3.1. Theoretical analysis and proposal of research hypotheses

3.1.1. Theoretical analysis.

1. The “3I” technological features of metaverse teaching and their application in higher education

The metaverse is an emerging virtual reality technology that is applied to teaching by creating an interactive three-dimensional digital environment. Due to its potential to provide immersive learning experiences, this technology has begun to be gradually promoted in universities in China [18]. When applied to teaching scenarios, the metaverse is believed to possess the “3I” technological features distinct from traditional teaching: Immersion, Interaction, and Imagination. Immersion refers to the highly realistic virtual environment that makes students feel as if they are physically present, enhancing their learning experience and emotional engagement. Interaction signifies the ability to provide interactive 3D dynamic scenes, strengthening the interaction between students and teaching content, teachers, and classmates. Imagination means stimulating students’ imagination and creativity through high-quality 3D visual effects and multisensory simulation [19–21].

Immersion enhances the emotional engagement and experience of learning through highly realistic environments; Interaction promotes communication between teachers and students through three-dimensional dynamic scenes; and Imagination utilizes 3D visual effects and multisensory simulation to stimulate students’ imagination and creativity. These technological advantages are not only applicable to courses with strong practicality such as sports, arts, and medicine but also to foreign language learning, helping students immerse themselves in the language environment they are learning and improving their language expression and cultural understanding abilities [22].

Despite the great potential of metaverse teaching, its implementation currently faces challenges, including issues of learning persistence, focus, and learning efficiency. In addition, the complexity of technological implementation, the need for teacher training to adapt to new technologies, and the potential cognitive load or physical discomfort that long-term immersion in a virtual environment may bring to students are all issues that metaverse teaching needs to overcome [23]. However, with continuous technological advancement and innovation, these challenges are expected to be gradually resolved, and metaverse teaching continues to show its value in innovating educational models and enhancing learning experiences. Metaverse teaching will become more personalized and intelligent, capable of adjusting teaching content and methods according to students’ learning habits and preferences. Advances in hardware equipment and software development innovation will make immersive experiences more realistic, interactions more natural, and imagination more widely applied. Ultimately, metaverse teaching is expected to be more closely integrated with existing educational technologies, forming a more comprehensive educational ecosystem that provides students with richer and more effective learning experiences, enhancing their learning efficacy.

1. 2. The role and importance of psychological imagery in metaverse teaching

The concept of psychological imagery originates from the Imagery Theory in cognitive psychology, which posits that individuals can mentally manipulate and transform these mental representations as if they were real objects [24].

As research deepens, the concept of psychological imagery has gradually been summarized, referring to the mental representation and experience of objects, events, or situations that do not currently exist without direct sensory input. This representation is usually vivid and concrete, and can evoke in the individual’s mind feelings and cognition similar to actual experiences [25]. In metaverse teaching, psychological imagery plays an extremely important role, acting as a bridge connecting teaching content with students’ cognitive structures, making the learning process more vivid and effective [26]. Specifically, psychological imagery promotes cognitive processing by providing a way to simulate learners’ actual experiences, helping students better understand and remember learning materials through visual internal representations. In addition, psychological imagery also stimulates students’ innovative thinking and spirit of exploration, enhancing their problem-solving abilities, thereby significantly improving learning efficacy in this immersive learning environment. This not only enriches students’ knowledge structures but also improves their mastery and application abilities of complex concepts, making metaverse teaching a powerful modern educational method.

1. 3. The importance and role of representational ability in metaverse teaching

Representational ability, as the capacity for individuals to construct, operate, and maintain concrete images on a psychological level, is an important part of cognitive psychology research. It is closely related to dual coding theory, constructivist learning theory, and psychological imagery theory, which emphasize the importance of non-verbal information processing and the proactivity of knowledge construction [27,28]. In the metaverse teaching scenario, representational ability plays a crucial role. It not only enhances students’ immersive learning experience and improves cognitive processing efficiency but also promotes the development of creativity and personalized learning. The stronger the representational ability, the stronger the students’ interaction and exploration capabilities in the metaverse environment. This ability subtly influences how students process and absorb information through psychological imagery, thereby affecting learning efficacy [29]. Specifically, students with strong representational ability can more effectively simulate complex scenes and concepts mentally, deepening their understanding and memory of learning materials, and can also more actively construct knowledge, forming deep cognitive connections. Therefore, when designing metaverse teaching strategies, it is crucial to cultivate students’ representational ability, which will help improve the quality of teaching and provide students with a richer and more meaningful learning experience, ultimately contributing to enhanced learning efficacy and the comprehensive achievement of educational goals.

1. 4. The role of metaverse teaching in enhancing learning efficacy

Learning efficacy is the embodiment of an individual’s confidence and ability to achieve learning objectives in educational activities, closely related to Bandura’s theory of self-efficacy [30,31]. Self-efficacy is the individual’s subjective assessment of their ability to successfully perform specific tasks, affecting learners’ motivation, persistence, and behavioral choices when facing challenges. In this cutting-edge field of metaverse teaching, students can gain more contextualized and practical intuitive experiences through simulation and interactive learning, which helps students build connections and bridges between theories, knowledge, skills, and practical application scenarios, thereby promoting a deep understanding and mastery of learning materials. With a significant enhancement of students’ self-efficacy, it not only strengthens their learning motivation and participation but also promotes their resilience and strategic application when facing learning challenges [32]. Therefore, metaverse teaching, by enhancing students’ self-efficacy, helps to improve learning efficacy, improve learning outcomes, and provides students with a learning platform full of potential. However, in practical applications, metaverse classrooms are not suitable for all courses [26], so it is also necessary to design appropriate teaching plans in combination with the specific nature of the course and teaching objectives to ensure that the advantages of metaverse teaching are maximized, while also considering how to complement traditional teaching methods to achieve the best teaching results.

