What Drives Hospitality Employees’ Trust in Service Robots?

Abstract

As service robots become more prevalent in hospitality settings, understanding what shapes employees’ trust in these technologies is essential for fostering effective human–robot collaboration. Despite extensive research on customer trust and robot-related attributes, employee perspectives have received limited and fragmented attention. The aim of this study is to examine how human, robot, and organizational factors collectively influence employees’ trust in service robots, thereby offering a more comprehensive understanding of trust formation in hospitality contexts. To address this aim, this study adopts a three-dimensional trust framework (human, robot, and organizational factors) and provides the first comprehensive empirical test in the hospitality sector. Drawing on survey data from 301 frontline hospitality workers in the United States, we investigated how various human-, robot-, and organization-related factors influence employees’ trust in service robots using bootstrap multiple regression analysis. The results reveal that human factors, particularly employees’ attitudes toward and comfort with robots, emerged as dominant trust predictors. Surprisingly, organizational factors showed minimal direct impact, suggesting complex trust dynamics unique to hospitality contexts. These findings significantly expand existing human–robot interaction (HRI) theory and offer critical practical insights for hospitality managers integrating robots into frontline service.

1. Introduction

Trust is a critical determinant in human–robot interaction (HRI) as it directly shapes how users perceive, accept, and collaborate with robotic technologies. Across diverse domains such as finance, healthcare, military, and transportation, research consistently highlights the pivotal role of trust in facilitating technology adoption (Choung et al., 2023; Gong, 2025; Leung et al., 2023; Song et al., 2022; Yan et al., 2022). Grounded in the Technology Acceptance Model (TAM), this body of work demonstrates that trust, along with perceived usefulness and ease of use, significantly impacts technology acceptance in contexts such as autonomous driving (Acerbo et al., 2025; Khan et al., 2025; S. Park et al., 2023), mobile banking (Arangarajan et al., 2024), military decision making (Schaefer et al., 2016), and AI-integrated systems (Choung et al., 2023; Davis, 1989).

In the hospitality field, trust in service robots similarly influences customers’ willingness to accept and use these technologies (Chi et al., 2023; S. Park, 2020; Tussyadiah et al., 2020; Wirtz et al., 2018). Humanoid robots, in particular, enhance service experiences through their social presence, often fostering greater customer trust and more favorable service evaluations (Mende et al., 2019; Qin et al., 2025). However, while much attention has been given to customer trust, employees’ trust in service robots is equally, if not more, critical. It directly affects job satisfaction, task performance, organizational adaptability, and the overall quality of customer service (Liu et al., 2024; Wolf & Stock-Homburg, 2025).

Importantly, trust in robots goes beyond basic comfort with technology; it is fundamental to effective collaboration between humans and robots in achieving shared goals (Alhaji et al., 2025; Bhargava et al., 2021; Le et al., 2023). While trust is often shaped by robot attributes such as appearance, performance, and reliability, it is also influenced by the broader environmental context and the characteristics of human collaborators involved (Hancock et al., 2011, 2021). While trust has been extensively studied across various domains such as finance, healthcare, and military, hospitality settings uniquely involve dynamic social interactions and high emotional labor. Existing literature in hospitality largely focuses on customer trust or robot attributes, leaving employee perspectives significantly understudied. In particular, previous meta-analyses (Hancock et al., 2021) have excluded hospitality contexts entirely, thus limiting their generalizability and leaving a critical theoretical gap concerning frontline employees.

Understanding this gap is essential because employees’ trust not only influences their collaboration with robots but also shapes customer experiences and organizational performance. Although prior studies have investigated individual human or organizational factors such as technology competence, anthropomorphism, attitudes toward robots, or leadership expectations (Della Corte et al., 2023; Li et al., 2024), these variables have seldom been examined collectively within an integrated framework. A comprehensive perspective that simultaneously considers human-related, robot-related, and organizational-related determinants is crucial for capturing the multifaceted nature of human–robot interaction in hospitality settings, where these factors operate interactively rather than independently. Accordingly, this study seeks to identify which human, robot, and organizational factors most strongly and consistently predict employees’ trust in service robots within the hospitality context. To this end, an exploratory predictive analysis is employed to assess the relative contribution of these factors and develop a hospitality-specific, employee-centered understanding of the antecedents of trust in service robots.

To address this research question, the present study systematically examines the relative influence of human-, robot-, and organization-related factors on hospitality employees’ trust in service robots. In doing so, it provides a more comprehensive and hospitality-specific understanding of the antecedents shaping trust within a service environment characterized by intensive human interaction, emotional labor, and high service quality expectations. Drawing on the three-dimensional trust framework proposed by Hancock et al. (2011, 2021) and Schaefer et al. (2016), this study adopts an integrative approach that concurrently considers a broad spectrum of trust antecedents across multiple dimensions.

The overarching aim of this research is to deepen understanding of how trust in service robots is developed among hospitality employees. Specifically, the study seeks to identify the most salient and influential combination of factors that predicts employee trust, thereby advancing a holistic understanding of trust formation. Guided by an exploratory predictive modeling approach, the study pursues three interrelated objectives: (1) to identify key human, robot, and organizational factors that significantly predict employee trust in service robots; (2) to assess the relative predictive contribution of these dimensions and their component variables; and (3) to provide an integrated model that captures how these factors interact to jointly forecast trust within real-world hospitality contexts. While prior research has often focused on individual determinants, such as robot attributes (e.g., reliability or performance) or human characteristics (e.g., attitude), in isolation, this study contributes by integrating these perspectives within a unified analytical framework. By doing so, it advances the understanding of employee trust in human–robot interaction, an area that has received considerably less attention than consumer trust. Beyond theoretical implications, the study also offers practical insights for hospitality organizations by highlighting employee-centered strategies for fostering trust, addressing diverse employee needs, aligning robot design and deployment with workplace realities, and promoting organizational environments that support sustainable human–robot collaboration.

The remainder of this paper is structured as follows. First, we present a comprehensive literature review on trust in HRI, highlighting key antecedents and examining how these factors have been positioned in the broader context of trust in automation and robotics. Next, we outline our study methodology, including the sample, data collection procedures, and measurement instruments. This is followed by regression analyses that identify the predictors of employee trust in service robots. Finally, we discuss the theoretical and practical implications of our findings, acknowledge the study’s limitations, and provide recommendations for future research.

