Designing Gamified Virtual Reality Intervention Based on Experiential Learning to Enhance Social Reciprocity in Children with Autism Spectrum Disorder

Abstract

This study investigates the effectiveness of designing a gamified Virtual Reality (VR) intervention, grounded in Kolb’s Experiential Learning Theory, in enhancing social reciprocity, affective expression, and daily living skills among Thai children with Autism Spectrum Disorder (ASD). Utilizing the Meta Quest 2 headset, the intervention simulated five real-world contexts—home routines, classroom behavior, street crossing, supermarket shopping, and fast-food interactions. A quasi-experimental pretest-posttest control group design was employed with 33 children aged 6–12, complemented by thematic analysis of caregiver and therapist interviews. Results from the Social Responsiveness Scale, Second Edition (SRS-2), indicated significant improvements in social communication (p = 0.001) and social motivation (p = 0.045), while changes in social awareness (p = 0.233) and repetitive behaviors (p = 0.169) were not statistically significant. However, an ANCOVA analysis controlling for pre-test scores found that the difference in post-test outcomes between the intervention and control groups did not reach statistical significance (p = 0.073), suggesting that observed improvements may be influenced by baseline variability. Qualitative feedback highlighted the program’s engagement value, cultural relevance, and usability, alongside suggestions for increased adaptivity and contextual nuance. These results underscore the potential of VR-based learning while highlighting the need for further research with larger samples and more adaptive systems. These findings offer practical insights for educators, researchers, teachers, and program developers by demonstrating how culturally grounded, gamified VR interventions when guided by experiential learning principles can effectively enhance engagement and support targeted social skill development in children with ASD.

1. Introduction

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition characterized by persistent deficits in social communication and interaction, alongside restricted and repetitive patterns of behavior, interests, or activities (American Psychiatric Association, 2013). These challenges often impede children with ASD from effectively participating in daily routines and community interactions. In Thailand, while inclusive education is increasingly promoted, many school-aged children with ASD continue to face substantial barriers in interpreting social cues, regulating affective states, and navigating typical community environments such as crossing roads or ordering food—leading to social exclusion, anxiety, and increased caregiver burden (Tawankanjanachot et al., 2024).

Recent advancements in immersive technologies have positioned Virtual Reality (VR) as a promising medium for ASD interventions. VR provides safe, repeatable, and controlled environments where children can practice complex real-life interactions and skills. The flexibility of VR allows for the creation of realistic social contexts and interactive tasks, which can be adapted to individual learning needs. Empirical evidence supports the use of VR to enhance emotional expression, social engagement, and daily living skills in children with ASD (H. H. S. Ip et al., 2022; Parsons & Cobb, 2011; Zhao et al., 2022). Despite these promising developments, research remains scarce in Southeast Asian contexts, particularly within Thai public school systems, where cultural norms, infrastructural constraints, and limited digital readiness can affect the feasibility and sustainability of such interventions (Tawankanjanachot et al., 2024).

Grounded in D. A. Kolb’s (1984) Experiential Learning Theory, this study applies a constructivist approach to guide the design of VR learning modules. According to Kolb, learning is a cyclical process involving four stages: concrete experience, reflective observation, abstract conceptualization, and active experimentation. These stages align naturally with immersive VR scenarios, where learners are immersed in simulated real-world tasks (e.g., shopping, navigating a crosswalk), guided through reflection and decision-making, and encouraged to apply new knowledge and skills in varied contexts. The learning process is further enhanced through gamification techniques, such as visual feedback, rewards, progress tracking, and interactive tasks, to increase motivation, reinforce learning, and support self-efficacy—key factors in learning among children with neurodiverse profiles (Hamari et al., 2014).

This study aims to address an important gap by designing and evaluating a VR-enabled training program tailored for Thai children with ASD aged 6 to 12. Using Meta Quest 2 headsets, the program spans five essential life contexts—home routines, classroom behavior, street crossing, supermarket shopping, and fast-food interactions—to foster the development of social reciprocity, affective expression, daily living skills, and emotion recognition. Delivered over a two-month period in real-world educational and public settings, the intervention not only explores the effectiveness of VR-based experiential learning but also highlights the practical design considerations needed to promote scalable, gamified ASD interventions in low- and middle-income countries. Accordingly, this study addresses the following research questions:

RQ1: To what extent does a VR-based training program enhance social reciprocity and affective expression in children with ASD?

RQ2: How do caregivers and therapists perceive the effectiveness, usability, and cultural relevance of the VR intervention?

RQ3: What are the challenges and limitations of using immersive VR to sustainably support the development of social skills in children with ASD in real-world contexts?

2. Related Works

2.1. ASD and Social Challenges

ASD is a lifelong neurodevelopmental condition characterized by persistent deficits in social communication and interaction, along with restricted, repetitive patterns of behavior, interests, or activities (American Psychiatric Association, 2013). Children with ASD often struggle with fundamental social skills such as joint attention, emotional reciprocity, and nonverbal communication, including facial expressions, eye contact, and gestures (Baron-Cohen et al., 2001; Golan & Baron-Cohen, 2006). These impairments significantly affect their ability to form and maintain social relationships, participate in group activities, and adapt to everyday social norms. Such challenges not only impact the child’s emotional development and peer integration but also contribute to higher levels of anxiety, behavioral outbursts, and withdrawal in unfamiliar or overstimulating environments (Lord et al., 2020). Social deficits in ASD are particularly complex due to their heterogeneous nature, with variations in severity, cognitive functioning, and co-occurring conditions such as sensory processing issues or attention difficulties.

In educational settings, especially in inclusive or mainstream classrooms, children with ASD may struggle to interpret social cues, follow group instructions, or respond appropriately to peer interactions, which can hinder both academic engagement and social participation. In Thailand, these challenges are compounded by structural and cultural barriers. Many regions face a shortage of trained specialists, limited access to early diagnosis services, and insufficient individualized support within public schools (Vorapanya & Dunlap, 2014). Furthermore, sociocultural stigma surrounding developmental disorders may lead to parental reluctance to seek formal assessments or interventions, further delaying support (Tawankanjanachot et al., 2024). As a result, children with ASD often receive delayed or fragmented intervention, reducing opportunities for early social skills development and functional independence. Addressing these challenges requires not only innovative therapeutic tools but also systems that are accessible, culturally sensitive, and tailored to the local educational landscape.