In summary: The relationship between the “3I” technological features, psychological imagery, representational ability, and learning efficacy in metaverse teaching is shown in Fig 2. The “3I” technological features, psychological imagery, and representational ability interact to jointly promote learning efficacy. The specific pathways and detailed logical mechanisms will be further elaborated in subsequent chapters. Among them, immersion (Immersion), interaction (Interaction), and imagination (Imagination) are themselves abstract concepts, but they are not convenient for operational measurement and empirical research. Therefore, this paper develops these three concepts into constructs that can be measured by questionnaire scales as substitute variables, namely immersion perception (Immersion perception), interaction perception (Interaction perception), and imagination perception (Imagination perception), facilitating future quantitative analysis and empirical verification of these abstract concepts, thereby more accurately assessing the impact of metaverse teaching on learning efficacy.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. The interactive model of’ 3I’ technical characteristics, mental image, imagery ability and learning efficiency in 2 yuan’s universe teaching.

https://doi.org/10.1371/journal.pone.0318548.g002

3.1.2. Proposing research hypotheses.

1. (1). Direct path

As outlined above, metaverse classrooms are characterized by the “3I” technological features, where immersion (Immersion) facilitates a highly realistic virtual environment that makes students feel as if they are physically present, thereby enhancing their learning experience and emotional engagement. Psychological imagery (Psychological image) refers to an individual’s mental representation and experience of objects, events, or situations that do not currently exist, without direct sensory input. Since immersion perception offers a way to simulate actual experiences, it can promote cognitive processing, aiding students in understanding and memorizing learning materials through vivid internal representations. Thus, immersion perception can positively foster the formation and development of psychological imagery by intensifying students’ emotional investment and learning experience.

Interaction (Interaction) provides interactive 3D dynamic scenes, strengthening the interaction between students and teaching content, teachers, and classmates. As a form of mental representation, psychological imagery can be enhanced and enriched through the interaction between the individual and the environment. In the metaverse teaching environment, interaction perception offers abundant opportunities for interaction, allowing students to participate more deeply in the learning process and form more vivid and concrete psychological imagery through this engagement. Therefore, interaction perception contributes to the positive development of psychological imagery by enhancing students’ interaction with the learning environment.

Imagination (Imagination) stimulates students’ imagination and creativity through high-quality 3D visual effects and multisensory simulation. Psychological imagery involves an individual’s mental representation of non-existent objects or situations, which requires a rich imagination. Imagination perception, by offering innovative and creative visual and sensory experiences, can ignite students’ imagination, prompting them to construct and explore new psychological imagery. Hence, imagination perception positively promotes the formation and enhancement of psychological imagery by fostering students’ creative thinking and imagination.

In summary, the following path hypotheses can be derived:

H1a: Immersion perception has a positive promoting effect on psychological imagery.

H1b: Interaction perception has a positive promoting effect on psychological imagery.

H1c: Imagination perception has a positive promoting effect on psychological imagery.

In metaverse teaching models, students experience deep immersion through highly realistic virtual environments. This immersive experience (immersion perception) encourages students to increase their emotional investment, thereby strengthening their focus and comprehension of learning materials. The enhancement of focus aids students in processing information more effectively, while the increase in emotional engagement helps to boost learning motivation. Together, these factors directly improve students’ confidence and ability to complete learning tasks, that is, learning efficacy.

Students are provided with interactive 3D dynamic scenes through metaverse courses, enhancing interaction (interaction perception) between students and teaching content, teachers, and classmates. This interaction not only promotes collaborative learning among students but also strengthens their exploration and participation in learning content. High-quality interactive experiences enable students to participate more actively in the learning process, solving problems through collaboration and sharing knowledge. This positive learning attitude and behavior directly improve students’ confidence and ability to achieve learning goals, that is, learning efficacy.

Metaverse courses stimulate students’ imagination and creativity through innovative and creative visual and sensory experiences (imagination perception). This perception allows students to construct and explore new concepts and scenarios mentally, thereby promoting the development of innovative thinking. The enhancement of imagination perception enables students to transcend the limitations of traditional learning methods and understand and apply knowledge in more innovative ways. This improvement in innovative ability helps students to show higher adaptability and problem-solving skills when facing new learning challenges, thereby directly improving their learning efficacy.

In summary, the following path hypotheses can be derived:

H1d: Immersion perception has a positive promoting effect on learning efficacy.

H1e: Interaction perception has a positive promoting effect on learning efficacy.

H1f: Imagination perception has a positive promoting effect on learning efficacy.

In the metaverse teaching environment, psychological imagery plays an essential role as the internal mental representation of students to the teaching content. Psychological imagery (Psychological image) allows students to have a vivid mental experience of non-existent objects, events, or situations without direct sensory input, stimulating feelings and cognition similar to actual experiences. This internal mental representation not only enriches students’ cognitive structures but also promotes cognitive processing by simulating actual experiences, helping students better understand and remember learning materials.

Furthermore, psychological imagery stimulates students’ innovative thinking and exploratory spirit, enhancing their problem-solving abilities. The enhancement of this ability is particularly important in the immersive learning environment because it significantly improves learning efficacy. The improvement of learning efficacy is closely related to the theory of self-efficacy; psychological imagery enhances students’ mastery and application ability of complex concepts, improving their self-efficacy and thus positively promoting learning efficacy.

Additionally, psychological imagery enriches students’ knowledge structures and improves their learning motivation and participation. As a result, students can participate more deeply in the learning process. This positive learning attitude and behavior directly improve students’ confidence and ability to achieve learning goals. At the same time, psychological imagery also enhances students’ resilience and strategic application when facing learning challenges because it provides a mental simulation that helps students rehearse possible solutions and coping strategies.

In summary, the following path hypothesis can be derived:

H1g: Psychological imagery has a positive promoting effect on learning efficacy.

1. (2). Moderating effect

Representational ability, as a key cognitive skill, enables individuals to construct, manipulate, and sustain concrete images mentally, which is crucial for learning and understanding complex concepts. In metaverse classrooms, immersion perception provides a highly realistic virtual environment that makes students feel as if they are truly part of it. This deep immersion experience can enhance students’ emotional engagement and learning experience. In this process, students with high representational ability can more effectively use the immersive experience to form specific psychological imagery because they can simulate and manipulate learning materials more precisely in their minds. Similarly, interaction perception, by offering interactive 3D dynamic scenes, enhances the interaction between students and teaching content, teachers, and classmates. In this interactive process, students with high representational ability can process information more deeply and construct knowledge, thus forming more vivid and specific psychological imagery. Furthermore, imagination perception stimulates students’ imagination and creativity through innovative and creative visual and sensory experiences. Students with high representational ability can better utilize this perception to build and explore new psychological imagery, further promoting the development of creative thinking and problem-solving skills. In summary, representational ability positively moderates the effects of immersion perception, interaction perception, and imagination perception on psychological imagery by enhancing mental simulation, promoting cognitive processing, and stimulating innovative thinking, playing a crucial moderating role in metaverse teaching and helping to improve students’ learning efficacy.

In summary, the following path hypotheses can be derived:

H2a: Representational ability can positively moderate the effect of immersion perception on psychological imagery.

H2b: Representational ability can positively moderate the effect of interaction perception on psychological imagery.

H2c: Representational ability can positively moderate the effect of imagination perception on psychological imagery.

Representational ability, as the psychological skill for individuals to build, operate, and maintain concrete images, plays a vital role in metaverse teaching. It is closely connected to dual coding theory and constructivist learning theory in cognitive psychology, emphasizing the importance of non-verbal information processing and the proactivity of knowledge construction. In the metaverse teaching environment, immersion perception enhances students’ emotional engagement and learning experience by providing a highly realistic virtual environment. Students with high representational ability can more deeply utilize this immersive experience, forming specific psychological imagery through mental simulation and manipulation of learning materials, thereby improving learning efficacy.

At the same time, interaction perception promotes communication between teachers and students and among students through three-dimensional dynamic scenes, enhancing students’ understanding and memory of learning content. In this process, students with high representational ability can participate in interactions more effectively, deeply process information and construct knowledge, further enriching psychological imagery and improving learning efficacy.

In addition, imagination perception stimulates students’ imagination and creativity through innovative 3D visual effects and multisensory simulation. Students with high representational ability can use this perception to build and explore new psychological imagery, promoting the development of creative thinking and problem-solving skills. The enhancement of innovative thinking and problem-solving skills helps students to show higher adaptability and problem-solving ability when facing new learning challenges, thereby improving learning efficacy.

In summary, the following path hypotheses can be derived:

H2d: Representational ability can positively moderate the effect of immersion perception on learning efficacy.

H2e: Representational ability can positively moderate the effect of interaction perception on learning efficacy.

H2f: Representational ability can positively moderate the effect of imagination perception on learning efficacy.

1. (3). Mediating effect

In the metaverse teaching environment, immersion perception, interaction perception, and imagination perception form the core of the “3I” technological features, which indirectly enhance learning efficacy by strengthening students’ psychological imagery. Specifically, immersion perception provides a highly realistic virtual environment that makes students feel as if they are truly present, and this deep immersion experience can enhance students’ emotional engagement and learning experience. The enhancement of emotional engagement and learning experience, according to H1a, can significantly promote the formation and development of students’ psychological imagery. Psychological imagery, as the internal mental representation of students to the teaching content, allows students to have a vivid mental experience of non-existent objects, events, or situations without direct sensory input, and this experience can stimulate feelings and cognition similar to actual experiences.

Furthermore, according to H1g, psychological imagery itself has a positive promoting effect on learning efficacy because it enriches students’ cognitive structures and improves their mastery and application abilities of complex concepts. Similarly, interaction perception and imagination perception also promote the formation of psychological imagery through similar mechanisms, respectively according to H1b and H1c, thereby affecting learning efficacy.

In summary, the following path hypotheses can be derived:

H3a: Immersion perception can have a positive impact on learning efficacy through psychological imagery (based on H1a and H1g).

H3b: Interaction perception can have a positive impact on learning efficacy through psychological imagery (based on H1b and H1g).

H3d: Imagination perception can have a positive impact on learning efficacy through psychological imagery (based on H1c and H1g).

1. (4). Total effect

In the metaverse teaching environment, immersion perception, interaction perception, and imagination perception constitute the core of the “3I” technological features, which positively impact learning efficacy through various mechanisms. Initially, immersion perception (H1d) provides a highly realistic virtual environment, making students feel as if they are truly present. This profound immersive experience enhances students’ emotional engagement and learning experience, directly increasing their confidence and ability to complete learning tasks, that is, learning efficacy. Moreover, immersion perception (H3a) can further promote the enhancement of learning efficacy through the mediating variable of psychological imagery.

Similarly, interaction perception (H1e) enhances the interaction between students and teaching content, teachers, and classmates through three-dimensional dynamic scenes. This interaction not only fosters collaborative learning among students but also strengthens their exploration and participation in learning content, thereby directly improving students’ confidence and ability to achieve learning objectives, that is, learning efficacy. Interaction perception (H3b) can also indirectly enhance learning efficacy through psychological imagery.

Lastly, imagination perception (H1f) stimulates students’ imagination and creativity through innovative and creative visual and sensory experiences, allowing students to mentally construct and explore new concepts and scenarios. This enhancement of innovative capabilities helps students exhibit higher adaptability and problem-solving skills when facing new learning challenges, directly improving their learning efficacy. Imagination perception (H3d) can also indirectly exert a positive impact on learning efficacy through the mediating variable of psychological imagery.

H4a: Immersion perception can have an overall positive impact on learning efficacy (based on H1d and H3a).

H4b: Interaction perception can have an overall positive impact on learning efficacy (based on H1e and H3b).

H4c: Imagination perception can have an overall positive impact on learning efficacy (based on H1f and H3d).

This paper intends to conduct a survey targeting college students in China who have experienced metaverse teaching scenarios to collect relevant data. We will design six demographic characteristics based on the characteristics of college students: Gender (0 represents male, 1 represents female), Age, Monthly Living Expenses (Expense), Academic Intention (0 indicates a preference for humanities and social sciences, 1 indicates a preference for natural sciences), and Grade. These demographic characteristics will be incorporated into the research framework. Integrating the aforementioned content, the conceptual model of this paper is depicted in Fig 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Conceptual model: the mechanism of the influence of’ 3I’ technical characteristics on learning efficiency in metauniverse teaching.

https://doi.org/10.1371/journal.pone.0318548.g003

3.2. Measurement and research methods of variables

This study plans to collect data through a survey questionnaire and, based on the research objectives and hypothetical paths proposed earlier, has summarized six core constructs: Immersion Perception (ISP), Interaction Perception (ITP), Imagination Perception (IGP), Psychological Image (PI), Representation Ability (RA), and Learning Efficiency (LE).

