Enhancing Cabin Crew Education Effectiveness through VR-based Training

Definitions of the functional characteristics of virtual reality (VR) have been discussed from two perspectives: technical mechanisms and user experience (Steuer, 1992; Slater and Wilbur, 1997). Slater and Wilbur defined VR as “a technology that enables users to become immersed in a digitally generated environment and perceive it as if they were physically present in that space” (Slater and Wilbur, 1997). This definition subsequently formed the conceptual foundation for VR research (Slater and Wilbur, 1997). In this study, VR Features are defined as the functional attributes provided by a virtual reality learning environment that enable immersion, interactivity, system responsiveness, learner control, and active engagement (Steuer, 1992; Radianti et al., 2020). Steuer defined the essential characteristics of VR as the combination of immersion and interactivity, explaining that “an environment that responds immediately to the user's actions enhances the sense of presence” (Steuer, 1992). This emphasizes that the core of the learning experience lies in the iterative process of immediate feedback between the user's intent and the environment, going beyond simple screen-based interaction.In addition, like cabin service training, in practice-based industrial education fields, realistic simulation-a key functional characteristic of VR-guides learners to perceive the virtual environment not merely as a visual medium but as an "actual experiential space." This shift in perception promotes learner immersion (Merchant et al., 2014).

In addition, these VR features are evaluated as key technological elements contributing to the economic and environmental sustainability of aviation education systems by enabling digital training infrastructure that reduces dependence on physical model training labs and decreases resource and energy consumption (Balcerzak, 2018; Karaman et al., 2018.)

Earlier studies have empirically demonstrated that the VR features consistently exert a positive influence on learners' sense of presence. Radianti et al.(2020) conducted a comprehensive review of 43 VR-based learning studies in higher education, revealing that the functional characteristics of immersive VR consistently tend to enhance learners' sense of presence (Radianti et al., 2020).

This suggests that VR is not merely a technology but a key factor determining the quality of the learning experience. Wirth et al. reported that specific functional elements of VR, such as “perspective design, spatial interaction, and system responsiveness,” significantly enhance learners' sense of presence. These findings were confirmed across diverse higher education fields, including science, medicine, and engineering (Wirth et al., 2007; Radianti et al., 2020). Furthermore, in the context of emergency response training, Kober et al.(2012) reported that “realistic environmental design and interactive scenarios” in VR training simulating fire evacuation situations increased learners' presence (Kober et al., 2012). These findings support the importance of immersive training for real emergency situations, given the nature of aviation service education. In the field of safety education, Makransky and Mayer proposed that “the multisensory immersive experience provided by VR induces a higher level of presence compared to traditional training methods.” They further stated that this aligns with neuroscientific evidence showing that combining multisensory stimuli (visual, auditory, tactile) activates the learner's nervous system and promotes event memory formation (Makransky and Mayer, 2022). Furthermore, a high level of presence that enables safe, repeated exposure to high-risk situations is also discussed as a means of human sustainability that can contribute to reducing actual accident rates and strengthening safety culture (Karaman et al., 2018; Talley and Howard, 2025).

Learners' sense of control and active engagement are core elements emphasized in constructivist learning theory. Empirical evidence has accumulated on how the interactivity and manipulability of VR environments enhance these elements. As proposed in Steuer's early research, the principle that “a structure allowing users to intervene in the environment and immediately observe consequences” promotes learners' active engagement has been reconfirmed in numerous subsequent studies (Steuer, 1992). Wirth et al. reported that the “manipulative freedom and immediate feedback” provided in VR environments enhance learners' active exploration and perceived control over learning (Wirth et al., 2007). Research in vocational education contexts demonstrates that these effects are even more pronounced. Merchant et al. reported that “VR-based learning significantly enhances learners' perceived control and learning engagement” in vocational education settings (Merchant et al., 2014). Chittaro and Ranon suggested in their VR learning environment study that “environmental designs enabling learners to perform tasks independently promote self-directed learning behaviors” (Chittaro and Ranon, 2007). In Groenestijn et al.'s study on embodied interaction, they reported that “VR environments supporting physical and cognitive interaction increase learners’ active learning behaviors” (Groenestijn et al., 2016). This suggests that learners' physical movements (e.g., cabin navigation, safety equipment operation) could be a crucial mechanism linking to learning outcomes in the aviation service training context. Furthermore, self-directed learning experiences based on high perceived control and active engagement are presented as a key factor in enhancing the human sustainability of aviation workforce development by promoting long-term job competency development and career continuity (Talley and Howard, 2025; Ochs et al., 2022).

