce platforms”, Journal of Retailing and Consumer Services, Vol. 68, 103000, doi: 10.1016/j.jretconser.2022.103000.
Gupta, K., Su, Y., Kunkel, T. and Funk, D.C. (2022), “The super engagers of freemium gamified services: using multimethod approach to examine why highly interactive consumers become paying consumers”, Internet Research, Vol. 32 No. 6, pp. 1891-1909, doi: 10.1108/intr-07-2021-0434.
Gupta, K., Su, Y., Kunkel, T. and Funk, D.C. (2024), “Paying while playing: examining the influence of interaction with gamified elements in fantasy sports on in-app spending”, European Sport Management Quarterly, pp. 1-22, doi: 10.1080/16184742.2024.2301970.
Hall, A. (2009), “Perceptions of the authenticity of reality programs and their relationships to audience involvement, enjoyment and perceived learning”, Journal of Broadcasting and Electronic Media, Vol. 53 No. 4, pp. 515-531, doi: 10.1080/08838150903310468.
Hamari, J., Alha, K., Järvelä, S., Kivikangas, J.M., Koivisto, J. and Paavilainen, J. (2017), “Why do players buy in-game content? An empirical study on concrete purchase motivations”, Computers in Human Behavior, Vol. 68, pp. 538-546, doi: 10.1016/j.chb.2016.11.045.
Hamari, J., Malik, A., Koski, J. and Johri, A. (2019), “Uses and gratifications of Pokémon Go: why do people play mobile location-based augmented reality games?”, International Journal of Human-Computer Interaction, Vol. 35 No. 9, pp. 804-819, doi: 10.1080/10447318.2018.1497115.
Häubl, G. and Trifts, V. (2000), “Consumer decision making in online shopping environments: the effects of interactive decision aids”, Marketing Science, Vol. 19 No. 1, pp. 4-21, doi: 10.1287/mksc.19.1.4.15178.
Hayes, A.F. (2017), Introduction to Mediation, Moderation and Conditional Process Analysis: A Regression-Based Approach, Guilford Press, New York, NY.
Hooker, R., Wasko, M., Paradice, D., Teigland, R. and Hofacker, C. (2019), “Beyond gaming: linking flow, brand attitudes, and purchase intent in realistic and emergent three-dimensional virtual environments”, Information Technology and People, Vol. 32 No. 6, pp. 1397-1422, doi: 10.1108/itp-11-2017-0407.
Hussain, Z. and Griffiths, M.D. (2008), “Gender swapping and socializing in cyberspace: an exploratory study”, CyberPsychology and Behavior, Vol. 11 No. 1, pp. 47-53, doi: 10.1089/cpb.2007.0020.
Hwang, Y., Ballouli, K., So, K. and Heere, B. (2017), “Effects of brand congruity and game difficulty on gamers' response to advertising in sport video games”, Journal of Sport Management, Vol. 31 No. 5, pp. 480-496, doi: 10.1123/jsm.2017-0022.
Hwang, Y., Watanabe, N.M. and Nagel, M. (2023), “Effects of in-game brand congruity on esports consumers' implicit and explicit memory”, International Journal of Sports Marketing and Sponsorship, Vol. 24 No. 1, pp. 108-128, doi: 10.1108/ijsms-03-2022-0085.
Jobanputra, K. (2023), “E-Commerce evolution: predictions and trends shaping the future”, available at: https://www.forbes.com/sites/forbesbusinesscouncil/2023/12/29/e-commerce-evolution-predictions-and-trends-shaping-the-future/?sh=523d5db07c6d (accessed 20 October 2024).
Kafai, Y.B., Fields, D.A. and Cook, M.S. (2010), “Your second selves: player-designed avatars”, Games and Culture, Vol. 5 No. 1, pp. 23-42, doi: 10.1177/1555412009351260.
Kang, H.J., Shin, J.H. and Ponto, K. (2020), “How 3D virtual reality stores can shape consumer purchase decisions: the roles of informativeness and playfulness”, Journal of Interactive Marketing, Vol. 49 No. 1, pp. 70-85, doi: 10.1016/j.intmar.2019.07.002.
Kim, C.K. and Chung, J.Y. (1997), “Brand popularity, country image and market share: an empirical study”, Journal of International Business Studies, Vol. 28 No. 2, pp. 361-386, doi: 10.1057/palgrave.jibs.8490105.
Kircaburun, K. and Griffiths, M.D. (2018), “The dark side of internet: preliminary evidence for the associations of dark personality traits with specific online activities and problematic internet use”, Journal of Behavioral Addictions, Vol. 7 No. 4, pp. 993-1003, doi: 10.1556/2006.7.2018.109.
Klimmt, C., Rizzo, A., Vorderer, P., Koch, J. and Fischer, T. (2009), “Experimental evidence for suspense as determinant of video game enjoyment”, CyberPsychology and Behaviour, Vol. 12 No. 1, pp. 29-31, doi: 10.1089/cpb.2008.0060.
