Understanding how loneliness relates to user behavior in an interactive virtual reality cafe

Abstract

Loneliness is a growing concern among young adults. Though loneliness is distinct from social isolation, studies indicate that causes can stem from diminishment in the parameters of social contact that underlie isolation. In this work, we analyze how pre-existing levels of loneliness for Generation Z adults relate to behavior, choices made, and frustration levels in response to wait times in a visually realistic virtual reality (VR)-based representation of a cafe, an environment involving social interactions and providing services to meet basic needs such as getting food. We conduct an experiment in which participants experience a VR cafe environment where they interact with a non-playable character representing a maître d’ and go through a selection of dining preferences and a selection on a digital menu. We introduce delay times to our virtual interaction as conditions in four randomized trials. We analyze data from a pilot study with 35 participants who completed all four trials of our experiment, providing raw positional data, as well as standardized surveys for measuring loneliness and frustration. Our results show participants self-identifying as female or Asian report ‘Moderate’ or higher categories of loneliness on the Revised UCLA-20 scale. Participants in the ‘Moderate’ or higher category had higher average frustration on the Frustration Discomfort Scale. Participants in the ‘Low’ category for loneliness were more likely to opt for a dine-in experience when there were no delays with the maître d’, while a higher percentage of participants in the ‘Moderate’ or higher category opted to exit when there were delays. In delayed scenarios, a higher percentage of participants in the ‘Moderate’ or higher category chose a non-table option after choosing to dine in. Our results indicate that loneliness may be associated with variations in individual behavior in social scenarios. Our findings enable informing future work on creating environments to assist individuals with overcoming the adverse effects of loneliness and frustration.

1 Introduction

Loneliness is a growing concern among young adults. Generation-Z or Gen-Z, the generation of adults that are in the age group 13–28 as of the time of this article, has been noted to be the loneliest generation (Bowler, 2020; Victor and Yang, 2013; Shovestul et al., 2020; Jarry et al., 2025). Though the relationship between the subjective experience of loneliness and social isolation, i.e., the objective evaluation of diminished social contact is complex (Jones et al., 1990), several studies note an inverse relationship (Bahr and Harvey, 1979; Russell et al., 1980; Jones et al., 1985; Levin and Stokes, 1986; Jones et al., 1990), suggesting that social isolation is a risk factor for loneliness (Twenge et al., 2019; Bu et al., 2020b). Lockdown measures during the COVID-19 pandemic formed one form of external social isolation that generated loneliness and heightened loneliness trajectories among young adults (Bu et al., 2020a; Hu and Gutman, 2021; Jarry et al., 2025). In interviews and self-reports, young adults frequently cite broader parameters of social isolation, e.g., presence of, quantity, and quality of social relationships, childhood exclusion, or “fake” portrayals of realities on social media (Fardghassemi and Joffe, 2022; Harries et al., 2025; Sundqvist and Hemberg, 2021; Hemberg et al., 2022) as causing them to feel lonely, so that young adults today are at risk of long-term loneliness especially when the quality of social ties is weak (Moore et al., 2023). Among young adults, loneliness is associated with negative states such as social anxiety (Lim et al., 2016; Eres et al., 2023; Jones et al., 1981), and higher depressive symptom scores, difficulties with regulating emotion, and higher risk of social isolation (Eres et al., 2023).

Young adults in university settings who report being lonely have been found to be more likely to perform activities such as eating dinner alone (Russell et al., 1980). Such activities are often conducted in a social setting on college campuses. Unfortunately, to the best of our analysis, work on evaluating how young adults who feel lonely or isolated behave in real world social settings is lacking. Work exists on understanding loneliness and its relationship to social media usage as systematically reviewed in O’Day and Heimberg (O’Day and Heimberg, 2021), with some studies showing that loneliness may predict social media usage. However, by its nature the work addresses behavior in online environments, to which young adults may turn if they perceive that the quality of their physical social networks is diminished. In this work we use virtual reality (VR) to provide a first-of-its-kind analysis of the behaviors of young adults reporting various levels of loneliness in a visually realistic VR-based representation of a cafe. We use a VR-based representation of a cafe as the real-world counterpart is an environment where individuals go to have social interactions as well as to have basic needs met, such as getting food. We use VR as it provides a scalable mechanism to study user behavior while tuning for variables such as wait time from the maître d’, seating layout, and table choices. The use of VR to investigate loneliness among young adults has been limited and targeted to how users participate in gameplay or online social platforms (Kenyon et al., 2023; Oh et al., 2023; Kalantari et al., 2023), rather than using it as a tool to understand how loneliness relates to behavior in visually realistic virtual representations of real world social environments such as our VR cafe.

We hypothesize that individuals who show higher loneliness scores on a standardized scale, such as the Revised UCLA-20 (Russell, 1996), are likely to engage in behaviors representative of social withdrawal, which in our case represents opting for take-out over dine-in or exiting if wait times in the social setting are long. Social withdrawal is characterized as the tendency to eschew social contact due to low confidence in ability to establish social relationships, and has been found to be related to loneliness Watson and Nesdale (2012). Informed by prior findings that individuals reporting lonely may have difficulty with emotion regulation Eres et al. (2023), we also hypothesize that they may show higher levels of frustration. We perform our analysis by collecting a dataset of 35 participants interacting in our immersive VR cafe called Flavor and Vine, as shown in Figure 1. Within our cafe, participants can choose between dining in or take out for their meal. Our cafe consists of four non-playable characters (NPCs), with one interactive maître d’ NPC who guides the participant through the dining experience and three customer NPCs already seated in the cafe. Typical cafe sounds, including ambient music, background conversation, and the clacking of utensils and plates are added to enable a sense of realism within the environment. During the interaction, a participant is presented with four randomized scenarios, or treatments, with either no delay or a 3-min delay in engaging with the maître d’ and after making the menu selection.

FIGURE 1

The goal of our study is to acquire subjective experience of loneliness from a user-centric perspective. We collect the subjective loneliness experience being informed by studies that survey loneliness (Blazer, 2020; Kenyon et al., 2023; Oppert et al., 2023; Abplanalp et al., 2024) in order to garner impact of social isolation, which is an objective measure of social contact. We are also interested in measuring participant mental state in terms of frustration. Frustration intolerance has been found to predict existential loneliness (Saricali and Guler, 2022). We collect interaction-oriented frustration measures to understand the relationship between loneliness and frustration when there exist deficiencies in the meeting of specific goals, such as provision of service or arrival of food on time. Prior studies demonstrate an association between social anxiety and loneliness (Jones et al., 1981; Vanhalst et al., 2013; Lim et al., 2016; Eres et al., 2023; Teo et al., 2013), and have collected social anxiety and loneliness measures in VR environments (Kenyon et al., 2023). Though we do not explicitly study social anxiety in this work, we have collected measures of social anxiety to support future investigation of the relationship between social anxiety, loneliness, and user choices. Prior to immersion, we collect loneliness, frustration intolerance, and social anxiety severity, and current anxiety levels using the Revised UCLA-20 (Russell, 1996), Frustration Discomfort Scale (FDS) (Harrington, 2005), Social Phobia Inventory (SPIN) (Connor et al., 2000), and State Trait Anxiety Inventory (STAI) (Speilberger et al., 1983) respectively. During the study, after each treatment, we ask participants to rate sense of isolation, frustration due to the delay, and anxiety. We measure isolation rather than loneliness as a more direct mechanism for the individual to assess social contact. We also ask participants to rate perception of proximity of other patrons, noise level, and the likelihood of making a similar choice in the real world. After completing the four treatments, we measure workload using NASA Task Load Index (Hart and Staveland, 1988), usability using the System Usability Scale (Brooke et al., 1996), virtual reality sickness (Kim et al., 2018), and re-apply the FDS to assess any post study changes in frustration and discomfort.

We observe significant differences in frustration across the four treatments when there is a service delay, and that participants with moderate or higher loneliness show higher frustration. Participants with lower loneliness levels were more likely to opt for dining in. Participants with moderate or higher loneliness levels opt to exit if the maître d’ introduces delay. More participants with moderate or higher loneliness scores opt for no standing table if choosing dining in, than lower loneliness. These baseline findings in our study indicate the potential for the use of VR representations of real world environments to perform extended investigations of loneliness, and provision of interventions that can be tailored to help counter the adverse health effects of loneliness.

2 Related work

2.1 Loneliness and social isolation

Our work focuses on young adults largely from Generation Z (Gen-Z), noted to be the loneliest generation (Bowler, 2020; Fardghassemi and Joffe, 2022). Though loneliness has been variously defined in the literature (Jones et al., 1990), the common identifiable thread among definitions is that loneliness is characterized as a subjective psychological experience of having deficiencies in the number or quality of social ties than desired by the individual (Perlman and Peplau, 1982; Mansfield et al., 2019; Abplanalp et al., 2024), making it distinct from physical isolation (Jones et al., 1990). Loneliness is considered distinct from social isolation, characterized as an umbrella term that encompasses the objective assessment of the quantity (structural component) and quality (functional component) of social relationships (Dickens et al., 2011; Teo et al., 2013), and includes definitions such as social disconnectedness or lack of social ties (Cornwell and Waite, 2009; Chou et al., 2011) and low strength of relationships (Holt-Lunstad et al., 2010). Social isolation may be induced by factors external to the individual who may want to make contact but cannot (Carpenito-Moyet, 2006; Kenyon et al., 2023), as occurred when lockdown measures were in place during the COVID-19 pandemic, However, social isolation may also occur due to internal factors (Teo et al., 2013).

Despite their distinction, the study of the impact of social isolation is often done by surveying individuals on their perceptions of loneliness (Blazer, 2020; Kenyon et al., 2023; Oppert et al., 2023; Abplanalp et al., 2024) in order to evaluate their subjective experience. Findings from studies demonstrate a complex relationship between loneliness and social contact (Jones et al., 1990). Some studies report an inverse relationship between loneliness and the number of casual and intimate friends, presence of romantic or marital partner, length of intimate relationships, recency of change in relationship, frequency of contact, and size of support network (Bahr and Harvey, 1979; Russell et al., 1980; Jones et al., 1985; Levin and Stokes, 1986; Jones et al., 1990). Others fail to find significant relationships (Austin, 1983; Baum, 1982; Berg et al., 1981; Cornwell and Waite, 2009), though most of these address older adults (Baum, 1982; Berg et al., 1981; Cornwell and Waite, 2009), whose influencing factors differ from those of the younger adult population considered in the present work. Among college students, loneliness has been found to be correlated to participants reporting that they more frequently perform activities considered social on their own (Russell et al., 1980; Jones et al., 1990). Such activities include dining (Russell et al., 1980; Jones et al., 1990), the scenario considered in our work. Even when a strong correlation is not found (Cornwell and Waite, 2009), perceived isolation and social disconnectedness have been found to be associated with poorer health (Cornwell and Waite, 2009; Algren et al., 2020), indicating value to collecting loneliness measures to address adverse health impacts. Loneliness has been found to have associations with social anxiety among young adults and adolescents (Jones et al., 1981; Vanhalst et al., 2013; Lim et al., 2016; Eres et al., 2023), as well as showing higher depressive symptom scores, and more difficulties with emotion regulation (Eres et al., 2023). Reports from young adults, garnered either through interviews (Sundqvist and Hemberg, 2021; Hemberg et al., 2022; Fardghassemi and Joffe, 2022) or through evaluations of user-created videos with loneliness themes (Harries et al., 2025) demonstrate that prevalent causes of loneliness include being disconnected, being unable to express themselves even when surrounded by people or be understood, and lack of face-to-face communication (Fardghassemi and Joffe, 2022), lack of quality relationships or non-existence of relationships (Harries et al., 2025), isolation (Sundqvist and Hemberg, 2021), and exclusion (Hemberg et al., 2022) showing that several user-expressed reasons for loneliness stem from diminishment in social contact. Similar to prior research (Blazer, 2020; Kenyon et al., 2023; Oppert et al., 2023; Abplanalp et al., 2024), our study uses loneliness measures to evaluate self-perceptions of social relationships.

2.2 Study of loneliness and social isolation in online environments and virtual reality

In the context of analyzing young adult behavior in online environments, most work lies in the area of online video game play. As discussed in the systematic review of Luo et al. (2022), relationships are complex and findings are mixed. Zhou and Leung (2012) find for individuals aged 18–22, loneliness is a predictor of addiction to gaming social networking sites. Kaya et al. (2025) find among university students with 77.6% of participants aged 18–23 that loneliness and social anxiety are associated with digital game addiction. Though Kowert et al. (2015) do not find an association between online video game play and loneliness among individuals aged 14–18 or 40+, they do find a significant association among individuals aged 19–39 (though subsequent path analysis did not reveal predictive relationships). Martončik and Lokša (2016) find using the UCLA loneliness scale and the SPIN scale that among World of Warcraft players aged 13–50 (mean 21), loneliness and social anxiety are found to be lower in online worlds than in real worlds. Mandryk et al. (2020) show using data from individuals aged 18–66 (mean 30.7, SD 7.7) playing World of Warcraft, that while harmonious passion (i.e., a balanced relationship with the beloved activity) decreases loneliness measured on the revised UCLA scale (Russell et al., 1980), obsessive passion (i.e., inflexible persistence toward the activity) is associated with higher loneliness. Unfortunately, an analysis over finer age bins is not provided in the works of Martončik and Lokša (2016) or Mandryk et al. (2020) to quantify the relationship among younger adults. For several studies, significant relationships between video game play and loneliness are not observed (Visser et al., 2013; Yang and Liu, 2017; Verheijen et al., 2019). Among more recent studies, Koban et al. (2022) find among online video game players aged 18–62 (mean 25.22, SD 6.68) that loneliness is a predictor of adverse gaming outcomes. Studies also investigate young adult involvement in virtual social networks. The study of Hernández Camblor and Moral Jiménez, 2024 shows that among Spanish adults aged 18–40 (mean 26.33, SD 5.91), higher levels of loneliness are associated with greater virtual social netHernández Camblor and Moral Jiménez, 2024 Hernández Camblor and Moral Jiménez, 2024 work use, with Instagram being the highest used. All these studies involve interactions in traditionally desktop and mobile environments rather than VR.

Though VR has been explored as a tool to address isolation, the work is largely targeted toward older adults. Among older adults, immersive VR has been found to reduce loneliness (Dinet et al., 2023), generate perceptions of ‘being with others’ in virtual tourism settings (Oppert et al., 2023), generate high levels of engagement (Kalantari et al., 2023), and provide promise to promote wellbeing (Veldmeijer et al., 2020). Studies investigating young adult connectedness in social VR appear to be limited. Maloney et al. (2021) report results from interviews with 20 individuals aged 13–18 to discover why teenagers are attracted to social VR. They find that entry points appear to be introduction to social VR through video streaming sites and offline friends, low barrier to entry for free games, unstructured mixed socializing and gaming interactions, and opportunities to engage in alternative activities such as crafting. The work of Kenyon et al. (2023) compares participant ratings of loneliness on the UCLA-20 scale (Russell, 1996) and social anxiety on the Social Phobia Inventory (SPIN) (Connor et al., 2000) scale for participants socializing during the COVID-19 pandemic within Microsoft’s Altspace VR platform during the lockdown period. Altspace VR, though now defunct, was an environment intended to enable users to interact in virtual spaces through avatars. The questionnaires were administered directly in a research space hosted within Altspace VR. Participants were distributed nearly evenly across the age spectrum between 18 and 65, and 3 participants were 66+. The authors found support for the hypotheses that lower levels of loneliness and social anxiety are reported by participants for being in the social VR platform than in the real world, by participants who have a friend group in the environment than not, and by participants who spend more time in Altspace VR. A breakdown by age bins is not provided in the study. The study of Oh et al. (2023) surveys participants aged 15–41 (mean 28.8, SD 8.89) playing one or both of the metaverse games Roblox and Zepeto to measure, among other variables, level of engagement in supportive interaction behaviors in the metaverse, social self-efficacy, and reported loneliness using 3 questions from the short scale of Hughes et al. (2004). Interestingly, though they find a negative relationship between social self-efficacy and loneliness, they find a significant positive relationship between supportive interaction behaviors and loneliness. Their study indicates that for their evaluated participants, the more confident people were in their own social skills, the less lonely they felt, whereas the more they engaged in supportive interaction, the more loneliness they experienced in daily life. Lai et al. (2023) find no significant difference in pre- and post-intervention UCLA-20 loneliness scores when 8 adolescent participants (mean age 16.6, SD 1.8) with disabilities recruited to interact with the online virtual platform VRChat.

In all studies discussed in prior paragraphs involving young adults engaging in online interactions, a key characteristic remains that surveys of loneliness are done by recruiting participants who are already engaged in online game play or social interaction environments. To the best of our knowledge, studies are lacking that investigate how new environments can assist young adults with overcoming the negative impacts of social isolation and perceptions of loneliness—particularly environments such as our VR cafe, that simulate real world environments where social interactions occur and where users go to have needs met such as picking up food. In providing a VR cafe platform through which investigations of loneliness can be made to inform interventions for loneliness, our study can be contextualized within studies that use VR for cognitive behavioral therapy (CBT). Exposure treatment through VR has been found to be promising when used to address social phobias and social anxiety (Bouchard et al., 2017; Freitas et al., 2021). CBT has been shown to have promise in being effective to reduce loneliness among young adults (Cawthorne et al., 2023), making the proposed VR environment valuable to assist individuals with overcoming the adverse effects of loneliness and social isolation.

3 Materials and methods

3.1 Flavor and vine: an interactive VR cafe

We design Flavor and Vine, a VR cafe, in Unity 6000.0.53f1 to simulate multi-person environments where the user is likely to be involved in inter-personal interactions that can be triggers for social anxiety Connor et al. (2000) as shown in Figure 2. Our VR cafe consists of an entryway, a waiting area, and a dining area. The entryway is decorated with a red carpet that guides the participant to the maître d’ podium and a wooden menu board display. The waiting area is decorated with a banquette and a digital tablet representing a take-out menu. The dining area consists of 5 sitting tables with chairs and 3 standing tables in a row with standing Table 1 closest to the entrance and standing Table 2, farthest, from the perspective of the participant. The dining area expands past standing Table 2 to a bar counter with stools and a rope stanchion to set a boundary in our scene. This not only keeps participants safe in our physical environment, but also aids in creating the illusion that the VR cafe is larger and more populated even if not visible to the participant. All tables are decorated with interactable white flower vases, table numbers, and condiments, with the exception of standing tables, which are equipped with a digital tablet displaying the menu items for the day. We limit the menu items on display to two selections to not create unnecessary distractions for the participant, being careful to list a broad range of food items that accommodate various dietary preferences. Reserved signs are placed only on the sitting tables that do not have seated NPCs to fit the narration that only standing tables are available for dining, conveyed by the maître d’ NPC to the participant. The walls of the dining area are decorated with framed photographs depicting themes of an urban cafe, which matches the external landscape, visible through the entrance windows of the cafe, and the metallic ceiling vents. The entrance corners of the cafe are decorated with green potted plants with the farthest corners of the cafe being decorated with utility carts and wine shelves.

FIGURE 2

3.1.1 Non-playable characters (NPCs)

We create a total of four NPCs using Reallusion’s Character Creator 4 and animated using iClone 8. Our sole interactive NPC who engages with the participant is the maître d’ NPC. The maître d’ NPC is attired in a white shirt, black pants, and apron and stands behind the welcome podium facing the participant. We also create three non-interactive NPCs, i.e., ones that do not actively engage with the participant, to represent the other customers in our cafe. Two seated customer NPCs are placed to the right of standing Table 2 in our scene and another seated customer NPC is placed to the left of standing Table 2 on a stool in front of the bar counter. In Figure 3, we show the four NPCs in our VR cafe.

FIGURE 3

3.1.2 Sounds

3.2 User study

3.2.1 Participant inclusion and exclusion criteria

3.2.2 Participant demographics

After participants completed our IRB approved Informed Consent form, they filled out a demographics survey consisting of the questions listed in Table 1. We provide participants with a freeform text box for their self-identified gender to ensure that we eliminate biases due to pre-defined labels, thus following best practices on eliciting participant gender (Hughes et al., 2016; Spiel et al., 2019; Scheuerman et al., 2020). For participant race and ethnicity, we use the categorizations set forth by the National Center for Education Statistics (National Center for Education Statistics, 2025). Given the relationship between socioeconomic status and isolation (Algren et al., 2020; Näher et al., 2020), we use education level as a measure of socioeconomic status (Oakes and Andrade, 2017). Given that our study is conducted in an immersive environment, we asked participants ‘How often do you use a VR device?’. For participants who responded with any option other than ‘Never’, we provided a freeform textbox with the prompt ‘List the devices you use/have used’. To understand if participants owned VR devices, we asked the question ‘Do you own a VR device?’ and provided a freeform text box to list their VR devices. To get an understanding of participant’s baseline dining preferences in the real world, we asked ‘How frequently do you use take-out services?’, ‘How frequently do you use dine-in services?’, and ‘Which of the following do you most often choose?’.

3.2.3 Pre-immersion surveys

After completing the demographics form, we asked participants to complete Harrington’s Frustration Discomfort Scale (FDS) to understand frustration intolerance (Harrington, 2005), Revised UCLA-20 for feelings of loneliness (Russell, 1996), Social Phobia Inventory (SPIN) to understand the severity of social anxiety (Connor et al., 2000), and State Trait Anxiety Inventory (STAI) to measure anxiety levels for the past month (Speilberger et al., 1983). We collect anxiety measures since loneliness and structural measures of isolation are known to have an association with social anxiety (Jones et al., 1981; Lim et al., 2016; Eres et al., 2023; Teo et al., 2013).

3.2.4 Treatment

3.2.5 Post-immersion surveys

After participants completed all four treatments, we administered the NASA Task Load Index (NASA TLX) to measure workload (Hart and Staveland, 1988), System Usability Scale (SUS) for perceived usability (Brooke, 1996), and virtual reality sickness questionnaire (VRSQ) to assess any motion sickness within the simulation (Kim et al., 2018), and re-administered the FDS (Harrington, 2005) to measure any post study differences in frustration and discomfort.

4 Results

As shown in Table 1, our participants range from 19 to 53 (average = 24.8 and standard deviation = 5.5). 31 of the 35 participants have age 27 and below, and therefore fall within our intended target population of Gen Z adults. 3 have ages 28, 30, and 31, and fall in the Zillenial category as of the study period. We find that the majority (N = 26) self-identified as male, with the remaining (N = 9) identifying as female. For our analysis, we focus on participant reported feelings on loneliness captured during pre-immersion using the Revised UCLA-20. We convert the raw UCLA-20 scores using the standard binning that places scores between 20–34 as Low, 35–49 as Moderate, 50–64 as Moderate High, and 65–80 as High Loneliness. In Section 4.1, we analyze the relationship between self-identified gender, race, and dining selection choices based on the UCLA-20 categories. In Section 4.2, we compare participant reported frustration due to the delay, sense of isolation, and anxiousness across each treatment. Finally, in Section 4.3, we analyze how participant selection choices to ‘Dine In’, ‘Take Out’, or ‘Exit’ along with follow up choices indicating ‘Yes’ or ‘No’ to standing tables when dining in are related to the UCLA-20 categories. Given the limited sample size within each self-identified gender and race bins, we opt against any hypotheses testing and only report differences based only on participant counts.

4.1 Participant perceived sense of loneliness pre-immersion

Overall, we find that 15 participants fall into the ‘Low’ category, 14 in ‘Moderate’, and 6 in ‘Moderate High’. We find that the majority (N = 7) of the participants self-identifying as female (N = 9) report either a Moderate or Moderate High level of loneliness. On the other hand, for participants self-identifying as male we see an even split (N = 13) for Low and Moderate/Moderate High levels of loneliness. We find a near even split for Low (N = 8) and Moderate/Moderate High (N = 7) levels of loneliness for those selecting ‘White’ for their race. In contrast, participants who selected ‘Asian’ were more likely to select Moderate/Moderate High (N = 11) over Low (N = 7). As shown in Table 3, when exploring participant responses to the question ‘Which of the following do you most often choose?’ with answer choices of ‘Dine in’, ‘Take out’, and ‘No preference’, we find that the majority (N = 8) of the participants selecting ‘Dine in’ (N = 12) fall into the Low category for loneliness. In contract, participants who selected ‘Take out’ (N = 13) were more likely to fall into the Moderate/Moderate High (N = 8) category. Finally, a majority (N = 8) of the participants who selected ‘No preference’ (N = 10) were in the Moderate/Moderate High category for loneliness. Due to the limited sample size, in the remainder of our analysis, we use the term ‘Moderate+’ to group participants in the ‘Moderate’ and ‘Moderate High’ bins together.

4.2 Participant sense of frustration

Our VR cafe presented participants with 4 different delay types—No wait for maître d’ and No wait for food (coded as 00), No wait for maître d’ and 3-min wait for food (coded as 03), 3-min wait for maître d’ and No wait for food (coded as 30), and 3-min wait for maître d’ and 3-min wait for food (coded as 33). Following each treatment, participants were ‘What was your frustration level due to the delay?’, ‘What was your frustration level due to feelings of isolation?’, and ‘What was your frustration level due to feeling anxious?’, with participants providing responses using a 5-point Likert scale with 1 being Low and 5 being High. Using Friedman’s Test (Friedman, 1940), we determine if there are statistically significant differences in frustration levels. We report effect size using Kendall’s W using the standard interpretations of values (Tomczak and Tomczak, 2014). We use Wilcoxon Signed Rank Tests (Wilcoxon, 1945) with Bonferroni correction for post hoc analysis when the Friedman’s Test is significant.

We find that the average participant reported levels of frustration due to the delay for the four treatments, ‘00’, ‘03’, ‘30’, and ‘33’, are 1.34 ± 0.80, 2.23 ± 1.40, 2.74 ± 1.46, and 2.63 ± 1.40 respectively. We find a statistically significant difference (, moderate effect size of ). Our post hoc analysis shows that treatment ‘00’ is significantly different from all other treatments and treatment ’03’ is significantly different from ‘30’. From Table 2, we find that participants in the ‘Moderate+’ category had higher average frustration scores for treatments ‘00’, ‘03’, and ‘30’. The average scores for the four treatments, ‘00’, ‘03’, ‘30’, and ‘33’, are 1.46 ± 0.95, 2.17 ± 1.38, 2.29 ± 1.45, and 2.20 ± 1.37 respectively for the participant report levels of frustration due to feelings of isolation. We find a statistically significant difference (, small effect size of ) between the four treatments. Our post hoc analysis reveals that treatment ‘00’ is significantly different from all other treatments. In Table 2, we see that participants in the ‘Moderate+’ category have higher average frustration scores for ‘00’ and ‘33’, however, this trend is reversed for the ‘03’ treatment. Participant reported frustration levels due to feeling anxious for the four treatments, ‘00’, ‘03’, ‘30’, and ‘33’, are 1.51 ± 0.98, 1.91 ± 1.27, 2.03 ± 1.38, and 2.03 ± 1.18 respectively. While the Friedman’s Test is statistically significant (, small effect size of ), the post hoc analysis shows marginal significance for treatment ‘00’ when compared to the other treatments. Similar to the feelings of frustration due to a sense of isolation, from Table 2 we see that participants in the ‘Moderate+’ category have higher average frustration scores for ‘00’ and ‘33’, with the trend reversed for the ‘03’ treatment.

Using a Wilcoxon Signed Rank Test (Wilcoxon, 1945) we find no statistically significant difference between the pre- and post-immersion Frustration Discomfort Scale (FDS) across all participants. Participants in the ‘Low’ category for UCLA-20 had an average FDS score of 71.67 ± 16.53 before immersion and 72.07 ± 18.89 after. In contrast, participants who were in the ‘Moderate+’ category had a higher score of 81.85 ± 22.60 before and 82.35 ± 19.48 after immersion.

4.3 Participant selection choices

We provided participants with three potential first selection choices once they were spawned into our cafe scene —‘Dine In’, ‘Take Out’, or ‘Exit’. Here, ‘Dine In’ indicates that the participant wants to ‘Dine In’ and ‘Take Out’ means the participant wants their food to go. We use ‘Exit’ to indicate the participant ending the simulation voluntarily. As shown in Table 4, in the ‘00’ and ‘03’ treatments, or the ones where the maître d’ did not introduce a delay in interacting with the participant, we observe that 25 out of 35 participants opted for the ‘Dine In’ option and 10 chose the ‘Take Out’. On the other hand, for the ‘30’ and ‘33’ treatments, where the maître d’ was delayed by 3-min, we find that 19 and 21 participants opted to ‘Dine In’, 5 and 6 pressed ‘Exit’, and 11 and 8 chose the ‘Take Out’ option respectively. Thus, when the maître d’ introduces a delay in engaging with the participant upon entry, we find that more participants are likely to select the ‘Exit’ option. In Table 4, we summarize the first selection choice made by the participant based on the UCLA-20 loneliness categories. We find that for treatments ‘00’ and ‘03’, or when there is no delay with the maître d’, a higher percentage of participants in the ‘Low’ loneliness category opt to ‘Dine In’ as opposed to those in the ‘Moderate+’ category. When the maître d’ is delayed by 3-min, or the ‘30’ and ‘33’ treatments, we find that more participants in the ‘Moderate+’ category opt to ‘Exit’ the simulation.

Participants who opted for the ‘Dine In’ option in the first selection were informed by the maître d’ that the cafe only had standing tables available. Participants had the option to either select ‘Yes Standing Table’ to indicate they were ok with going to a standing table or ‘No Standing Table’ to indicate that they did not want a standing table. Participants who selected ‘No Standing Table’ were asked if they wanted a take out menu, to which they could select ‘Yes Take Out’ or ‘No Take Out’. From Table 5, we observe that for all treatments the majority of the participants opted to take the standing table option when they were informed that there were no other seating options. However, we find a higher percentage of participants who fall into the ‘Moderate+’ category for UCLA-20, i.e., Moderate and Moderate High, opted for ‘No Standing Table’ for the ‘30’ and ‘33’ treatments when compared to those in the ‘Low’ category.

Participants who selected ‘Yes Standing Table’ were informed by the maître d’ NPC that they could select any of the three standing tables. In our simulation, the tables are numbered ‘1’, ‘2’, and ‘3’, with 1 being closest to the entrance of the cafe. From Table 6, we find that most participants select Standing Table 3 (Tbl2). We find that participants in the ‘Low’ loneliness category select Standing Table 2, albeit only in a single instance in ‘00’, ‘30’, and ‘33’, while no participant in the ‘Moderate+’ selects Standing Table 2.

5 Discussion

In our dataset, the majority (N = 30) of our participants selected rarely or never for their frequency of use of a VR device. Despite the lack of experience with VR, we find that our immersive environment receives an average score of 82.14 on the SUS, or an ‘A’ grade (Lewis and Sauro, 2018). Similarly, our average NASA TLX score was 21.34, indicating a relatively low overall workload on the participant. The high SUS and low NASA TLX score indicates that our immersive environment did not have usability or taskload issues that may have impacted how participant behavior during each treatment. The low NASA TLX score may also indicate that participants did not face mental or physical fatigue as a result of completing each treatment and their associated surveys.

Our analysis in Section 4.1 shows the relationship between self-identified gender, race, and loneliness, with more participants identifying as female or Asian reporting ‘Moderate’ or higher categories of loneliness. We provide our analysis using participant counts as opposed to conducting any hypotheses testing due to the limited sample size within self-identified gender and race bins. Section 4.2 shows that we find a statistically significant difference in frustration levels across the four treatments when there is a delay in service. We observe that participants in the ‘Moderate’ or higher category of loneliness had a higher average frustration on the standardized FDS scale. Though our work addresses the specific case of service, it aligns with the findings of Sarcali and Guler (Saricali and Guler, 2022) that broad-scale existential loneliness relates to frustration intolerance. In Section 4.3, we observe that when there is no delay with the maître d’, participants categorized as ‘Low’ on the loneliness scale were more likely to opt for a dine in experience. When a delay is introduced by the maître d’, we find more participants in the ‘Moderate’ or higher category opt to exit the environment. In the delayed scenario, a higher percentage of participants in the ‘Moderate’ or higher category opted for the ‘No Standing Table’ option after opting to dine in when compared to those in the ‘Low’ category. These observations reflect the findings on young adults (Russell et al., 1980) reporting lonely dining alone. Overall, our findings indicate that individuals reporting moderate or higher loneliness scores may portray behaviors indicative of being frustrated with being in the social settings, or making choices that enable avoiding prolonged stay in the social environment. In our analysis we focused on participant responses gathered through the survey instruments. As part of future work, we will explore how behavior data recorded by the VR device, such as eye gaze, head position, and head orientation correlate with participant choices made within the environment, such as choosing to exit the simulation.

In our study we opted for immediate service, i.e., no wait, or a short delay, i.e., 3-min wait. In the real world, waits can get longer if there is a queue waiting for the maître d’ or if the kitchen is understaffed or the cafe is busy. Future work can explore studies that involve longer delays, for example a 10-min wait to be seated at the table or a 15-min wait to receive the food. Our study explored participant behavior when surrounded by one NPC maître d’ and three customers NPCs in a typical cafe setting. Future research may explore broader identities for the maître d’ and other NPCs as well as the social setting, for example a formal restaurant, bar, or casual dining place. We did not investigate how participant behavior changes with changing crowd density or scenarios where another patron requests to join the table due to the lack of space. When the participant was standing in line, we did not include NPCs either ahead or behind the participants. We also did not include awkward scenarios such as a patron storming out because of the delay. Future work may explore how these dynamics shape participant choices and behavior. Similar to work by Alvarez Igarzábal et al. (2021), which explored participant perception of waiting in real and virtual waiting room, future work can also explore how participant behavior changes between real and virtual cafes.

Persistent avoidance of a social setting such as a cafe that also serves the purpose of meeting basic needs can be detrimental to users in the long run. Our VR cafe provides the opportunity to enable cognitive behavioral therapy (CBT) to enable individuals to overcome loneliness (Cawthorne et al., 2023), shown to be effective for loneliness in real world environments and for social phobias in VR spaces (Bouchard et al., 2017; Freitas et al., 2021). Our goal in this work has been to provide a baseline evaluation of behavior in social service-providing settings for users reporting various levels of loneliness. In future work, we plan to work with CBT specialists to incorporate feedback responses that encourage the user toward more productive coping behaviors. Our current work primarily targets baseline evaluation of loneliness scores and their relationship to user behavior in terms of frustration and choices made in the VR environment. Our environment provides the opportunity to study relationships of loneliness and user choices to other mental health conditions such as social anxiety and depression. The relationships of loneliness to social anxiety have been extensively studied Teo et al. (2013); Jones et al. (1981); Vanhalst et al. (2013); Eres et al. (2023), with studies showing consistent and positive associations among young adults and adolescents (Jones et al., 1981; Vanhalst et al., 2013). Eres et al. (2023) find that participants reporting being more lonely also reported feeling more socially anxious, and having higher depressive symptom scores, more difficulties with emotion regulation, and greater risk of social isolation. As part of future work, we are interested in evaluating the relationship of anxiety measures we have collected using the SPIN and STAI scales to loneliness and user behavior in VR. In addition to mental health conditions, it would also be of interest to study the relationships of user choices and reports of loneliness to personality traits. The relationship between loneliness and the Big Five traits of openness, conscientiousness, extraversion, agreeableness, and neuroticism has been studied, with studies showing that individuals reporting lonely may on average show low extraversion Buecker et al. (2020); Teppers et al. (2013) and high neuroticism Buecker et al. (2020), and that low extraversion and low agreeableness predict social isolation Pucker et al. (2019). Leveraging VR to study the relationship of mental state and user traits on interaction preferences has the benefit to support informed use of VR for therapy cognizant of the natural leanings of individual users.

5.1 Limitations

We acknowledge that participants are skewed towards Non Hispanic or Latino for ethnicity and Asian and White for race. Such a skew is expected given the collection was conducted predominantly on a university campus and its surrounding community. Similarly, our collection is skewed towards participants self-identifying as male with no participant providing a non-binary gender. The participants in our dataset fall into the ‘Low’, ‘Moderate’, or ‘Moderate High’ bins for loneliness on the UCLA-20 scale, with only 6 participants in the ‘Moderate High’ category. To perform meaningful analysis, we combined the ‘Moderate’ and ‘Moderate High’ categories into a single ‘Moderate+’ category.

As future work, we will broaden our recruitment strategies to encompass a wider participant diversity and incorporate longitudinal studies through the typical 4 years of undergraduate education. A larger sample size can enable the use of demographic variables, such as self-identified gender, ethnicity, and/or race, as independent variables to predict loneliness bins on the UCLA-20 scale. Our current study relies on the same participant undertaking all four treatments, which leads to a testing internal validity threat. An expanded collection will enable participants to be randomly assigned to a single treatment, which results in a more robust experimental design. As the first of its kind study on understanding how loneliness shapes user behavior within an immersive representation of a social scenario, the limitations of our work are expected.

6 Conclusion

Our work investigates the relationship of user scores of loneliness to frustration and choices made by users when visiting a VR cafe, a social setting that addresses basic needs. Our results indicate that moderate to high loneliness levels relate to users being frustrated, and making avoidance decisions, such as opting for not dining in especially if delay is introduced in service. Future work can expand on our VR cafe to introduce interventions for users who feel isolated to manage feelings of frustration and avoidance choices.

References

• 1

  AbplanalpS. J.GreenM. F.WynnJ. K.EisenbergerN. I.HoranW. P.LeeJ.et al (2024). Using machine learning to understand social isolation and loneliness in schizophrenia, bipolar disorder, and the community. Schizophrenia10, 88. 10.1038/s41537-024-00511-y

  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  AlgrenM. H.EkholmO.NielsenL.ErsbøllA. K.BakC. K.AndersenP. T. (2020). Social isolation, loneliness, socioeconomic status, and health-risk behaviour in deprived neighbourhoods in Denmark: a cross-sectional study. SSM-population Health10, 1. 10.1016/j.ssmph.2020.100546

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  Alvarez IgarzábalF.HrubyH.WitowskaJ.KhoshnoudS.WittmannM. (2021). What happens while waiting in virtual reality? A comparison between a virtual and a real waiting situation concerning boredom, self-regulation, and the experience of time. Technol. Mind, Behav.2, 117–127. 10.1037/tmb0000038

  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  AustinB. A. (1983). Factorial structure of the ucla loneliness scale. Psychol. Reports53, 883–889. 10.2466/pr0.1983.53.3.883

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  BahrH. M.HarveyC. D. (1979). Correlates of loneliness among widows bereaved in a mining disaster. Psychol. Rep.44, 367–385. 10.2466/pr0.1979.44.2.367

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  BaumS. K. (1982). Loneliness in elderly persons: a preliminary study. Psychol. Reports50, 1317–1318. 10.2466/pr0.1982.50.3c.1317

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  BergS.MellströmD.PerssonG.SvanborgA. (1981). Loneliness in the swedish aged. J. Gerontology36, 342–349. 10.1093/geronj/36.3.342

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  BiswasD.LundK.SzocsC. (2019). Sounds like a healthy retail atmospheric strategy: effects of ambient music and background noise on food sales. J. Acad. Mark. Sci.47, 37–55. 10.1007/s11747-018-0583-8

  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  BlazerD. (2020). Social isolation and loneliness in older adults—a mental health/public health challenge. JAMA Psychiatry77, 990–991. 10.1001/jamapsychiatry.2020.1054

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  BouchardS.DumoulinS.RobillardG.GuitardT.KlingerE.ForgetH.et al (2017). Virtual reality compared with in vivo exposure in the treatment of social anxiety disorder: a three-arm randomised controlled trial. Br. J. Psychiatry210, 276–283. 10.1192/bjp.bp.116.184234

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  BowlerA. (2020). Isolation among generation z in the United States. Ballard Brief.4, 1–16.

  • Google Scholar
  • View reference in article

• 12

  BrookeJ. (1996). Sus-a quick and dirty usability scale. Usability Evaluation Industry189, 4–7.

  • Google Scholar
  • View reference in article

• 13

  BuF.SteptoeA.FancourtD. (2020a). Loneliness during a strict lockdown: trajectories and predictors during the covid-19 pandemic in 38,217 United Kingdom adults. Soc. Science and Medicine265, 113521. 10.1016/j.socscimed.2020.113521

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  BuF.SteptoeA.FancourtD. (2020b). Who is lonely in lockdown? Cross-cohort analyses of predictors of loneliness before and during the covid-19 pandemic. Public Health186, 31–34. 10.1016/j.puhe.2020.06.036

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  BueckerS.MaesM.DenissenJ. J.LuhmannM. (2020). Loneliness and the big five personality traits: a meta–analysis. Eur. J. Personality34, 8–28. 10.1002/per.2229

  • CrossRef
  • Google Scholar

• 16

  Carpenito-MoyetL. J. (2006). Nursing Diagnosis: Application to Clinical Practice. Lippincott Williams and Wilkins.

  • Google Scholar
  • View reference in article

• 17

  CawthorneT.KällA.BennettS.BakerE.AnderssonG.ShafranR. (2023). The development and preliminary evaluation of cognitive behavioural therapy (cbt) for chronic loneliness in young people. Behav. Cognitive Psychotherapy51, 414–431. 10.1017/s1352465823000231

  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  ChanT.FriedmanD. S.BradleyC.MassofR. (2018). Estimates of incidence and prevalence of visual impairment, low vision, and blindness in the United States. JAMA Ophthalmology136, 12–19. 10.1001/jamaophthalmol.2017.4655

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 19

  ChouK.-L.LiangK.SareenJ. (2011). The association between social isolation and dsm-iv mood, anxiety, and substance use disorders: wave 2 of the national epidemiologic survey on alcohol and related conditions. J. Clinical Psychiatry72, 1468–1476. 10.4088/JCP.10m06019gry

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  ConnorK. M.DavidsonJ. R.ChurchillL. E.SherwoodA.WeislerR. H.FoaE. (2000). Psychometric properties of the social phobia inventory (spin): new self-rating scale. Br. J. Psychiatry176, 379–386. 10.1192/bjp.176.4.379

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  CornwellE. Y.WaiteL. J. (2009). Social disconnectedness, perceived isolation, and health among older adults. J. Health Social Behavior50, 31–48. 10.1177/002214650905000103

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  DickensA. P.RichardsS. H.GreavesC. J.CampbellJ. L. (2011). Interventions targeting social isolation in older people: a systematic review. BMC Public Health11, 647. 10.1186/1471-2458-11-647

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  DinetJ.MorèreY.NouchiR.SakakiK.ClancheF.CasteranM. (2023). “Breaking social isolation for older people living alone with technology,” in Proceedings of the European Conference on Cognitive Ergonomics 2023, 1–5. 10.1145/3605655.3605660

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  EresR.LimM. H.BatesG. (2023). Loneliness and social anxiety in young adults: the moderating and mediating roles of emotion dysregulation, depression and social isolation risk. Psychol. Psychotherapy Theory, Res. Pract.96, 793–810. 10.1111/papt.12469

  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  FardghassemiS.JoffeH. (2022). The causes of loneliness: the perspective of young adults in london’s most deprived areas. Plos One17, e0264638. 10.1371/journal.pone.0264638

  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  FreitasJ. R. S.VelosaV. H. S.AbreuL. T. N.JardimR. L.SantosJ. A. V.PeresB.et al (2021). Virtual reality exposure treatment in phobias: a systematic review. Psychiatr. Q.92, 1685–1710. 10.1007/s11126-021-09935-6

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  FriedmanM. (1940). A comparison of alternative tests of significance for the problem of m rankings. Annals Mathematical Statistics11, 86–92. 10.1214/aoms/1177731944

  • CrossRef
  • Google Scholar
  • View reference in article

• 28

  HarriesB.ZenoneM.HartwellG. (2025). Feeling lonely in the online crowd: what tiktok tells us about young people and loneliness. Health Promotion International40, daaf094. 10.1093/heapro/daaf094

  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  HarringtonN. (2005). The frustration discomfort scale: development and psychometric properties. Clin. Psychol. Psychotherapy Int. J. Theory Pract.12, 374–387. 10.1002/cpp.465

  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  HartS. G.StavelandL. E. (1988). Development of nasa-tlx (task load index): results of empirical and theoretical research. Adv. Psychology52, 139–183.

  • Google Scholar
  • View reference in article

• 31

  HembergJ.ÖstmanL.KorzhinaY.GroundstroemH.NyströmL.Nyman-KurkialaP. (2022). Loneliness as experienced by adolescents and young adults: an explorative qualitative study. Int. J. Adolesc. Youth27, 362–384. 10.1080/02673843.2022.2109422

  • CrossRef
  • Google Scholar
  • View reference in article

• 32

  Hernández CamblorC.Moral JiménezM. V. (2024). Use of virtual social networks and its relationship with the perception of loneliness and social skills among spanish young adults of generation z and y. Acta Colomb. Psicol. 231–247.

  • Google Scholar

• 33

  Holt-LunstadJ.SmithT. B.LaytonJ. B. (2010). Social relationships and mortality risk: a meta-analytic review. PLoS Medicine7, e1000316. 10.1371/journal.pmed.1000316

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  HuY.GutmanL. M. (2021). The trajectory of loneliness in UK young adults during the summer to winter months of covid-19. Psychiatry Res.303, 114064. 10.1016/j.psychres.2021.114064

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  HughesM. E.WaiteL. J.HawkleyL. C.CacioppoJ. T. (2004). A short scale for measuring loneliness in large surveys: results from two population-based studies. Res. Aging26, 655–672. 10.1177/0164027504268574

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  HughesJ. L.CamdenA. A.YangchenT. (2016). Rethinking and updating demographic questions: guidance to improve descriptions of research samples. Psi Chi J. Psychol. Res.21, 138–151. 10.24839/2164-8204.jn21.3.138

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  JarryF.DorfmanA.Pelletier-DumasM.LinaJ.-M.StolleD.LacourseÉ.et al (2025). Understanding the interplay between social isolation, age, and loneliness during the covid-19 pandemic. Sci. Rep.15, 239. 10.1038/s41598-024-81519-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  JonesW. H.FreemonJ.GoswickR. A. (1981). The persistence of loneliness: self and other determinants 1. J. Personality49, 27–48. 10.1111/j.1467-6494.1981.tb00844.x

  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  JonesW. H.CarpenterB. N.QuintanaD. (1985). Personality and interpersonal predictors of loneliness in two cultures. J. Personality Soc. Psychol.48, 1503–1511. 10.1037/0022-3514.48.6.1503

  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  JonesW. H.RoseJ.RussellD. (1990). “Loneliness and social anxiety,” in Handbook of Social and Evaluation Anxiety. Editor LeitenbergH. (Springer science and business media), 247–266.

  • Google Scholar
  • View reference in article

• 41

  KalantariS.XuT. B.MostafaviA.KimB.DilanchianA.LeeA.et al (2023). Using immersive virtual reality to enhance social interaction among older adults: a cross-site investigation. Innovation Aging7, igad031. 10.1093/geroni/igad031

  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  KayaF.BayzanŞ.BhaisA. Z. B.GünA. (2025). Lonely, anxious, and unhappy: the psychosocial toll of digital game addiction in university students. Int. J. Ment. Health Addict., 1–26. 10.1007/s11469-025-01568-1

  • CrossRef
  • Google Scholar
  • View reference in article

• 43

  KenyonK.KinakhV.HarrisonJ. (2023). Social virtual reality helps to reduce feelings of loneliness and social anxiety during the covid-19 pandemic. Sci. Rep.13, 19282. 10.1038/s41598-023-46494-1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 44

  KimH. K.ParkJ.ChoiY.ChoeM. (2018). Virtual reality sickness questionnaire (vrsq): motion sickness measurement index in a virtual reality environment. Appl. Ergonomics69, 66–73. 10.1016/j.apergo.2017.12.016

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 45

  KobanK.BiehlJ.BornemeierJ.OhlerP. (2022). Compensatory video gaming. gaming behaviours and adverse outcomes and the moderating role of stress, social interaction anxiety, and loneliness. Behav. and Inf. Technol.41, 2727–2744. 10.1080/0144929x.2021.1946154

  • CrossRef
  • Google Scholar
  • View reference in article

• 46

  KowertR.VogelgesangJ.FestlR.QuandtT. (2015). Psychosocial causes and consequences of online video game play. Comput. Hum. Behav.45, 51–58. 10.1016/j.chb.2014.11.074

  • CrossRef
  • Google Scholar
  • View reference in article

• 47

  LaiB.YoungR.CraigM.ChavianoK.Swanson-KimaniE.WozowC.et al (2023). Improving social isolation and loneliness among adolescents with physical disabilities through group-based virtual reality gaming: feasibility pre-post trial study. JMIR Form. Res.7, e47630. 10.2196/47630

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 48

  LevinI.StokesJ. P. (1986). An examination of the relation of individual difference variables to loneliness. J. Personality54, 717–733. 10.1111/j.1467-6494.1986.tb00422.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 49

  LewisJ. R.SauroJ. (2018). Item benchmarks for the system usability scale. J. Usability Studies13, 1–10.

  • Google Scholar
  • View reference in article

• 50

  LimM. H.RodebaughT. L.ZyphurM. J.GleesonJ. F. (2016). Loneliness over time: the crucial role of social anxiety. J. Abnormal Psychology125, 620–630. 10.1037/abn0000162

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 51

  LuoY.MoosbruggerM.SmithD. M.FranceT. J.MaJ.XiaoJ. (2022). Is increased video game participation associated with reduced sense of loneliness? A systematic review and meta-analysis. Front. Public Health10, 898338. 10.3389/fpubh.2022.898338

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 52

  MaloneyD.FreemanG.RobbA. (2021). “Stay connected in an immersive world: why teenagers engage in social virtual reality,” in Proceedings of the 20th Annual ACM Interaction Design and Children Conference, 69–79.

  • Google Scholar
  • View reference in article

• 53

  MandrykR. L.FrommelJ.ArmstrongA.JohnsonD. (2020). How passion for playing world of warcraft predicts in-game social capital, loneliness, and wellbeing. Front. Psychology11, 2165. 10.3389/fpsyg.2020.02165

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 54

  MansfieldL.DaykinN.MeadsC.TomlinsonA.GrayK.LaneJ.et al (2019). A Conceptual Review of Loneliness Across the Adult Life Course (16+ Years). London, United Kingdom:What Works Centre for Wellbeing.

  • Google Scholar
  • View reference in article

• 55

  MartončikM.LokšaJ. (2016). Do world of warcraft (mmorpg) players experience less loneliness and social anxiety in online world (virtual environment) than in real world (offline)?Comput. Hum. Behav.56, 127–134. 10.1016/j.chb.2015.11.035

  • CrossRef
  • Google Scholar
  • View reference in article

• 56

  MooreG.FardghassemiS.JoffeH. (2023). Wellbeing in the city: young adults’ sense of loneliness and social connection in deprived urban neighbourhoods. Wellbeing, Space Society5, 100172. 10.1016/j.wss.2023.100172

  • CrossRef
  • Google Scholar
  • View reference in article

• 57

  NäherA.-F.Rummel-KlugeC.HegerlU. (2020). Associations of suicide rates with socioeconomic status and social isolation: findings from longitudinal register and census data. Front. Psychiatry10, 898. 10.3389/fpsyt.2019.00898

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 58

  National Center for Education Statistics (2025). Definitions for New Race and Ethnicity Categories. Available online at: https://nces.ed.gov/ipeds/report-your-data/race-ethnicity-definitions (Accessed December 15, 2025).

  • Google Scholar
  • View reference in article

• 59

  OakesJ. M.AndradeK. E. (2017). The measurement of socioeconomic status. Methods Social Epidemiology18, 23–42.

  • Google Scholar
  • View reference in article

• 60

  OhH. J.KimJ.ChangJ. J.ParkN.LeeS. (2023). Social benefits of living in the metaverse: the relationships among social presence, supportive interaction, social self-efficacy, and feelings of loneliness. Comput. Hum. Behav.139, 107498. 10.1016/j.chb.2022.107498

  • CrossRef
  • Google Scholar
  • View reference in article

• 61

  OppertM. L.NgoM.LeeG. A.BillinghurstM.BanksS.TolsonL. (2023). Older adults’ experiences of social isolation and loneliness: can virtual touring increase social connectedness? A pilot study. Geriatr. Nursing53, 270–279. 10.1016/j.gerinurse.2023.08.001

  • CrossRef
  • Google Scholar
  • View reference in article

• 62

  O’DayE. B.HeimbergR. G. (2021). Social media use, social anxiety, and loneliness: a systematic review. Comput. Hum. Behav. Rep.3, 100070. 10.1016/j.chbr.2021.100070

  • CrossRef
  • Google Scholar
  • View reference in article

• 63

  PerlmanD.PeplauL. A. (1982). Loneliness: A Sourcebook of Current Theory, Research and Therapy, 36. Wiley.

  • Google Scholar
  • View reference in article

• 64

  PuckerH. E.TemesC. M.ZanariniM. C. (2019). Description and prediction of social isolation in borderline patients over 20 years of prospective follow-up. Personality Disord. Theory, Res. Treat.10, 383–388. 10.1037/per0000337

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 65

  RussellD. W. (1996). Ucla loneliness scale (version 3): reliability, validity, and factor structure. J. Personality Assessment66, 20–40. 10.1207/s15327752jpa6601_2

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 66

  RussellD.PeplauL. A.CutronaC. E. (1980). The revised ucla loneliness scale: concurrent and discriminant validity evidence. J. Personality Social Psychology39, 472–480. 10.1037//0022-3514.39.3.472

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 67

  SaricaliM.GulerD. (2022). The mediating role of psychological need frustration on the relationship between frustration intolerance and existential loneliness. Curr. Psychol.41, 5603–5611. 10.1007/s12144-022-02866-w

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 68

  ScheuermanM. K.SpielK.HaimsonO. L.HamidiF.BranhamS. M. (2020). Hci guidelines for gender equity and inclusivity. UMBC Fac. Collect.1, 1.

  • Google Scholar
  • View reference in article

• 69

  ShovestulB.HanJ.GermineL.Dodell-FederD. (2020). Risk factors for loneliness: the high relative importance of age versus other factors. PloS One15, e0229087. 10.1371/journal.pone.0229087

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 70

  SpeilbergerC. D.GorsuchR.LusheneR.VaggP.JacobsG. (1983). Manual for the state-trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists.

  • Google Scholar

• 71

  SpielK.HaimsonO. L.LottridgeD. (2019). How to do better with gender on surveys: a guide for hci researchers. Interactions26, 62–65. 10.1145/3338283

  • CrossRef
  • Google Scholar

• 72

  SundqvistA.HembergJ. (2021). Adolescents’ and young adults’ experiences of loneliness and their thoughts about its alleviation. Int. J. Adolesc. Youth26, 238–255. 10.1080/02673843.2021.1908903

  • CrossRef
  • Google Scholar

• 73

  Sonos Pro Editorial Team (2026). Background music for cafes: how to amplify your business and customer experience. Available online at: https://www.sonos.com/en/blog/background-music-for-cafes (Accessed March 3, 2026).

  • Google Scholar

• 74

  TeoA. R.LerrigoR.RogersM. A. (2013). The role of social isolation in social anxiety disorder: a systematic review and meta-analysis. J. Anxiety Disorders27, 353–364. 10.1016/j.janxdis.2013.03.010

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 75

  TeppersE.KlimstraT. A.DammeC. V.LuyckxK.VanhalstJ.GoossensL. (2013). Personality traits, loneliness, and attitudes toward aloneness in adolescence. J. Soc. Personal Relat.30, 1045–1063. 10.1177/0265407513481445

  • CrossRef
  • Google Scholar

• 76

  TomczakM.TomczakE. (2014). The need to report effect size estimates revisited. an overview of some recommended measures of effect size. Trends Sports Sci.1, 19–25.

  • Google Scholar

• 77

  TwengeJ. M.SpitzbergB. H.CampbellW. K. (2019). Less in-person social interaction with peers among us adolescents in the 21st century and links to loneliness. J. Social Personal Relationships36, 1892–1913. 10.1177/0265407519836170

  • CrossRef
  • Google Scholar

• 78

  VanhalstJ.LuyckxK.ScholteR. H.EngelsR. C.GoossensL. (2013). Low self-esteem as a risk factor for loneliness in adolescence: perceived-but not actual-social acceptance as an underlying mechanism. J. Abnormal Child Psychology41, 1067–1081. 10.1007/s10802-013-9751-y

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 79

  VeldmeijerL.WartenaB.TerlouwG.Van’t VeerJ. (2020). Reframing loneliness through the design of a virtual reality reminiscence artefact for older adults. Des. Health4, 407–426. 10.1080/24735132.2020.1848976

  • CrossRef
  • Google Scholar

• 80

  VerheijenG. P.BurkW. J.StoltzS. E.van den BergY. H.CillessenA. H. (2019). Associations between different aspects of video game play behavior and adolescent adjustment. J. Media Psychol. 32, 27–39.

  • Google Scholar

• 81

  VictorC. R.YangK. (2013). “The prevalence of loneliness among adults: a case study of the United Kingdom,” in Loneliness Updated (London, United Kingdom:Routledge), 88–107.

  • Google Scholar

• 82

  VisserM.AntheunisM. L.SchoutenA. P. (2013). Online communication and social well-being: how playing world of warcraft affects players’ social competence and loneliness. J. Appl. Soc. Psychol.43, 1508–1517. 10.1111/jasp.12144

  • CrossRef
  • Google Scholar

• 83

  WatsonJ.NesdaleD. (2012). Rejection sensitivity, social withdrawal, and loneliness in young adults. J. Appl. Soc. Psychol.42, 1984–2005. 10.1111/j.1559-1816.2012.00927.x

  • CrossRef
  • Google Scholar

• 84

  WilcoxonF. (1945). Individual comparisons by ranking methods. Biom. Bulletin1, 80–83. 10.2307/3001968

  • CrossRef
  • Google Scholar

• 85

  YangC.LiuD. (2017). Motives matter: motives for playing pokémon go and implications for well-being. Cyberpsychology, Behav. Soc. Netw.20, 52–57. 10.1089/cyber.2016.0562

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 86

  ZhouS. X.LeungL. (2012). Gratification, loneliness, leisure boredom, and self-esteem as predictors of sns-game addiction and usage pattern among chinese college students. Int. J. Cyber Behav. Psychol. Learn. (IJCBPL)2, 34–48.

  • Google Scholar