In educational psychology research examining the roles of learning motivation and emotion, enjoyment is recognized as a key emotional variable that enhances learners' sustained engagement and learning effectiveness (Pekrun, 2006). The relationship between the immersive experience provided by VR-based learning environments and enjoyment has been elucidated through the following prior studies. According to Makransky and Mayer's review, immersive experiences in VR-based learning environments function as “a key factor enhancing learners' enjoyment and positive emotions” (Makransky and Mayer, 2022). This can be interpreted as a psychological mechanism that deepens immersion in learning, going beyond a simple increase in positive affect. Through their analysis, de Freitas and Oliver explained that “presence in VR environments influences the learning experience by mediating learner enjoyment” (de Freitas and Oliver, 2006). Merchant et al. reported in their safety education study that “multisensory VR environments and gamification elements significantly enhance learners' interest and enjoyment” (Merchant et al., 2014). Makransky and Mayer presented that “immersive VR learning consistently improves learners' enjoyment and learning motivation compared to traditional educational methods” (Makransky and Mayer, 2022). Lucardie emphasized that in adult learning contexts, “realistic VR environments are factors that sustain learners' emotional engagement and enjoyment” (Lucardie, 2014). This enjoyment-based sustained engagement can reduce dropout rates and support long-term competency development in fields requiring repetitive learning, such as aviation services. Consequently, it contributes to human sustainability in terms of talent retention and organizational human capital management (Karaman et al., 2018; Talley and Howard, 2025).

The impact of presence on learning outcomes has been explained based on cognitive load theory and attention theory. Slater and Wilbur proposed that high presence “focuses learners' attention on learning tasks, enhancing cognitive processing efficiency” (Slater and Wilbur, 1997). Schrader and Bastiaens also reported that presence “enhances learners' task immersion and improves cognitive processing levels” (Schrader and Bastiaens, 2012). Slater et al., who directly measured learning performance improvements in VR environments, reported that “in VR environments inducing high presence, learners' comprehension and performance levels significantly improve” (Slater et al., 2006). Hartmann et al. further proposed in the field of affective learning that “presence exerts a particularly strong influence on affective learning outcomes such as empathy and attitude change” (Hartmann et al., 2016). In aviation safety and service training, presence can be understood as a mechanism that promotes the transfer of safety behaviors and customer response capabilities by more realistically reproducing actual emergency situations or service interactions. This ultimately enhances on-site safety and service quality, thereby strengthening the social sustainability of the aviation industry (Karaman et al., 2018; ATAG, 2022).

Constructivist learning theory and self-determination theory emphasize the positive impact of learners' active engagement and experiences of autonomy on learning outcomes (Ryan and Deci, 2000; Hmelo-Silver, 2004). Makransky and Lilleholt reported in VR learning research that “the sense of control and active behavior experienced by learners simultaneously enhance learning satisfaction and learning effectiveness” (Makransky and Lilleholt, 2018). Hmelo-Silver, grounded in cognitive learning theory, stated that “proactive exploration and manipulation experiences are significantly associated with cognitive learning outcomes” (Hmelo-Silver, 2004). Parong and Mayer, in their cognitive design study, explained that “the more learners actively engage in task performance and experience immediate feedback, the greater the improvement in knowledge transfer and performance outcomes” (Parong and Mayer, 2018). In aviation service education, Control & Active experiences can contribute to enhancing on-site judgment and response capabilities by enabling active practice of complex emergency procedures and service situations. This suggests a potential link to strengthening human sustainability and organizational safety culture, as discussed in prior research (Ochs et al., 2022; Talley and Howard, 2025).

