Introduction
The purpose of practical learning is to enable students to better apply their theoretical knowledge in the real environment. Practice learning and theory learning complement each other, so that students can achieve better learning outcomes. With the continuous progress of science and technology, an increasing number of new digital devices and applications have been used for the purpose of practical learning (Zawacki-Richter & Latchem,
2018). Among these new technologies, devices and applications based on desktop virtual reality (VR) technology are gradually valued and applied in different educational domains (Radianti et al.,
2020).
Desktop VR technology has been widely utilized in learning. Currently, many learning systems based on desktop VR technology have emerged around the world and a large number of learners are using them. Labster is one of the most well-known and widely used desktop VR platforms, offering a variety of different desktop VR systems for learners around the world. According to the data from it, up to March 2022, Labster has provided virtual lab products to more than 5 million students from high schools and universities around the world. It has developed different desktop VR products for 39 major professional categories and has been used by more than 2,000 institutions and schools. In addition, according to the data from ilab-x, another large desktop VR platform, in November 2020, there were 353 national virtual simulation laboratory teaching programs in China. One year later, the number of national virtual simulation experimental teaching projects grew to 687. The annual growth rate reached 94.61%. According to the site, the most popular desktop VR system experiment has had more than 250,000 experiment visits, and the second and third most popular ones both have more than 100,000 visits. All data suggest that desktop VR technology is being used more and more extensively in education. Therefore, this study investigates whether learning with desktop VR technology can enhance learners' learning outcomes.
With the advantages and rapid development of desktop VR technology, it has been widely applied in different areas of education. As displayed in Table
1, many scholars have studied the application of desktop VR systems in different domains. For example, Makransky and Petersen (
2019) analyzed the relationship between VR features, presence, self-efficacy, knowledge and other variables of 199 first-year medical undergraduates when using a desktop VR system called
medical genetics simulation. Barrett and Blackledge (
2013) analyzed the relationship between variables such as immersion, perceived usefulness (PU), perceived ease of use (PEOU), presence, motivation of 87 final year undergraduate students studying electricity when using a desktop VR called ‘VES’.
Table 1
A selected review of research on desktop VR technologies in different application domain
| Biology | A bacterial isolation virtual lab simulation | Perceptions of assessment; INTRINSIC MOTIVATION; SELF-EFFICACY; TRANSFER | Paired samples t-tests |
| Biology | A desktop VR program called 'V-Frog™' | Performance achievement; spatial ability; learning mode | Descriptive statistics |
| Biology | A desktop VR program called 'V-Frog™' | VR features; PU; PEOU; presence; motivation; learning outcomes | SEM analysis |
| Chemistry | Second Life | VR features; usability; self-efficacy; learning outcome | SEM analysis |
| Construction | Desktop-based VR, immersive VR | | Summary |
Blackledge and Barrett ( 2013) | Electricity | Prototype model 'VES' | Immersion; PU; PEOU; motivation; intention to use; satisfaction | Case study |
| Engineering | A small VR system called 'PI-casso' | | Experiment |
| IT | A curriculum delivery application called 'lotus learning space' | Time; place; space; interaction; technology; learner control | Descriptive Statistics |
| Mathematics | A set of mathematics instructional games called Dimension™ | | Experiment |
Pasqualotti and Freitas ( 2002) | Mathematics | A virtual environment called 'MAT3D' | Performance | Case study |
Makransky and Petersen ( 2019) | Medicine | Desktop VR medical genetics simulation | VR features; PE; active learning; intrinsic motivation; self-efficacy; | SEM analysis |
| Medicine | A genetics simulation developed by Labster | Intrinsic motivation; self-efficacy; transfer | Paired samples t-tests |
| Nursing | The pharmacology inter-leaved learning VR | Learning outcome (course score) | Descriptive statistics |
Through the review and analysis of all the articles in Table
1, we can conclude that there is almost no research regarding changes in students' self-efficacy, goal orientation, technology acceptance and learning behavior before and after VR technology is used, and few scholars have studied the effectiveness of desktop VR systems in business learning. In business learning, it is impractical for learners to directly participate in the actual business decisions. The application of desktop VR technology enables business students to experience simulated business operations and the decision-making process before entering the workplace. Therefore, it is of practical importance to take desktop VR technology into business learning. Hence, this research aims to analyze the relationship among students' self-efficacy, goal orientation, technology acceptance (e.g., PEOU and PU), learning behavior and learning outcome, so as to clarify whether the use of desktop VR technology can enhance learners' learning outcomes. Besides, this research also measures the changes of these variables as well as gender difference in the early and late stages of the course study when desktop VR technology is applied in business simulation learning.