2.2. VR Interventions for ASD

VR provides immersive, interactive, and controlled environments that are particularly suitable for training children with ASD in social and emotional competencies (Parsons & Cobb, 2011; Strickland, 1997). These virtual settings allow children to repeatedly engage in structured scenarios that simulate real-life interactions without the unpredictability or anxiety that may arise in natural environments. By offering visual and auditory cues, real-time feedback, and adaptive difficulty levels, VR systems enable children with ASD to build foundational skills such as facial emotion recognition, turn-taking, and situational awareness (Souto et al., 2020; Shahmoradi & Rezayi, 2022). Research has demonstrated that these platforms are especially effective when the learning tasks are embedded within socially relevant contexts. For instance, Kandalaft et al. (2013) reported significant improvements in social cognition and reasoning abilities following VR-based training in high-functioning individuals with ASD, indicating the potential of immersive technology to address higher-order socio-cognitive functions.

Despite these benefits, the accessibility and contextual relevance of VR interventions remain a concern, particularly in non-Western settings. Most VR-based ASD programs originate in North America and Europe, with cultural assumptions embedded in language, gestures, and social scripts that may not align with local norms in Southeast Asian contexts (Li et al., 2023). While a growing number of studies, such as H. H. S. Ip et al. (2022), show that VR enhances social reciprocity and emotional expression in school environments, such interventions must be adapted to reflect regional communication styles, environmental cues, and family dynamics to be truly effective. A recent meta-analysis by Altın et al. (2025) confirms the general efficacy of VR in promoting social responsiveness among children with ASD but also highlights a disproportionate representation of studies from high-income countries. This gap underscores the need for culturally informed VR programs that are co-developed with local educators, therapists, and families to ensure both usability and sustainability in diverse educational and therapeutic settings.

2.3. Gamification and Engagement in ASD Training

Gamification, adding game elements such as rewards, progress bars, and challenges, has been shown to enhance user engagement and promote learning persistence among neurodiverse learners (Hamari et al., 2014; Koivisto & Hamari, 2019). From a theoretical perspective, Self-Determination Theory (SDT) provides a valuable framework for understanding how gamified interventions support intrinsic motivation by addressing three basic psychological needs: autonomy, competence, and relatedness (Ryan & Deci, 2000). In the context of ASD training, gamified systems can foster competence by offering achievable challenges and immediate feedback, autonomy through meaningful choices within tasks, and relatedness by simulating socially interactive experiences.

Empirical findings support these theoretical claims. Wang et al. (2025) found that gamified VR interventions significantly improved communication and cooperative behavior among children with ASD, suggesting an enhancement of relatedness and social motivation. Recent systems such as HSVRS (Yu et al., 2023) and Eggly (Lyu et al., 2023) integrate real-time feedback, augmented reality, and adaptive challenges to improve attention and gaze fixation—key components of social engagement. However, the design of gamified elements must be carefully calibrated. Poorly structured feedback loops or excessive stimulation may undermine user autonomy or overwhelm sensory-sensitive learners, potentially increasing anxiety (Toda et al., 2018). Thus, gamification in ASD interventions should be intentionally aligned with SDT principles to ensure that motivational benefits are balanced with emotional and cognitive appropriateness.

2.4. Experiential Learning Theory in Immersive Training

D. A. Kolb’s (1984) Experiential Learning Theory describes a four-stage process—concrete experience, reflective observation, abstract conceptualization, and active experimentation—that aligns naturally with interactive VR modules (A. Y. Kolb & Kolb, 2005). H. H. Ip et al. (2017) applied this model to develop VR interventions that cycle learners through virtual tasks with embedded coaching and feedback. These learning loops were effective in fostering emotional awareness and interaction confidence. Recent work by Kim et al. (2024) emphasizes combining experiential learning with physiological monitoring to improve emotional regulation outcomes in ASD training.

2.5. VR Platforms for Intervention

Although evidence increasingly supports the effectiveness of VR-based interventions for enhancing social skills in children with Autism Spectrum Disorder (ASD), limited attention has been given to cross-cultural adaptation. Most existing platforms—such as VR-SS Training (Moon & Ke, 2019) and VRESS (Kourtesis et al., 2023)—are developed in Western contexts and do not reflect local languages, social values, or interaction norms. Similarly, platforms like VR Rehabilitation Therapy (Zhang et al., 2025) and VRI (Frolli et al., 2022) demonstrate significant improvements in cognition and emotion recognition, yet lack cultural customization for diverse learning environments. Many VR platforms are developed based on Western cultural norms (Moon & Ke, 2019; Frolli et al., 2022; Kourtesis et al., 2023; Maddalon et al., 2024), which may not be easily adaptable to Southeast Asian cultures. As Chan et al. (2020) noted, the absence of culturally adapted systems may hinder the real-world applicability and acceptance of these interventions across different societies.

While progress has been made in Asia through platforms such as HSVRS (Yu et al., 2023) and VR in Schools (H. H. Ip et al., 2017), the integration of deep cultural features remains minimal. For example, although H. H. Ip et al. (2017) incorporated Kolb’s Experiential Learning Model within a Hong Kong school setting, the social scenarios and emotional expressions remained relatively generic. A summary of existing virtual reality-based interventions for social skills is presented in

Table 1. Summary of Platforms for Virtual Reality-Based Interventions for Social Skills in ASD.

