Next Article in Journal
Smart Grid Intrusion Detection System Based on Incremental Learning
Previous Article in Journal
A Machine Learning Approach to Investigating Key Performance Factors in 5G Standalone Networks
Previous Article in Special Issue
NRXR-ID: Two-Factor Authentication (2FA) in VR Using Near-Range Extended Reality and Smartphones
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Electronics
Volume 14
Issue 19
10.3390/electronics14193818
electronics-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Detailed Review of the Design and Evaluation of XR Applications in STEM Education and Training

by
Magesh Chandramouli
1,
Aleeha Zafar
1,* and
Ashayla Williams
1,2
1
Department of Computer Information Technology and Graphics, Purdue University Northwest, Hammond, IN 46323, USA
2
PIA XR LLC, Merillville, IN 46410, USA
*
Author to whom correspondence should be addressed.
Electronics 2025, 14(19), 3818; https://doi.org/10.3390/electronics14193818
Submission received: 19 August 2025 / Revised: 13 September 2025 / Accepted: 23 September 2025 / Published: 26 September 2025
(This article belongs to the Special Issue Emerging Technologies in Augmented, Virtual, and Mixed Reality: Advancing Human Experience in Digital and Physical Environments)
Browse Figures
Review Reports Versions Notes

Abstract

Extended reality (XR) technologies—including augmented reality (AR), virtual reality (VR), mixed reality (MR), and desktop virtual reality (dVR)—are rapidly advancing STEM education by providing immersive and interactive learning experiences. Despite their potential, many XR applications lack consistent design grounded in human–computer interaction (HCI), leading to challenges in usability, engagement, and learning outcomes. Through a comprehensive analysis of 50 peer-reviewed studies, this paper reveals both strengths and limitations in current implementations and suggest improvements for reducing cognitive load and enhancing engagement. To support this analysis, we draw briefly on a dual-phase learning model (L1–L2), which distinguishes between interface learning (L1) and conceptual or procedural learning (L2). By aligning theoretical insights with practical HCI strategies, the discussions from this study are intended to offer potentially actionable insights for educators and developers on XR design for STEM education. Based on a detailed analysis of the articles, this paper finally makes recommendations to educators and developers on important considerations and limitations concerning the optimal use of XR technologies in STEM education. The guidelines for design proposed by this review offer directions for developers intending to build XR frameworks that effectively improve presence, interaction, and immersion whilst considering affordability and accessibility.

1. Introduction

1.1. Background

XR technologies have gained prominence in STEM (Science, Technology, Engineering, and Mathematics) education by offering immersive, spatially enriched learning environments [1]. These technologies represent a natural evolution of experiential learning through interaction [2] by allowing learners to visualize complex systems, manipulate 3D models, and perform procedural tasks that foster hands-on experience without the constraints of physical labs [3,4]. Whether through MR-based engineering simulations or AR-supported molecular visualization, XR applications are increasingly positioned as scalable solutions for addressing instructional gaps in science and engineering disciplines [3]. However, the effective integration of XR into education demands more than immersive visuals. It requires thoughtful attention to instructional design, user interaction, and cognitive load. Studies highlight the importance of key affordances—such as presence, agency, and multisensory feedback—in driving learning outcomes in immersive environments [5,6]. At the same time, HCI design principles, including intuitive design, contribute to the overall usability of XR systems. Visual clarity, gesture-based controls, and intuitive navigation are essential to ensuring that learners can focus on the instructional material rather than on managing interface complexity [7]. This reflects the intersection of cognitive theory and interface design, underscoring how both shape educational outcomes and inform the role of XR in modern STEM learning.

1.2. Gaps and Challenges

Despite widespread interest, many XR systems used in STEM learning environments exhibit inconsistent design, lack instructional scaffolding, and risk overwhelming learners with cognitive demands [6]. Studies show that poor interface design can divert attention from core learning goals, thereby compromising the transfer of procedural knowledge [5,8]. Moreover, frameworks like the Cognitive Affective Model of Immersive Learning (CAMIL) [5] have shown that presence and agency must be intentionally designed into XR systems—not assumed as default outcomes of immersion. Recent work also emphasizes the need to separate the cognitive processes involved in using an XR interface from those involved in domain-specific learning. To support this distinction, we draw lightly on the L1–L2 model of learning introduced in prior XR training research [8]. This model distinguishes between L1 (learning how to use the system) and L2 (learning the actual STEM content). Although not a central framework of this paper, this distinction helps identify where XR applications succeed or fall short in supporting deep learning. A failure to manage the L1 demands often leads to increased cognitive load, reducing the learner’s ability to engage meaningfully with the L2 content [6]. In addition, the field lacks a consolidated synthesis of how different XR platforms (AR, MR, VR, and dVR) address—or neglect—core learning goals like spatial reasoning, procedural mastery, and engagement. This paper seeks to address this gap. Unlike prior reviews that primarily catalog XR outcomes, this paper advances the field by synthesizing findings through the dual-phase L1–L2 learning framework and HCI principles. This combined lens provides a unique perspective, emphasizing how usability and interface design mediate the learner’s ability to transition from surface-level interaction (L1) to domain-level learning (L2).

1.3. Research Questions

This paper is guided by the following research questions:
  • How are the principal aspects of XR including presence, immersion, and interaction applied in current XR applications for STEM education?
  • What challenges do learners face due to cognitive load, interface design or limited adaptability in XR systems.
  • What design guidelines can be derived to support accessible, engaging, and cognitively effective XR learning environments?
In the context of this study, it is envisaged that the integration of theoretical concepts and practical applications will help educators and developers design XR environments that are both meaningful as well as impactful.

2. Theoretical Framework

The design of XR systems for STEM education can be better understood and explained through a combination of learning theories, cognitive models, and user-centered design principles. These frameworks help contextualize how immersive environments support—or hinder—learner engagement, skill acquisition, and knowledge transfer. This section outlines key theoretical perspectives relevant to the evaluation and development of XR-based learning environments.

2.1. Cognitive Affective Model of Immersive Learning (CAMIL)

Immersive Virtual Reality (IVR) has been shown to enhance learning outcomes; however, there needs to be a proper theoretical foundation regarding how this occurs. This gap is addressed through the CAMIL, developed by Makransky and Petersen [5]. CAMIL emphasizes that instructional design should leverage the most effective affordances of the medium to influence learning outcomes, rather than relying solely on the degree of immersion. The model identifies two key affordances that drive immersive learning: presence and agency. Presence refers to the sensation of “being there” in the environment being studied, enabling interaction with others and fostering heightened concentration. Agency describes the learner’s ability to influence and enact meaningful changes within the environment, which supports motivation. It allows learners to manipulate virtual objects, make informed decisions, and track the results of their actions. These affordances operate through six cognitive and affective drivers of learning: interest, intrinsic motivation, self-efficacy, embodiment, cognitive load, and knowledge acquisition. While CAMIL has significantly advanced understanding of immersive learning, it may be limited in contexts where complex interactions or usability barriers divert cognitive resources. As this review shows, many XR systems struggle to balance sensory immersion with user control, leading to high extraneous load alongside strong visual presence.