In emotion research within educational psychology, enjoyment is classified as a representative positive emotion that promotes learners' utilization of cognitive resources and learning persistence (Pekrun, 2006; Pekrun et al., 2007). Pekrun's control-value theory explains that positive emotions related to learning enhance attention to learning and strategic learning (Pekrun, 2006). Linnenbrink-Garcia et al. stated that “enjoyment significantly predicts learning outcomes” in the context of self-regulated learning (Linnenbrink-Garcia et al., 2016). Lucardie's meta-analysis study explained that “in adult learning contexts, enjoyment is a key factor enhancing learning engagement and learning effectiveness” (Lucardie, 2014). Ke reported in the field of game-based learning that “enjoyment is closely associated with training outcomes” (Ke, 2008). Pekrun and Linnenbrink-Garcia presented in achievement emotions research that “enjoyment positively influences learning achievement and intention to participate” (Pekrun and Linnenbrink-Garcia, 2014). In aviation service education, the enjoyment and interest provided by VR environments can mitigate fatigue from repetitive training and increase the intention to continue learning. In the long term, this contributes to improved training completion rates, skill proficiency, and talent retention, functioning as an emotional resource that supports the human sustainability of aviation workforce development systems (Karaman et al., 2018; Talley and Howard, 2025).

Learning Outcomes are conceptualized as cognitive, affective, and behavioral changes manifested in learners following an educational experience. Biggs and Tang defined learning outcomes as “clear, measurable states that learners should attain through an educational program” (Biggs and Tang, 2011). In the context of aviation service education, these learning outcomes extend beyond mere knowledge acquisition to include behavioral competencies such as adherence to safety procedures, emergency response capabilities, and service attitudes. These competencies serve as critical indicators of an airline's safety and service quality and can be considered one of the key factors evaluated when assessing the social sustainability of the industry as a whole (Karaman et al., 2018; ATAG, 2022).

Based on the preceding literature review, this study establishes the following research hypotheses.

This study established its research design based on the research model presented in Fig. 1 and applied quantitative survey methods, developing a total of 19 survey questions, summarized in Table 1. The survey questions were developed by creating measurement items for each variable based on prior research from various existing fields. The developed survey items were then refined to align with the objectives of this study and were designed to be answered using a 5-point Likert scale. Prior to the main survey, a pilot test was conducted with 15 aviation industry trainees to ensure the readability and clarity of the survey items; based on the feedback, the items were polished before proceeding with the main survey.

Fig. 1. Research model

Download Original Figure

Sample extraction employed convenience sampling, a non-probabilistic method based on voluntary participation. The survey was administered via Google Survey Form to university students and graduates majoring in aviation services, and the final analysis was performed based on a total of 289 completed responses.

Data were collected using a Google Survey Form targeting university students and graduates majoring in aviation services, resulting in a final sample of 289 valid responses. Data analysis was conducted using SPSS version 30.0, performing various basic statistical analyses, including descriptive statistics and internal reliability analysis. Subsequently, to test the hypotheses established in this study, confirmatory factor analysis (CFA) was performed using AMOS 30.0. This process included examining convergent validity, discriminant validity, and model fit, as well as conducting path analysis through structural equation modeling (SEM).

Furthermore, correlation analysis was performed to check for multicollinearity among the independent variables and was used as a preliminary verification procedure to ensure the suitability of the SEM analysis. All structural equation modeling analyses were conducted using AMOS 30.0.