Platform/Study	Target Skills	Hardware	Country	Findings
VR-SS Training (Moon & Ke, 2019)	Social communication, task adherence	Not specific	USA	VR-based social skills training using OpenSimulator showed high treatment integrity across 90 sessions for 15 children with HFA. Treatment integrity (Jaccard Index) varied by scenario: higher in game-based and restaurant-roleplay activities, lower in formal interview settings.
VR Rehabilitation Therapy (Zhang et al., 2025)	Cognition, imitation, social interaction	VR HTC VIVE	China	Improved imitation and cognition in ASD children after immersive VR therapy; statistically significant in multiple domains.
VRI (Frolli et al., 2022)	Emotion recognition, situational understanding	Not specific	Italy	The VR group showed significantly shorter acquisition times for complex social-emotional tasks, though both groups performed similarly in primary emotion recognition.
HSVRS (Yu et al., 2023)	Gaze fixation, attention training	VR + Eye	China	VR-based hide-and-seek significantly improved eye-gaze behavior; supported by attention
VRESS (Kourtesis et al., 2023)	Social skills, executive function	HTC Vive Pro Eye	Greece	Performance correlated with executive functioning and memory; adaptive scenarios encouraged participation.
(Sideraki & Drigas, 2023)	General social skills	Not specific	Greece	VR is more effective than traditional therapy (e.g., role-play, group sessions) in improving social skills. VR provides a safe, customizable, and immersive environment with higher engagement.
VR in Schools (H. H. Ip et al., 2017)	Affective expression, social reciprocity	Oculus Rift	Hong Kong	Large-scale field trial in schools showed affective expression and reciprocity gains using Kolb’s experiential model.
VR-based Social Skills Training (Maddalon et al., 2024)	Social communication, emotion recognition	Meta Quest VR	USA	Adaptive VR with biosignal-driven feedback personalized the experience and improved learning outcomes.

. In Thailand, where digital health and education technologies are expanding, most interventions still rely on foreign-developed systems with limited contextual relevance. To address this gap, the present study introduces a VR-based program co-developed with Thai educators, therapists, and cultural experts. This platform incorporates localized language, culturally familiar social situations, and visual elements aligned with Thai communication patterns and family dynamics, thereby offering a more culturally responsive approach to ASD intervention.

3. Virtual Reality System Design

The VR intervention used in this study was designed to support the development of social reciprocity, affective expression, emotion recognition, and daily living skills in children with ASD. The design was grounded in Kolb’s Experiential Learning Theory (H. H. S. Ip et al., 2022) and incorporated gamification elements to create a structured yet engaging and interactive learning environment. The VR system was specifically developed for Thai children aged 6 to 12 years and deployed using the Meta Quest 2 headset, which provided an immersive, untethered experience with six degrees of freedom.

The virtual environments were developed using Unity 2021.3 as the core game engine, with 3D modeling and animation created in Autodesk Maya 2020. Each scenario simulated a real-world daily living setting relevant to the children’s everyday experiences, including the home, school, supermarket, street-crossing area, and fast-food restaurant, as shown in Figure 1, Figure 2 and Figure 3. The content and design of these environments were created through close collaboration with interdisciplinary researchers from the Faculty of Humanities, Faculty of Public Health, and the College of Arts, Media, and Technology, Chiang Mai University. Additionally, the development process was informed by the expertise of child therapy specialists at the Banprajan Child Center, ensuring that the virtual content was developmentally appropriate, contextually realistic, and culturally sensitive.

3.1. Learning Scenario Design

Each of the five daily contexts was structured into three sequential stages: instruction, practice (coaching), and assessment. In the instruction phase, children received visual and auditory guidance introducing the social tasks or routines. The practice phase allowed children to interact with animated agents (e.g., teachers, peers, store clerks) with varying degrees of support, such as speech prompts, gestures, or visual cues. The assessment phase required the child to complete the scenario independently, with system-logged data capturing task success, completion time, and behavioral cues.

The scenarios’ embedded core skills aligned with the study’s objectives. For example, in the supermarket task, children practiced identifying items, greeting a cashier, handling money, and expressing gratitude capturing daily living and affective interaction. In the crosswalk scenario, children practiced safety behaviors, eye contact, and gesture-based reciprocity (e.g., waving to drivers). The summary of Learning Scenario Design shown in

Table 2. Virtual reality learning modules, scenarios, objectives, and number of sessions designed for children with ASD.

Learning Module	Scenario	Design	Objective	No. of Sessions
1. Orientation and Familiarization	VR Playground	Safe open area with simple interactive objects for controller practice and avatar exploration.	Learn to navigate, interact with objects, and feel comfortable in the VR environment.	1
2. Daily Living Skills	Home Routine	Simulated home with tasks such as brushing teeth, packing a bag, and setting the table.	Learn routines and daily tasks using visual and verbal prompts.	4
	Supermarket	Simulated shopping trip with money handling and grocery choices.	Practice purchasing items, understanding currency, and social interaction with cashiers.	4
3. Affective and Emotional Skills	School Classroom Setting	Children interact with teachers and classmates in a quiet rule-based environment	Learn social norms, self-regulation, and respectful behavior.	4
	Fast Food Restaurant	Ordering food, waiting in line, and responding to staff.	Practice affective expression, waiting turn, and polite responses.	3
4. Social Reciprocity	Crosswalk and Street	Street with traffic lights, vehicles, and pedestrian cues.	Learn road safety, eye contact, and social gestures like waving or thanking.	4
	School Recess	Interacting with peers in a playground setting.	Develop reciprocal communication and shared play behavior.	4

.

3.2. Gamification and Feedback Features

To enhance motivation and reinforce learning, the VR system incorporated a variety of gamified elements tailored to engage children with ASD in a structured yet enjoyable way (Wang et al., 2025). These included achievement badges for completing scenarios, points awarded for task accuracy and socially appropriate responses, a progress bar to visualize task completion, and time pressure to encourage decision-making under mild constraints. In selected modules, a simple leaderboard was introduced to foster friendly competition. Positive reinforcement was embedded throughout, with encouraging feedback. At the end of each session, a performance dashboard summarized progress, promoting repeated practice and sustained motivation. These gamification features align with prior research demonstrating the effectiveness of points, progress tracking, and competitive feedback in enhancing user engagement and learning in digital learning environments.