2.2. Constructivist and Cognitive Load Theories

The constructivist learning paradigm emphasizes that learners actively construct their own knowledge through interaction and exploration [9]. XR technologies align naturally with this approach, allowing learners to manipulate virtual objects, observe outcomes, and test hypotheses in safe, controlled environments. Systematic reviews of AR in STEM education confirm this alignment, showing its effectiveness in promoting spatial reasoning and inquiry-based learning, while also noting challenges with scalability and instructional integration [10]. Such experiential engagement fosters critical thinking and procedural learning, particularly within STEM disciplines. Cognitive Load Theory [6] further highlights the importance of managing the learner’s working memory. Poorly designed XR systems can introduce excessive extraneous load—through confusing navigation or overly complex visuals—which in turn reduces germane load, the mental effort directly devoted to learning. To optimize cognitive processing, effective XR environments incorporate multimodal cues, gradually increasing task difficulty, and providing scaffolded feedback.

2.3. Human–Computer Interaction and Usability

XR learning environments with the potential to deliver effective learning experiences must be grounded in sound HCI principles [7]. Elements such as natural navigation, responsive sensor inputs, and multimodal feedback are essential to reducing user frustration and maintaining flow. When XR interfaces lack intuitive, responsive controls, learners tend to focus disproportionately on L1 learning—figuring out how the system operates—rather than engaging with L2 learning, which centers on the STEM content itself [8].

2.4. L1–L2 Learning in XR Systems

A recent work [8] in XR-based training environments introduced a dual-phase learning model—L1–L2 Learning Theory—to conceptualize how users learn within immersive systems [8]. L1 refers to learning how to navigate and interact with the XR interface, while L2 denotes the acquisition of domain-specific procedural or conceptual knowledge. Although not a replacement for broader theories, the L1–L2 distinction highlights a practical challenge: if XR systems are poorly designed, L1 demands dominate and detract from effective L2 learning.
In this review, we use the L1–L2 perspective to support our evaluation of interface complexity, cognitive demands, and instructional effectiveness across XR studies. A visual diagram (Figure 1) later in the paper illustrates how these phases interact to influence learning outcomes.

3. Methodology

This review adopts a structured and transparent methodology to examine how XR technologies—including AR, MR, and VR—are applied in STEM education. The primary aim was to evaluate how XR systems support learning through presence, affordances, agency, and HCI principles, while identifying recurring design challenges such as cognitive overload and interface complexity. A comprehensive literature search was conducted across major academic databases including IEEE Xplore (https://ieeexplore.ieee.org), SpringerLink (https://link.springer.com), PubMed (https://pubmed.ncbi.nlm.nih.gov), ScienceDirect (https://www.sciencedirect.com), Google Scholar (https://scholar.google.com). ResearchGate (https://www.researchgate.net) was used as a supplementary platform to access full texts and identify relevant studies, not as a bibliographic database. The search focused on studies published between 2010 and 2024 and used keyword combinations such as “XR in STEM education,” “AR/VR affordances,” “immersive learning and cognitive load,” and “HCI in extended reality systems.” Studies were included if they were published in English, peer-reviewed, and explicitly addressed educational applications of XR in STEM disciplines. Only articles that reported learning outcomes or analyzed key elements of XR design—such as interaction, usability, or presence—were retained for review.
From an initial pool of 250 articles, 150 remained after the removal of duplicates and non-relevant titles. Following abstract and full-text screening, a final sample of 50 articles was selected based on the relevance of their research questions, the clarity of their design evaluation, and their contribution to understanding XR-supported STEM learning. Studies focused solely on non-educational XR applications (e.g., entertainment, marketing) or lacking methodological rigor were excluded. Although the selection process was informed by general PRISMA—Preferred Reporting Items for Systematic Reviews and Meta-Analyses—guidelines [11] to ensure transparency and replicability, the emphasis of this review remains interpretive and design-driven rather than strictly meta-analytical. Figure 2 provides a visual summary of the article selection workflow. For each included study, key characteristics were extracted and thematically analyzed. These included the type of XR platform (AR, MR, VR, or desktop VR), the subject domain (e.g., engineering, chemistry, medicine), reported learning outcomes, interface design strategies, presence or agency indicators, and observed barriers such as cognitive overload or interaction breakdowns. These insights inform the synthesis tables and thematic discussions in the following sections.

4. Findings

This section presents a thematic synthesis of 50 peer-reviewed studies examining the use of XR technologies in STEM education. Findings are organized according to three primary constructs that emerged across literature: presence, affordances, and agency. These constructs are also evaluated considering HCI principles and the L1–L2 learning framework discussed earlier.

4.1. Presence and Immersion

Presence—the sense of “being there” in the XR environment—was frequently associated with increased learner engagement, motivation, and spatial comprehension [12,13,14]. High-fidelity environments that integrated visual, auditory, and haptic cues supported greater immersion and more effective concept visualization, particularly in domains requiring 3D spatial reasoning, such as molecular chemistry or mechanical engineering [15,16]. However, multiple studies emphasized that presence alone does not ensure learning gains. This finding echo broader literature on immersive learning’s influence on user motivation and focus [17]. In several cases, heightened sensory immersion led to cognitive overload when not paired with intuitive interface design [18,19]. This aligns with the L1–L2 framework, as learners often became stuck at the interface level (L1), unable to engage deeply with domain content (L2). For instance, recent XR demonstrations involving realistic gestures and interactive virtual components such as in Figure 3, significantly improved learners’ presence by combining intuitive interfaces and structured feedback [20]. XR systems that combined high presence with streamlined navigation and structured feedback were more effective at improving both task accuracy and knowledge transfer.

4.2. Affordances and Interaction

Affordances—defined as the actionable properties of virtual objects and environments—play a central role in how learners interact with XR content. Well-designed affordances allow learners to manipulate 3D objects, receive immediate feedback, and engage in exploratory learning [21,22,23]. Studies in procedural training and engineering simulation consistently found that gesture-based inputs, context-aware overlays, and progressive hints enhanced cognitive engagement and knowledge retention [24,25]. Contemporary analyses have further compared AR, VR, and MR modalities, suggesting that affordance design should be tailored to modality-specific strengths for maximum effectiveness [26]. Conversely, affordances that were either too rigid or overly complex increased extraneous cognitive load and contributed to user frustration [27,28]. For example, Chandramouli et al. demonstrated how clearly defined virtual interaction methods in digital manufacturing modules effectively reduced user errors and improved engagement in procedural tasks as Figure 4 exemplifies these task-aligned affordances from a digital-manufacturing VR module, showing how interaction design scaffolds complex procedures [29]. Novice users benefitted from scaffolded affordances—gradual exposure to interaction options that aligned with learning objectives. As summarized in Table 1, the most effective XR environments offered affordances that directly supported the transition from L1 to L2 learning by minimizing ambiguity, reducing error rates, and enhancing learner confidence.

4.3. Agency and Autonomy

Agency—the learner’s perceived control over their learning journey—was another recurring theme across high-performing XR systems [30,35]. Environments that allowed for user-driven pacing, variable manipulation, or scenario-based branching were strongly correlated with higher intrinsic motivation and deeper engagement. These features were especially effective in simulations used for medical training, lab safety, or equipment assembly [36,37]. This finding is consistent with the meta-analyses showing that immersive VR environments can both motivate and challenge learners, depending on the balance between autonomy and structured guidance [38]. For instance, recent XR training environments designed for active user participation as demonstrated in Figure 5, enabled learners to make real-time decisions, significantly enhancing their engagement and sense of control over the learning process [39]. Importantly, multiple studies found that agency was tightly coupled with interface fluency: when users felt burdened by unfamiliar controls or awkward interaction mechanics, their sense of autonomy diminished, and learning outcomes suffered [2,37]. The L1–L2 theory provides a helpful explanation—interface learning barriers at L1 reduce the learner’s ability to fully engage with domain tasks at L2. XR systems that promoted agency often achieved this by combining adaptive interfaces with clear user feedback, reducing unnecessary complexity and preserving decision-making control. A synthesis of these patterns appears in Table 2.