Table 2 presents the demographic characteristics of the 289 respondents. The gender distribution was 90.0% female and 10.0% male, indicating an overwhelmingly high proportion of female respondents. This is likely due to occupational characteristics and industry workforce composition, which reflect a significantly higher proportion of female students majoring in Cabin Service Management. This pattern can be interpreted as the combined result of preferences for aviation cabin service roles, traditional career choice tendencies, and the gender ratio structure within educational institutions. Therefore, the 90% female respondent rate in this study is not considered indicative of sample bias; rather, it is thought to contribute to the reliability and validity of this research by demonstrating close alignment with the actual composition of students majoring in Cabin Service Management.

The largest age group was 21-23 years old at 66.4%, followed by 18-20 years old at 19.4%, 24-26 years old at 13.1%, and 27 years old and above at 1.0%. Regarding educational background, the largest number of respondents were currently enrolled in (or had dropped out of) their second year of university, accounting for 38.4%. This was followed by university/college graduates at 31.1%, those enrolled in (or having dropped out of) their first year of university at 16.6%, those enrolled in (or having dropped out of) their third year at 6.9%, high school students or graduates at 3.8%, and university seniors (or dropouts) at 3.1%.

The number of VR training sessions participants had attended was categorized into four groups. Among the 289 respondents, those with four or more sessions accounted for the highest proportion at 46.4%, followed by 19.4% who participated once, 18.7% twice, and 15.6% three times. This distribution indicates that VR-based training is being utilized within aviation service and flight operations curricula as a form of repetitive and cumulative learning. Notably, the fact that respondents who experienced VR training four times or more accounted for nearly half (46%) of the total suggests that a sustained, practice-based VR learning environment has been established, rather than VR being used merely as an experiential training tool. This can be interpreted as a result of subject matter characteristics where repetitive learning is crucial, such as aviation safety, cabin service procedures, and emergency response manuals, and it also serves as indirect evidence of the high educational utility of VR content.

Prior to hypothesis testing, this study examined the validity of the measurement model through confirmatory factor analysis (CFA). Analysis of the squared multiple correlation (SMC) values for each item revealed that all measurement items had SMC values ranging from 0.563 to 0.878, meeting the recommended threshold of 0.4. Therefore, the construct validity and explanatory power of the survey items used in this study were confirmed to be sufficiently established.

Furthermore, to confirm internal consistency, Cronbach's α was computed. The results ranged from 0.876 to 0.938 for all variables—VR Features, Presence, Control & Active Engagement, Enjoyment, and Learning Outcomes—exceeding the benchmark of 0.70. These results demonstrate that the survey items possess a high level of reliability and also meet the criteria necessary for assessing convergent validity. All relevant indicators (Cronbach’s α, SMC, Average Variance Extracted [AVE], and Construct Reliability [CR]) are summarized in Table 3.

The discriminant validity analysis results are shown in Table 4. The benchmark values of AVE≥0.5 and CR≥0.7 were satisfied. After confirming the reliability and validity of each construct, VR Features was found to sufficiently meet the criteria with CR=0.938 and AVE=0.791. Presence (CR=0.878, AVE=0.643), Control & Active Engagement (CR=0.915, AVE=0.729), Enjoyment (CR=0.938, AVE=0.835), and Learning Outcomes (CR=0.936, AVE=0.784) also exceeded the required levels.

As presented in Table 5, various model fit indices were calculated to verify the suitability of the research model. The results showed that most fit indices met, or were very close to, the recommended criteria, thereby confirming the structural validity of the model.

Fig. 2 and Table 6 present a summary of the proposed research model for this study, including the path analysis results.

Fig. 2. SEM analysis results

Download Original Figure

The path analysis results of the structural equation model (SEM) in this study revealed that VR Features exert a strong influence on learners' experiential factors. Specifically, VR Features had statistically significant effects on Presence (β=0.907, p<.001), Control & Active Engagement (β=0.913, p<.001), Enjoyment (β=0.863, p<.001), confirming the acceptance of hypotheses H1, H2, and H3. Furthermore, Learning Outcomes were significantly influenced by Presence (β=0.209, p=.005), Control & Active Engagement (β=0.552, p<.001), and Enjoyment (β=0.232, p=.005), supporting H4, H5, and H6.