3.3. Interface and Usability Considerations

The VR system was designed to support young users with limited prior experience in virtual environments, particularly children with ASD, while providing real-time coaching tools for adults. As shown in Figure 4, the user interface was intended for teachers, caregivers, or therapists to monitor the child’s performance and deliver appropriate support. The interface included clearly labeled buttons such as Objective, Coaching Method, Assessment Method, and Rate the Response alongside a time-tracking display and session control options (e.g., repeat, main menu). Avatars in the VR environment used exaggerated facial expressions and clear gestures to aid emotion recognition. To minimize sensory overload, scene complexity was reduced and visual/auditory stimuli were selected based on known sensitivities in children with ASD, such as avoiding high-contrast reds and flashing lights. As illustrated in Figure 4, sessions were conducted in a supervised 5-m safe space, allowing free movement. Pilot testing with therapists and caregivers confirmed that the system supported independent child interaction, with adult facilitators using the tablet interface to observe progress and deliver in situ coaching when needed.

3.4. Theoretical Framework

This study is grounded in an integrative theoretical framework that combines Kolb’s Experiential Learning Cycle with gamification strategies to guide the design of immersive VR interventions for children with ASD. Kolb’s model highlights the cyclical process of concrete experience, reflective observation, abstract conceptualization, and active experimentation, which supports active engagement and knowledge transfer across different contexts. By structuring VR scenarios around this cycle, the intervention aims to foster meaningful opportunities for practicing social skills, reflecting on interactions, and generalizing these experiences into real-life communication.

In parallel, gamification elements such as feedback, rewards, and progressive challenges were incorporated to sustain motivation, engagement, and emotional regulation throughout the VR experience (Wang et al., 2025; Sipone et al., 2025). The framework (Figure 5) illustrates how application design features VR scenarios situated in daily life contexts and gamified elements are linked to theoretical contributions from Kolb’s cycle and psychological outcomes. Together, these components are expected to reduce social anxiety, enhance social reciprocity, and improve communication skills, thereby supporting the overall learning outcomes of the intervention.

4. Research Design

This study employed a quasi-experimental pretest-posttest control group design (Rogers & Révész, 2019) combined with a convergent mixed-methods approach to evaluate the effectiveness of a gamified VR intervention in enhancing social reciprocity, affective expression, and daily living skills in children with ASD. The intervention was designed according to Kolb’s Experiential Learning Theory and incorporated gamification elements to enhance engagement and reinforce behavioral learning. Quantitative data were collected using a standardized behavioral rating scale, while qualitative insights were gathered through semi-structured interviews with caregivers and therapists. An overview of the research design and procedure is shown in Figure 6.

4.1. Participants

A total of 45 children aged 6 to 12 years with a formal diagnosis of ASD were assessed for eligibility. Of these, 7 children were excluded due to time constraints (n = 4), loss of contact (n = 1), or declining to participate (n = 2). The remaining 38 participants were randomized into two groups: an intervention group (n = 19) and a waitlist control group (n = 19). Children in the intervention group received the VR-based learning intervention, while those in the control group continued with standard school activities without exposure to the VR system. During the 8-week intervention period, 2 participants from the intervention group and 3 from the control group withdrew due to personal reasons, resulting in final analyzed samples of 17 participants in the intervention group and 16 in the control group. The age range of 6–12 years was deliberately selected as it represents a developmental stage in which children with ASD are actively acquiring foundational social communication and reciprocity skills, while also possessing sufficient cognitive and motor abilities to interact with immersive technologies. This choice is consistent with prior VR-based interventions in autism research, where small to medium samples are common in feasibility studies and allow careful observation of individual engagement and outcomes.

All participating children were enrolled in inclusive classrooms or special education programs within Thai public schools. Inclusion criteria included a formal diagnosis of ASD, school enrollment, basic verbal communication abilities, and adequate cognitive and motor function to interact with the VR system, while children with severe intellectual disabilities or physical impairments that would hinder VR use were excluded. None of the participating children had prior experience with immersive VR, so a structured familiarization session was provided at the beginning of the program. To strengthen ecological validity, the intervention was implemented in the school setting, embedding VR activities into familiar routines and ensuring cultural relevance. Collaboration with Thai therapists and teachers during design and implementation helped tailor the content to local contexts, increasing the likelihood of real-world skill transfer. In addition to child participants, therapists and caregivers who regularly interacted with the children were also involved in behavioral assessments and post-intervention interviews, providing a multi-informant perspective on learning outcomes.

4.2. Instruments

4.2.1. Social Responsiveness Scale

The SRS-2 (Constantino & Gruber, 2012) was used as a quantitative instrument to assess changes in social behavior before and after the intervention. The form was completed by both therapists and caregivers, capturing behaviors in daily life contexts. The SRS-2 consists of 65 items rated on a 4-point Likert scale (0 = Not True to 3 = Almost Always True), measuring five subdomains: Social Awareness, Social Cognition, Social Communication, Social Motivation, and Restricted Interests and Repetitive Behavior. It yields a Total Score indicating overall severity of social impairment. In this study, a Thai-translated version of the SRS-2 was developed using a forward–backward translation method, followed by expert review and pilot testing for linguistic and cultural equivalence.

4.2.2. Semi-Structured Interviews

Semi-structured interviews with caregivers and therapists were conducted after the VR training program to explore perceptions of its effectiveness, design, and real-world applicability. The interviews focused on four key themes: (1) perceived improvements in social interaction and skill transfer, (2) feedback on the content and clarity of social scenarios, (3) reflections on the usability and engagement of the VR system, and (4) acceptance and long-term sustainability. Each 20–30-min session was conducted individually, recorded with consent, and thematically analyzed to identify key patterns and insights. The semi-structured interviews are presented in Appendix A.