2. Literature Review

2.1. Defining Trust in Human–Robot Interaction

Trust in HRI is widely recognized as a critical factor influencing the effectiveness, acceptance, and long-term success of human–robot collaboration. Despite the importance of trust, however, there is no consensus on its definition, and various ways to measure trust have been used in studies investigating trust in robots (Schäfer et al., 2024). Drawing from multidisciplinary research in psychology, computer science, and organizational behavior, trust in robots is often conceptualized as an individual’s belief or expectation that a robotic agent will act in ways that support their goals, particularly in contexts characterized by uncertainty and vulnerability (Malle & Ullman, 2021). This conceptualization builds upon foundational definitions of trust in human–automation interaction by J. Lee and See (2004) and aligns closely with the interpersonal trust model developed by Mayer et al. (1995), which frames trust in organizational settings as a relationship involving two parties: the trustor (the party placing trust) and the trustee (the party being trusted). Central to this model is a willingness to accept vulnerability, essentially a readiness to take risk in the relationship. Extending this framework to human–robot contexts, trust in robots can be broadly understood as a human’s willingness to rely on a robot’s actions, decisions, or recommendations in situations where uncertainty or perceived risk is present (Kohn et al., 2021).

To further unpack trust in robots, it is essential to consider the unique factors and types of trust that influence human perceptions and behaviors in these interactions. Unlike traditional human-to-human trust, trust in robots is shaped not only by perceptions of the robot’s reliability or performance but also by its predictability, transparency, fairness, and perceived intent (De Visser et al., 2016, 2018; Hancock et al., 2011). Scholars have identified multiple types of trust, including dispositional trust (a person’s general tendency to trust automation), situational trust (trust in a specific task or context), and learned trust (trust formed over time through repeated interactions) (Madhavan & Wiegmann, 2007). However, this study focuses on dispositional trust as an initial trust belief, given its relevance to understanding the antecedents of employees’ trust in service robots in hospitality settings.

In hospitality settings, where robots are increasingly integrated into frontline service roles, trust tends to center on perceptions of reliability, fairness, and safety in service delivery. This reflects not only employees’ perceptions of the robot’s performance but also their overall evaluation of the robot’s trustworthiness (Malle & Ullman, 2021). This aligns with prior literature that defines trust in robots as both a cognitive and affective judgement, encompassing beliefs about the robot’s competence, reliability, and intent (De Visser et al., 2018; Kraus et al., 2024). In this study, trust in robots is conceptualized as a global evaluation that includes both general beliefs about the robot’s trustworthiness and confidence in its ability to perform tasks reliably and without error. This construct focuses on the robot as the trustee, reflecting the cognitive judgment employees make when assessing whether the robot can fulfill its assigned function effectively. Such trust is essential for shaping employees’ willingness to engage with robotic systems, adapt to technological changes, and maintain high-quality service delivery in hospitality environments (Gefen et al., 2003; S. S. Park et al., 2021). Accordingly, this study adopts a belief-based definition of trust in robots: employees’ belief and confidence that robots will perform their assigned tasks reliably and without error (Kopp, 2024).

2.2. Factors Influencing Trust in Robots

Trust in robots is shaped by a complex interplay of factors that include not only robot-specific characteristics but also the surrounding environment and human collaborators’ attributes (Hancock et al., 2011). A wide range of antecedents influence trust in robots, including robot attributes (e.g., appearance, behavior, level of autonomy), human-related factors (e.g., prior experience, attitudes), and organizational elements (e.g., task relevance, management preference). In service-oriented settings such as hospitality, trust in robots extends beyond mere task performance (e.g., accuracy or reliability) to encompass social and emotional dimensions, including perceived safety, fairness, and empathy. This is because service robots in hospitality often assume frontline, interactive roles where trust is not only based on functional outcomes but also on how well the robot supports human-centered service experiences (Mende et al., 2019; Simon et al., 2020). Unlike manufacturing or logistics contexts where trust is primarily grounded in task efficiency and technical precision, hospitality settings require trust in the robot’s ability to facilitate positive social interactions and uphold service quality expectations.

Understanding these broader influences on trust is essential for fostering effective human–robot collaboration in hospitality workplaces, where both functional performance and the robot’s ability to support human-centered service experiences are critical to service success. Although Hancock et al. (2011, 2021) provide a robust three-dimensional trust model encompassing human, robot, and environmental factors, their meta-analyses largely omitted service robots in hospitality contexts. Hospitality environments significantly differ from other settings due to employees’ constant interpersonal interactions and high emotional labor demands. These unique contextual characteristics likely influence trust differently, necessitating tailored empirical investigation. Thus, the present study addresses this gap by empirically validating the three-dimensional trust framework within hospitality frontline service, exploring new combinations of trust antecedents, particularly human attitudes, replacement concerns, workload perceptions, robot anthropomorphism, and organizational expectations. Figure 1 outlines key antecedents of trust in HRI within the hospitality context, categorized into three broad dimensions: human-related, robot-related, and organization-related factors. The three-dimensional framework was adapted from foundational work by Hancock et al. (2011, 2021) and Schaefer et al. (2016). However, the specific variables within each dimension were identified a priori based on a comprehensive literature review to reflect the hospitality context and research focus of this study.

Human-related factors refer to individual attributes of hospitality employees that influence their perceptions of and interactions with service robots. In this study, these include perceived workload change, replacement concern, attitudes toward robots, comfort with robots, technology competence, and demographic characteristics such as age and gender. These personal variables are modeled as independent variables, reflecting employees’ predisposition toward robotic systems and their general openness to technology in the workplace.

Robot-related factors refer to the perceived characteristics of the robot itself that shape employees’ trust and engagement. These include attributes such as the robot’s reliability, functionality, appearance, anthropomorphic features, and the transparency of its actions. In HRI research, robot-related factors are frequently examined due to their direct influence on users’ perceptions of the robot’s competence, predictability, and social presence. Attributes like robot performance and anthropomorphism (the extent to which robots exhibit human-like features in appearance or behavior) are particularly relevant in service contexts where frontline interactions require both functional effectiveness and an ability to engage socially with humans.

Organizational-related factors represent workplace-level influences that affect how employees perceive and work with service robots. These encompass leadership expectations, organizational culture, training support, policies, and the broader workforce context in which human–robot interactions occur. In the context of service robot adoption in the hospitality sector, organizational-related factors have received relatively limited attention compared to robot-related or human-related factors, despite their potential significance in shaping trust dynamics. For this study, we focus specifically on management expectations and perceived task relevance, selected for their theoretical relevance and practical measurability within the scope of our research. These organizational elements may significantly influence employees’ trust by shaping perceptions of institutional support and task coordination.

By organizing these antecedents into three dimensions and adopting an exploratory approach to empirically examine the relative influence of human-, robot-, and organizational-related factors on employees’ trust in service robots, this study provides a more holistic framework for understanding employee trust in the hospitality context. The following sections examine each category in greater detail, offering insights into how these factors interact to influence trust in service robots.