As shown in Table 1, the aforementioned six constructs all have similar defined constructs and have formed validated questionnaire scales. Therefore, this study has referred to the scale items in this existing literature and combined them with the research subjects of this paper, as well as the research context localized to China. To facilitate the understanding of the subjects in the local context, appropriate revisions have been made. When designing the questionnaire, the reliability and validity of the scale must be fully considered. Given that this paper intends to use Structural Equation Modeling (SEM) for modeling, the measurement of reliability will be assessed using Cronbach’s alpha coefficient and composite reliability. In terms of validity, it is divided into content validity and structural validity, with a focus on ensuring that the scale items can fully represent the meaning of the constructs [33], and considering the scale’s understandability and applicability. To this end, we have invited experts in the field of urban environmental research with experience in questionnaire surveys to evaluate the content validity and make multiple revisions based on feedback. Structural validity usually involves statistical measurement methods. In the context of SEM, it is further divided into convergent validity and discriminant validity, which can be assessed through methods such as the Average Variance Extracted (AVE) value, factor loadings, and the Fornell-Larcker criterion. The subsequent parts of this paper will elaborate on these aspects in detail.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Measurement items included in questionnaire.

https://doi.org/10.1371/journal.pone.0318548.t001

3.3. Data source

In December 2023, the Education Department of Zhejiang Province, China, developed a series of virtual reality technology application courses for undergraduate college students across the province: “Trauma First Aid Practice Based on Virtual Reality,” “Virtual Simulation Practice Teaching of ‘Long March Source’ Revolutionary History,” “Virtual Simulation Experiment of Railway Track Dynamics,” and “College English Listening, Speaking, Reading, Writing, and Translation” metaverse teaching platforms. Among the aforementioned courses, “College English Listening, Speaking, Reading, Writing, and Translation” is a public course aimed at undergraduate students from colleges across the province, hence it has the highest enrollment. This course integrates AI technology and offers the most comprehensive metaverse features, allowing students to engage in role-playing and converse with virtual entities in virtual settings. Instructors play the role of organizers and managers in this course and do not directly participate in teaching. According to backend data, the number of enrollees exceeded ten thousand. Other courses, due to their very specific professional focus, have fewer enrollees and offer more singular functionalities, making them less suitable as samples for this study. Considering the breadth and representativeness of the research, we have chosen the student group from the “College English Listening, Speaking, Reading, Writing, and Translation” metaverse course as our target sample for the survey. In collaboration with the Zhejiang Provincial Education Department, we distributed online survey questionnaires through the students’ registered email addresses. Students could choose whether to respond, and the entire process was voluntary, thus not infringing on their personal privacy and freedom. To increase the response rate, we also offered cash incentives in the questionnaire. The web address for the online questionnaire is: https://www.wjx.cn/vm/mtBYfOx.aspx#. The survey lasted from March to May 2024, and a total of 973 questionnaires were collected. After manual quality review, we excluded those that were carelessly answered or severely lacking data, retaining 930 valid questionnaires, achieving an effective recovery rate of 95.58%.

Given that this study collects data through a questionnaire survey, it is necessary to clarify the following: In accordance with the principles of the Declaration of Helsinki, an application for ethical review of scientific research was submitted to the Academic Ethics Committee of Wenzhou University of Technology and was granted approval by the committee (Report Number: WZUT2024003261). The questionnaires provided in this study were completed voluntarily by the participants after a full explanation of the research objectives, procedures, potential risks, and benefits. All participants have confirmed that informed consent has been obtained from all subjects and/or their legal guardians.

The demographic characteristics of the sample are shown in Table 2 and include undergraduate students of different genders, ages, living standards (with monthly living expenses used as a proxy variable for living standards), academic inclinations (Intention), and grades. It is evident that the gender distribution is relatively balanced, with males accounting for 52.04% and females for 47.96%. The age range covers from 36.99% of those under 19 years old to 18.60% of those over 23 years old. Living standards vary from 21.83% earning less than 1,000 RMB per month to 11.83% earning over 5,000 RMB. The academic inclination is significantly skewed towards natural sciences, representing 56.02%. The distribution of grades is more dispersed, with 46.24% freshmen, 20.32% sophomores, 18.17% juniors, and 15.27% seniors. This may be because freshmen are more curious about new technologies and teaching methods, thus more likely to choose metaverse courses and participate in the survey due to their pursuit of new experiences and exploratory spirit. Additionally, as students progress to higher grades, they might face increasing academic and career planning pressures, which could affect their willingness and time to participate in the questionnaire survey.

Overall, the sample demonstrates diversity in gender, age, economic status, academic interests, and learning stages, meeting the requirements for extensive and representative analysis and providing a comprehensive perspective of the student population for the study.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Sample statistical characteristics.

https://doi.org/10.1371/journal.pone.0318548.t002

3.4. Modeling method

In the process of constructing the analytical model, this study encountered the complexity of numerous concepts and the multiple pathways of potential variables between them. Traditional regression analysis may not fully capture the deep-level interactions among these variables [42]. In light of this, Structural Equation Modeling (SEM) was chosen as the analytical framework for this study due to its capability in handling complex models involving multiple variables and pathways, as well as its advantages in measuring latent variables and theoretical development [43]. SEM can simultaneously process multiple regression equations and control the interrelationships between latent variables, more accurately determining the degree and direction of the relationships between variables [44]. This method helps to comprehensively understand the impact of metaverse teaching technology characteristics on students’ learning efficacy, as well as the mediating role of psychological imagery and the moderating role of representational ability, thereby providing a scientific basis for optimizing metaverse teaching strategies and improving students’ learning outcomes.

Specifically, this study integrates demographic characteristics into SEM for analysis, noting that most of these characteristics are categorical variables, and the gender variable only contains two values, 0 and 1, which does not meet the assumption of normal distribution. Therefore, we selected the Generalized Structural Equation Model (GSEM) to accommodate these categorical variables [45]. In terms of model parameter estimation, we employed the Partial Least Squares-SEM (PLS-SEM) method. SEM parameter estimation is typically divided into two types: Covariance-Based SEM (CB-SEM) and PLS-SEM, where CB-SEM requires data to meet the normal distribution, while PLS-SEM is not limited by this. Given that this study used GSEM and included binary categorical variables such as gender, PLS-SEM became the more appropriate choice. Moreover, since the study explores the variable of urban value perception, which is exploratory in nature, the flexibility and iterative capability of PLS-SEM in model building are particularly important. It allows researchers to adjust the model based on data feedback to more accurately reflect the characteristics of the data [46]. Therefore, we decided to use PLS-SEM for parameter estimation.