4.2.3. Procedure

The VR intervention was implemented over an 8-week period, with each child receiving two sessions per week, totaling 16 sessions. Each session lasted approximately 30 min and was conducted individually using the Meta Quest 2 headset at the Banprajan Child Center. The VR modules simulated five daily life contexts: home routines, school participation, supermarket shopping, street crossing, and fast-food dining as shown in Figure 7. Gamification elements—including performance feedback, achievement stars, points, progress tracking, and interactive decision-making—were integrated into each module to enhance engagement and reinforce learning. A simple performance dashboard provided visual feedback on task completion and rewards to motivate continued participation. Therapists facilitated brief debriefing discussions after each session to support reflection and transfer of learning. Pre- and post-intervention assessments using the Social Responsiveness Scale–Second Edition (SRS-2) were completed by both caregivers and therapists, and semi-structured interviews were conducted post-intervention with a subsample of caregivers and therapists to gather qualitative feedback on behavioral changes and system usability.

4.2.4. Data Analysis

Quantitative data from the SRS-2 were analyzed using paired-sample t-tests to assess within-group changes and ANCOVA to compare post-test outcomes between the intervention and control groups, controlling for baseline scores. Effect sizes (Cohen’s d) were calculated to determine the magnitude of change. Qualitative data from the interviews were analyzed using thematic analysis, following Braun and Clarke’s (2006) six-step approach. Transcripts were coded inductively to identify recurring themes related to social behavior, emotional development, and VR experience.

5. Result

5.1. Quantitative Findings from the Social Responsiveness Scale

The SRS-2 was used to assess changes in five domains of social reciprocity: communication, motivation, awareness, cognition, and repetitive behaviors. Pre- and post-test comparisons were conducted to examine within group improvements, and ANCOVA was applied to compare post-test outcomes between the intervention and control groups while adjusting for baseline differences.

5.1.1. Pre- and Post-Test Comparisons

The results of the SRS-2 are presented in

Table 3. Pre- and post-intervention SRS-2 total and subscale scores analyzed using paired-sample t-tests.

Dimension	Group	Pre-Test (SD)	Post-Test (SD)	Mean Difference	p-Value	Cohen’s d
Social Awareness	Intervention	15.82 (3.41)	14.97 (3.22)	−0.85	0.233	0.25
	Control	15.67 (3.59)	15.42 (3.71)	−0.25	0.686	0.07
Social Communication	Intervention	22.14 (4.22)	17.47 (4.30)	−4.67	0.001 **	1.11
	Control	21.88 (4.09)	21.94 (4.01)	+0.06	0.939	0.01
Social Motivation	Intervention	18.76 (3.11)	17.10 (3.05)	−1.66	0.045 *	0.53
	Control	18.53 (3.08)	18.71 (3.14)	+0.18	0.745	0.06
Repetitive Behaviors	Intervention	30.63 (5.27)	29.58 (5.04)	−1.05	0.169	0.20
	Control	29.87 (5.18)	29.48 (5.45)	−0.39	0.682	0.08
Total Score	Intervention	87.35 (12.48)	79.58 (11.92)	−7.77	0.020 *	0.62
	Control	85.94 (11.92)	84.81 (12.67)	−1.13	0.612	0.09

Note: * p < 0.05, ** p < 0.01.

and visualized in Figure 8. Pre- and post-intervention comparisons revealed significant improvements in the intervention group across several domains. The total score significantly decreased from a pre-test mean of 87.35 (SD = 12.48) to a post-test mean of 79.58 (SD = 11.92), t(16) = 2.53, p = 0.020, indicating a medium effect size (Cohen’s d = 0.62). A substantial improvement was observed in social communication, which showed the largest effect size (Cohen’s d = 1.11), with scores decreasing from 22.14 (SD = 4.22) to 17.47 (SD = 4.30), p = 0.001. Social motivation also improved significantly (p = 0.045, d = 0.53), whereas changes in social awareness (p = 0.233, d = 0.25) and repetitive behaviors (p = 0.169, d = 0.20) were not statistically significant. These findings suggest that the VR-based intervention had a meaningful impact on social communication and motivation in children with ASD, while improvements in awareness and repetitive behavior may require a longer or more targeted intervention. While the within-group results highlight individual improvements, ANCOVA was conducted to assess whether these effects remained significant when comparing the intervention and control groups.

5.1.2. ANCOVA Between the Intervention and Control Groups

To further assess the effectiveness of the VR intervention, an ANCOVA was conducted on post-test scores while controlling for pre-test scores (

Table 4. Results of ANCOVA on the post-tests of the Intervention and Control groups.

Source	Type III Sum of Squares	df	Mean Square	F	Sig.	Partial Eta Squared
Corrected Model	2019.021 a	2	1009.510	30.767	<0.001	0.323
Intercept	830.106	1	830.106	25.299	<0.001	0.164
pretest	1933.825	1	1933.825	0.514	<0.001	0.001
Group	107.265	1	107.265	3.269	0.073	0.011
Error	4232.718	129	32.812
Total	62,046.644	132
Corrected Total	6251.739	131
R squared = 0.323 (Adjusted R Squared = 0.312)

Note. ^a = R Squared values are based on the adjusted model.

). The overall model was statistically significant, F(2, 131) = 30.767, p < 0.001, accounting for approximately 32.3% of the variance in post-test scores (R² = 0.323; Adjusted R² = 0.312). However, the main effect of group (intervention vs. control) did not reach statistical significance, F(1, 129) = 3.269, p = 0.073, with a small effect size (partial η² = 0.011). This indicates that after controlling for baseline differences, the improvement in post-test scores between the groups was not statistically significant. Notably, pre-test scores were a significant covariate, F(1, 129) = 25.299, p < 0.001, demonstrating a strong predictive influence on post-test outcomes. Collectively, these results suggest that although the intervention group showed observable gains, the difference was not large enough to achieve statistical significance once pre-test performance was accounted for.

5.2. Qualitative Findings from Semi-Structured Interviews

Thematic analysis of the semi-structured interviews with caregiver (n = 8) and therapist (n = 3) revealed consistent positive perceptions of the VR training program across four key themes.