4.4. Summary of Findings

The review identified clear patterns across studies examining the use of XR in STEM education. High presence environments supported learner engagement and spatial reasoning, but only when paired with intuitive interface design that minimized cognitive load. Affordances aligned with task objectives—such as gesture-based interaction, contextual feedback, and scaffolded guidance—were consistently linked to improved understanding and task performance. Learner agency also emerged as a key factor in motivation and autonomy, particularly in simulations that enabled adaptive control and personalized learning pathways. These findings highlight the interconnected nature of presence, affordances, and agency in effective XR learning systems. Well-designed environments that prioritize both interface fluency (L1) and domain-level learning outcomes (L2) provide learners with immersive, interactive, and self-directed educational experiences. The following section examines how these insights inform theoretical understanding and the design of XR-assisted STEM learning environments.

5. Discussion

This part critically looks at the findings in view of learning theory and human–computer interface design, especially the L1–L2 framework presented above. It demonstrates the combinations of presence, affordances, and agency to shape learning during XR-based STEM education, while also outlining implications for future research and design within practical applications.

5.1. Theoretical Implications

The synthesis presented here reinforces the proposed advantages of the L1–L2 model within XR learning environments. When learners lack sufficient training on unfamiliar interfaces or diverse interaction mechanisms, much of their cognitive capacity is consumed by L1-level tasks—such as operating interface functions or navigating virtual spaces. This, in turn, reduces their ability to engage with domain-specific content at the L2 level, where the conceptual understanding of material and skill acquisition occurs [33,40]. For example, studies using AR scaffolding demonstrated that gradual task sequencing reduced L1 load and improved conceptual reasoning [31]. Similarly, VR-based engineering simulations reported higher transfer of procedural knowledge when interface complexity was minimized [22,35]. Systems that streamline interface learning and offer intuitive feedback tend to facilitate a smoother transition to deep learning tasks, as supported by several studies reviewed [16,18,24]. In practical classroom contexts, the L1–L2 framework can guide the balance between interface learning (L1) and domain-specific learning (L2). For instance, structured tutorials for AR or VR interfaces can reduce the cognitive demands of navigation and control, enabling students to dedicate more cognitive resources to disciplinary reasoning and problem-solving tasks. Scaffolding strategies, such as stepwise introductions to XR affordances or adaptive prompts, will ensure that learners move beyond surface-level interaction into deeper conceptual engagement. As previously discussed, studies of XR-based scaffolding in online and blended learning environments support this approach, showing that gradual exposure to XR features improves both usability and knowledge transfer [22,31]. In this sense, the L1–L2 framework is not only a theoretical construct but also a pedagogical tool for educators designing XR-enabled classrooms. This pattern is consistent with principles from cognitive load theory, particularly Mayer’s distinction between extraneous and germane load [27]. High-immersion systems must carefully balance realism with interaction usability to avoid cognitive overload, as emphasized by Bowman et al. in their exploration of immersion’s effect on spatial understanding and learning outcomes [41]. Thus, presence must not be treated as an isolated goal but rather as a supportive feature that enhances, rather than complicates, the learning process. While presence, affordances, and agency are well-established in XR research, their significance in the current synthesis lies in their interaction with emerging directions. The convergence of these constructs with AI-driven adaptive feedback, neuroadaptive XR that responds to biometric signals, and standardized evaluation frameworks for learning impact represents an evolution of XR research beyond traditional immersive metrics [42,43].

5.2. Design Implications

The findings suggest several critical considerations for designing effective XR systems in STEM education. Heuristic evaluation methods for VR emphasize clarity, navigation, and interaction coherence [44]. Interface simplicity must be prioritized, particularly in the early stages of learner engagement. When users can easily interpret gestures, navigate scenes, and receive clear feedback, their cognitive resources are freed for domain learning [23,30,34]. Scaffolded interaction strategies have proven to mitigate overload and improve knowledge retention [31]. Affordances must also be well-aligned with the user’s skill level. Studies showed that beginners benefit more from guided sequences, while more experienced learners perform better with systems that offer flexibility and open-ended exploration [21,24,36]. Learner agency emerged as a significant contributor to sustained motivation and task completion. Environments that allowed participants to manipulate variables, adjust pacing, or follow branching scenarios fostered greater cognitive engagement and a sense of ownership over the learning process [37,45]. However, this sense of control is dependent on interface fluency; when interaction is hindered, agency is quickly diminished. Systems that balance adaptive guidance with user autonomy are therefore more likely to support effective L1–L2 progression.

5.3. Limitations in Current XR Research

Despite increasing adoption, XR research in education still exhibits notable limitations. Many studies emphasize short-term performance indicators without examining long-term learning gains or transferability to real-world contexts [15,35]. A decade-long review of empirical studies similarly emphasized the need for more nuanced outcome metrics [37]. Very few investigations evaluate both interface usability and domain mastery simultaneously, leaving an incomplete understanding of how interaction design mediates learning [40,46]. Additionally, user diversity remains insufficiently addressed. Learners with limited exposure to spatial computing, or those with accessibility needs, are often excluded from evaluation cohorts [47]. This lack of inclusivity limits the generalizability and equity of current findings. Addressing these gaps will require research designs that integrate usability testing, accessibility assessments, and longitudinal tracking into XR learning evaluations. Another challenge that is faced with is that, despite the promise of XR, adoption in educational institutions faces practical barriers such as high hardware and facility costs, lack of technical infrastructure, and limited faculty expertise. However, centralized XR labs, shared device programs, and targeted professional development have shown promise in mitigating these challenges [48,49,50].

5.4. Future Research Directions

While this review consolidates current insights into XR-based STEM learning, several research gaps remain. Future studies should evaluate the long-term learning outcomes of immersive XR environments rather than focusing solely on short-term task performance. The transferability of skills learned in XR to real-world contexts remains underexplored and requires longitudinal investigations. Additionally, comparative research across XR modalities—such as AR, VR, and spatial augmented reality (SAR)—is needed to determine which technologies are most effective for specific competencies or learning domains. Standardized criteria of evaluation can be obtained with meta-analyses and benchmark frameworks [51,52]. There are also other emerging areas that have been around long enough to be explored, including AI-enhanced XR, multimodal interaction, and neuroadaptive interfaces. The innovations present the possibility of real-time alteration, depending on the response of learners or their emotional state but need to be verified in various learner samples and real educational environments [32,53]. Future work should also aim to advance the field through quantitative meta-analyses that compare effect sizes across modalities (AR vs. VR vs. MR), and empirical validation of the L1–L2 framework in real-world educational settings. Moreover, systematic comparative studies between XR-enhanced and traditional or blended learning models would deepen our understanding of added value and scalability. The idea of inclusive design must become the basis of research in XR, exposing the technology to users with various physical capabilities and neurodiversity, as well as digital literacy. In addition to this, looking ahead, XR education research must also expand into AI-augmented environments where intelligent agents adaptively scaffold tasks in real time, tailoring support to learner needs. Neuroadaptive XR, informed by eye-tracking, EEG, or physiological signals, has the potential to dynamically adjust system demands to reduce cognitive overload. Additionally, the field would benefit from standardized frameworks for evaluating XR effectiveness, including shared metrics for presence, usability, and learning transfer. Such innovations promise to extend the foundational constructs of presence, affordances, and agency into new frontiers of evidence-based practice [42,43].