These results indicate that H4 through H6 were also statistically supported, showing that each experiential factor positively influenced learning outcomes to varying degrees within the VR-based learning environment. The statistical analysis results of this study show that VR Features, the learning experiences conducted based on virtual reality, ultimately showed a significant positive relationship with the trainees' learning outcomes.

The analysis showed that Presence has a significant positive effect on learning outcomes (β=0.209, p<.05). This implies that the more learners perceive the VR environment as a “realistic experiential space” similar to an actual cabin situation, the more effectively they acquire safety procedures and service-related knowledge and skills. Furthermore, immersive learning experiences based on high levels of Presence hold the potential to enhance response capabilities in real emergency situations and service performance. This can serve as a foundation for strengthening the human and social sustainability of the aviation industry by reducing human error and protecting passengers.

Control & Active Engagement was also identified as the variable exerting the strongest influence on learning outcomes (β=0.552, p<.001). This suggests that the more learners actively manipulate situations and control their own learning process within the VR environment, the deeper their understanding and retention of the knowledge and procedures required for actual job performance. In the long term, such experiences of active control can enhance self-directed learning and job adaptability, supporting human sustainability—the ability to continuously update competencies even after training. They are also expected to mitigate workforce replacement costs and organizational operational risks by preventing safety incidents and maintaining service quality.

Enjoyment was likewise found to have a significant positive effect on learning outcomes (β=0.232, p<.01). The enjoyment and interest experienced in the VR learning environment function as emotional resources that support more stable learning achievement by increasing learners' willingness to engage with tasks and alleviating fatigue from repetitive learning. Particularly in contexts requiring repeated practice of identical procedures and scenarios, such as aviation service training, high Enjoyment reduces training dropout rates and promotes long-term competency development. This, in turn, enhances human sustainability by maintaining and strengthening human capital, while simultaneously contributing to optimized educational resource utilization through efficient digital training compared to physical practice.

Taken together, these results indicate that H4 through H6 are all statistically supported, showing that each experiential factor positively influences learning outcomes to varying degrees within VR-based learning environments. The statistical analysis results of this study confirm that VR Features significantly contribute to trainees' academic performance through learning experiences conducted in virtual reality.

This study was conducted to empirically analyze the impact of education utilizing virtual reality (VR) in a practice-centered aviation service training environment on learners' experiences and learning outcomes, based on the premise that the aviation industry is a high-reliability industry (HRI) that demands both safety and service quality. Aviation service training necessitates high-level facilities, equipment, and safety management systems to fully replicate actual aircraft cabin environments. However, constructing and maintaining physical mock-ups involves substantial costs, resource consumption, and spatial constraints, which have been identified as factors threatening the long-term sustainability of training systems (Balcerzak, 2018). Against this backdrop, this study examined whether VR-based practical training can support human and environmental sustainability in aviation service education. To do this, we structurally analyzed the relationship between experiential factors, such as presence and enjoyment, and learning outcomes (Merchant et al., 2014; Yang et al., 2023).

Specifically, through empirical analysis targeting the students and graduates majoring in aviation services, this study examined the potential for VR to function as an alternative training environment that compensates for the limitations of existing lecture-based and model-centered educational programs. On this basis, it explored whether digital-based educational strategies can establish themselves as a sustainable development pathway for aviation service workforce training systems. This problem awareness aligns with recent discussions that aviation workforce education should contribute to protecting human capital and establishing resource-efficient training systems, going beyond simple technical training. This is particularly relevant to achieving the SDGs' goals of Quality Education (SDG 4), Decent Work and Economic Growth (SDG 8), Responsible Consumption and Production (SDG 12), and Climate Action (SDG 13) (ATAG, 2022; Karaman et al., 2018).