4.1. Reliability and validity test of external model

This study utilizes the PLS-SEM approach to estimate the structural equation model depicted in Fig 2. In accordance with the PLS-SEM validation process, the assessment of the outer model is a primary step. The outer model examination predominantly pertains to the relationship between the survey items and the constructs they are intended to reflect, gauging whether these items are capable of representing the constructs. To scrutinize the outer model, several indicators must be evaluated: First, reliability, which measures the relationship between the measurement indicators (survey items) and the latent variables (also known as constructs), essentially determining if the indicators are sufficient to represent the latent variables. Five indicators are assessed for this purpose (as shown in Fig 4):

1. (1). Cronbach’s α, a statistical measure used to appraise the consistency and dependability of a set of observed variables (i.e., metrics or items) in relation to a latent variable (a variable not directly observable).
2. (2). Composite reliability (CR), an indicator that evaluates the measurement consistency of multiple observed variables (items) concerning a single latent variable. CR is considered a stricter measure of reliability as it accounts for both the intercorrelations and the individual variances of the observed variables.
3. (3). Average variance extracted (AVE), which gauges the extent to which a latent variable explains the variance of its corresponding observed indicators. AVE represents the proportion of variance in the observed indicators that is accounted for by the latent variable.
4. (4). Factor loadings, indicators that measure the strength and direction of the relationship between observed variables and their corresponding latent variables. Factor loadings reflect the degree to which an observed variable contributes to its associated latent variable and are key in assessing whether the survey items effectively represent the composition and characteristics of the latent variable within a measurement model.
5. (5). The Fornell-Larcker criterion, an instrument for assessing and comparing the discriminatory validity of different latent variables. This criterion was introduced by Fornell and Larcker in their research to ensure that each latent variable in the model is distinctly identifiable and possesses uniqueness.

Specifically, the analysis begins with Cronbach’s α to evaluate the internal consistency of the survey items, with α values ranging from 0 to 1. Generally, an α value below 0.6 suggests poor internal consistency, values between 0.7 and 0.8 indicate moderate reliability, and values between 0.8 and 0.9 signify high reliability [47]. Following this, composite reliability (CR) is calculated to assess the uniformity of the measurement indicators for the latent variables; a higher CR value implies more stable and reliable measurement, with a value of 0.7 or above typically considered satisfactory [48]. Subsequently, the average variance extracted (AVE) is evaluated, which reflects the variance explanatory capacity of the latent variable’s measurement indicators and is an important measure of convergent validity. An ideal AVE value exceeds 0.5, while a range of 0.36 to 0.5 is also acceptable [49]. Moreover, the assessment of factor loadings is crucial, as it indicates the degree of contribution of each item to the latent variable, with loadings greater than 0.5 generally considered effective [50]. The results of the tests, as shown in Table 3, indicate that the scales used in this study have met acceptable standards of reliability.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Reliability and convergence testing of external models.

https://doi.org/10.1371/journal.pone.0318548.t003

Finally, the Fornell-Larcker criterion is applied to assess the discriminatory validity, which is based on comparing the square of the AVE values of the latent variables with the squares of their correlations with other latent variables [49]. If a variable’s AVE value is higher than the square of its correlation coefficients with other variables, it suggests that the variable has good discriminatory validity, indicating a clear distinction between them.

The advantage of this assessment method is that it provides a clear, quantifiable standard for judging the discriminatory validity between constructed variables, which helps ensure that the variables in the study are independent and have discriminating power. This is particularly important for multivariate analysis methods such as Structural Equation Modeling (SEM), as these methods rely on clear and independent relationships between constructed variables.

The Fornell-Larcker criterion, known for its systematic and straightforward nature, has been widely adopted [45]. This study also employs this method of examination, with the specific results presented in Table 4. The square root of AVE values on the diagonal are all greater than the other values within the boxes, indicating good discriminatory validity.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Fornell-Larcker criterion.

https://doi.org/10.1371/journal.pone.0318548.t004

4.2. Internal model structure test

In this study, the process of testing the internal model structure involves verifying the interrelationships between latent variables as depicted in Fig 3, as well as empirically evaluating the theoretical hypotheses proposed in Section 3.1. The results of the examination are presented in detail in Table 5, which is divided into five main sections corresponding to the four research hypotheses put forward in this paper and the role of demographic characteristics:

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Hypothesis test results.

https://doi.org/10.1371/journal.pone.0318548.t005

1. (1). Direct Path: This section validates the single-path hypotheses, assessing the direct effect relationships between the latent variables. This test aims to determine whether there is an expected direct impact between the latent variables, providing a basis for understanding the fundamental mechanisms at work within the model.
2. (2). Moderating Effect: This section delves into the potential influence of moderating variables on the relationships between variables. Specifically, the study examines how Representation Ability (RA) moderates the impact of the “3I” technological features of metaverse teaching—Immersion Perception (ISP), Interaction Perception (ITP), and Imagination Perception (IGP) on Psychological Image (PI), as well as the impact of these technological features on Learning Efficiency (LE). The presence of a moderating effect reveals the complexity of the interplay between variables and aids in understanding the variability in relationships under different conditions.
3. (3). Mediating Effect: In this part, the paper evaluates the potential mediating effects between variables, exploring whether Psychological Image (PI) might influence the relationship between the “3I” technological features of metaverse teaching and Learning Efficiency (LE) as a mediating variable. The test of mediating effects helps to uncover the complex causal chains between variables and clarify the underlying mechanisms.
4. (4). Total Effect: This section focuses on the overall effects between the latent variables, including the sum of direct effects and indirect effects through mediating variables. The assessment of total effects provides a comprehensive perspective for understanding the overall relationship between variables and helps to evaluate the combined impact of the “3I” technological features of metaverse teaching on learning efficacy.
5. (5). Demographic Influence: The final section analyzes the impact of demographic characteristics on the relationships between variables in the model. The study particularly focuses on how demographic characteristics such as gender, age, monthly living expenses, academic inclination, and grade affect learning efficacy (LE), aiming to reveal the potential influence of different population characteristics on the acceptance and learning outcomes of metaverse teaching.

1. (1). Direct Path

To ensure conciseness and clarity in presentation, this study employs a graphical format to illustrate the validation results of the direct path hypotheses and moderating effect hypotheses (as shown in Fig 5). The hypotheses that were supported are H1a, H1b, H1c, H1d, H1e, and H1g, while the hypothesis that was not supported is H1f.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Validation of metaverse classroom impact hypotheses.