Desktop virtual reality
Virtual reality refers to "a specific type of reality simulation system constructed by the combination of hardware and software systems" (Biocca & Delaney,
1995). Depending on the level of immersion provided by VR equipment, VR technology can be categorized as immersive VR and non-immersive VR (Radianti et al.,
2020). When using immersive VR devices, users feel like they are immersed in a virtual world and do not perceive that they are interacting with a screen and a set of devices. Example of such VR technology includes HMDs (head-mounted displays) such as HTC Vive and enhanced VR such as data gloves (Khalifa & Shen,
2004; Martín-Gutiérrez et al.,
2017). On the contrary, when using non-immersive VR devices, users can still perceive that they are looking at a screen or interacting with the devices. The most commonly used non-immersive VR is desktop VR system (Biocca & Delaney,
1995; Robertson et al.,
1997). Desktop VR is a non-immersive VR consisting of a personal computer and software applications that can be interacted with by using common devices such as a keyboard and mouse (Ausburn & Ausburn,
2004; Chen et al.,
2004; Lee & Wong,
2014).
Immersive VR and non-immersive VR have their own advantages and disadvantages, and therefore can be applied to different contexts. Immersive VR has the characteristics of high immersion, so it has the potential to maximize learning efficiency. However, immersive VR also has many disadvantages in practice. First of all, long-term use of immersive VR devices may cause users to experience symptoms such as dizziness, nausea and vomiting dizziness. Secondly, the high equipment and maintenance costs of immersive VR prevent it from being widely used in practice learning (Chuah et al.,
2010; Lee & Wong,
2014; Merchant et al.,
2014). Dalgarno et al. (
2002) suggest that "immersion in virtual environment is caused by user's control over environment, interaction with the environment, not just the nature of the environment itself". Although it cannot make users experience complete immersion, non-immersive VR such as desktop VR also has its incomparable advantages. With the rapid iteration of computer chips and network technology, many desktop VR software was introduced, allowing users to use their personal computers or mobile phones to access desktop VR system whenever and wherever they want, which increases the convenience of using the technology (Dickey,
2005). Besides, compared with immersive VR, desktop VR has lower use and maintenance costs, which makes it more likely to be widely used (Srivastava et al.,
2019).
Methodology
Context
A course called Business Decision Simulation from SHU-UTS SILC Business School, Shanghai University was selected for this study. The course is offered in the third year of the undergraduate business administration program and lasts for 10 weeks. It was conducted using a desktop virtual simulation system called CESIM Global Challenge, in which students were required to form a team of four or five people to virtually run the business of a multinational company. At the end of the 10-week course, each team was ranked according to 10 rounds of business simulation. This course is designed to help management students understand and learn how to make corporate decisions in a dynamic business environment.
Participants
Participants of this study are 94 junior and senior students majoring in Business Administration from SILC Business School of Shanghai University (N = 94; 26 males, 68 females).
Instruments
The questionnaire used in this study is composed of four sub-questionnaires with a total of 44 questions, all of which have been used in existing studies to measure college students.
Self-efficacy was measured with the translated new general self-efficacy (NGSE) scale, which is a 5-point Likert scale consisting of 8 question items (Chen et al.,
2001). The Cronbach's
\(\alpha\) value was 0.86.
PU and PEOU was measured using a sub-questionnaire from a TAM scale, which is a 5-point Likert scale consisting of 8 question items (Venkatesh,
2000). The Cronbach's
\(\alpha\) values were 0.87 for PU and 0.86 for PEOU.
A translated version of the validated Achievement Goals Questionnaire was adapted to measure students' goal orientation (Elliot & McGregor,
2001), which is a 5-point Likert scale consisting of 12 items with each of the four goal orientations consisting of three questions. The Cronbach's
\(\alpha\) values were 0.87 for mastery approach, 0.89 for mastery avoidance, 0.92 for performance approach and 0.83 for performance avoidance.
A translated version of the R-SPQ-2F was adapted to measure students' learning behavior (John et al.,
2001). this scale is a 5-point Likert scale consisting of 20 items, with 10 question items for deep and surface learning approach, respectively. The Cronbach's
\(\alpha\) values were 0.73 for deep learning approach and 0.64 for surface learning approach.