5.5. Practical Implications

The findings and framework presented in this paper offer practical value for both educators and XR developers. For instructors, understanding how interface fluency (L1) supports conceptual learning (L2) enables more informed choices when integrating XR into the curriculum. This is particularly relevant for competency-based education, where immersive environments can simulate real-world challenges with high fidelity [54,55]. Educators can also tailor XR use based on learner profiles, emphasizing guided sequences for novices and open-ended exploration for advanced students. From an institutional perspective, aligning XR integration with accreditation outcomes—such as those set by ABET—can strengthen program-level assessment and curricular innovation [56]. For developers, the implications point toward the need for learner-centered design. Systems must balance immersion with usability, embed adaptive affordances, and offer scaffolds that promote learner agency without overwhelming novices. Incorporating multisensory interaction, ethical safeguards, and feedback mechanisms will improve both user satisfaction and learning efficacy [57]. Furthermore, aligning XR experiences with constructivist goals will support deeper learning outcomes and real-world applications [58].

6. Proposed Design Framework

Based on the synthesis of reviewed studies and grounded in the L1–L2 learning theory, this paper proposes a design framework for effective XR-based STEM education. The framework recognizes the layered nature of learning in immersive systems and positions interface fluency (L1) as a prerequisite for meaningful domain engagement (L2). It emphasizes five interdependent pillars—presence, affordances, agency, cognitive load reduction, and inclusivity—that collectively shape how learners progress from system navigation to conceptual mastery.

6.1. Enhancing Presence

Presence is not treated as an isolated outcome but as a facilitator of cognitive and emotional engagement. XR systems that employ realistic textures, directional audio, and immersive visuals can significantly improve users’ spatial presence and contextual relevance [15,17]. Tools such as head-mounted displays (HMDs) combined with multi-sensory cues—including haptic feedback—further increase learner engagement by simulating real-world tasks [15,46,54]. For example, applications in virtual surgery, engineering simulations, and racket sports drills demonstrate that fidelity and realism can improve focus and task performance [53,55]. Importantly, presence must be supported by intuitive navigation and task-aligned interactions, as excessive visual complexity or sensory stimulation may lead to distraction and cognitive overload [24,40,46].

6.2. Strengthening Agency

Agency enables learners to feel in control and play an active role in shaping their experience. XR systems that support interactive decision-making and hypothesis testing—such as scenario-based VR and hands-on AR experiments—lead to increased motivation and deeper conceptual learning [15,37,45]. Responsive systems that adapt task complexity to learner input (via algorithms or branching pathways) allow for a personalized pace of engagement, enhancing persistence and reducing frustration [23,56]. Furthermore, collaborative XR environments, where learners take on roles or solve problems jointly, promote both social and metacognitive development [57]. This approach aligns with constructivist principles and has been shown effective in cadet training [52], medical education [51], game-based learning [25] and STEM classrooms [35].

6.3. Optimizing Affordances

XR systems incorporate affordances—features or cues that support learner interaction and influence learning outcomes. When designed intuitively, these affordances reduce unnecessary cognitive load and improve task fluency. Elements such as voice commands, gesture-based controls, and haptic feedback make interactions more natural and interfaces easier to navigate [17,30,32]. Evidence from robotics training and engineering simulation programs shows that well-structured information and object interactivity enhance learners’ understanding of complex spatial relationships [32]. Moreover, dynamic affordances that appear or disappear based on user progress help scaffold the learning process while preventing overload [12,56]. Recent work in design education highlights how AR affordances can support conceptual modeling but also raise unique challenges in managing complexity [59]. Aligning instructional elements with educational objectives (e.g., Bloom’s taxonomy) ensures learners progress through increasingly complex levels of thinking [60].

6.4. Reducing Cognitive Load

The Cognitive Load theory requires prioritizing the levels of mental effort in order to achieve effective learning. This is where XR environments may be very relevant because they can simplify interfaces, scaffold task complexity and provide timely feedback. These strategies make it possible to keep the learners within the core learning targets without overwhelming them. XR systems should avoid visual and auditory clutter and other extraneous features that are not linked to the learning material in order to avoid distraction. Simplified designs with identifiable parts (e.g., engine parts highlighted; circuit elements labeled) ideally contribute to attracting the attention of the learners and alleviating mental burden when working on a complex task [6,16,27]. Figure 6 shows an annotated step in SPEXTRA where minimal on-screen elements, progressive prompts, and clear labels reduce extraneous load during part placement, aligning with our cognitive-load guidelines [20]. In addition, real-time feedback also increases the learning process by enabling the user to observe their improvement and give corrective actions in course of undertaking the tasks. Success signals, progress bar, and contextual prompts are ways to keep learners engaged or motivate them to spend less mental energy or cognitive effort [18]. To achieve the best outcomes, this feedback needs to be closely aligned with user behaviors in AR and VR settings to facilitate iterative learning capabilities and strengthen knowledge [16]. The importance of gradual complexities on tasks cannot also be overlooked in the management of cognitive loads. Good XR learning environments provide simple instructions or tasks first and allow an easy transition to more complex ones. This scaffolding will aid the skills course and allow the learners to find confidence gradually [31,47]. Studies confirm that AR’s effect on cognitive load is context-dependent, sometimes alleviating extraneous effort but in other cases creating overload if not carefully designed [61]. Learner persistence and success is likely to occur when the task demands, and their mental preparations are matched. On the other hand, inexpertly designed user-interfaces may create undue load—particularly in the case of new users. As one of the solutions to this situation, the XR systems will need to incorporate usability-focused concepts like the ease of navigation, clarity of visuals, and multimodal interactions that combine visual, auditory, and tactile feedback. In addition to making comprehension easier, these design features also minimize the chances of cognitive overload, creating a more accessible and productive learning environment [27,44,62].

6.5. Ensuring Accessibility and Inclusivity

For XR to be genuinely effective in STEM learning, it must accommodate varied learner needs and be made accessible in terms of physical, sensory, cognitive diversities. Accessibility is more than an ethical standard or requirement, but it is also a design aspect that widens the scope of immersive learning technologies and their efficacy. Accessibility should be considered at the design phase itself and not as a patch-up exercise. Concrete accessibility design guidelines are needed for XR educational tools. Strategies such as ensuring compatibility with screen readers, implementing alternative input methods (e.g., voice or switch-based controllers), offering simplified UI overlays for neurodiverse users, and leveraging low-cost mobile XR systems can significantly broaden access—especially in under-resourced contexts [64,65,66]. Moreover, features such as text to speech, ability to change the levels of contrast, and different ways of interacting make sure that learners with disabilities do not feel left out [47]. Also noteworthy is the involvement of learners belonging to underrepresented or marginalized groups in the design and testing of XR platforms. Involving these users helps identify potential barriers early and leads to more thoughtful, user-informed solutions. Beyond technical accessibility, content should reflect diverse cultural contexts and educational backgrounds. Localized interfaces, multilingual options, and contextually relevant scenarios foster stronger learner engagement and promote equity across learning environments [67]. A participatory design approach therefore, enhances system relevance, usability, and cultural alignment, fostering systems that reflect and respect diverse learner needs and experiences, as illustrated in prior research involving accessibility-oriented XR design depicted in Figure 7, highlights accessibility-minded choices (clear typography, captions, and modular pacing) that support a wider range of learners [63]. In addition, recent research has demonstrated the potential of AR in informal science learning to broaden access for diverse learner groups outside traditional classrooms [68].