Note 1: *** p < 0.001.** p < 0.01. * p < 0.05.“NS” means no significant effect. Note 2: shows the path is significant; shows the path is not significant.

https://doi.org/10.1371/journal.pone.0318548.g005

The findings indicate that the three characteristics of metaverse courses—Immersive Perception (ISP), Interaction Perception (ITP), and Imagination Perception (IGP)—each have a direct impact on Psychological Image (PI), and PI in turn has a direct effect on Learning Efficiency (LE). This establishes a link between the metaverse classroom and learning efficacy, confirming the potential mechanisms through which metaverse classrooms can influence learning efficacy. For courses such as foreign language learning, the immersion in metaverse classrooms, facilitated by realistic 3D environments, makes learners feel as if they are physically present, enhancing their emotional engagement and learning experience, and enabling them to more easily form clear mental images of language use scenarios. Interaction Perception, through interactive 3D dynamic scenes, strengthens the interaction between learners and the teaching content, teachers, and peers, allowing learners to practice language skills in an authentic context and build psychological imagery of language knowledge. Imagination Perception stimulates learners’ imagination and creativity through innovative 3D visual effects and multisensory simulation, encouraging them to explore and construct new language structures through mental simulation, thereby better mastering grammar and vocabulary.

Psychological imagery plays a positive role in students’ learning efficacy in foreign language learning on multiple levels: firstly, it enhances cognitive processing, aiding students in forming a deep understanding and memory of learning materials; secondly, it stimulates students’ innovative thinking, prompting them to actively explore and construct knowledge; additionally, psychological imagery can increase students’ emotional engagement, making the learning process more attractive and enhancing motivation and participation. Importantly, psychological imagery boosts students’ self-efficacy, giving them more confidence to face learning challenges, thereby improving learning efficacy. It also enhances students’ problem-solving abilities and adaptability, helping them effectively cope with new learning situations. Lastly, psychological imagery promotes immersion in the language environment, helping students mentally place themselves in the target language environment, which is crucial for improving language proficiency and cultural understanding.

The direct effects of ISP and ITP on LE are also confirmed because immersion perception, by providing an immersive learning environment, allows learners to deeply participate and emotionally engage, thereby enhancing their focus and comprehension of learning materials. Cultural immersion also deepens the understanding of the social and cultural context of language use. Interaction perception, through real-time interaction and collaborative learning, increases the frequency and quality of language practice and allows learners to adjust their learning strategies based on immediate feedback. These combined factors strengthen learners’ self-efficacy and intrinsic motivation, promote the depth and breadth of cognitive processing, enabling learners to more effectively process and integrate new information, and comprehensively enhance their mastery and application of language knowledge.

However, the non-significant direct effect of IGP on LE may stem from its indirect nature and the multidimensionality of learning efficacy. Imagination Perception primarily promotes innovative thinking and imagination, which may need to indirectly enhance learning efficacy through mediating variables such as psychological imagery, rather than directly. Moreover, individual differences among learners, the quality of instructional design, the choice of assessment methods and indicators, and the diversity of sample characteristics may all affect the effectiveness of Imagination Perception. If instructional activities do not effectively integrate imagination perception or if learners struggle to apply it to specific learning tasks, its role in promoting learning efficacy may not be directly apparent. Therefore, while Imagination Perception has its unique value, its contribution to learning efficacy may be more complex and conditional.

1. (2). Moderating effect

In this study, the hypotheses regarding the moderating effects include H2a, H2b, H2c, H2d, H2e, and H2f. The hypotheses supported by the findings are H2c and H2f, while those not supported are H2a, H2b, H2d, and H2e. The support for H2c indicates that Representational Ability (RA) moderates the impact of Imagination Perception (IGP) on Psychological Image (PI), with a negative moderating effect, as indicated by β_H2c = −0.102. The simple slope chart of this moderating effect is shown in Fig 6. The support for H2f suggests that RA has a positive moderating effect on the role of IGP in enhancing Learning Efficiency (LE), with β_H2f =  0.071. The simple slope chart illustrating this positive moderating effect is depicted in Fig 7.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Negative moderating effect of representation ability on imaginative perception and psychological image.

https://doi.org/10.1371/journal.pone.0318548.g006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Positive moderating effect of representation ability on imaginative perception and learning efficiency.

https://doi.org/10.1371/journal.pone.0318548.g007

The negative moderating effect of RA on the relationship between IGP and PI implies that when learners possess strong representational skills, they inherently have the capacity to construct concrete images and scenarios. However, when the content presented in metaverse classrooms deviates from their preconceived notions, it may lead to cognitive dissonance. This phenomenon is commonly observed between readers’ psychological expectations of literary works and the visual representation in film adaptations, often resulting in adaptations that fail to meet viewers’ expectations, leading to reduced acceptance and negative evaluations.

The positive moderating effect of RA on the impact of IGP on Learning Efficiency (LE) can be understood as follows: without the mediating role of psychological imagery, imagination perception may not have a significant direct effect on learning efficacy. However, when learners have high representational ability, they can more effectively utilize imagination perception in the metaverse environment to foster creative thinking and innovative learning strategies, thereby enhancing learning efficacy. In other words, representational ability acts as a catalyst in this process, accelerating the transformation of imagination perception into improved learning efficacy.

Representational ability does not moderate other pathways, possibly due to the complexity and diversity of the learning process. Firstly, RA does not moderate the influence of immersion perception and interaction perception on psychological imagery, as these perceptions directly affect learners’ emotional and cognitive engagement in a relatively direct and autonomous manner, independent of individual representational ability. In essence, immersion and interaction perceptions can independently promote the formation of psychological imagery because they provide rich experiences and immediate feedback that learners can directly perceive and respond to. Secondly, RA does not moderate the impact of immersion and interaction perceptions on learning efficacy, as these perceptions may already provide learners with sufficient motivation and engagement, thereby directly enhancing learning efficacy without the need for additional moderation by representational ability. These perceptions foster active participation and deep cognitive processing by learners by creating engaging learning environments and interactive learning opportunities, both of which are key factors in improving learning efficacy. Moreover, learning efficacy, as a comprehensive construct, may be influenced by various factors, including but not limited to individual traits, learning strategies, teaching methods, and learning environments. While representational ability is an important component of these factors, it may not be the sole or primary moderating variable.

1. (3). Mediating effect

The mediating effect examines the role of psychological imagery (PI) in the influence of the metaverse classroom’s “3I” characteristics on learning efficacy. This study involves three hypotheses, namely H3a, H3b, and H3c, which represent the potential impacts of ISP, ITP, and IGP on learning efficacy (LE) through PI, respectively. All three hypotheses are supported, confirming that psychological imagery serves as a significant mediator between the “3I” characteristics of metaverse classrooms and learning efficacy. Specifically, immersion perception enhances students’ emotional engagement and learning experience by providing a deep sense of participation and cultural immersion; interaction perception stimulates students’ language practice and learning motivation through effective interactivity; and imagination perception promotes the construction of language knowledge and enhancement of learning experience through creative thinking and imagination. These perceptual dimensions significantly improve learning efficacy by strengthening students’ psychological imagery, that is, their mental representation and cognitive structure of language knowledge, scenarios, and communication processes. The research findings reveal three key dimensions that should be emphasized in the design of metaverse classrooms and provide educators with a theoretical basis for optimizing metaverse teaching strategies to achieve more effective educational outcomes.

1. (4). Total effect

This paper explores the overall impact of the “3I” characteristics of metaverse classrooms—Immersion Perception (ISP), Interaction Perception (ITP), and Imagination Perception (IGP)—on learning efficacy. The total effect test results show that H4a and H4b are supported, indicating that both immersion perception and interaction perception significantly positively affect learning efficacy. In contrast, the insignificance of H4c indicates that the impact of imagination perception on learning efficacy is not significant.

Immersion Perception enhances students’ emotional engagement and confidence in completing tasks by creating an in-depth simulated learning environment that provides an immersive experience. This high level of cultural immersion and sense of participation strengthens students’ emotional investment. Interaction Perception, through real-time interaction in three-dimensional dynamic scenes, strengthens the interaction between students, teaching content, teachers, and classmates, promoting collaborative learning and deepening exploration and engagement with the content, further enhancing the confidence and ability to achieve learning objectives. Specifically, the coefficients for immersion perception (βH4a = 0.161) and interaction perception (βH4b = 0.189) suggest that in language learning, the importance of communication and interaction may surpass mere situational immersion, with interaction perception being particularly crucial for enhancing language practical skills.

However, the overall impact of Imagination Perception on learning efficacy is not significant, implying that creative thinking and imagination do not independently and significantly affect learning efficacy. Further analysis reveals that imagination perception needs the moderating effect of representational ability and the mediating effect of psychological imagery to significantly enhance learning efficacy.

Therefore, the implications of this study for metaverse language teaching are that immersion perception and, especially, interaction perception should be emphasized, as they play a vital role in enhancing language practical skills. At the same time, to fully leverage the role of imagination perception, guidance and cultivation through specific language use contexts are needed, which may involve innovative teaching methods and rich metaverse environment designs.

1. (5). Demographic influence

As indicated in Table 5, age and grade level have the most significant negative impact on learning efficacy, while other demographic characteristics do not show a noticeable effect. The explanation for this finding is that in this study, younger students may exhibit greater learning efficacy due to their stronger adaptability to new technologies and curiosity. Additionally, students in lower grade levels might demonstrate higher motivation and engagement in learning due to the novelty of the content and environment, as well as less academic pressure. However, other demographic characteristics such as gender and living expenses may have less direct connection to the learning process, or their impact may be obscured by mediating variables like learning motivation and learning strategies, thus their direct influence on learning efficacy is not significant in this study.

4.3. Comprehensive discussion on the research results

To gain a deeper understanding of the impact of metaverse teaching on student learning efficacy, this paper thoroughly explores the “3I” technological characteristics of metaverse teaching (Immersion Perception, Interaction Perception, and Imagination Perception) and the mechanisms by which psychological imagery affects learning efficacy. This study incorporates theories from cognitive psychology, including Dual Coding Theory, Constructivist Learning Theory, and Mental Imagery Theory, and constructs an integrated model that also considers demographic characteristics such as gender, age, consumption level, academic inclination, and grade. It investigates how metaverse teaching characteristics, psychological imagery, and representational ability collectively enhance learning efficacy and how demographic characteristics influence these relationships.

The study’s findings indicate that Immersion Perception and Interaction Perception have significant positive effects on learning efficacy, while the impact of Imagination Perception is not significant. Psychological imagery plays a partial mediating role between metaverse teaching characteristics and learning efficacy, and representational ability moderates certain pathways. Additionally, age and grade among demographic characteristics significantly affect learning efficacy, while other characteristics like gender and living expenses do not. These findings provide a theoretical and empirical foundation for the practical application and strategic optimization of metaverse teaching.

Psychological imagery’s significant mediating role aligns with the perspective of Embodied Cognition Theory, which posits that cognitive processes result from the interaction of bodily experiences and the environment [14]. However, research based on Embodied Cognition Theory has not identified the key mediating mechanism of psychological imagery as a variable. This paper confirms the crucial mediating role of psychological imagery, revealing how metaverse teaching characteristics influence learning efficacy through the internal representation of psychological imagery, thus offering a new perspective and empirical support for personalized and efficient metaverse teaching.

Constructivist Learning Theory posits that learning is achieved through active construction and social practice. The metaverse teaching environment, with its highly interactive and immersive space, facilitates learners’ active exploration and construction of knowledge. Therefore, metaverse teaching can enhance learning motivation, participation, and outcomes by promoting learners’ practical operations and social interactions in a virtual environment [12]. This paper further explores the mediating role of psychological imagery between metaverse teaching characteristics and learning efficacy and analyzes how representational ability moderates this mediating relationship. The construction in the metaverse teaching environment pertains to the formation of psychological imagery, making this analysis more specific and in-depth than the general active construction discussed in Constructivist Learning Theory, providing more direct guidance for practice and offering empirical and theoretical support for the formulation and optimization of metaverse teaching strategies.

Flow Theory suggests that the immersive learning environment provided by metaverse classrooms can promote complete engagement and focus among students. Metaverse teaching, utilizing VR, AR, and MR technologies, creates a learning environment that aligns with the immersive experience described in Flow Theory, where students are more likely to achieve a state of flow under conditions of balanced challenges and skills, clear goals, and immediate feedback. This state enhances students’ intrinsic motivation, promotes cognitive development and deep learning, and stimulates emotional engagement, improving the quality of the learning experience [13]. This study confirms the direct effect of Immersion Perception on learning efficacy, corroborating some of the achievements of Flow Theory regarding immersive experiences enhancing learning outcomes. Moreover, we have further established that Immersion Perception can influence learning efficacy through psychological imagery, providing more specific and in-depth insights than the general active construction discussed in Constructivist Learning Theory, and offering direct guidance for practice, as well as empirical and theoretical support for the formulation and optimization of metaverse teaching strategies.

Additionally, the core contribution of this study lies in revealing the critical moderating role of representational ability in metaverse classrooms. Specifically, representational ability has a negative moderating effect on the relationship between Imagination Perception and psychological imagery, while it positively affects the link between Imagination Perception and learning efficacy. This finding highlights the importance of individual traits in the metaverse teaching environment and guides us to delve into how individual characteristics like representational ability shape the learning process and outcomes.