5.2.1. Effectiveness of Training Social Reciprocity Skills

Feedback from both caregivers and therapists highlighted noticeable improvements in children’s social reciprocity and affective expression following participation in the VR training program. Children were reportedly more willing to initiate greetings, maintain eye contact, and engage in basic social exchanges. One caregiver shared, “After the sessions, my son started saying hello to our neighbors without being prompted.” (C2). A therapist echoed this, stating, “He used to avoid eye contact, but now he can respond when someone talks to him, especially in class.” (T3). These changes were most evident in structured settings, such as classrooms or therapy sessions. However, several caregivers noted that generalizing these skills to home or community environments remained inconsistent. “At school, she’s better at following turn-taking, but at home, she still struggles to apply what she learned,” explained one caregiver (C6).

Therapists expressed interest in the VR system as an engaging and supportive supplement to traditional interventions. “VR is a promising tool to introduce social concepts in a controlled, repeatable way,” commented one therapist (T1). However, they emphasized that VR should not be viewed as a primary method for building social skills. “It can’t replace live interaction or real-world practice,” added another (T2), noting the importance of caregiver and therapist reinforcement outside the virtual environment. While VR was described as effective in capturing attention and encouraging participation, long-term success still depended on consistent follow-up and support in daily life settings.

5.2.2. Feedback on Content Design

Both caregivers and therapists generally appreciated the clarity, structure, and relevance of the VR content. Scenarios involving familiar social contexts—such as classroom participation, turn-taking, and family routines—were seen as especially helpful in reinforcing appropriate behavior. “The scenarios were really helpful, especially the ones about taking turns in class,” said one therapist (T1). A caregiver similarly noted, “It helped him understand what to say when someone give something or when he needs to wait his turn.” (C4). These structured, role-based interactions were described as effective in helping children recognize and practice socially appropriate responses in a safe and repeatable environment.

However, several participants pointed out areas for improvement. Some felt that the content lacked variety and flexibility, limiting its ability to address a broader range of real-world situations. “There need to be more types of activities—the current ones feel too limited,” commented a caregiver (C3). Another added, “The content is a bit too rigid; some scenes need to adapt more naturally to different reactions or situations.” (C1, C2, C5). Additionally, cultural adaptation was highlighted as an area for development. “Some of the gestures and settings didn’t quite reflect Thai culture,” noted a therapist (T2), while a caregiver recommended, “Using examples like wai greetings or school lunch routines would make it more relatable.” (C8). These insights suggest a need for both expanded scenario diversity and improved contextual flexibility to better reflect users’ lived experiences.

5.2.3. Feedback on VR Design

The immersive and interactive design of the VR system was widely praised for its ability to capture children’s attention and sustain their engagement throughout the sessions. Caregivers and therapists highlighted that the visual and auditory elements were effective in making the learning experience more enjoyable and focused. “My daughter usually loses interest quickly, but she stayed focused during the entire VR session,” shared a caregiver (C1). A therapist added, “The transitions and instructions were easy to follow, and even kids with minimal tech experience could handle it.” (T3). The intuitive interface and structured scene progression contributed to a smooth user experience for most participants.

Despite these strengths, several concerns were raised about sensory and technical aspects of the system. Some caregivers reported instances of sensory overload, particularly in scenes with rapid movements or loud background sounds. “One of the busier scenes with loud background sounds made him anxious,” noted a caregiver (C5). Another said, “The visuals were too fast in one game—it overwhelmed my son and made him take off the headset.” (C3). Participants suggested adding customizable settings for audio, visual effects, and scene pacing to accommodate children with different sensory sensitivities. In terms of hardware, a few users mentioned physical discomfort, especially for younger children. “The headset was a bit too big and kept sliding down during the session,” commented a therapist (T1, T2, T3), referring to the Meta Quest device. This suggests a need for adjustable or child-friendly hardware designs to enhance comfort and accessibility for smaller users.

5.2.4. Acceptance and Sustainability

Overall, caregivers and therapists expressed strong acceptance of the VR program, largely attributing its appeal to the immersive, game-like experience. Many children reportedly viewed the sessions as play rather than therapy, which helped reduce resistance and increase motivation. “He asked to play the VR game again the next day. That’s rare for him,” shared a caregiver (C7). A therapist similarly noted, “The kids didn’t feel like they were doing a therapy session—it felt more like a fun activity, which made them more engaged.” (T3). This sense of novelty and enjoyment made it easier to integrate the program into daily routines, with several participants expressing interest in using it regularly. “If we could use this once or twice a week, I think it would really help in the long run,” said a therapist (T2).

Despite these positive perceptions, concerns were raised regarding the sustainability of the intervention. Several participants pointed out limitations in both the variety of training missions and the availability of physical space for implementation. “There aren’t enough tasks to keep using it long-term—it might get repetitive,” commented a caregiver (C1, C2, C8). A therapist added, “We had to rotate children in and out of a small room to run the session. It’s not easy when space is limited.” (T1). Additionally, questions about technical maintenance and the need for more structured content progression emerged. “If more levels or goals were added, it would help keep the kids motivated,” suggested a caregiver (C5). These concerns underline the importance of expanding content, providing adequate space, and offering ongoing technical support to ensure the program’s continued use and impact.

6. Discussion

6.1. Social Reciprocity and Affective Expression

Regarding RQ1, the findings suggest that the VR-based training program was effective in improving specific aspects of social functioning—namely, social communication and social motivation—as evidenced by significant gains on these two subscales of the SRS-2. Participants demonstrated increased willingness to initiate greetings, respond in conversations, and engage more confidently during structured social tasks. These improvements may reflect the program’s alignment with Kolb’s Experiential Learning Theory, which emphasizes learning through iterative cycles of concrete experience, reflective observation, abstract conceptualization, and active experimentation. In this intervention, children engaged in realistic VR scenarios (e.g., ordering food, crossing streets), received feedback through coaching prompts, and practiced behaviors in increasingly independent settings—creating opportunities for experiential learning to reinforce social behaviors. However, ANCOVA results indicated that post-test improvements were not statistically different between groups after controlling for baseline scores, suggesting that the observed effects may have been partially influenced by pre-existing differences.