Kowert, R., Domahidi, E., Festl, R. and Quandt, T. (2014), “Social gaming, lonely life? The impact of digital game play on adolescents' social circles”, Computers in Human Behavior, Vol. 36, pp. 385-390, doi: 10.1016/j.chb.2014.04.003.
Latoschik, M.E., Roth, D., Gall, D., Achenbach, J., Waltemate, T. and Botsch, M. (2017), “The effect of avatar realism in immersive social virtual realities”, Proceedings of the 23rd ACM Symposium on Virtual Reality Software and Technology, pp. 1-10, doi: 10.1145/3139131.3139156.
Lazarus, R.S. (1991), Emotion and Adaptation, Oxford University Press, New York, NY.
Lazarus, R.S. and Folkman, S. (1984), Stress, Appraisal, and Coping, Springer publishing company, New York, NY.
Lee, M. and Williams, R. (2019), “The impact of in-game shopping on media communication: a cross-disciplinary examination”, Media Studies Journal, Vol. 15 No. 1, pp. 78-94.
Lewis, M.L., Weber, R. and Bowman, N.D. (2008), “They may be pixels, but they’re MY Pixels: developing a metric of character attachment in role-playing video games”, CyberPsychology and Behaviour, Vol. 11 No. 4, pp. 515-518, doi: 10.1089/cpb.2007.0137.
Lombart, C., Millan, E., Normand, J.M., Verhulst, A., Labbé-Pinlon, B. and Moreau, G. (2020), “Effects of physical, non-immersive virtual, and immersive virtual store environments on consumers' perceptions and purchase behavior”, Computers in Human Behavior, Vol. 110, 106374, doi: 10.1016/j.chb.2020.106374.
Mitchell, A.A. and Olson, J.C. (2000), “Are product attribute beliefs the only mediator of advertising effects on brand attitude?”, Journal of Marketing Research, Vol. 18 No. 3, pp. 318-332, doi: 10.1353/asr.2000.0010.
Mofokeng, T.E. (2021), “The impact of online shopping attributes on customer satisfaction and loyalty: moderating effects of e-commerce experience”, Cogent Business and Management, Vol. 8 No. 1, 1968206, doi: 10.1080/23311975.2021.1968206.
Moon, J.H., Kim, E., Choi, S.M. and Sung, Y. (2013), “Keep the social in social media: the role of social interaction in avatar-based virtual shopping”, Journal of Interactive Advertising, Vol. 13 No. 1, pp. 14-26, doi: 10.1080/15252019.2013.768051.
Neustaedter, C. and Fedorovskaya, E.A. (2009), “Presenting identity in a virtual world through avatar appearances”, in Gooch, A. and Tory, M. (Eds) Graphics Interface 2009, Canadian Information Processing Society, Toronto, ON, Canada, pp. 183-190.
Norton, M.I., Mochon, D. and Ariely, D. (2012), “The IKEA effect: when labor leads to love”, Journal of Consumer Psychology, Vol. 22 No. 3, pp. 453-460, doi: 10.1016/j.jcps.2011.08.002.
Oh, S., Kim, W.B. and Choo, H.J. (2023), “The effect of avatar self-integration on consumers' behavioral intention in the metaverse”, International Journal of Human-Computer Interaction, Vol. 40 No. 23, pp. 1-14, doi: 10.1080/10447318.2023.2279416.
Petty, R.E. (2018), Attitudes and Persuasion: Classic and Contemporary Approaches, Routledge, New York, NY.
Petty, R.E. and Cacioppo, J.T. (1986), “The elaboration likelihood model of persuasion”, Advances in Experimental Social Psychology, Vol. 19, pp. 123-205, doi: 10.1016/s0065-2601(08)60214-2.
Peukert, C., Pfeiffer, J., Meißner, M., Pfeiffer, T. and Weinhardt, C. (2019), “Shopping in virtual reality stores: the influence of immersion on system adoption”, Journal of Management Information Systems, Vol. 36 No. 3, pp. 755-788, doi: 10.1080/07421222.2019.1628889.
Pitardi, V., Bartikowski, B., Osburg, V.S. and Yoganathan, V. (2023), “Effects of gender congruity in human-robot service interactions: the moderating role of masculinity”, International Journal of Information Management, Vol. 70, 102489, doi: 10.1016/j.ijinfomgt.2022.102489.
Prentice, D.A., Gerrig, R.J. and Bailis, D.S. (1997), “What readers bring to the processing of fictional texts”, Psychonomic Bulletin and Review, Vol. 5 No. 3, pp. 416-420, doi: 10.3758/bf03210803.
Rai, P. (2024), “NBA 2K25 reveals new features coming to the city”, available at: https://gamerant.com/nba-2k25-the-city-new-features/ (accessed 10 October 2024).
Rogers, R. (2017), “The motivational pull of video game feedback, rules, and social interaction: another self-determination theory approach”, Computers in Human Behavior, Vol. 73, pp. 446-450, doi: 10.1016/j.chb.2017.03.048.
Saleem, T., Ishaq, M.I., Raza, A. and Junaid, M. (2024), “Exploring the effect of telepresence and escapism on consumer post-purchase intention in an immersive virtual reality environment”, Journal of Retailing and Consumer Services, Vol. 81, 104014.
Schwind, V., Knierim, P., Tasci, C., Franczak, P., Haas, N. and Henze, N. (2017), “‘These are not my hands!’ Effect of gender on the perception of avatar hands in virtual reality”, Proceedings of the 2017 CHI conference on human factors in computing systems, pp. 1577-1582, doi: 10.1145/3025453.3025602.
Shafer, D.M., Carbonara, C.P. and Popova, L. (2014), “Controller required? The impact of natural mapping on interactivity, realism, presence and enjoyment in motion-based video games”, Presence: Teleoperators and Virtual Environments, Vol. 23 No. 3, pp. 267-286, doi: 10.1162/pres_a_00193.
Shah, H., Carrel, A.L. and Le, H.T. (2024), “Impacts of teleworking and online shopping on travel: a tour-based analysis”, Transportation, Vol. 51 No. 1, pp. 99-127, doi: 10.1007/s11116-022-10321-9.
Sim, M., Kim, S.Y. and Suh, Y. (2022), “Sample size requirements for simple and complex mediation models”, Educational and Psychological Measurement, Vol. 82 No. 1, pp. 76-106, doi: 10.1177/00131644211003261.
Sirgy, M.J. (1982), “Self-concept in consumer behavior: a critical review”, Journal of Consumer Research, Vol. 9 No. 3, pp. 287-300, doi: 10.1086/208924.
Smith, C.A. and Ellsworth, P.C. (1985), “Patterns of cognitive appraisal in emotion”, Journal of Personality and Social Psychology, Vol. 48 No. 4, pp. 813-838, doi: 10.1037//0022-3514.48.4.813.
Smith, A., Johnson, B. and Martinez, C. (2021), “In-game shopping and user experience: an interdisciplinary analysis”, Journal of Interactive Entertainment, Vol. 8 No. 2, pp. 45-62.
Starcevic, V., Eslick, G.D., Viswasam, K., Billieux, J., Gainsbury, S.M., King, D.L. and Berle, D. (2023), “Problematic online behaviors and psychopathology in Australia”, Psychiatry Research, Vol. 327, 115405, doi: 10.1016/j.psychres.2023.115405.
Steinkuehler, C. and Duncan, S. (2008), “Scientific habits of mind in virtual worlds”, Journal of Science Education and Technology, Vol. 17 No. 6, pp. 530-543, doi: 10.1007/s10956-008-9120-8.
Suh, K.S., Kim, H. and Suh, E.K. (2011), “What if your avatar looks like you? Dual-congruity perspectives for avatar use”, MIS Quarterly, Vol. 35 No. 3, pp. 711-729, doi: 10.2307/23042805.
Tabachnick, B.G. and Fidell, L.S. (2013), Using Multivariate Statistics, Pearson, Boston, MA.
Tate, R.F. (1954), “Correlation between a discrete and a continuous variable. Point-biserial correlation”, The Annals of Mathematical Statistics, Vol. 25 No. 3, pp. 603-607, doi: 10.1214/aoms/1177728730.
Tavinor, G. (2019), “Towards an analysis of virtual realism”, Proceedings of DiGRA Conference: Game, Play and the Emerging Ludo-Mix.
Tian, Y., Zuo, T., Sun, Q., Sun, L., Cao, S. and Qin, N. (2021), “The association between generalized and specific problematic internet use and its gender differences across different educational levels”, Frontiers in Psychology, Vol. 12, 634581, doi: 10.3389/fpsyg.2021.634581.
Tiedens, L.Z. and Linton, S. (2001), “Judgment under emotional certainty and uncertainty: the effects of specific emotions on information processing”, Journal of Personality and Social Psychology, Vol. 81 No. 6, pp. 973-988, doi: 10.1037/0022-3514.81.6.973.
Trepte, S. and Reinecke, L. (2010), “Avatar creation and video game enjoyment”, Journal of Media Psychology, Vol. 22 No. 4, pp. 171-184, doi: 10.1027/1864-1105/a000022.
Trepte, S., Reinecke, L. and Behr, K.M. (2009), “Creating virtual alter egos or superheroines? Gamers' strategies of avatar creation in terms of gender and sex”, International Journal of Gaming and Computer-Mediated Simulations, Vol. 1 No. 2, pp. 52-76, doi: 10.4018/jgcms.2009040104.
Watson, L. and Spence, M.T. (2007), “Causes and consequences of emotions on consumer behaviour: a review and integrative cognitive appraisal theory”, European Journal of Marketing, Vol. 41 No. 5, pp. 487-511, doi: 10.1108/03090560710737570.
Zhao, X., Lynch, G.J. and Chen, Q. (2010), “Reconsidering Baron and Kenny: myths and truths about mediation analysis”, Journal of Consumer Research, Vol. 37 No. 2, pp. 197-206, doi: 10.1086/651257.