Data collection procedure
Students' self-efficacy, goal orientation, technology acceptance, and learning behavior are obtained through questionnaire research. In the second class of the first week (time point A), the course instructor first introduced the basic information and rules of the VR system to all students, and then asked them to form groups. Each group needed to determine their company name, slogan, and each individuals’ role in the company. After that, the course assistant distributed the questionnaire and took the first measurement. In the last course of the tenth week (time point B), when all course instruction has been completed and the simulation is finished, a questionnaire is distributed by the course assistant for a second measurement.
Student's learning outcomes are measured by the final score of the Business Decision Simulation course, as course scores are the most accurate and quantitative indicator that can reflect students’ learning outcomes directly. The total score is 100 points which is composed of students' course participation score, simulated decision making score, and presentation score.
Data analysis
Descriptive statistical analysis was used to get a basic overview of the sample. The scales all passed a reliability test (Cronbach's alpha). Paired sample T tests were conducted to determine if there were statistically significant differences between different goal orientation, self-efficacy, PU, PEOU and learning behaviors in the early and late stages of the ten weeks of study. The relationships between variables were determined by correlation analysis. Independent sample t-tests were conducted to reflect differences in variables between genders.
Results
The reliability of each scale at time point A and B is shown in Table
3. All scales had Cronbach's alphas greater than 0.7 at both time points, which indicates good reliability of the scales.
Table 3
Cronbach's alpha at time point A and B
Self-efficacy | .864 | .888 | 8 |
Goal orientation | | | |
Performance-approach | .898 | .848 | 3 |
Performance-avoidance | .812 | .792 | 3 |
Mastery-approach | .841 | .755 | 3 |
Mastery-avoidance | .740 | .755 | 3 |
PEOU | .794 | .808 | 4 |
PU | .874 | .818 | 4 |
Learning behavior | | | |
Deep learning | .887 | .897 | 10 |
Surface learning | .818 | .822 | 10 |
Changes of variables over time
The means and standard deviations of the variables at time point A and B are shown in Table
4. The tested student group exhibited high level of self-efficacy at time point A and B (M = 3.98; 4.15). For goal orientation, compared with the scores on the variables of performance avoidance (M = 3.48; 3.16) and mastery avoidance (M = 3.44; 3.28), the tested sample showed higher levels of performance approach (M = 3.91; 3.88) and mastery approach (M = 4.11; 3.83). In addition, PU and PEOU was reported on an below average level at time point A (M of PU = 3.75; M of PEOU = 3.50), while both scores increased at time point B(M of PU = 4.00; M of PEOU = 3.78). For learning behaviors, the respondents showed moderate levels of deep learning behaviors (M = 3.70; 3.71) and lower levels of surface learning behaviors (M = 2.81; 3.07) on both time point A and B. Learning outcomes were measured by the students' course grade, with an average score of 86.38 for all students involved in the study.
Table 4
Means and standard deviations of the variables (5-point scale) at time point A and B
Self-efficacy | 3.98 | .53 | 4.15 | .57 |
Goal orientation | | | | |
Performance-approach | 3.91 | .75 | 3.88 | .81 |
Performance-avoidance | 3.48 | .84 | 3.16 | .90 |
Mastery-approach | 4.11 | .61 | 3.83 | .66 |
Mastery-avoidance | 3.44 | .82 | 3.28 | .79 |
PEOU | 3.50 | .71 | 3.78 | .68 |
PU | 3.75 | .66 | 4.00 | .62 |
Learning behavior | | | | |
Deep learning | 3.70 | .58 | 3.71 | .61 |
Surface learning | 2.81 | .58 | 3.07 | .62 |
Learning outcome | | | 86.38 | 5.14 |
Paired-sample t-tests were used to determine whether the variables were significantly different after ten weeks of learning using desktop VR. The results of the paired-samples t-test are provided in Table
5. Significant increase in students' self-efficacy, PEOU and PU occurred from time point A to B. For goal orientation, mastery approach and mastery avoidance decreased significantly, while performance approach and performance avoidance did not change significantly. Surface learning behaviors decreased significantly from time point A to B, while no significant changes occurred for deep learning behaviors.