In contrast, the program did not yield statistically significant improvements in social awareness or repetitive behaviors. These domains typically involve more complex cognitive and emotional processes, such as interpreting subtle facial cues, understanding others’ mental states, or managing sensory-driven routines—skills that may not be sufficiently activated through the scripted and goal-driven nature of the current VR tasks (Kandalaft et al., 2013; Leekam et al., 2011). The absence of adaptive, emotionally nuanced interactions and the lack of content explicitly targeting sensory regulation may have limited the intervention’s broader behavioral impact. To strengthen future outcomes, VR-based programs should incorporate dynamic response mechanisms, emotion recognition features, and flexible scenario paths that allow learners to respond to varying social cues. Additionally, integrating sensory-friendly design elements and personalized pacing may enhance the intervention’s relevance for children with diverse ASD profiles (Dechsling et al., 2021).

6.2. Content Relevance and Cultural Adaptation

In response to RQ2, caregivers and therapists generally perceived the VR training program as relevant, engaging, and effective in supporting structured social interactions for children with ASD. Designed around Kolb’s Experiential Learning Model, the program offered five core scenarios home routines, classroom behavior, street crossing, supermarket shopping, and fast-food interactions that provided realistic and goal-oriented tasks to reinforce social and daily living skills. These scenarios were positively received, particularly for their clarity, repetition, and alignment with familiar routines such as queuing, turn-taking, and ordering food. Caregivers noted that these structured interactions helped children better understand appropriate responses in predictable situations. However, while the content offered practical relevance, it remained largely linear and lacked flexibility to accommodate a broader range of individual behaviors or unexpected reactions.

A recurring theme in participant feedback was the limited cultural specificity of the VR content. Although the scenarios were based on common daily routines, they did not fully reflect localized Thai social norms, such as performing the traditional “wai” greeting, participating in morning flag ceremonies, or using culturally embedded classroom routines like collective bowing or hierarchical language when speaking to teachers. Therapists and caregivers explicitly noted these omissions, suggesting that their absence reduced the cultural resonance and contextual realism of the experience for some children. As a result, the presence of therapists and caregivers remained essential—not only to facilitate navigation within the VR system but also to help interpret cultural nuances, bridge contextual gaps, and reinforce culturally appropriate behaviors during and after VR sessions.

This limitation highlights the need for more intentional cultural adaptation beyond surface-level translation or environment setting. As prior studies have shown, embedding culturally familiar gestures, expressions, and environmental cues enhances learner engagement and the transferability of skills (H. H. Ip et al., 2017). Future iterations of the program should prioritize flexible, context-sensitive design that allows scenarios to be modified or extended based on regional practices. In addition, incorporating user-customizable content and input from local educators and therapists during development may help ensure the intervention is both inclusive and representative of the lived experiences of the target population (Kandalaft et al., 2013; Dechsling et al., 2021).

6.3. Usability, Technical Challenges, and Sustainability

Regarding RQ3, feedback from caregivers and therapists highlighted both the strengths and limitations of the VR system’s usability. The immersive design and gamified elements were generally effective in maintaining children’s engagement and attention during sessions. However, several usability concerns were raised, particularly regarding the physical fit of the Meta Quest 2 headset for younger users. Some children experienced discomfort due to the headset’s size and weight, which required frequent adjustment. This issue is consistent with previous concerns about age-appropriateness and device ergonomics in pediatric populations, especially given manufacturer guidelines recommending VR use for ages 13 and older (Peli, 1990; Rushton & Riddell, 1999). While no serious adverse effects were observed in this study, minor sensory fatigue and eye strain were reported—underscoring the need for continued caution and close monitoring when applying VR in children (Li et al., 2018; Tychsen & Foeller, 2019).

In addition to physical comfort, the children’s experiences within the virtual environment itself were generally positive. Participants appeared calm, engaged, and able to complete the scenarios without distress, and no sessions were interrupted due to discomfort inside the virtual world. Importantly, the VR design incorporated several sensory-sensitive features informed by related research on immersive environments for children with ASD. Audio elements were carefully calibrated to avoid overlapping sounds and excessive volume, while visual stimuli were designed to minimize flickering lights and abrupt transitions. The user interface emphasized simplicity and clarity, enabling intuitive navigation with minimal cognitive demand. Scene composition also avoided visual clutter, crowded spaces, and overstimulating environments, all of which have been reported as potential triggers for sensory overload in ASD populations (Soltiyeva et al., 2023; Zhang et al., 2025). Despite these precautions, individual differences in sensory processing suggest that some children may still experience cognitive load or emotional fatigue, particularly in more visually or auditorily complex scenarios.

From a technical standpoint, the current VR system lacked adaptability and dynamic content. Caregivers and therapists noted that scenarios often felt repetitive, and the system offered limited responsiveness to varied user behaviors. The scripted nature of interactions reduced opportunities for personalized learning or spontaneous social exchanges—an important factor for children with ASD, who often benefit from flexible, responsive environments. The absence of adjustable audio-visual settings also posed challenges for children with sensory sensitivities. These limitations highlight the need for future systems to incorporate adaptive mechanisms, such as real-time branching paths or AI-driven scenario adjustments, to better align with individual learning trajectories (Lv, 2023; Feuerriegel et al., 2023).

Sustainability emerged as another key challenge. While the program was successfully piloted in several Thai schools, broader implementation requires more than just access to hardware. Participants cited practical barriers such as limited physical space, session management logistics, and a lack of trained staff to facilitate VR use. Although some schools expressed willingness to reinvest in the system, long-term adoption will depend on structural support. This includes policy backing, government subsidies for equipment, and integration into national special education frameworks. Sustainability also hinges on continuous content development and support infrastructure. As noted by therapists, the novelty effect may diminish over time if scenarios are not regularly refreshed. Integrating generative AI may offer a solution by enabling dynamic content creation and scenario updates tailored to user needs (Lv, 2023; Feuerriegel et al., 2023). However, its implementation must be carefully aligned with pedagogical goals and ethical guidelines to ensure meaningful and equitable outcomes.

6.4. Limitations and Future Work

While this study provides promising evidence for the use of immersive VR to support social skill development in children with ASD, several limitations must be acknowledged. First, the relatively small sample size and quasi-experimental design limit the generalizability of the findings. The absence of random assignment or blinding introduces the possibility of selection bias and expectancy effects. Additionally, although pre-post comparisons revealed significant within-group improvements, ANCOVA results indicated that differences between groups were not statistically significant after controlling for baseline scores suggesting that some observed effects may be partially attributable to pre-existing differences.

Second, the intervention scenarios were largely scripted and lacked adaptive interactivity, which may have constrained opportunities for personalized learning or spontaneous behavior. While gamification and coaching tools were integrated to enhance engagement, the system did not dynamically respond to individual performance or behavioral variability. Furthermore, although the program was co-developed with Thai professionals, it did not incorporate deeper cultural markers such as specific gestures, school customs, or communication styles unique to Thai contexts—limiting its cultural resonance and real-world applicability.

Third, the accessibility and cost of immersive VR remain notable barriers to widespread implementation. Although the use of standalone headsets like the Meta Quest 2 offers a more affordable alternative to earlier, tethered systems, the initial investment in hardware, software development, and training may still be prohibitive for under-resourced schools or clinical settings. Ensuring equitable access to such interventions will require future research to explore cost-effective delivery models, such as mobile VR platforms, shared-use frameworks, or integration with existing technological infrastructure. Partnerships with governmental agencies, NGOs, and educational institutions could also support broader deployment and long-term sustainability.

Fourth, the scripted and linear nature of the intervention scenarios presents limitations in terms of pedagogical responsiveness and learner personalization. While the modules were designed to align with Kolb’s Experiential Learning Theory, the system did not support dynamic branching, real-time decision-making, or adaptive feedback based on learner behavior. This constrains the extent to which the experience reflects the full experiential learning cycle—particularly the stages of reflective observation and active experimentation. For neurodiverse learners, such as children with ASD, the absence of tailored interaction pathways may limit engagement and reduce opportunities for spontaneous or exploratory learning. Future research should investigate the integration of adaptive learning technologies, such as AI-driven scenario branching and real-time behavior tracking, to improve pedagogical responsiveness and support individualized learning trajectories.

Another important limitation is the absence of consideration for individual player profiles and personality traits in relation to gamification. Prior research demonstrates that learner typologies and personality dimensions can significantly shape engagement, motivation, and learning outcomes in game-based applications (Vergara et al., 2022; Sipone et al., 2025) Incorporating player profiles into VR-based interventions could therefore improve personalization and better align gamified features with learner needs. While this study did not explicitly address such factors, future work should explore personality-informed and HEXAD-based design approaches to strengthen adaptability, inclusivity, and learner experience in immersive learning systems.

Finally, while this study reported generally positive outcomes, it is important to acknowledge that not all VR interventions in the literature yield uniformly favorable results. Prior research has noted challenges such as sensory overstimulation, attentional fatigue, and motion sickness in some users, particularly among children with heightened sensory sensitivities (Li et al., 2018; Dechsling et al., 2021; Tychsen & Foeller, 2019). Although these effects were not observed in our sample—likely due to session duration, careful design, and therapist support—future work should further investigate the conditions under which VR may present cognitive or emotional strain. The positive outcomes observed here should therefore be interpreted within the context of a controlled and culturally tailored environment, and may not generalize to all settings or user profiles.

Future research should address these limitations through more robust experimental designs, including larger and more diverse samples, randomized assignment, and longitudinal follow-up to assess sustained learning and real-world generalization. Technically, future systems should explore adaptive and emotion-sensitive VR platforms potentially supported by generative AI that can tailor content based on user responses and contextual needs. From a sociocultural perspective, structured frameworks for cultural adaptation should be applied to ensure that interventions align with the lived experiences and norms of target populations. Additionally, investigating the integration of VR into multi-tiered educational or clinical models could inform sustainable pathways for broader implementation in low- and middle-income countries.

7. Conclusions

This study explored the use of an immersive, gamified VR intervention grounded in Experiential Learning Theory to support the development of social reciprocity, affective expression, and daily living skills in children with ASD. The findings demonstrated that the program significantly improved social communication and social motivation, particularly in structured social scenarios. However, ANCOVA results indicated that between-group differences in post-test scores were not statistically significant after adjusting for baseline scores, suggesting that the observed improvements may have been partially influenced by pre-existing differences. These outcomes suggest that well-designed VR experiences featuring goal-oriented tasks and feedback-driven learning cycles can provide valuable scaffolding for children with ASD, especially in promoting active participation and behavioral engagement.

However, the intervention showed limited effects in more complex domains such as social awareness and repetitive behaviors, pointing to the need for emotionally adaptive interactions and targeted sensory regulation strategies. Additionally, while participants perceived the content as relevant and engaging, concerns were raised about the rigidity of scenarios and limited cultural specificity. These findings underscore the importance of designing VR systems that are both flexible and contextually grounded, incorporating dynamic content, culturally familiar elements, and personalized pacing to meet the diverse needs of neurodiverse learners.

To ensure future effectiveness and generalizability, researchers should adopt more robust designs with larger, randomized samples and explore adaptive VR systems that respond to users’ behavioral variability. Looking ahead, the long-term success and scalability of VR-based interventions in educational and clinical settings will require a multi-stakeholder approach combining technical innovation with policy support, teacher training, and integration into inclusive education frameworks. Advances in generative AI offer promising opportunities for enhancing scenario variability and user responsiveness, but their application must be guided by pedagogical intent and cultural sensitivity. Ultimately, immersive VR has the potential to become a complementary tool within broader ASD support systems when embedded within culturally responsive, sustainable, and learner-centered design ecosystems.