6.6. Summary of the Design Framework

This design framework positions XR-based STEM learning as a layered, learner-centered process that begins with interface fluency (L1) and progresses toward deep conceptual engagement (L2). Drawing on the interdependent pillars of presence, affordances, and agency, the framework emphasizes the need for intuitive interaction, cognitively aligned feedback, and meaningful learner control. It also underscores the importance of reducing cognitive load and ensuring accessibility to accommodate diverse user needs. This design approach is summarized in Figure 8. By synthesizing insights from the literature and grounding the approach in the L1–L2 theory, the framework offers a coherent and adaptable guide for developers and educators seeking to maximize the pedagogical impact of XR environments in STEM education.

7. Conclusions

The present paper introduced an evidence-based literature review of XR, including AR, MR, and VR in the STEM teaching and learning contexts. Using a combination of constructivist learning theory, the cognitive load theory, and the learning framework presented in L1 and L2, the paper synthesized the information presented in 50 peer-reviewed sources to determine how XR systems influence learner engagement, concept comprehension, and skill acquisition. The design framework outlined in this study underlines the fact that interface fluency (L1) is the baseline to domain-level knowledge (L2) supported by the three pillars of designing: presence, affordances, and agency. Together, these elements support a learner-centered approach to integrating XR. The review demonstrates that XR technologies, when properly designed, can significantly enhance STEM learning outcomes. Environments that are immersive yet intuitive, visually rich yet cognitively efficient, and interactive yet adaptive can scaffold users’ progression from surface-level interaction to deep conceptual engagement. Presence alone is not a guarantee of learning—rather, it must be supported by thoughtfully designed affordances and meaningful learner agency to sustain motivation and promote self-regulated learning. However, the analysis also reveals major design and evaluation shortcomings. Issues such as limited longitudinal assessment, insufficient support for accessibility and inclusivity, and inconsistent alignment with learning goals persist across many studies. The current body of work frequently fails to holistically address both system usability and domain mastery, often overlooking the nuanced interplay between interface learning and educational outcomes.

7.1. Key Findings

The central finding of this review is that educational effectiveness in XR hinges not merely on technological sophistication, but on pedagogical alignment and user-centered design. When learners encounter steep interface learning curves, their cognitive resources are disproportionately consumed at the L1 level, leaving little room for engagement with STEM content. Systems that simplify navigation, provide timely and contextual feedback, and gradually escalate task complexity enable learners to transition more efficiently to L2-level engagement. Studies reviewed also show that affordances must be carefully tailored to task complexity and learner expertise. Well-structured XR environments enhance problem-solving skills, promote collaboration, and foster critical thinking when agency is enabled through adaptive guidance rather than overwhelming freedom. The success of such systems ultimately rests on the designer’s ability to integrate cognitive, perceptual, and emotional aspects of learning into the XR interface.

7.2. Contributions

Incorporating strengths and limitations of the design and implementation of XR technologies, this scholarly review enriches the understanding of their possibilities in STEM learning contexts. Considering HCI, the framework finds challenges in cognitive load, usability, and accessibility. This systematic approach enables educators and developers of XR systems to enhance awareness regarding interaction, engagement, and learning impact. The conclusions identified in this study support the effectiveness of XR in STEM education as a tool that facilitates otherwise impossible, engaging, and progressive learning. These elements are explained within the study context, demonstrating how XR technologies eliminate the gap between theory and practice, enhance problem-solving skills, and encourage teamwork and enthusiasm. They provide crucial directional advice for implementing XR to address the educational needs of the 21st century. In summary, this paper makes significant contributions to the evolving field of XR in STEM education by addressing issues related to conceptual systematic review, framework development, and educational implications. This is based on a systematic analysis of 50 peer-reviewed articles, wherein the present investigation examines how XR applications align with cognitive load and constructivist learning theories, offering a novel framework that bridges educational theory with XR design practice.

7.3. Implications for STEM Education

The findings of this review offer several actionable insights for educators, developers, and policymakers working at the intersection of XR and STEM education. For instructors, the integration of XR into the curriculum must be guided by a nuanced understanding of the learner’s journey—starting from interface fluency and progressing toward conceptual mastery. XR platforms can serve as powerful tools to simulate real-world challenges, encourage problem-solving, and provide learners with low-risk, high-fidelity training environments. Developers must prioritize user-centered design principles that accommodate diverse learning needs, including adaptive progression, multimodal interaction, and compatibility with assistive technologies. Ethical considerations such as cognitive overload, cybersickness, and user consent must also be addressed from the outset of design. Policymakers and institutional leaders should consider these frameworks when implementing XR-based instructional strategies, ensuring that initiatives are inclusive, scalable, and aligned with accreditation and workforce. For instance, prior research has demonstrated the successful implementation of XR in STEM education settings, highlighting how well-designed XR experiences can effectively bridge theoretical learning and practical application [20]. Figure 9 provides a practical example of how production-style instructional assets can be embedded in XR curricula to scale skills training while preserving clarity and consistency.

7.4. Final Remarks

The review presented in this paper offers a structured roadmap for the effective integration of XR in STEM education. By synthesizing insights from cognitive load theory, constructivist learning theory, and the CAMIL framework with the L1–L2 learning model, the study proposes a layered framework that emphasizes interface fluency, adaptive affordances, and learner agency. As XR technologies evolve in fidelity, accessibility, and reach, their success will depend not on novelty but on the quality of learning they facilitate. The framework introduced here offers a blueprint for designing XR systems that are immersive, inclusive, and pedagogically sound. Recent meta-reviews emphasize the nuanced differences between AR and VR, particularly regarding trade-offs between immersion, cognitive load, and long-term learning impact [69]. By closing the gap between theory and application, this study lays a foundation for future research and development, ensuring that XR continues to serve as a transformative force in preparing learners for the complex challenges of STEM fields in the 21st century.

Author Contributions

Writing—original draft, M.C., A.Z. and A.W.; Writing—review & editing, M.C. and A.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Purdue University Northwest, College of Technology Faculty Professional Development Fund and it is sincerely acknowledged.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Conflicts of Interest

Author Ashayla Williams was employed by the company PIA XR LLC. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References