Table 5
Paired sample T test of variables from time point A to B
Self-efficacy | A | 3.98 | .53 | − 2.227 | 93 | .028** |
| B | 4.15 | .57 | | | |
Goal orientation | | | | | | |
Mastery-approach | A | 4.11 | .61 | 2.914 | 93 | .004* |
| B | 3.83 | .66 | | | |
Mastery-avoidance | A | 3.44 | .84 | 3.310 | 93 | .001* |
| B | 3.03 | .90 | | | |
PEOU | A | 3.50 | .71 | − 2.790 | 93 | .006* |
| B | 3.78 | .68 | | | |
PU | A | 3.75 | .66 | − 2.640 | 93 | .010* |
| B | 4.00 | .62 | | | |
Learning behavior | | | | | | |
Surface learning | A | 2.81 | .59 | − 2.960 | 93 | .004* |
| B | 3.07 | .62 | | | |
The relationships between variables
Correlation analysis was used to analyze the relationship between self-efficacy, different goal orientations, PEOU, PU, and learning behaviors.
As shown in Table
6, self-efficacy is positively related to performance approach and mastery approach goal orientations at both time points A and B, and positively related to performance-avoidance goal orientation at time point A. In addition, self-efficacy showed significant positive correlations with PEOU and PU at both time points.
Table 6
Significant relationships between the concepts at time point A and B
Self-efficacy—goal orientation | | |
Self-efficacy—performance-approach | | |
| A | .427* |
| B | .487* |
Self-efficacy—performance-avoidance | | |
| A | .273* |
Self-efficacy—mastery-approach | | |
| A | 284* |
| B | \(.387*\) |
Self-efficacy—PEOU | | |
| A | .287* |
| B | .324* |
Self-efficacy—PU | | |
| A | .330* |
| B | .446* |
Goal orientation—Learning behavior | | |
Performance-approach—Deep learning | | |
| A | .474* |
| B | .467* |
Mastery-approach—deep learning | | |
| A | .677* |
| B | .441* |
Performance-avoidance—surface learning | | |
| A | .518* |
| B | .491* |
Mastery-approach—surface learning | | |
| A | − .268* |
| B | − .239* |
Mastery-avoidance—surface learning | | |
| A | .457* |
| B | .366* |
PEOU—learning behavior | | |
PEOU—deep learning | | |
| A | .211** |
| B | .468* |
PEOU—surface learning | | |
| B | .213** |
PU—learning behavior | | |
PU—deep learning | | |
| A | .395* |
| B | .589* |
Learning behavior—learning outcome | | |
Deep learning—learning outcome | | |
| B | .312* |
Surface learning—learning outcome | | |
| B | − .263** |
For the relationship between goal orientation and learning behavior, performance approach and mastery approach are positively correlated with deep learning behavior at both time point A and B, performance avoidance and mastery avoidance are positively correlated with surface learning behavior, while mastery approach is negatively correlated with surface learning behavior at two time points. The correlation between PEOU and deep learning is significant at both time points, and the correlation between PEOU and surface learning behavior is significant only at time point B; PU showed positive correlation with deep learning behavior at A and B, but not significant correlated with surface learning behavior.
The results of the correlation analysis also showed that students' learning behaviors affect their learning outcomes. Deep learning behaviors positively affect their learning outcomes, while surface learning behaviors negatively affect their learning outcomes.
Gender differences between variables
Independent samples t-tests were conducted to compare the variables on gender (see Table
7).
Table 7
Significant differences with respect to gender
Self-efficacy | | | | | | | |
A | 4.23 | .56 | 3.88 | .50 | 2.737 | 40 | .009 |
B | 4.40 | .64 | 4.06 | .52 | 2.447 | 38 | .019 |
PEOU | | | | | | | |
A | 3.79 | .83 | 3.39 | .63 | 2.523 | 92 | .013 |
PU | | | | | | | |
A | 4.07 | .74 | 3.63 | .59 | 2.731 | 37 | .010 |
B | 4.25 | .65 | 3.91 | .59 | 2.345 | 41 | .024 |
There were significant differences between the mean scores of men and women in terms of self-efficacy, PEOU and PU. Specifically, men showed higher levels of self-efficacy and PU than women at both time points A and B. Moreover, at time point A, PEOU was also significantly higher for men than for women, while this significant relationship disappeared at time point B.
Conclusions and discussion
Seven hypotheses and two research questions were formulated in this study. These hypotheses and research questions were confirmed and answered by collecting questionnaires and analyzing the data. Hypotheses formulated namely:
The results of the hypotheses validation are shown in Table
8.
Table 8
Research hypotheses results
H1 | Self-efficacy | PU | .330* | .446* | Confirmed |
H2 | Self-efficacy | PEOU | .287* | .324* | Confirmed |
H3 | Self-efficacy | Goal orientation | | | Partially confirmed |
| Self-efficacy | Performance approach/avoidance | .427*/.273* | .487*/.141 | |
| Self-efficacy | Mastery approach/avoidance | .284*/.032 | .387*/−.138 | |
H4 | PU | Learning behavior (Surface/Deep approach) | − .198/.395* | −.001/589* | Partially confirmed |
H5 | PEOU | Learning behavior (Surface/Deep approach) | −.085/.211* | .213**/.468* | Partially confirmed |
H6 | Goal orientation | Learning behavior (Surface/Deep approach) | | | Confirmed |
| Performance approach/avoidance | Surface approach | −.202/.518* | −.109/.491* | |
| Performance approach/avoidance | Deep approach | .474*/−.053 | .467*/−.100 | |
| Mastery approach/avoidance | Surface approach | −.268*/.457* | −.239**/.366 | |
| Mastery approach/avoidance | Deep approach | .677*/.104 | .441*/−.155 | |
H7 | Learning behavior (Surface/Deep approach) | Learning outcome | | −.263**/.312* | Confirmed |
H1 and H2 were confirmed at both time points A and B. The results of the correlation analysis showed that self-efficacy was positively associated with perceived ease of use and perceived usefulness at both time points A and B.
H3 was partially confirmed. It was found that Self-efficacy was significantly and positively related to performance approach, performance avoidance, and mastery approach goal orientations at time point A, and correlated with performance approach and mastery approach at time point B.
H4 was partially confirmed by the significant positive correlation between PEOU and deep learning at time points A and B and the correlation with surface learning at time point B.
H5 was partially confirmed by the results of the correlation analysis which indicated that PU was significantly correlated with surface learning at both time points A and B but not significant correlated with deep learning behavior.
H6 was confirmed by the significant positive correlations between approach goal orientations and deep learning behavior and between avoidance goal orientations and surface learning behavior. Besides, a negative correlations between mastery approach and surface learning had been found.
H7 was confirmed by the significant positive correlations between deep learning behavior and learning outcome as well as the significant negative correlations between surface learning behavior and learning outcome.
Changes and relations of variables during the use of desktop VR
Self-efficacy, goal orientations, technology acceptance (PEOU and PU), and learning behaviors all changed before and after desktop VR is used for learning. In addition to the changes in the variables themselves, the interrelationships between these variables also changed.
First, students' self-efficacy increased significantly after using desktop VR, but performance avoidance goal orientation did not change, which led to a change in the relationship between these two variables. A significant positive correlation between self-efficacy and performance avoidance goal orientation was found at time point A, but this correlation was no longer significant after ten weeks of study.
Second, both PEOU and surface learning behaviors increased significantly after using the desktop VR for ten week. No significant correlation existed between PEOU and surface learning behaviors at time point A, but a significant positive correlation was found at time point B.
Possible explanations for these changes in a short time span are the students' perceptions and attitudes toward the learning environment and course requirements, and the regulation of their own behavior in response to these objective conditions. In this study, self-efficacy was positively related to approach goal orientations, which in turn was positively related to deep learning behaviors. This fits with the results of related studies which indicate that students with high self-efficacy are more probable to become active participants (i.e., develop a mastery approach and/or performance approach goal orientation) and exhibit positive learning behaviors that may lead to good learning outcomes (Caraway et al.,
2010; Elliot,
1999).
The current study showed that there was a significant positive correlation between self-efficacy and technology acceptance, and technology acceptance then have a positive impact on deep learning behavior. This is in agreement with the findings of some earlier studies. Chen et al. (
2001) suggested that individuals with high self-efficacy are more willing to try and learn a new system or device and exhibit higher level of perceived ease of use and perceived usefulness. Furthermore, Zheng and Li (
2020) pointed out that self-efficacy affects users' technology acceptance, which further influences their intention to use as reflected in deep learning behavior.
Besides, this research also found that different learning behaviors result in different learning outcomes. Goh (
2005) proposed that deep learning behaviors are positively associated with positive learning outcomes and surface learning behaviors are negatively associated with learning outcomes. Magdalena (
2015) suggested that deep learning behaviors are associated with high performance. Consistent with these studies, the present study found that learners' deep learning behaviors are positively associated with learning outcomes and surface learning behaviors are negatively associated with learning outcomes. This result suggests that differences in learners' learning behaviors have an impact on their learning outcomes when learning with desktop VR.
In short, learners' learning strategies are constantly adjusted as learners' self-efficacy, technology acceptance, and goal orientation change. Liem et al. (
2008) argued that students choose to change their learning strategies when they believe they can achieve the same or even higher scores when they switch from deep learning strategies to surface learning strategies, and vice versa. Students' self-efficacy, goal orientation, and learning behaviors change as they become familiar with environmental and curricular requirements. Therefore, it is essential that educators are aware of how students' self-efficacy, goal orientation, and learning behaviors change over a relatively short time span, especially over a course cycle. In addition, for courses that require the use of desktop VR systems for learning, system developers need to provide students with some guidance and basic instruction to help them become familiar with the system and enter the learning process more quickly, as acceptance of the VR system enhances students' learning behaviors and further enhances their learning outcomes.
Gender difference and self-efficacy, PEOU and PU
In the present study, male and female students showed significant differences. According to the results of the independent samples t-tests in Table
7, Males scored significantly higher than females on self-efficacy, PEOU and PU. This is consistent with the results of some earlier studies. Numerous studies on self-efficacy have found that females' self-efficacy can be slightly lower than that of males (Turner & Schieman,
2008). Ong and Lai (
2006) have shown that males scored significantly higher than females on computer self-efficacy, perceived usefulness, perceived ease of use. In line with this research, Yukun et al. (
2013) also confirmed the differences between males and females in self-efficacy and perceived ease of use.
Social and status differences associated with both females and males may explain the gender differences in self-efficacy (Ma et al.,
2015; Schwarzer et al.,
1999). Such innate gender differences and social status distinctions can lead to cognitive differences between males and females. Females may be more susceptible to stress, less able to cope with their environment, and therefore more likely to experience negative emotions. Males, on the other hand, are relatively more ambitious and independent and therefore exhibit a higher sense of self-efficacy. The difference in self-efficacy further influences the difference in technology acceptability between men and women.
In addition, the differences between self-efficacy, and technology acceptance between men and women may also be caused by differences in thinking patterns between men and women. Men may show higher interest and greater receptivity when using a new system or technology (Braak,
2004; Schumacher & Morahan-Martin,
2001; Teo & Lim,
1996). However, differences between men and women were not found in goal orientation and learning behavior. This suggests that there are no significant differences in goal orientation and learning behaviors between genders when desktop VR is used for learning, and that high scores in self-efficacy and technology acceptance among males do not result in different goal orientation and learning behaviors compared to females. A better understanding of gender differences in user attitudes toward desktop VR systems could help researchers consider gender factors when developing and testing desktop VR learning systems in the future. In addition, administrators and tutors can become aware that the same desktop VR system may be perceived in a different way depending on gender and then improve user acceptance by making targeted adjustment of the course schedule and content.
Limitations and future research
This study contributes to the continuously increasing research on self-efficacy, goal orientation, technology acceptance, and learning behavior. particularly in environments using desktop VR, as much of the research is conducted in more conventional course environment. Nevertheless, this study has some limitations that have to be considered. Firstly, the relatively small sample size of this study does not allow for testing of concepts. The opportunity to use path analysis (e.g., structural equation modeling) to test conceptual models is prohibited. The advantage of path analysis is that interrelationships can be identified. Secondly, two measurements were administered over a 10-week time span. Students may be less motivated to fill out the questionnaires. Finally, all measures were self-report measures.
The theoretical model presented in this research was partially confirmed, but future planned studies should also be conducted for different desktop VR systems and increase the sample size to increase the applicability of findings. Since self-efficacy, goal orientation, technology acceptance and learning behavior were found to be subject to change within a short time span, it is necessary for future research to investigate specific methods and pathways to influence these variables and ultimately affect learning behavior in a positive way (i.e., achieve deep learning).
Desktop VR is a practical learning technology that gives students more hands-on opportunities and increases their sense of presence compared to traditional learning; it does not require high equipment costs and does not cause adverse reactions from users compared to immersive VR. Therefore, using desktop VR for learning is a compromise solution. As desktop VR is increasingly used in different teaching and learning areas, its impact on students should also be studied in more depth. Future research ought to consider the role of desktop VR in learning from more perspectives and investigate its impact paths and change paths in greater depth.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.