2.2.1. Human-Related Factors

Workload Change: When service robots are perceived as reducing human workload (i.e., taking over repetitive or tedious tasks), employees tend to view them more favorably and are more likely to trust their value as team members (Hopko et al., 2022; Pan et al., 2025; Qin et al., 2025; Xu et al., 2023). For instance, recent studies of hotel staff found that employees anticipated robots would reduce their physical and mental workload and improve work efficiency, leading to positive attitudes toward working with robots (Kim, 2023; Paluch et al., 2022). On the other hand, when robots introduce new responsibilities such as training, maintenance, supervision, they can significantly increase employees’ workload, leading to frustration and reduced trust. Studies in hotel settings have found that excessive workload due to robot-related tasks (e.g., managing malfunctions, teaching coworkers how to use robots, assisting guests with robotic interfaces) triggers negative emotional arousal and employee resistance to continued collaboration (Almokdad & Lee, 2024; Hopko et al., 2022; Liu et al., 2024; Osawa et al., 2017).

Replacement Concern: This refers to employees’ fear of losing their jobs to robots, a particularly salient issue in the hospitality industry. This concern has been studied under terms such as “robot-induced unemployment” or service robot risk awareness, both referring to perceived job insecurity caused by the adoption of service robots (Parvez et al., 2022; Yu et al., 2022). Recent studies have shown that perceived job insecurity significantly affected job engagement and organizational commitment, often leading to increased burnout or turnover intentions (Horpynich et al., 2025; Kong et al., 2021; Koo et al., 2021). Hotel frontline employees often express concern that adopting service robots poses a direct threat to their job security (Brougham & Haar, 2018; Reis et al., 2020; Rosete et al., 2020). Studies report that employees who perceive robots as competitors for their job tend to exhibit more negative attitudes and are less likely to trust or collaborate with robots (Ivanov et al., 2020; V. N. Lu et al., 2020; Wirtz et al., 2018). Prior research highlights an inverse relationship between job security and openness to robots—employees with greater job insecurity (e.g., fear of job loss) are significantly more cautious and resistant, while those who feel secure in their positions are more accepting (Arslan et al., 2022; Leung et al., 2023; Nam et al., 2021). Beyond the hospitality sector, individuals who fear robots, AI, and technology they do not understand—“technophobes”—are more likely than non-technophobes to report having anxiety-related mental issues and to fear unemployment and financial insecurity (McClure, 2018). Granulo et al. (2019) found that people generally react negatively to the idea of being replaced by robots, which can erode trust in robotic systems.

Attitude Toward Robots: An individual’s attitude toward robots significantly influences trust in specific HRI scenarios. In the hospitality context, this factor is not always measured directly, but emerging evidence suggests it plays a role comparable to other domains. Researchers have categorized individuals as “robophiles” and “robophobes” in tourism and hospitality studies (Jin, 2024; Kazandzhieva & Filipova, 2019). A robophile, someone with a positive attitude toward robots, is more likely to trust and accept service robots readily. Conversely, a robophobe, who feels uncomfortable with or threatened by robots, approaches service robots with low initial trust. Studies in tourism have found that individuals with a higher propensity to trust technology also report higher trust in service robots (Ivanov et al., 2018; Tussyadiah et al., 2020). In practice, an employee or guest who believes that “robots can be helpful” is more likely to place immediate trust in a concierge robot. In contrast, someone with a skeptical view of robots will need more evidence of the robot’s reliability and usefulness before developing trust.

Comfort with Robots: Comfort with robots specifically refers to the degree to which individuals feel at ease, as a subjective emotional or psychological ease, when interacting with robotic technologies. Comfort with robots has been shown to significantly influence use intention or acceptance in robotic technologies (Becker et al., 2023; Tussyadiah et al., 2020). Recent studies suggest that trust functions as a mediator between beliefs about robots and intention to use service robots (Kraus et al., 2024). In hospitality settings, hotel employees who have prior exposure to robots or receive relevant training are more likely to feel comfortable during service interactions (Paluch et al., 2022; Xu et al., 2023). Taken together, these studies indicate that greater comfort with robots may lead to increased trust among employees.

Technology Competence: An individual’s technology competence, defined as their skill level and familiarity with modern technology, can significantly influence their trust in service robots. Although this factor has not been deeply explored in hospitality research, related concepts such as technology readiness have been discussed in the general technology acceptance contexts (Blut et al., 2021; Walczuch et al., 2007). In the hospitality context, Ma et al. (2022) found that customers with higher technology readiness were more likely to feel comfortable with and positively evaluate robot-assisted restaurant service experiences, suggesting that readiness can enhance perceived usefulness and acceptance. In hospitality settings, employees who possess strong technical skills or prior experience with automation may be better equipped to understand how service robots function, which could facilitate a higher level of trust. In contrast, those with lower technological competence might experience uncertainty or hesitation, making them more likely to approach robots with initial skepticism. Thus, technology competence may play a critical role in shaping employees’ trust in service robots.

Age: Hospitality research suggests that employee perceptions and attitudes toward service robots or technology adoption may vary by age (Romero & Lado, 2021; Vayghan et al., 2023). In general, younger generations such as Millennials and Gen Z tend to show more positive attitude and greater acceptance of emerging technologies and innovation. Younger individuals are typically more familiar with and comfortable using technology, which contributes to their quicker trust and acceptance of service robots (Ayyildiz et al., 2022; Binesh & Baloglu, 2023). In contrast, older adults often require more evidence of a robot’s reliability before developing trust. Su et al. (2022) found that elderly hotel guests prioritize emotional attributes such as empathy when evaluating a robot’s trustworthiness. Similarly, Tussyadiah et al. (2020) noted that negative attitudes toward technology, more prevalent among older users, can hinder the development of trust.

Gender: Numerous studies have explored gender as an important factor in human–robot interaction (Firmino De Souza et al., 2025). Gender differences in the context of service robots have been investigated from multiple perspectives, including preference, trust, acceptance, collaboration, and role expectations (Li et al., 2024; Pitardi et al., 2023; Roesler et al., 2022; Winkle et al., 2023). However, findings across studies remain mixed. Some research reports no significant gender-based differences in attitudes toward service robots (Naneva et al., 2020), while others suggest that men and women may interact with and respond to service robots differently depending on the context (Gallimore et al., 2019; Ivanov et al., 2018). For example, Wagner-Hartl et al. (2022) found that woman assessed collaboration with robots more positively than men, and younger women expressed greater trust in service robots more than their male counterparts. Conversely, K.-H. Lee and Yen (2023) reported that males tend to show a stronger preference for robotic services than females. Despite these varied findings, it is evident that gender remains a significant factor that needs to be considered in research on human–robot interaction in general, and employee–robot collaboration in particular.

2.2.2. Robot-Related Factors

Robot Performance: Out of a wide range of antecedents that influence trust in HRI, robot performance factor represented by reliability, accuracy, and efficiency has proven to be the most influential factor in forming trust in robotics and automation (Hancock et al., 2011, 2021; Kim, 2023; L. Lu et al., 2019; Schaefer et al., 2016). Research indicates that trust in automation and robotics hinges on factors such as how reliable and predictable the system is and how transparent it is about what it’s doing (Chiou & Lee, 2023; Hancock et al., 2011, 2021; Langer & Landers, 2021; Schmidt et al., 2020). For example, in military contexts, trust often depends on the consistent performance of automated systems (Schaefer et al., 2019; Troath, 2025). In healthcare, trust in robotic assistants is closely linked to their ability to perform tasks accurately and reliably (Ghazali et al., 2018; Kyrarini et al., 2021; Nambiappan et al., 2022). The ability to complete the task accurately for the employee could affect their ability to trust and accept them in the workplace (Emaminejad et al., 2024).