  1. Milgram, P.; Kishino, F. A taxonomy of mixed reality visual displays. IEICE Trans. Inf. Syst. 1994, 77, 1321–1329. [Google Scholar]
  2. Lee, K. Augmented reality in education and training. TechTrends 2012, 56, 13–21. [Google Scholar] [CrossRef]
  3. Antoni, V.; Maurer, F.; Cesari, O.; Eichhorn, C.; Vilhjálmsson, H.H. AR in molecular visualization for chemistry education. Chem. Educ. Res. Pract. 2023, 23, 12–28. [Google Scholar]
  4. Carlson, P.; Hays, J.; May, J. Virtual training: Learning transfer of assembly tasks. IEEE Trans. Vis. Comput. Graph. 2015, 21, 770–782. [Google Scholar] [CrossRef]
  5. Makransky, G.; Petersen, G.B. The cognitive affective model of immersive learning (CAMIL): A theoretical research-based model of learning in immersive virtual reality. Educ. Psychol. Rev. 2021, 33, 937–958. [Google Scholar] [CrossRef]
  6. Sweller, J. Cognitive load theory. In Psychology of Learning and Motivatio; Mestre, J.P., Ed.; Academic Press: San Diego, CA, USA, 2011; Volume 55, pp. 37–76. [Google Scholar]
  7. Yang, H.; Li, X.; Sun, W. Gesture-based interactions in educational VR environments. IEEE Trans. Learn. Technol. 2020, 13, 432–440. [Google Scholar]
  8. Chandramouli, M.; Nere, K.; Williams, A.; Shimpi, J. Multimodal virtual reality & digital twin-based training for specialized process equipment in biotech/biopharma manufacturing. In Proceedings of the International Conference on Intelligent Computing, Virtual Augment Reality Simulations, Birmingham, UK, 25–27 July 2025. [Google Scholar]
  9. Hein, G.E. Constructivist Learning Theory; Institute for Inquiry: Fitzroy, VI, USA, 1991. [Google Scholar]
  10. Ibáñez, M.-B.; Delgado-Kloos, C. Augmented reality for STEM learning: A systematic review. Comput. Educ. 2018, 123, 109–123. [Google Scholar] [CrossRef]
  11. Page, M.J.; McKenzie, J.E.; Bossuyt, P.M.; Boutron, I.; Hoffmann, T.C.; Mulrow, C.D.; Shamseer, L.; Tetzlaff, J.M.; Akl, E.A.; Brennan, S.E.; et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ 2021, 372, 71. [Google Scholar] [CrossRef]
  12. Slater, M.; Wilbur, S. A framework for immersive virtual environments (FIVE): Speculations on the role of presence in virtual environments. Presence 1997, 6, 603–616. [Google Scholar] [CrossRef]
  13. Makransky, G.; Lilleholt, L. A structural equation modeling investigation of the emotional value of immersive virtual reality in education. Educ. Technol. Res. Dev. 2018, 66, 1141–1164. [Google Scholar] [CrossRef]
  14. Dede, C. Immersion, presence, and distributed learning. J. Comput. High. Educ. 2009, 21, 95–113. [Google Scholar]
  15. Radianti, J.; Majchrzak, T.A.; Fromm, J.; Wohlgenannt, I. A systematic review of immersive virtual reality applications for higher education. Comput. Educ. 2020, 147, 103778. [Google Scholar] [CrossRef]
  16. Parong, J.; Mayer, R.E. Learning science in immersive virtual reality. J. Educ. Psychol. 2018, 110, 785–797. [Google Scholar] [CrossRef]
  17. Freina, L.; Ott, M. A literature review on immersive virtual reality in education: State of the art and perspectives. In Proceedings of the International Scientific Conference eLearning and Software for Education, Bucharest, Romania, 23–24 April 2015; Volume 1, pp. 133–141. [Google Scholar]
  18. Makransky, G.; Terkildsen, T.S.; Mayer, R.E. Adding immersive virtual reality to a science lab simulation causes more presence but less learning. Learn. Instr. 2019, 60, 225–236. [Google Scholar] [CrossRef]
  19. Freeman, A.; Becker, S.A.; Cummins, M.; Davis, A.; Giesinger, C.H. Ananthanarayanan, NMC Horizon Report: 2017 Higher Education Edition; New Media Consortium: Austin, TX, USA, 2017. [Google Scholar]
  20. SPEXTRA NSF. Demo Mode. Available online: https://spextra-nsf.org/demo-mode (accessed on 30 July 2025).
  21. Bacca, J.; Baldiris, S.; Fabregat, R.; Graf, S.; Kinshuk. Augmented reality trends in education: A systematic review of research and applications. J. Educ. Technol. Soc. 2014, 17, 133–149. [Google Scholar]
  22. Dunleavy, M.; Dede, C. Augmented reality teaching and learning. In Handbook of Research on Educational Communications and Technology; Spector, J.M., Merrill, M.D., Elen, J., Bishop, M.J., Eds.; Springer: New York, NY, USA, 2014; pp. 735–745. [Google Scholar]
  23. Akçayır, M.; Akçayır, G. Advantages and challenges associated with augmented reality for education: A systematic review. Educ. Res. Rev. 2017, 20, 1–11. [Google Scholar] [CrossRef]
  24. Radu, I. Augmented reality in education: A meta-review and cross-media analysis. Pers. Ubiquitous Comput. 2014, 18, 1533–1543. [Google Scholar] [CrossRef]
  25. Chen, C.-H.; Wang, K.-C.; Lin, Y.-H. The comparison of solitary and collaborative modes of game-based learning on students’ science learning and motivation. J. Educ. Technol. Soc. 2015, 18, 237–248. [Google Scholar]
  26. Tene, T.; Marcatoma Tixi, J.A.; Palacios Robalino, M.D.L.; Mendoza Salazar, M.J.; Vacacela Gomez, C.; Bellucci, S. Integrating immersive technologies with STEM education: A systematic review. Front. Educ. 2024, 9, 1410163. [Google Scholar] [CrossRef]
  27. Mayer, R.E. Multimedia Learning, 2nd ed; Cambridge University Press: New York, NY, USA, 2009. [Google Scholar]
  28. Milgram, P.; Takemura, H.; Utsumi, A.; Kishino, F. Augmented reality: A class of displays on the reality-virtuality continuum. In Proceedings of the Telemanipulator and Telepresence Technologies, Boston, MA, USA, 31 October–1 November 1994; Volume 2351, pp. 282–292. [Google Scholar]
  29. Chandramouli, M.; Jin, G.; Heffron, J.; Fidan, I.; Cossette, M.; Welsch, C.A.; Merrell, W. Virtual reality education modules for digital manufacturing instruction. In Proceedings of the 2018 ASEE Annual Conference & Exposition, Salt Lake City, UT, USA, 24–27 June 2018. [Google Scholar]
  30. Jensen, L.; Konradsen, F. A review of the use of virtual reality head-mounted displays in education and training. Educ. Inf. Technol. 2018, 23, 1515–1529. [Google Scholar] [CrossRef]
  31. Zheng, B.; Lin, C.H.; Kwon, J.B. The impact of scaffolding on learners’ cognitive load in an online learning environment. Interact. Learn. Environ. 2020, 28, 450–465. [Google Scholar]
  32. Chandramouli, M.; Lei, T.; Chou, Y.M.; Huang, S.R.; Kolanuvada, S. Desktop virtual reality (VR) & ispace for GIS education through interactive virtual learning environments. ASEE Comput. Educ. J. 2016, 16, 91–104. [Google Scholar]
  33. Dalgarno, B.; Lee, M.J.W. What are the learning affordances of 3-D virtual environments? Br. J. Educ. Technol. 2010, 41, 10–32. [Google Scholar] [CrossRef]
  34. Dunleavy, M.; Dede, C.; Mitchell, R. Affordances and limitations of immersive participatory augmented reality simulations for teaching and learning. J. Sci. Educ. Technol. 2009, 18, 7–22. [Google Scholar] [CrossRef]
  35. Merchant, Z.; Goetz, E.T.; Cifuentes, L.; Keeney-Kennicutt, W.; Davis, T.J. Effectiveness of virtual reality-based instruction on students’ learning outcomes in K–12 and higher education: A meta-analysis. Comput. Educ. 2014, 70, 29–40. [Google Scholar] [CrossRef]
  36. Alhalabi, W. Virtual reality systems enhance students’ achievements in engineering education. Behav. Inf. Technol. 2016, 35, 919–925. [Google Scholar] [CrossRef]