These findings also remind us that when designing and implementing metaverse teaching strategies, it is essential to consider students’ individual differences to ensure that each student can have an optimized learning experience. Furthermore, the moderating role of representational ability offers significant insights for the formulation of teaching strategies. It suggests that when constructing metaverse teaching activities, careful design is needed to balance students’ Imagination Perception and representational ability, thereby stimulating more efficient learning and cognitive growth.

This further requires us to reflect on how to integrate customized learning support within the metaverse teaching framework and how to cleverly utilize instructional design to maximize the potential of each student. In summary, this study not only provides new perspectives for understanding the complexity of metaverse teaching but also offers a theoretical foundation and practical guidance for achieving personalized and effective teaching practices.

5.1. Conclusion

This study delves into the impact of metaverse classroom “3I” technological features (Immersion Perception, Interaction Perception, and Imagination Perception) and their relationship with psychological imagery and representational ability on learning efficacy. A comprehensive theoretical model framework was constructed, proposing and validating a series of hypotheses regarding the relationships between the metaverse’s “3I” characteristics, psychological imagery, representational ability, and learning efficacy. The study reaches the following main conclusions:

Firstly, the positive impact of the “3I” characteristics of metaverse teaching on learning efficacy: The findings reveal that Immersion Perception and Interaction Perception in metaverse teaching significantly enhance learning efficacy. This indicates that through highly realistic virtual environments and enhanced interactivity, metaverse teaching can improve students’ emotional engagement, focus, and comprehension, thereby increasing learning efficacy.

Secondly, the mediating role of psychological imagery: Psychological imagery serves as a partial mediator between the characteristics of metaverse teaching and learning efficacy. This implies that the internal mental representations formed through metaverse teaching can facilitate students’ in-depth understanding and retention of learning materials, stimulate innovative thinking and exploratory spirit, and thus enhance learning efficacy.

Lastly, the moderating role of representational ability: Representational ability moderates certain pathways within metaverse teaching. Specifically, representational ability negatively moderates the impact of Imagination Perception on psychological imagery, while positively moderating the effect of Imagination Perception on learning efficacy. This reveals the importance of individual traits such as representational ability in the metaverse teaching process and how it affects students’ absorption of teaching content and learning outcomes.

These conclusions provide a theoretical and empirical basis for the practical application and strategic optimization of metaverse teaching, emphasizing the importance of considering individual differences and creating immersive, interactive learning environments when designing and implementing metaverse teaching activities. The study also points to the potential of Imagination Perception in enhancing learning efficacy and the need for further research and innovative teaching methods to fully leverage its role.

5.2. Countermeasure and suggestion

Metaverse teaching, as an emerging educational model, relies on technologies such as VR (Virtual Reality), AR (Augmented Reality), and MR (Mixed Reality) to provide an immersive learning environment and has the potential to revolutionize traditional education models. However, to fully leverage the advantages of metaverse teaching, teaching strategies must be carefully designed and implemented. Based on the research findings, this paper proposes the following strategies for metaverse teaching:

1. (1). Enhance the immersive learning experience by utilizing metaverse technology to create highly realistic learning environments that promote emotional engagement and learning experiences. Design student-centered teaching activities that encourage active exploration and learning. Specifically, highly realistic learning environments can be created: for example, using virtual reality technology to develop immersive learning scenarios that match the course content, such as historical reenactments, virtual laboratories, or language learning environments, to enhance students’ situational awareness and cultural understanding.
2. (2). Promote interactivity and collaborative learning by enhancing interactions among students through interactive 3D dynamic scenes, encouraging collaborative learning, and improving learning motivation, participation, and outcomes. Establish virtual classrooms and discussion areas to encourage real-time communication and collaboration among students, and promote knowledge sharing through team projects and group discussions. Utilize the social features of metaverse platforms to organize virtual academic conferences or debates to cultivate students’ critical thinking and communication skills.
3. (3). Stimulate creative thinking and personalized learning by encouraging students to use innovative tools and activities to stimulate creative thinking, while also considering individual differences among students, providing personalized learning pathways and resources. For example, combining the imaginative characteristics of the metaverse, design innovative tasks such as virtual architectural design, digital art creation, etc., to stimulate students’ creativity and imagination. Guide students to conduct innovative experiments through virtual scenarios, explore new solutions to problems, and develop the ability to solve complex problems.

In addition, it is necessary to focus on improving learning efficacy by ensuring that learning activities match the challenges and skills of students and by promoting students’ focus and learning efficiency through clear goals and immediate feedback. Recognize that there are differences in students’ adaptability to new technologies and provide necessary support and training for different students. Finally, combine metaverse teaching with traditional teaching methods to leverage their respective advantages and achieve the best teaching outcomes. These strategic recommendations aim to fully utilize the advantages of metaverse teaching, while considering the challenges that may be encountered in the implementation process, with the goal of achieving personalized and efficient education. Through these comprehensive strategies, a richer, more effective, and attractive learning experience can be provided for students, thereby promoting innovation and development in the field of education.

5.3. Limitation and future work

This study delves into the “3I” technological characteristics of metaverse classrooms (Immersion Perception, Interaction Perception, and Imagination Perception) and their relationship with psychological imagery and representational ability, as well as their impact on learning efficacy. While this study offers valuable insights and empirical support, there are limitations that suggest directions for future research.

Firstly, the sample of this study is drawn from a specific group of students in a metaverse language learning course, which may limit the generalizability of the findings. Different subjects or teaching content may require distinct teaching strategies and technological applications; hence, future research should verify the conclusions of this study across a broader range of academic disciplines and teaching scenarios.

Secondly, the study primarily relies on survey questionnaires to collect data, which may be subject to the participants’ subjectivity. Future research could employ mixed methods, combining quantitative and qualitative data for a more comprehensive understanding.

Thirdly, although this study takes into account the moderating role of representational ability, there are other individual difference factors, such as learning styles and motivation levels, that could affect the effectiveness of metaverse teaching. These factors warrant further exploration in future research.

Lastly, with the rapid development of metaverse technology and the emergence of new tools and platforms, future research could focus on how these new technologies can further optimize metaverse teaching and their potential impact on educational practices.

In summary, this study provides a theoretical and empirical foundation for the practical application and strategic optimization of metaverse teaching and points the way for future research to promote the theoretical and practical development of related fields.