Anthropomorphism: Anthropomorphism has received considerable attention as a key antecedent of human–robot trust and use intention. The positive effect of anthropomorphic features on trust is well-supported in the literature (Blut et al., 2021; Qin et al., 2025; Van Pinxteren et al., 2019). Increased trust is partly attributed to anthropomorphism enhancing perceptions of a robot’s competence and social presence. Anthropomorphic design also improves a robot’s likability—its “social attraction”—which helps customers feel more comfortable and fosters trust (Christoforakos et al., 2021; Van Doorn et al., 2025; Yoganathan et al., 2021). While most hospitality research has focused on customers’ perceptions, reactions, and satisfaction with anthropomorphic robots (Li et al., 2024; Qin et al., 2025; So et al., 2024), much less is known about how hospitality employees perceive such robots as colleagues or whether anthropomorphic features influence employee trust.

2.2.3. Organization-Related Factors

Management Expectation: Management expectation can be broadly defined as employees’ perception of the extent to which their supervisors or organizational leadership anticipate their acceptance of and effective collaboration with service robots. In that sense, employees may be more inclined to trust robots if management actively promotes using robots and demonstrates confidence in them. Van Looy (2022) found that employees are more likely to develop trust in intelligent robots when management reinforces corporate values, clearly communicates the benefits of robot adoption, and expresses confidence in their implementation. As many studies emphasize the organization’s role in shaping the psychological climate surrounding service robot integration (V. N. Lu et al., 2020; Paluch et al., 2022), it is important to examine whether employees’ perceptions of management expectations influence the development of trust in service robots.

Task Relevance: Perceived task relevance of service robots refers to the degree to which an employee considers the presence or use of service robots to be important and relevant for carrying out their job duties. Research suggests that higher perceived task relevance is associated with greater acceptance of service robots, particularly when robots perform routine or physically demanding tasks (Belanche et al., 2020; Ivanov et al., 2020). Grounded in task-technology fit theory, Belanche et al. (2020) found that when consumers perceive robots as relevant to the task, such as in repetitive or low-complexity service contexts, they are more likely to accept and appreciate their use. Similarly, Ivanov and Webster (Ivanov et al., 2020) demonstrated that willingness to pay for robot-delivered services increases when robots are seen as functionally appropriate for the task. While most task relevance studies have focused on customer-facing contexts, addressing functional appropriateness and satisfaction, the connection between task relevance and employee trust in service robots remains underexplored. From the employee’s perspective, robots that align with job duties can enhance perceived job fit, acceptance, and satisfaction (Bhargava et al., 2021). Therefore, it is worth investigating how perceived task relevance influences employees’ trust in service robots.

3. Materials and Methods

3.1. Study Sample and Data Collection

The study population targeted frontline hospitality employees aged 18 or older who were currently working in hotels or restaurants in the United States. To enhance clarity, we note that our definition of “frontline hospitality employees” was based on respondents’ self-reported daily, direct customer interaction, rather than formal job titles or organizational rank. This functional definition reflects the study’s emphasis on frontline service experience in human–robot interaction contexts. As such, respondents who identified as owners, executives, or managers were included in the analytic sample if they confirmed routine, direct interactions with customers. This approach is particularly relevant in small or operationally flexible hospitality settings where managerial staff often perform frontline duties. This research context was selected because of the growing use of service robots in U.S. hospitality settings, making it a relevant and practical setting to investigate employee trust in such technologies. To recruit participants, a convenience sampling method was employed using the Qualtrics national panel. Although convenience sampling inherently limits generalizability to the broader population, it was deemed appropriate given the exploratory nature of the study and the need to reach a specific population segment efficiently.

An online survey was administered using Qualtrics, yielding 333 usable responses collected during a two-week period from 7 December to 23 December 2022. These responses reflected the panel provider’s rigorous screening process, which excluded nonstandard responses exhibiting inconsistent patterns (e.g., identical answers for all items). Using Qualtrics branching logic, respondents were routed to one of three wording sets based on their robot-experience status (current, former, or none). Survey items referenced the appropriate timeframe or hypothetical framing (current use, past use, or prospective use), ensuring that evaluations of trust and related constructs were framed in a manner consistent with respondents’ experience. We also treated robot-experience status as a control in robustness checks. When comparing respondents with and without prior robot experience, we found no significant differences in their perceptions of service robots including trust, and all substantive results remained unchanged. To ensure clarity, service robots were defined at the beginning of the survey as self-movable agents that interact with humans and perform tasks with some level of autonomy without continuous human intervention, and this definition was presented to all respondents prior to answering the subsequent questions.

For the final analysis, we focused on 301 respondents who reported daily, direct interactions with customers, excluding those without such interactions. The sample size was considered adequate based on established guidelines for regression-based analysis and the number of predictors in the model, ensuring sufficient statistical power. According to Hair et al. (2018), the recommended ratio of observations to independent variables ranges from 15 to 20, indicating that 15–20 observations are required per variable to generalize the results of multiple regression analysis. Given that our model included 12 independent variables, a sample size between 180 and 240 was considered sufficient. Our final sample of 301 exceeds the recommended threshold. To further confirm adequacy, a prior power analysis using G*Power 3.1 was conducted with a power of 0.95, medium effect size of 0.15, and 12 variables (Faul et al., 2009). This analysis indicated that 184 responses were required to achieve the desired power. Together, these indices confirm that the sample size was sufficient. Data were analyzed using IBM SPSS Statistics 30. Multiple regression analysis with bootstrapping was employed to identify and assess the most significant and stable combination of human-, robot-, and organization-related factors that collectively predict employee trust in service robots.

For ethical rigor, the first page of the online survey presented an informed consent form describing the purpose of the study, voluntary participation, anonymity, and data usage. Participants could proceed only after affirming consent. Otherwise, the survey was designed to end automatically. The study received an exemption from Institutional Review Board (IRB) review at the corresponding author’s institution.

3.2. Measurement

To examine the direct relationships between the identified antecedents and trust in robots, this study employed a multiple regression model. Trust in robots was specified as the dependent variable, while eleven independent variables were included: robot performance, preference for human-like and machine-like robots, management expectation, task relevance, workload change, replacement concern, attitude toward robots, comfort with robots, technology competence, age, and gender. All variables, except for gender and age, were measured using a five-point Likert scale (perception-based), where 1 represented “strongly disagree” and 5 represented “strongly agree”. Workload change was measured on a scale ranging from “significantly reduced” and “significantly increased”.