  37. Mikropoulos, T.A.; Natsis, A. Educational virtual environments: A ten-year review of empirical research (1999–2009). Comput. Educ. 2011, 56, 769–780. [Google Scholar] [CrossRef]
  38. Acevedo, P.; Magana, A.J.; Benes, B.; Mousas, C. A systematic review of immersive virtual reality in STEM education: Advantages and disadvantages on learning and user experience. IEEE Access 2024, 12, 189359–189386. [Google Scholar] [CrossRef]
  39. SPEXTRA NSF. Active Training Mode. Available online: https://spextra-nsf.org/active-training-mode (accessed on 30 July 2025).
  40. Makransky, G.; Mayer, R.E. Benefits of taking a virtual field trip in immersive virtual reality: Evidence for the immersion principle in multimedia learning. Educ. Psychol. Rev. 2021, 33, 1533–1549. [Google Scholar] [CrossRef]
  41. Bowman, D.A.; McMahan, R.P. Virtual reality: How much immersion is enough? Computer 2007, 40, 36–43. [Google Scholar] [CrossRef]
  42. Liu, Y.; Chopra, A.; Dhakal, P.; Banerjee, A.; D’Mello, S.K. Toward neuroadaptive XR: Biometric sensing and adaptive interfaces in immersive learning. Comput. Educ. 2023, 194, 104673. [Google Scholar]
  43. Kivikangas, J.M.; Vogt, J.; Ravaja, N. A review of presence and measurement in virtual reality research. Front. Psychol. 2021, 12, 643. [Google Scholar]
  44. Sutcliffe, A.; Gault, B. Heuristic evaluation of virtual reality applications. Interact. Comput. 2004, 16, 831–849. [Google Scholar] [CrossRef]
  45. Davis, N.; Childs, M.; Howard, Y.; Cobb, S.; Hall, C.; Sanchez-Vives, A.; Newall, E.; Koren, N.; Howard-Jones, P.; Boniface, M. Immersive virtual reality as a tool to enhance experiential learning: A theoretical review. Comput. Educ. 2020, 144, 103710. [Google Scholar]
  46. Fowler, C. Virtual reality and learning: Where is the pedagogy? Br. J. Educ. Technol. 2015, 46, 412–422. [Google Scholar] [CrossRef]
  47. Almutairi, A.; Utray, F.M. Accessibility in virtual reality: A systematic literature review. Virtual Real. 2022, 26, 1065–1083. [Google Scholar]
  48. Smith, J. XR in Higher Education: Adoption Considerations and Recommendations. EDUCAUSE Review. January 2024. Available online: https://er.educause.edu/articles/2024/1/xr-in-higher-education-adoption-considerations-and-recommendations (accessed on 14 October 2024).
  49. Chen, R. Addressing XR Access in Schools via Shared Device Programs. Jobs for the Future. 2023. Available online: https://www.jff.org/vr-in-education-addressing-challenges-and-unlocking-possibilities/ (accessed on 14 October 2024).
  50. Thompson, L. Barriers to XR adoption for educators. In Proceedings of the National Conference on Science Education (NSTA), New Orleans, LA, USA, 6–9 November 2024; pp. 204–210. [Google Scholar]
  51. Renganayagalu, S.K.; Mallam, S.; Nazir, S. Effectiveness of VR head-mounted displays in professional training: A systematic review. Technol. Knowl. Learn. 2021, 26, 999–1041. [Google Scholar] [CrossRef]
  52. Tuta, J.; Luić, L.; Car, Ž. A conceptual model of augmented virtual and reality in cadet training. In Proceedings of the European Conference on Electrical Engineering and Computer Science (EECS), Athens, Greece, 28–30 December 2019; pp. 128–133. [Google Scholar]
  53. Li, T.; Liu, S. Key technology of laparoscopic surgery training system based on virtual reality technology. IOP Conf. Ser. Mater. Sci. Eng. 2020, 750, 012145. [Google Scholar] [CrossRef]
  54. Antoni, V.; Maurer, F.; Cesari, O.; Eichhorn, C.; Vilhjálmsson, H.H. Augmented virtuality training for special education teachers. In Proceedings of the IEEE International Symposium on Mixed and Augmented Reality Adjunct (ISMAR-Adjunct), Singapore, Singapore, 17–21 October 2022; pp. 327–332. [Google Scholar]
  55. Liu, H.; Wang, Z.; Mousas, C.; Kao, D. Virtual reality racket sports: Virtual drills for exercise and training. In Proceedings of the IEEE Conference Virtual Reality 3D User Interfaces (VR), Porto de Galinhas, Brazil, 9–13 November 2020; pp. 566–576. [Google Scholar]
  56. ABET. Engineering Accreditation Commission. 2 November 2019. Available online: https://www.abet.org/wp-content/uploads/2020/09/EAC-Criteria-2020-2021.pdf (accessed on 14 October 2024).
  57. Fussell, S.R.; Setlock, L.D. Computer-mediated communication. In The Oxford Handbook of Language and Social Psychology; Holtgraves, T., Ed.; Oxford University Press: Oxford, UK, 2014; pp. 471–490. [Google Scholar]
  58. Dede, C. The role of digital technologies in deeper learning. In Students at the Center: Deeper Learning Research Series; Jobs for the Future: Boston, MA, USA, 2014. [Google Scholar]
  59. Kim, J.Y.; Choi, J.K. Effects of AR on cognitive processes: An experimental study on object manipulation, eye-tracking, and behavior observation in design education. Sensors 2025, 25, 1882. [Google Scholar] [CrossRef]
  60. Bloom, B.S. Taxonomy of Educational Objectives: The Classification of Educational Goals: Handbook I, Cognitive Domain; Longmans: New York, NY, USA, 1956. [Google Scholar]
  61. Buchner, J.; Buntins, K.; Kerres, M. The impact of augmented reality on cognitive load and performance: A systematic review. J. Comput. Assist. Learn. 2022, 38, 285–303. [Google Scholar] [CrossRef]
  62. Rizzo, A.S.; Koenig, S.T. Is clinical virtual reality ready for primetime? Neuropsychology 2017, 31, 877–899. [Google Scholar] [CrossRef] [PubMed]
  63. Chandramouli, M.; Zahraee, M.; Winer, C.; Fidan, I. MOOC videos in Project MANEUVER. In Proceedings of the 2020 ASEE Annual Conference & Exposition., Montréal, QC, Canada, 21–24 June 2020. [Google Scholar]
  64. Calvillo-Gámez, E. Accessible immersion: Inclusive design strategies in extended reality. Univers. Access Inf. Soc. 2023, 22, 789–805. [Google Scholar]
  65. Hamed, M.; Wright, L. Designing XR for accessibility: Voice & switch control solutions. arXiv 2024, arXiv:2412.16321. [Google Scholar]
  66. Patel, J. Participatory design for XR in inclusive classrooms. Educ. Technol. Arch. 2023, 24, 125–142. [Google Scholar]
  67. Richter, F. Video Games Are More than Child’s Play. Statista. 12 December 2022. Available online: https://www.statista.com/chart/9821/us-gamer-demography/ (accessed on 14 October 2024).
  68. Yun, S.T.; Olsen, S.K.; Quigley, K.C.; Cannady, M.A.; Hartry, A. A review of augmented reality for informal science learning: Supporting design of intergenerational group learning. Visit. Stud. 2022, 26, 1–23. [Google Scholar] [CrossRef]
  69. Poupard, M.; Larrue, F.; Sauzéon, H.; Tricot, A. A systematic review of immersive technologies for education: Effects of cognitive load and curiosity state on learning performance. Br. J. Educ. Technol. 2025, 56, 5–41. [Google Scholar] [CrossRef]
Electronics 14 03818 g001
Figure 1. L1–L2 Learning Theory in XR. L1 involves learning to use the XR interface, while L2 focuses on domain-specific knowledge. Effective XR design reduces L1 load to support deeper L2 learning.
Figure 1. L1–L2 Learning Theory in XR. L1 involves learning to use the XR interface, while L2 focuses on domain-specific knowledge. Effective XR design reduces L1 load to support deeper L2 learning.
Electronics 14 03818 g001
Electronics 14 03818 g002