First, by validating the relationship between VR's functional characteristics, experiential factors, and learning outcomes using a structural equation model within the aviation service education context, this study extends the interactive VR learning mechanism proposed by Yang et al. (2023) into the service and safety education domains. While existing aviation service research has primarily focused on operational outcomes such as service quality, customer satisfaction, and safety management systems (Thomas et al., 2023; National Fire Agency, 2024), this study aims for academic expansion by presenting an integrated model that simultaneously considers both the design variables of the training environment itself (VR features) and the learner's psychological experience (Steuer, 1992; Makransky and Lilleholt, 2018; Makransky and Mayer, 2022; Pekrun, 2006).

Second, this study distinguishes itself from existing VR education research by examining aviation service training from the perspective of human and environmental sustainability. Similar to Karaman et al. (2018) and Talley and Howard (2025), who argue that aviation sustainability reporting and training must simultaneously consider protecting human capital and reducing environmental impacts, this study suggests that VR-based education can contribute not only to learning outcomes but also to building resource-efficient, low-carbon training systems, thereby integrating a sustainability framework into educational technology research.

Third, by comparing within a single model the differential influence of experiential factors (Presence, Control & Active Engagement, and Enjoyment) on learning outcomes, the study provides theoretical clues for prioritizing which experiential variables should be designed and managed in future aviation service education and similar service-industry training. This can be interpreted as applying and empirically validating, in the context of aviation service education, the discussions of Radianti et al. (2020) and Pekrun (2006), which emphasized the roles of presence and emotional enjoyment.

Aviation service education is practice-centered training directly linked to aviation safety. However, conducting repeated practice under conditions identical to actual flight operations or the cabin environment faces significant constraints in terms of cost, safety, and operations. The results of this study suggest that VR is an educational medium that mitigates these constraints while enhancing learners' presence, sense of control, and enjoyment, and that these factors are significantly linked to learning outcomes (Merchant, Goetz, Cifuentes, Keeney Kennicutt, and Davis, 2014; National Fire Agency, 2024). These findings indicate that aviation training institutions should strategically introduce and expand VR not merely as an experiential tool, but as a sustainable core educational infrastructure applicable in actual training and operational settings. Based on the empirical findings of this study, the following key practical implications are presented for designing and operating VR-based education in aviation service training environments and across the broader aviation industry's workforce development and training systems.

This study has a few limitations. First, the sample was restricted to students and graduates majoring in airline service from several educational institutions in Korea that have introduced VR equipment. For this reason, it is important to remember that these findings may not be the same for current airline crews or learners in other countries. Second, as the analysis was based on cross-sectional survey data, it could not directly verify the long-term persistence of VR learning experiences on learning outcomes or the transfer of training results to actual job performance (Jensen and Konradsen, 2018). Third, while addressing sustainability, this study limited its analytical scope to presenting theoretical and practical implications for human and educational sustainability in aviation workforce development and educational infrastructure design. This was achieved by examining the relationship between VR functionality, experiential factors, and learning outcomes, rather than directly quantifying environmental and economic performance indicators such as carbon emissions, energy consumption, or cost savings. Therefore, as this study presents sustainability discussions at the educational psychology and educational technology level as a first step, quantitative evaluations of environmental performance and cost/resource efficiency remain tasks to be addressed in subsequent research.

Furthermore, as the data are based on a single-point self-report survey, the possibility of common method variance (CMV) cannot be entirely ruled out. Future research should strengthen procedural and statistical controls, such as combining survey and behavioral data, collecting time-lagged data, and designing marker variables.

Future research should collaborate with various aviation training institutions and airlines to track and verify the long-term impact of VR training experiences on safety behaviors, service quality, and accident/error rates in actual flight operations. Additionally, by looking at both human factors and environmental engineering to measure how VR training affects energy use and carbon emissions compared to real models, we can estimate the actual impact of this technology on the aviation industry's sustainability (Balcerzak, 2018; Karaman, Kilic, and Uyar, 2018). Finally, by studying students' feelings—such as anxiety and tiredness—through different research methods, future studies can clearly find the best conditions for VR training to help both people and the planet (Radianti, Majchrzak, Fromm, and Wohlgenannt, 2020; Stanney et al., 2020).