The trust in robots scale consisted of two items adapted from S. S. Park et al. (2021), J. Lee and See (2004), and Gefen et al. (2003). Robot performance was measured using five items modified from L. Lu et al. (2019) and Kim (2023). Technology competence was assessed through three items adapted from Cheng and Guo (2021). Replacement concern was measured using three items based on Huang and Rust (2018). Attitude toward robots was evaluated with three items adapted from S. S. Park et al. (2021) and Curran and Meuter (2005). Task relevance was assessed using three items adapted from Zhang et al. (2022), while management expectation was measured with two items adapted from L. Lu et al. (2019). The remaining variables were developed with reference to multiple sources including Van Pinxteren et al. (2019), L. Lu et al. (2019), and Zhang et al. (2022). All items were carefully reviewed and modified to ensure alignment with the current research context.

4. Results

4.1. Descriptive Results

Table 1. Demographic Profile of the Study Sample (N = 301).

Variable	Category	Distribution
Gender	Male	208 (69.1%)
	Female	92 (30.6%)
	Other	1 (0.3%)
Age	18–29	110 (36.5%)
	30–39	160 (53.2%)
	40–49	22 (7.3%)
	50–59	7 (2.3%)
	60 and over	2 (0.7%)
Annual personal income	Less than $34,999	44 (14.6%)
	$35,000~$74,999	87 (28.9%)
	$75,000~$99,999	74 (24.6%)
	$100,000~$149,999	74 (24.6%)
	$150,000 or above	19 (6.5%)
	Prefer not to answer	3 (1.0%)
Education	Less than high school	1 (0.3%)
	High school	15 (5.0%)
	Associate degree	50 (16.6%)
	Bachelor’s degree	172 (57.1%)
	Graduate degree	60 (19.9%)
	Other	2 (0.7%)
	Prefer not to answer	1 (0.3%)
Ethnic Background	White/Caucasian	189 (62.8%)
	Hispanic or Latino	11 (3.7%)
	African American	75 (24.9%)
	Asian	3 (1.0%)
	Native Hawaiian or Pacific Islanders	1 (0.3%)
	Native American	11 (3.7%)
	Mixed Race	11 (3.7%)
Work Location (Top Five States)	1. California	79 (26.2%)
	2. New York	65 (21.6%)
	3. Texas	33 (11.0%)
	4. Illinois	21 (7.0%)
	5. Florida	20 (6.6%)
Work Location (Top Five Cities)	1. Los Angeles	67 (22.3%)
	2. New York City	34 (11.3%)
	3. Albany, NY	18 (6.0%)
	4. Dallas	17 (5.6%)
	5. Chicago	15 (5.0%)

provides an overview of the sample’s demographic characteristics. The majority of respondents identified as male (69.1%), aged between 30 and 39 years (53.2%), and predominantly White/Caucasian (62.8%). More than half of the respondents held bachelor’s degrees (57.1%), and the majority of respondents (53.5%) reported an annual personal income between $35,000 and $99,999. The majority of respondents work in California (26.2%), followed by New York (21.6%), Texas (11.0%), Illinois (7.0%), and Florida (6.6%), rounding out the top five states. Top five cities include Los Angeles (22.3%), New York City (11.3%), Albany (NY), Dallas, and Chicago.

Table 2. Professional Characteristics of the Study Sample (N = 301).

Variable	N	Category	Distribution
Directly worked with service robots	301	Currently working with service robots	100 (33.2%)
		Previously worked with service robots, not currently	30 (10.0%)
		Never worked with service robots	171 (56.8%)
Service business	301	Hotels	161 (53.5%)
		Restaurants	140 (46.5%)
Type of workplace	301	Multi-national chain hotel/restaurant	79 (26.2%)
		National/regional chain hotel/restaurant	80 (26.6%)
		Independent hotel/restaurant	142 (47.2%)
Employment status	301	Full-time	235 (78.1%)
		Part-time	64 (21.3%)
		Contract/Temporary	2 (0.7%)
Current job position	301	Owner	14 (4.7%)
		Executive	46 (15.3%)
		Manager	133 (44.2%)
		Non-managerial employee	136 (45.2%)
Work Length in current position	293		4.49 years
Work Length in current company	301	Less than 6 months	3 (1.0%)
		6 months–12 months	5 (1.7%)
		1–2 years	68 (22.6%)
		3–5 years	150 (49.8%)
		6 years and over	75 (24.9%)
Work length in the hospitality industry	301	Less than a year	7 (2.3%)
		1–3 years	73 (24.3%)
		4–6 years	129 (42.9%)
		7–9 years	66 (21.90%)
		10 years and over	26 (8.6%)
Business size	301	Micro-sized business (less than 10 employees)	4 (1.3%)
		Small-sized business (10–49 employees)	127 (42.2%)
		Medium business (50–249 employees)	148 (49.2%)
		Large-sized business (more than 250 employees)	22 (7.3%)
Experiencing labor shortage in current workplace	301	Yes	143 (47.5%)
		No	158 (52.5%)

provides an overview of the sample’s professional characteristics. In terms of work experience, 56.8% of participants reported having never worked with service robots. 33.2% reported they are currently working with robots, while 10.0% had previously worked with service robots but were not currently using them. The sample was evenly distributed across hotel (53.5%) and restaurant (46.5%) employment settings. Regarding business size, most respondents worked in small to medium-sized businesses, with 42.2% employed in small businesses (10–49 employees) and 49.2% in medium-sized businesses (50–249 employees). A smaller proportion worked in large businesses (7.3%), while 1.3% were employed in micro-sized businesses (fewer than 10 employees). In terms of employment status, 78.1% of respondents were employed full-time, while 21.3% were working part-time, and less than 1% held contract or temporary positions. It was also observed that many participants held manager positions (44.2%) or non-managerial roles (45.2%). Executives comprised 15.3% of the sample, while 4.7% were business owners. Most respondents work in independent hotels/restaurants (47.2%). The average length of time respondents have worked in their current position is 4.5 years. Half of the respondents have been with the company for 3–5 years (49.8%), and 42.9% of the respondents have been in the hospitality field for 4–6 years. A total of 47.5% of the respondents reported that their workplace was experiencing labor shortage.

Table 3. Descriptive Statistics of Variables for Multiple Regression Analysis.

Variable 1	Statement	Mean (S.D.)	Cronbach’s α
Trust 2 (4.17) 3	Service robots are trustworthy.	4.29 (0.811)	0.786 (r = 0.667)
	I trust service robots to perform without any error.	4.05 (1.033)
Robot-Related Factors
Robot Performance (4.25) 3	Service robots provide consistent and reliable service to customers.	4.23 (0.789)	0.844
	Service robots provide accurate service to customers.	4.25 (0.822)
	Service robots reduce mistakes or errors.	4.17 (0.935)
	Service robots help human workers concentrate on more engaging tasks by taking over the physically demanding and repetitive tasks.	4.28 (0.771)
	Service robots increase efficiency of workflows as they don’t need breaks or experience fatigue.	4.33 (0.806)
Human-looking robot preference	How much do you prefer working with human-looking robots?	4.34 (0.900)	N/A
Machine-looking robot preference	How much do you prefer working with machine-looking robots?	4.07 (1.050)	N/A
Human-Related Factors
Workload Change 4	If your workplace hired service robots, how do you think your workload would change?	2.4 (1.283)	N/A
Replacement Concerns (3.43) 3	Service robots will replace my position in the future.	3.43 (1.341)	0.879
	Service robots will replace what I do now in my job.	3.28 (1.343)
	Service robots will replace human employees in this industry.	3.58 (1.300)
Attitude toward Robots (4.23) 3	I like working with service robots.	4.24 (0.822)	0.870
	Working with service robots is a pleasant experience for me.	4.19 (0.848)
	Working with service robots is a positive experience for me.	4.27 (0.790)
Comfort with Robots	Working with service robots makes me feel: Uncomfortable/comfortable	4.11 (1.156)	N/A
Technology Competence (4.22) 3	Other people come to me for advice on new technology.	4.13 (0.783)	0.812
	I can usually figure out high-tech products without help.	4.26 (0.778)
	I understand how most technology works.	4.27 (0.803)
Organization-Related Factors
Management expectation (4.22) 3	When service robots are employed, the company would like me to collaborate with service robots.	4.24 (0.751)	0.684 (r = 0.523)
	It will give a good impression to my supervisor if I work with service robots.	4.20 (0.850)
Task Relevance (4.02) 3	Working with service robots is important to my job.	3.99 (0.983)	0.827 (r = 0.705)
	Working with service robots is relevant to my job.	4.05 (0.953)

¹ All variables were measured by a 5-point Likert scale. ² Dependent Variable. ³ The average of the items within each factor. ⁴ It was measured from (1) significantly reduced to (5) significantly increased.

presents the statements and descriptive statistics of variables used in the multiple regression analysis. To assess the internal consistency of items within the variables—trust, robot performance, replacement concerns, attitude towards robots, technology competence, management expectation, and task relevance, we conducted reliability tests. Cronbach’s alpha values indicate that most variables exhibit an acceptable level of internal consistency, with alpha coefficients exceeding the commonly recommended threshold of 0.70 (Nunnally & Bernstein, 1994).

For variables measured with only two items, reliability was also assessed using inter-item correlation, which is considered a more appropriate measure than Cronbach’s Alpha in such cases (Eisinga et al., 2013). It is important to note that alpha tends to be lower when the number of items is small, and inter-item correlation provides a more reliable indicator of internal consistency in this context. The inter-item correlation coefficients (r) for all variables with two items fall within the acceptable range of 0.30 to 0.70 for measuring a single construct (Clark & Watson, 2016). The correlation coefficients between the two statements indicate a moderate positive correlation, suggesting that, while the two statements are related, they are not redundant.

4.2. Model Fit and Predictive Utility

A bootstrap multiple regression analysis was conducted to assess the ability of the combined human, robot, and organizational factors to collectively predict employee trust in service robots. Since our data was obtained through convenience sampling, we assessed the assumption of normality of residuals using both graphical and statistical tests. A Q-Q plot and histogram of the residuals indicated that they approximated a normal distribution, supporting the normality assumption. However, the Kolmogorov–Smirnov test yielded a significant result (p < 0.001), suggesting a deviation from normality. Given this concern, a bootstrapped regression analysis was conducted to provide more robust estimates of the regression coefficients.

As presented in

Table 4. Results of Bootstrap Multiple Regression Analysis.

	Bootstrap Regression Coefficients				Tolerance	VIF	BCa 95% Confidence Interval
Independent Variables	B	Bias	S.E	Sig.			Lower	Upper
Constant	0.188	0.003	0.316	0.544	-	-	−0.425	0.889
Robot Performance **	0.280	−0.002	0.086	<0.001	0.392	2.551	0.121	0.450
Human-looking Robot Preference	−0.028	−0.002	0.044	0.500	0.657	1.522	−0.104	0.051
Machine-looking Robot Preference	0.069	−0.002	0.037	0.059	0.784	1.275	0.005	0.130
Workload Change *	−0.050	0.000	0.024	0.043	0.813	1.231	−0.098	−0.003
Replacement Concerns	0.012	0.000	0.027	0.657	0.769	1.300	−0.040	0.061
Attitude toward Robots **	0.373	0.001	0.092	<0.001	0.319	3.134	0.191	0.556
Comfort with Robots **	0.234	0.002	0.044	<0.001	0.602	1.660	0.144	0.325
Technology Competence	0.084	−0.001	0.055	0.129	0.684	1.461	−0.016	0.194
Management Expectation	0.031	−0.001	0.064	0.608	0.502	1.990	−0.098	0.148
Task Relevance	−0.014	0.002	0.052	0.782	0.476	2.102	−0.124	0.095
Age	−0.032	−0.002	0.039	0.410	0.920	1.087	−0.107	0.038
Gender *,a	−0.154	0.004	0.073	0.031	0.908	1.102	−0.300	−0.004
	R2 = 0.637; Adjusted R2 = 0.621
	F-value = 42.038 (12, 288), p < 0.001
	SSTotal = 218.694, PRESS = 88.028, RMSE = 0.51
	N = 301

^a A dummy variable was defined as follows: Female = 0, Male = 1. * significant at p < 0.05; ** significant at p < 0.001.

, the model incorporating all 12 predictor variables was statistically significant, F(12, 288) = 42.038, p < 0.001. The model yielded an R² of 0.637, indicating that the set of factors accounted for 63.7% of the variance in employee trust. Crucially, the Adjusted R² was 0.621. The small difference between R² and Adjusted R² suggests the model is robust and that the inclusion of the 12 predictors is highly efficient, minimizing the loss of predictive power. To establish the generalizability required for an exploratory predictive analysis, the model’s out-of-sample performance was assessed using two key metrics. The Root Mean Square Error (RMSE) was calculated to be 0.51. Given that employee trust was measured on a 5-point scale, this low RMSE indicates a highly accurate model, with the average prediction error being only about half a scale point. Furthermore, the Predicted Residual Sum Of Squares (PRESS) statistic was 88.028. Since this value is significantly lower than the Total Sum of Squares for the dependent variable (218.694), it strongly confirms the model’s robustness and strong cross-validated predictive capability with minimal risk of overfitting. This finding reinforces the model’s reliability and practical utility for forecasting employee trust. Evaluation of regression model assumptions confirmed the model’s statistical appropriateness. The Durbin–Watson statistic was 2.094, which is close to the ideal value of 2.0, indicating that the assumption of independent errors (i.e., lack of autocorrelation) was met. Levene’s Test of Homogeneity of Variance (F = 0.684, p = 505) confirmed no significant differences in variances, supporting the assumption of homoscedasticity.

4.3. Individual Predictor Stability and Significance

The results from the bootstrap multiple regression analysis revealed that five of the twelve factors were statistically significant predictors of employee trust in service robots (See Table 4 for full coefficients). Attitude toward robots (B = 0.373, p < 0.001) was the strongest positive predictor. Robot performance (B = 0.280, p < 0.001) and comfort with robots (B = 0.234, p < 0.001) were also highly significant positive predictors of employee trust in service robots. In contrast, organizational factors (management expectation, task relevance) and concerns about robot replacement showed minimal and statistically non-significant effects, contradicting common assumptions in the existing literature. Interestingly, workload increase (B = −0.050, p < 0.05) negatively affected trust, suggesting that predicted trust decreases as the perceived workload changes due to the robot. Gender was also a significant negative predictor, indicating a difference in predicted trust based on gender (B = −0.154, p < 0.05). The use of bootstrapping (N = 1000) provided evidence of the stability and robustness of the parameter estimates. For all five significant predictors, the Bias-Corrected and accelerated (BCa) 95% confidence intervals did not contain zero, confirming that the predictive effects of these factors are stable and reliable across resampling of the data.

Preference for machine-looking robots (B = 0.069, p = 0.059) showed a positive and marginally significant effect on employee trust, indicating a trend where employees who prefer more mechanical, utilitarian robot designs tend to report higher trust in service robots. While the effect did not reach conventional levels of statistical significance (p < 0.05), the direction and magnitude of this relationship suggest that design preferences rooted in functional aesthetics, rather than human-like appearance, may positively influence trust perceptions in hospitality settings. Conversely, preference for human-looking robots (B = −0.028, p = 0.500) exhibited no significant association with trust, indicating that a preference for anthropomorphic designs does not appear to influence employees’ trust in service robots in this context. Factors such as replacement concern, technology competence, management expectation, and task relevance did not emerge as significant predictors of employee trust in this model, suggesting their influence is either negligible or absorbed by other factors in this specific hospitality context.

The assumption of non-multicollinearity was assessed to ensure the stability and reliability of the parameter estimates, which is crucial for a predictive model. The Variance Inflation Factor (VIF) values for all 12 predictors ranged from a minimum of 1.087 to a maximum of 3.134, which are well below the commonly accepted threshold of 5 (Hair et al., 2018), indicating that multicollinearity is not a concern in this model. The lowest VIF was observed for Age (VIF = 1.087), and the highest was observed for Attitude Toward Robots (VIF = 3.134). These results suggest that the independent variables are sufficiently independent of one another, and the regression estimates are not adversely affected by multicollinearity. This confirms the high internal consistency and stability of the predictive model’s coefficients.

5. Discussion

5.1. Theoretical Implications

The overall analysis confirmed that the combination of human, robot, and organizational factors provides a strong and stable foundation for predicting employee trust in service robots. The collective group of 12 factors was highly effective in modeling trust, and our selection of factors from the literature is highly efficient. Our model is not overfit to the current sample and is robust enough to be applied to other similar hospitality settings, securing its utility for broad forecasting. This study advances the theoretical understanding of trust in human–robot interaction (HRI) within the hospitality context by underscoring the influential role of human-related factors in shaping employees’ trust in service robots. While earlier research predominantly emphasized robot-related attributes (e.g., performance and reliability) as the primary drivers of trust (Blut et al., 2021; Hancock et al., 2011), more recent meta-analytic work (Hancock et al., 2021) has demonstrated the growing importance of human-related and contextual factors. Our findings strongly align with and reinforce this evolving perspective that human- and environmental-related variables have gained prominence, while robot-related factors have shown relatively diminished influence over time. Employee attitudes toward robots emerged as the most significant determinant of trust, illustrating that general dispositions and affective orientations toward robotic technologies are central in trust formation (S. Park, 2020; Xu et al., 2023). Similarly, comfort with robots was a significant predictor of trust, further corroborating prior research that highlights the role of user familiarity and emotional ease in HRI (Cain et al., 2019; Christ-Brendemühl, 2022). Unlike Hancock et al. (2011), however, our study found that organizational factors (parallel to Hancock et al.’s environmental dimension) did not show significant effects.

Interestingly, this study challenges some key prevailing assumptions in the trust literature. Contrary to earlier work emphasizing the threat of job replacement and the enabling role of technology competence in trust formation (Ma et al., 2022; Venkatesh et al., 2003), neither replacement concerns nor technology competence significantly predicted employees’ trust in service robots in this study. It is noteworthy, however, that the mean score for replacement concern (M = 3.43) indicates that employees are still mindful of potential job displacement. Rather than suggesting an absence of concern, these findings imply that replacement anxiety, while present, may not directly influence employees’ trust in robots. Similarly, general technology competence did not emerge as a significant factor, suggesting that trust in service robots may be more influenced by context-specific interactions than by overarching technological capabilities. These results highlight a shift where trust in robots is not necessarily tied to job security fears or general tech-savviness, but rather to perceptions of the robot’s functional role within the hospitality environment.

The non-significance of technology competence in this study may reflect widespread technological familiarity among contemporary hospitality employees, reducing its differentiating effect on trust formation. Moreover, general technology proficiency may not directly translate into trust towards service robots, given that trust building often involves more complex and context-specific relational dynamics. Employees might, therefore, rely more heavily on experiential cues or relational interactions, such as direct robot behaviors or peer experiences, rather than their technical proficiency alone. These findings underscore the importance of future research exploring the nuanced moderating roles, such as prior robot exposure, task types, and targeted training experiences.

Likewise, the absence of a significant relationship between age and trust in service robots challenges common assumptions that older workers are generally less receptive to automation (Binesh & Baloglu, 2023; Tussyadiah et al., 2020). Our results suggest that age may not exert a direct influence on trust. Instead, it may function as a mediator or moderator in trust formation. However, this interpretation warrants caution due to the demographic composition of our sample. Approximately 90% of participants were under the age of 40, with Millennials and Gen Z cohorts overwhelmingly represented. This lack of age diversity may have limited the variability needed to detect age-based differences in trust, thereby constraining the generalizability of our findings to older populations, underscoring the importance of further investigation with more age-balanced samples.

Another theoretical contribution lies in our finding that organizational factors, namely management expectation and task relevance, did not significantly predict trust. This result contrasts with prior studies that emphasize the importance of organizational support and leadership in shaping employees’ acceptance of new technologies. One possible explanation is that in the hospitality industry, service robot adoption is still in its early stages. At this stage, employees may rely more on their personal attitudes and direct experiences with robots than on organizational cues. In addition, cultural and contextual contingencies such as the limited institutionalization of robot-related policies or the variability of management practices across firms may delay the influence of organizational factors on trust formation. The results suggest that organizational influences may become more salient as robot adoption becomes more mature and standardized within the industry. Future research should therefore examine longitudinal or cross-cultural contexts to better capture when and how organizational factors translate into employee trust.

5.2. Practical Implications

The findings of the study carry several actionable insights for hospitality organizations integrating service robots into their operations. First and foremost, building employee trust requires more than simply improving robot functionality. Given the dominant influence of human-related factors, particularly attitude and comfort, hospitality managers should prioritize efforts that foster positive employee perceptions of and emotional ease toward service robots. Specifically, targeted training programs, interactive robot demonstrations, and clear communication strategies emphasizing robots’ roles as collaborative partners rather than potential replacements are recommended. Additionally, considering the minimal direct influence of organizational factors such as managerial expectations, hospitality organizations should shift from top-down, policy-centric approaches toward fostering relational trust through experiential initiatives, including regular robot-interaction training and informal robot familiarization sessions. While recent research indicates that employees are increasingly beginning to view robots as collaborative tools rather than replacements (Li et al., 2024; Yin et al., 2024), targeted interventions that demystify robotic systems and explicitly frame them as supportive tools can further reinforce this perspective, helping to build familiarity and reduce any lingering skepticism.

Moreover, since perceived workload increase negatively affects trust, hospitality managers should clearly communicate how tasks will be redistributed and ensure robot integration does not add stress or ambiguity to employees’ roles. Transparent change management processes addressing these concerns can help mitigate trust erosion and facilitate smoother transitions. Additionally, considering gender-based differences identified in this study, hospitality organizations should design tailored trust-building interventions to effectively address varying employee concerns and expectations. For example, male employees, who exhibited lower trust in this study, may benefit from trust-building interventions that directly address concerns related to reliability, control, or role clarity. Since female employees reported higher trust levels in this study and prior research indicates emotional responses may more strongly influence women’s trust in robots (Gallimore et al., 2019), strategies emphasizing empathy, social design, and relational qualities may be particularly effective for this group.

Finally, anthropomorphism in robot design does not universally enhance trust, especially among employees. While consumers may favor social presence in humanoid robots, employees might prioritize functionality, clarity of role, and predictability, which could be more strongly associated with machine-like appearances. The marginal significance for machine-looking preference suggests that this is a relevant area for further exploration, especially in job roles where trust in robot performance is critical. Consistent with this, Wang et al. (Wang et al., 2024) found that employees experience greater enjoyment and less stress when working with non-humanoid robots, indicating that human-like features may not always be beneficial. Developers and hospitality managers should therefore consider context-specific design principles that prioritize functional design and psychological comfort over human likeness, particularly when employee trust and task performance are at stake.

6. Conclusions

The overall analysis of this study confirmed that the integrated model encompassing human, robot, and organizational factors was statistically significant, offering a strong and stable framework for predicting employee trust in service robots. The study also revealed that attitudes toward robots, perceived comfort, and evaluations of robot performance as the key predictors of employee trust. In contrast, organizational variables demonstrated limited or nonsignificant effects. The findings challenge assumptions about the significance of technology competence and job replacement concerns, suggesting that trust is shaped more by experiential and relational dynamics than by general technological familiarity. Beyond theoretical contributions, these results offer practical insights for hospitality managers by emphasizing the need to cultivate positive employee attitudes, provide experiential training to enhance comfort with robotic technologies, and tailor trust-building strategies to diverse workforce groups. Together, these findings reinforce the significance of employee-centered approaches in fostering sustainable and effective service robot integration within hospitality operations.

Limitations and Future Research

This study has several limitations that pave the way for future exploration. First, the cross-sectional design of this study limits the ability to infer causal relationships, as data were collected at a single point in time. While this study focused on dispositional trust (an individual’s general tendency to trust robots), it did not examine situational trust (trust in a specific task or context) or learned trust (trust developed over time through repeated interactions). Given that trust in HRI is inherently dynamic, these forms of trust may evolve as employees gain direct experience working with service robots, potentially influencing their perceptions and behaviors in ways that differ from initial dispositional trust. Future research should adopt longitudinal designs to capture how trust fluctuates before, during, and after interactions with robots, allowing for a more comprehensive understanding of trust development over time. Additionally, examining how task-specific factors (e.g., complexity of robot-assisted tasks, service context) shape situational trust would provide valuable insights into trust calibration processes in hospitality environments, where trust must be continually built and maintained through experience.

Second, this study relied exclusively on self-reported measures, which may not fully reflect actual behavior in workplace settings. While self-perceptions offer valuable insights, they can be affected by biases such as social desirability or limited self-awareness. Future research should consider incorporating behavioral or experimental approaches, such as trust games or task-based simulations, to validate self-reported trust and reveal discrepancies between perceptions and observable behaviors (Kohn et al., 2021).

Third, our measurement of trust focused solely on general trust and did not capture other potential aspects such as task-specific trust, cognitive trust, or integrity-based trust. Given that some studies conceptualize trust as a multidimensional construct (Malle & Ullman, 2021; S. Park, 2020), future research should adopt more comprehensive trust measures and assess trust at multiple time points to better reflect its dynamic and evolving nature.

Fourth, the study’s generalizability may be constrained by the geographic distribution of respondents. Although our survey included participants from 38 U.S. states, a substantial proportion were concentrated in five states. This uneven distribution may partly reflect the actual diffusion of service robots across the United States, as adoption is not uniform, and certain states and metropolitan areas have more actively implemented service robots in hospitality settings. At the same time, because our sample included employees both with and without direct experience working with service robots, the study captures a range of perspectives that mitigates some of the concerns associated with geographic disproportionality. Nevertheless, the potential overrepresentation of employees from adoption-leading regions may limit generalizability. While employees’ perceptions of service robots are not expected to vary substantially across states in the United States, future research should include more balanced samples across regions to strengthen external validity.

Fifth, the study’s generalizability is further limited by the sampling frame, which included only hospitality employees in the United States. As prior research suggests that cultural context plays a significant role in shaping attitudes toward automation and trust in technology (Chi et al., 2023; Li et al., 2024), future studies should examine how factors such as national cultural values, prior exposure to service robots, and industry norms may shape trust formation differently across global settings. Future research should incorporate cross-cultural comparisons to explore how cultural differences affect trust in service robots, thus extending the generalizability of findings across diverse contexts.

Finally, this study focused solely on identifying antecedents of trust, rather than exploring how trust functions within broader model of effective human–robot collaboration. Prior literature in HRI and technology adoption has examined trust as a mediator or moderator in broader models of human–robot collaboration, influencing outcomes such as technology acceptance, usage, and resistance (Kraus et al., 2024). Building on this foundation, future research should explore how trust interacts with other psychological and organizational constructs to advance a more integrative understanding of trust’s role in the effective implementation of service robots in hospitality and related sectors. Moreover, the data were collected in December 2022. While the theory-driven relationships tested here are unlikely to have changed considerably in the short term, contemporaneous generalizability may be affected by ongoing diffusion of service robots. Future research should replicate these analyses with newer and/or longitudinal samples to assess progressive stability.