Figure 2. Selection workflow for reviewed XR studies in STEM education.
Figure 2. Selection workflow for reviewed XR studies in STEM education.
Electronics 14 03818 g002
Electronics 14 03818 g003
Figure 3. Demonstration of immersive presence in XR illustrating the integration of multimodal sensory cues (visual, auditory, and haptic) that facilitate increased learner engagement and spatial comprehension (adapted from [20]).
Figure 3. Demonstration of immersive presence in XR illustrating the integration of multimodal sensory cues (visual, auditory, and haptic) that facilitate increased learner engagement and spatial comprehension (adapted from [20]).
Electronics 14 03818 g003
Electronics 14 03818 g004
Figure 4. Interactive virtual reality modules used in digital manufacturing training showcasing affordances such as real-time object manipulation and gesture-based interactions, enhancing cognitive engagement and knowledge retention (adapted from [29]).
Figure 4. Interactive virtual reality modules used in digital manufacturing training showcasing affordances such as real-time object manipulation and gesture-based interactions, enhancing cognitive engagement and knowledge retention (adapted from [29]).
Electronics 14 03818 g004
Electronics 14 03818 g005
Figure 5. Active training mode in XR systems demonstrating learner-driven exploration and autonomy, enabling self-paced and scenario-based interactions critical for intrinsic motivation and deeper learning (adapted from [39]).
Figure 5. Active training mode in XR systems demonstrating learner-driven exploration and autonomy, enabling self-paced and scenario-based interactions critical for intrinsic motivation and deeper learning (adapted from [39]).
Electronics 14 03818 g005
Electronics 14 03818 g006
Figure 6. A simplified XR interface using clearly marked components and context-sensitive feedback mechanisms, effectively managing cognitive load and facilitating learner focus on complex tasks (adapted from [63]).
Figure 6. A simplified XR interface using clearly marked components and context-sensitive feedback mechanisms, effectively managing cognitive load and facilitating learner focus on complex tasks (adapted from [63]).
Electronics 14 03818 g006
Electronics 14 03818 g007
Figure 7. Inclusive design considerations in XR education, demonstrating accessible functionalities such as adjustable interfaces and assistive technology integration to accommodate diverse learner needs (adapted from [63]).
Figure 7. Inclusive design considerations in XR education, demonstrating accessible functionalities such as adjustable interfaces and assistive technology integration to accommodate diverse learner needs (adapted from [63]).
Electronics 14 03818 g007
Electronics 14 03818 g008
Figure 8. Summary of the proposed design framework.
Figure 8. Summary of the proposed design framework.
Electronics 14 03818 g008
Electronics 14 03818 g009
Figure 9. Practical implementation of XR environments in STEM education, illustrating user-centered design principles, multimodal interaction, and alignment with real-world workforce development needs (adapted from [63]).
Figure 9. Practical implementation of XR environments in STEM education, illustrating user-centered design principles, multimodal interaction, and alignment with real-world workforce development needs (adapted from [63]).
Electronics 14 03818 g009
Table 1. XR affordances and their effects on learning in STEM education. This table summarizes key interactive features reported across reviewed studies, highlighting their benefits and design-related challenges.
Table 1. XR affordances and their effects on learning in STEM education. This table summarizes key interactive features reported across reviewed studies, highlighting their benefits and design-related challenges.
Affordance TypeDescriptionPositive ImpactObserved Limitations
Gesture-Based InteractionUse of hand or body movements for control (e.g., grabbing, rotating)Increased realism, engagement, and fine motor mapping [7,21]Learning curve; fatigue in longer tasks
Contextual OverlaysReal-time instructional cues or visual aids embedded in the XR environmentImproved procedural clarity and task sequencing [22,23]Risk of screen clutter if poorly designed
Multi-Modal FeedbackCombines visual, auditory, and haptic responsesEnhances error detection and reinforcement of concepts [17,30]Can cause sensory overload without adaptive control
Scaffolding MechanicsStepwise interaction guidance based on learner progressReduces extraneous load, supports novices in complex tasks [31,32]May restrict autonomy for more experienced users
Real-Time Object ManipulationAbility to dynamically explore and modify virtual objectsDeepens conceptual understanding, especially in STEM domains [33,34]Performance lag in lower-end XR systems
Table 2. Features supporting learner agency in XR learning environments. The table outlines agency-enhancing strategies, their educational impact, and considerations for effective implementation.
Table 2. Features supporting learner agency in XR learning environments. The table outlines agency-enhancing strategies, their educational impact, and considerations for effective implementation.
Agency FeatureExample ImplementationEffect on Learning OutcomesDesign Considerations
User-Driven NavigationNon-linear scene access or timeline controlEncourages exploration and self-paced learning [30,35]Needs clear cues to prevent disorientation
Adaptive Difficulty LevelsTasks adjust based on real-time performanceMaintains flow and reduces frustration [18,36]Calibration required to avoid under/over-challenging users
Scenario-Based BranchingLearners make decisions that alter the learning pathwayIncreases motivation and critical thinking [22,37]Must align with learning objectives
Real-Time Variable ManipulationModify system parameters in simulations (e.g., pH, force)Enhances experimental understanding and autonomy [25,40]Can be overwhelming without adequate domain knowledge
Feedback on Decision OutcomesImmediate response to learner choicesReinforces cause-effect reasoning, supports L2 learning [2,37]Should balance guidance with exploratory freedom
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Chandramouli, M.; Zafar, A.; Williams, A. A Detailed Review of the Design and Evaluation of XR Applications in STEM Education and Training. Electronics 2025, 14, 3818. https://doi.org/10.3390/electronics14193818

AMA Style

Chandramouli M, Zafar A, Williams A. A Detailed Review of the Design and Evaluation of XR Applications in STEM Education and Training. Electronics. 2025; 14(19):3818. https://doi.org/10.3390/electronics14193818

Chicago/Turabian Style

Chandramouli, Magesh, Aleeha Zafar, and Ashayla Williams. 2025. "A Detailed Review of the Design and Evaluation of XR Applications in STEM Education and Training" Electronics 14, no. 19: 3818. https://doi.org/10.3390/electronics14193818

APA Style

Chandramouli, M., Zafar, A., & Williams, A. (2025). A Detailed Review of the Design and Evaluation of XR Applications in STEM Education and Training. Electronics, 14(19), 3818. https://doi.org/10.3390/electronics14193818

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop