Virtual Reality in astronomy research and space-related education: a systematic review

Virtual Reality (VR) offers visualization capabilities particularly suited to astronomy and to Space-related education, yet its two roles, educational tool and research instrument, had not been synthesized jointly. This systematic review analyses 20 studies published between 2020 and 2025 to identify the benefits of VR in astronomy research and in education for Space-related fields. Following the PRISMA 2020 protocol, Scopus, Web of Science and Teacher Reference Center were searched. Of the included studies, 90% addressed educational applications and 10% (two studies) examined VR as a working instrument for astronomical research. Fifteen studies are empirical (30% of the corpus followed mixed methods, 25% quantitative and 20% qualitative designs) and five (25%) are technical or descriptive contributions; the quality of the empirical studies was appraised with the Mixed Methods Appraisal Tool (MMAT). The reviewed evidence indicates that VR can enhance students' performance and engagement and can help scientists to visualize and measure complex three-dimensional phenomena; controlled comparisons reported significantly higher quiz scores with VR than with traditional learning (8.31 versus 7.55) and 16% less time needed in the physical laboratory after virtual preparation. The evidence on VR for professional astronomical research, in contrast, is still too scarce to support robust conclusions. The review identifies theoretical and methodological gaps concerning longitudinal analysis, standardized assessment instruments and the integration of physiological measures. Future investigations should focus on a wider use of VR in astronomy research, on developing more robust metrics for assessing its impact, and on broadening the analysis of the skills involved in the use of such technologies.

Virtual Reality (VR) is defined in the Oxford Advanced Learner’s Dictionary as “images and sounds created by a computer that seem almost real to the user, who can interact with them by using sensors” (Virtual reality noun, 2026). Definition, Pictures, Pronunciation and Usage Notes | Oxford Advanced Learner’s Dictionary at Oxford Learners Dictionaries.Com). In practice, the technology combines stereoscopic visualisation, spatial audio and natural interaction to place the user inside a computer-generated environment. These properties make VR particularly well suited to space science, a discipline whose objects of study are remote, three-dimensional and often impossible to reproduce in a laboratory, and whose professionals must train for environments that cannot be safely rehearsed on Earth.

The concept predates the technology by several decades. Its first articulation is usually attributed to Stanley G. Weinbaum, whose 1935 science fiction story “Pygmalion’s Spectacles” describes a pair of goggles through which the protagonist experiences an entirely fictional world involving sight, smell, taste and touch. Two decades later, Morton Heilig built a machine that translated part of this vision into hardware: a multisensory cinema device that immersed the viewer in experiences such as riding a motorcycle, complete with wind on the face and the smells of the street. The decisive technical step came when Ivan Sutherland and his student Bob Sproull developed the first head-mounted display (HMD) capable of presenting moving three-dimensional imagery, a system now widely regarded as the direct ancestor of contemporary VR devices.

The space sector adopted the technology early. NASA, in collaboration with VPL Research, Inc., developed the Virtual Interface Environment Workstation (VIEW), a system comprising an HMD, a pair of instrumented gloves (DataGloves) and a full-body suit (DataSuit) that allowed users to visualise and interact with artificial environments generated by computer, or with real environments transmitted through video cameras. A succession of projects then brought space closer both to the public and to the professionals of the sector, including astronauts, engineers and scientists. The VR MIR Space Station (1995) and the VR International Space Station application at the Yu. A. Gagarin Cosmonaut Training Center (GCTC) (2012) were developed for astronaut training. The German ETS-VII Experiment (GETEX) (1999) tested the technique known as Projective Virtual Reality, in which actions performed by a user in the virtual world are projected onto the real world, an approach of particular value for space missions involving robotic control (Freund and Rossmann, 1999). EARTHX 2018, developed by students at the University of Birmingham, proposed a concept for astronaut survival training on another planet. Today, the NASA Johnson Space Center Virtual Reality Laboratory (VRL) operates as a training facility for Extravehicular Activity (EVA) and robotics operations (Garcia, 2023).

This trajectory shows that attempts to apply VR in the space sector have accompanied the technology since its origins. Most of these initiatives, however, have concentrated on astronaut training or on science outreach. Far less attention has been paid to VR as a working instrument for scientific and engineering practice, or to the way immersive environments shape the learning experience and the knowledge that users acquire. Clarifying these two dimensions is the purpose of the present review.

Rationale and objectives of the systematic review

VR has become widely accessible over the past decade, following the commercial release of the first Oculus Rift headsets. Although the technology emerged from the videogame industry, it has since spread across many domains, and applications in the space sector are now numerous. The research literature has not developed at the same pace in all fields. Most empirical evaluations of VR concentrate on medicine and health education, whereas comparatively few studies examine its impact in the space field, whether in education or in professional scientific and engineering activity. Previous systematic reviews confirm this imbalance and, at the same time, delimit the novelty of the present work. Radianti et al. (2020) mapped immersive VR applications in higher education and found that learning theories were rarely considered in the design of the applications. Hamilton et al. (2021) restricted their synthesis to experimental studies comparing head-mounted displays with less immersive methods and reported positive but heterogeneous learning outcomes. In the STEM domain, Pellas et al. (2020) reviewed VR-supported instructional design practices in K-12 and higher education, while Di Natale et al. (2020) examined a decade of empirical research on immersive VR in educational settings. In the health field, where the empirical literature is most abundant, Kyaw et al. (2019) meta-analysed the effect of VR on health professions education. None of these reviews addresses the space sector specifically, and none considers VR simultaneously as an educational tool and as a working instrument in astronomy research. The present review articulates both dimensions within a single framework, which, to the best of our knowledge, has not been attempted before.

The scope of the review is deliberately delimited. It covers articles published between 2020 and 2025, written in English or Spanish, that examine VR either as an educational tool in areas whose results can be transferred to the Space field or as a working instrument in astronomy and astrophysics research. The search strategy mirrors this delimitation: the strings combined Virtual Reality with education and with astronomy and astrophysics, so the instrument dimension of the review is confined to astronomy research and does not extend to other space-sector activities such as aerospace engineering, astronaut training, planetary science or mission operations, which were not searched for specifically. Domains such as medicine and health education, which concentrate most of the empirical literature on VR but whose findings cannot be applied to that context, fall outside this scope.

This review addresses that gap through four objectives: To integrate the main uses of VR in the space sector, both as a tool for scientific and engineering work and as an educational tool.

To evaluate the use and effectiveness of VR in these two roles.

To systematise the analytical methodologies employed in empirical studies of knowledge retention, understanding, engagement, emotion and adaptability.

To identify knowledge gaps and opportunities for future research on VR as a tool for education and for scientific and engineering work.

Research questions

Two research questions guide this review:

RQ1. How does VR influence users' knowledge acquisition in Space-related education, and how effective is it as a working instrument in astronomy research?

RQ2. Which methodological approaches allow this influence and effectiveness to be evaluated comprehensively and with empirical validation?

Each research question is linked to a specific synthesis strategy. RQ1 is answered through the thematic synthesis of the results, organized in three domains (astronomy research, education and users’ health) and condensed in the evidence matrix of Supplementary Table S2, which allows the findings of the individual studies to be compared across themes. RQ2 is answered through the methodological coding of the included studies, described in the Typology of the Studies section, and through the quality appraisal reported in the Quality Appraisal section, which together characterize the analytical approaches employed in the field and their empirical robustness. The Conclusions return explicitly to both questions.

PRISMA statement (Preferred Reporting Items for Systematic reviews and Meta-Analyses) was first published in 2009 as a guideline to report systematic reviews ensuring its transparency in why the review was done, what methods were used and what results were found. Due to the latest improvements in methodology and terminology of systematic reviews, PRISMA statement was updated in 2020 (PRISMA 2020), to reflect the recent progress in identification, selection, assessment and synthesis of studies (Page et al., 2021).

Therefore, to ensure transparency and scientific rigor in this systematic review, PRISMA 2020 statement is applied and its checklist is followed: (1) describing the rationale for the review and providing an explicit statement of the questions the review addresses; (2) specifying the criteria for the review and the information sources, defining the search strategy, the data collection and assessment process, the synthesis method applied to the studies and data collected, and the methods to assess risk of bias and certainty during the review; (3) exposing the results and describing the study characteristics, the risk of bias and the certainty of evidence; (4) providing a general interpretation of the results and discussing any limitation of the review process and the implications of the results for practice, policy and future research; (5) declaring additional information, like competing interests, sources of financial support, etc.

The protocol of this review was not registered in PROSPERO, since that registry only accepts systematic reviews with health-related outcomes; this circumstance is acknowledged in the Limitations of the Review. The search protocol, the data extraction sheet and the quality appraisal records are available from the corresponding author upon reasonable request.

This review focuses on empirical studies, previous reviews and theoretical articles addressing the impact of Virtual Reality technology in education and other activities related to Space science.

Development of the search protocol

The literature search was conducted in July and October 2025, using recognized academic databases in the field of science, technology and education. The search done in July was more generic, including in the analysis articles related to other branches of science, like medicine; while in October, the developed search was more focused in solely Space related articles. The used databases are briefly described below: Scopus: is a large database of peer-reviewed literature that covers science, technology, medicine, social science, arts and humanities.

Web of Science (WoS): is a scientific citation indexing service that provides access to a collection of databases in the fields of science, social science, arts and humanities.

Teacher Reference Center (TRC): is a database of peer-reviewed literature that covers topics from Assessment to Teacher Education.

These three databases were selected because, in combination, they cover the two dimensions of the review. Scopus and WoS are classified as principal search systems suitable for evidence synthesis on account of their coverage and of the precision, recall and reproducibility of their query interfaces (Gusenbauer and Haddaway, 2020), and comparative analyses document the breadth of their journal coverage, with the journal collection of WoS very largely contained within that of Scopus (Singh et al., 2021). TRC adds specialized coverage of teaching practice, and PubMed/MEDLINE was deliberately not consulted because the health field falls outside the scope defined in the Introduction. Two features of the eligibility criteria further limit the risk associated with not searching IEEE Xplore, the ACM Digital Library and ERIC directly. First, this review only includes peer-reviewed journal articles, and the content that most distinguishes those three sources from Scopus and WoS is precisely of the types the criteria exclude: conference proceedings in the case of IEEE Xplore and the ACM Digital Library, and grey literature such as reports and working papers in the case of ERIC, while the journal literature they index is to a very large extent also indexed by Scopus and WoS. Second, the two search strings were deliberately built from the broadest terms of the field to maximize recall, shifting the selection effort to the manual screening of 844 records rather than to the sensitivity of the query. The residual risk, namely journal articles published in venues indexed only by those sources, cannot be excluded, and its implications for the validity of the findings are discussed in the Limitations of the Review.

The wording applied on the databases search tab consisted of few Boolean operators connecting specific terms related to our research purpose. The search was done in two steps, with the wording “Virtual Reality AND ((education OR astrophysics) OR (education AND astrophysics))” as a first search, which took place in July, and “Virtual reality AND astronomy” as a second search, in October.

The two strings deliberately combined a small number of broad terms in order to maximize recall within the selected databases, at the cost of not incorporating narrower synonyms (e.g. head-mounted display, immersive environment) or controlled vocabulary such as thesaurus descriptors; this restriction is likewise acknowledged in the Limitations of the Review.

Later, to all the articles obtained under the wording described above, certain criteria were applied to filter the articles more relevant to the systematic review. The same criteria were used for the first and second wording search and is described below: Language: Spanish or English.

Period: between 2020 and 2025.

Document type: articles.

Besides, for the first wording search an additional criterion was applied: the category. The purpose of this filter was to select only the articles with the subject area related to our research. Since each database had the category options named differently, the filter selection varied on each database. In Web of Science database, the category filter applied was: ‘education scientific disciplines’, ‘engineering aerospace’ and ‘astronomy astrophysics’; while in Scopus database, the categories selected were: ‘physics and astronomy’ and ‘Earth and planetary sciences’. In the case of the database Teacher Reference Center, no category filter was needed since all the articles were related to education and the wording in the first search was enough to screen the topics we were looking for.

Analysis and screening

Once the search of articles was completed, the two authors analysed the records independently in the three phases described below, completing the process in February 2026. Inter-rater agreement at the screening stage was quantified on the 210 records assessed in duplicate by both authors: agreement was reached on 205 of them, corresponding to Cohen’s kappa = 0.89 (Cohen, 1960), a strong level of agreement (McHugh, 2012). The five discrepant records were resolved by discussion until consensus was reached. The three phases were the following:

Phase 1: title and abstract screening

In this first stage of the analysis of the articles, title and abstracts were reviewed to determine if they were related or interesting for the research. In this phase, duplicated articles were removed and unrelated studies excluded.

Phase 2: full-text reading and application of inclusion/exclusion criteria

Later, inclusion and exclusion criteria were applied to the related studies to select the articles for full-text reading.

The articles that met at least one of the following inclusion criteria were selected to full-text reading: Application of Virtual Reality in science and engineering education.

Application of Virtual Reality in science and engineering.

Studies on the emotional and cognitive impact of Virtual Reality in users.

On the other hand, the exclusion criteria were the following: Articles without clear empirical or methodological support.

Studies focused on specific topics whose results cannot be applied to the Space field.

Phase 3: coding and thematic analysis

After excluding the unrelated articles and full-text reading the remaining ones, the following thematic coding was carried out: Cognitive and emotional impact of Virtual Reality on students at all ages and, specifically, when learning about topics that can hardly be tested in a laboratory.

Education techniques used.

Limitations of Virtual Reality and negative effects of long exposure to it.

Improvements in work performance and efficiency when using Virtual Reality as a tool.

This last phase of the analysis allowed to extract patterns, trends and theoretical gaps for our research.

In parallel with the thematic coding, the following information was extracted from every included study into a structured sheet: authors and year, country of the participating institutions, disciplinary area, methodological design, sample and participants, VR hardware or software employed, variables measured and main findings. This information is summarized in Supplementary Table S1 and supports both the demographic analysis and the cross-study synthesis presented later.

Finally, the methodological flow of this systematic review is shown in Figure 1.

Quality appraisal of the included studies

The methodological quality of the included studies was appraised with the Mixed Methods Appraisal Tool (MMAT), version 2018 (Hong et al., 2018). This instrument was chosen because the corpus combines quantitative, qualitative and mixed methods designs, and the MMAT provides specific criteria for each of these categories within a single tool. Five of the twenty studies are technical or descriptive contributions that do not correspond to any MMAT category: the presentations of the E0102-VR (Baracaglia and Vogt, 2020) and iDaVIE (Jarrett et al., 2021) applications, the argumentative review of Soliman et al. (2021), the development of the VR-CFD training tool (Ren et al., 2024) and the digital twin validation for astronaut training (Piñal and Arguelles, 2024). Their eligibility is not affected by this circumstance: all five are peer-reviewed journal articles that meet every inclusion criterion, and they passed the full-text screening because they provide clear methodological support, in the form of documented system architectures, demonstrations on real observational datasets, validation procedures against reference data, expert assessment or literature-based argumentation. Their evidentiary status is nevertheless different from that of the fifteen empirical studies, and the synthesis treats the two groups differently. The five technical contributions provide capability evidence, that is, evidence of what VR systems can currently do, at what cost and under which hardware constraints; this evidence feeds the visualization, limitations and work performance themes of the evidence matrix (Supplementary Table S2) and the part of RQ1 concerning VR as a working instrument. They were not used to support effectiveness claims: every statement in the Results, Discussion and Conclusions about learning gains, engagement or performance differences rests exclusively on the appraised empirical studies, and where the five technical contributions appear in interpretative passages they are cited as technical demonstrations or as arguments, never as outcome evidence. In Supplementary Table S3 they are documented as not applicable. The remaining fifteen studies were rated by both authors, and disagreements were resolved by consensus. Each study was rated on the two screening questions and on the five criteria of its corresponding methodological category, and the result was recorded as the number of criteria met, from 0 to 5. Following the recommendation of the tool’s authors, no overall numerical score was used to exclude studies; instead, the appraisal informed the interpretation of the findings, giving greater interpretative weight in the Discussion to the studies meeting four or five criteria. The complete appraisal is provided in Supplementary Table S3. In summary, five studies met four or five criteria of their category, nine met three, and one met two or fewer; the most frequent shortcomings were the limited reporting of the integration between quantitative and qualitative strands in mixed methods studies and the limited control of confounding variables in the non-randomized designs.

Demographics of the publications

The demographic analysis of the publications considered for this systematic review discloses relevant information regarding its temporal distribution, study typology, geographical origin of the research and disciplinary areas.

Temporal distribution

All the studies selected (twenty studies) for this systematic review were published between 2020 and 2025, showing that the topic of this review is a live issue and new studies will be published in the next years (see Figure 2). Besides, it is worth mentioning that the year with the highest number of published studies is 2024, with 6 studies (30%). This peak coincides with a period of renewed commercial visibility of the technology, marked by the release of new consumer headsets, but the available data do not allow any causal relation to be established between both facts. It must also be noted that the literature search was closed in October 2025 and the screening and analysis were completed in February 2026, so the number of published studies in 2025 does not cover the full year and may additionally be affected by database indexing delays.

Typology of the studies

Concerning the methodological nature of the selected studies, they can be classified into the following groups (Figure 3): 25% of the studies (five) attended to a quantitative methodology based on statistics and close-ended surveys to measure variables like vividness or locomotive interactivity, among others.

20% of the studies (four) adopted a qualitative methodology focused, mostly, on users’ experience and thoughts related to the use of Virtual Reality as an innovative technology.

30% of the studies (six) followed a mixed methodology, combining statistics and open-ended surveys, with the aim to gather and validate quantitatively the users’ experience and thoughts.

The remaining 25% (five) are technical or descriptive contributions rather than empirical outcome studies: presentations of VR applications, a system validation and an argumentative review. Their evidentiary status and their treatment in the synthesis are described in the Quality Appraisal section.

Disciplinary areas

The selected studies can be classified into two main groups attending to their target field (Figure 4): Studies with an educational purpose. 90% of the selected studies are enclosed in this category, but inside of it we can distinguish between studies talking about education in science (30%), in engineering (25%), or miscellaneous (35%). This last subgroup encompasses studies talking about education in general, for industry workers, in aviation field, and astronauts training.

Studies with a scientific/engineering purpose. Only 10% of the selected studies (two studies) focused on using VR with a scientific/engineering objective. Both discuss the use of VR for the exploration and visualization of data in the disciplinary area of astronomy and, more specifically, astrophysics. This proportion is too small to support robust conclusions about VR as a working instrument for astronomical research, so the findings concerning this dimension must be read as preliminary indications rather than as consolidated evidence, a point taken up again in the Discussion and the Conclusions.

Geographical location

40% of the selected studies were carried out as a collaboration between universities from different countries and, in most cases, even from different continents.

According to the continents involved in the selected studies, the following statistics were obtained: Europe (36%), highlighting Netherlands, Spain and United Kingdom as the countries involved in more studies.

North America (12%), principally United States.

South America (6%), with Brazil and Chile.

Africa (3%), with South Africa.

Asia (33%), mainly China and South Korea.

Oceania (9%), predominantly Australia.

It is worth noting that Europe and Asia are the continents that took part in most of the selected studies. However, in the case of Europe, most of its studies’ contributions were as part of collaboration with other continents.

Figure 5 shows a more detailed geographical distribution of the selected articles. The colours’ distribution corresponds to the continents previously listed.

The results of this systematic review are organized in three main sections attending to their impact to different fields of the society: (1) the impact of the use of VR in astronomy research; (2) the impact of the use of VR in education; and (3) the impact of the use of VR in the users’ health.

Each of the mentioned sections gathers the most relevant theoretical and empirical contributions found in the selected studies. Supplementary Table S1 summarizes the characteristics of the twenty included studies, and Supplementary Table S2 presents the evidence matrix that maps each study onto the themes of the synthesis.

Impact of VR in astronomy research

Regarding the impact of VR in astronomy research, only two of the included studies address the use of VR for scientific analysis in astrophysics (Baracaglia and Vogt, 2020; Jarrett et al., 2021), so the findings of this subsection describe an emerging line of work rather than a consolidated body of evidence. Results showed the vast potential and largely still hidden capabilities of VR technology in the field, which allows to exploit human capacity to visually comprehend and perceive complex phenomena (Jarrett et al., 2021), as well as to characterize the size of structures by visualizing and measuring distances and angles in 3D (Baracaglia and Vogt, 2020).

Besides, studies on the field also pointed out the limitations and challenges of VR technology in general and as an astrophysics tool, highlighting the importance of a robust, polyvalent and generic visualization, exploration and measurement tool for 3D datasets (Baracaglia and Vogt, 2020). To this context, iDaVIE owners (Jarrett et al., 2021) are focused on developing the features of stream data and calculations from a cloud or server within the VR environment, and suggest further improvements in VR applications for the field, as it is the possibility of multi-user capability, a correct choice of colours for data transformation and visualization, and the selection of a correct rendering engine to avoid core memory (CPU) limitations.

Additionally, iDaVIE team also proposed an interesting usage of the VR in the planetarium environment, by projecting in real-time onto the dome surface when a presenter is seeing in a VR environment (Jarrett et al., 2021).

Impact of VR in education

Concerning the impact of VR in education, we organized the results of the related studies in three different sub-sections according to the educational area they talk about: education in general, in engineering and in science.

Impact of VR in education in general

In relation to the results of VR in education in general, it is worth noting that some of the studies pointed out the increase of effectiveness with the use of this technology (Kwon, 2020; de Moraes Rossetto et al., 2023; Predescu et al., 2023). Besides, de Moraes Rossetto et al. also highlighted the importance of teachers trained in VR when applying the technology in their lessons for better results, while Predescu et al. concluded the importance of combining VR with traditional educational methods to really improve the teaching and learning. In this regard, according to Ng and Chu results, educators should not blindly follow technology but link its content with pedagogy to develop better problem-solving skills, visualize scientific phenomena and consolidate science concepts (Ng and Chu, 2021).

Moreover, a research focused on the use of Metaverse in South Korea education (Yoo, 2024), stated that VR should not only focus on improving academic achievements, but also on developing creative competence and critical thinking.

On the other hand, the study from Ren et al. commented the enhanced training system that provides VR, which allows better decision making and, in the case of their VR-CFD-Based tool for dust exposure mitigation in underground tunnelling operations, the improvement of decision making resulted in better health outcome (Ren et al., 2024).

Finally, a single system-validation study with digital twins for astronauts training reported overlapping activation of the same brain areas whether the experience was real or simulated (Piñal and Arguelles, 2024). In relation to this result, Kwon already suggested in 2020 to measure brain wavelengths to probe in a more objective way the improvement of effectiveness when using VR since, in his experiment, the closer to reality the VR experience was, the better the learning effectiveness observed (Kwon, 2020).

Impact of VR in education in engineering

According to a study carried by Soliman et al., VR increases students’ performance and grades in engineering while, at the same time, it reduces costs by replacing labs with this technology (Soliman et al., 2021). In this context, Vahdatikhaki et al. add the shorter time spend in the VR labs in comparison to a traditional one, as well as an enhanced performance in the VR laboratory (Vahdatikhaki et al., 2024). In fact, although Motejlek and Alpay did not measure any difference in effectiveness of using VR compared to traditional methods, they also pointed out the advantages of VR in educational logistics, allowing a more flexible and self-directed learning, along with an increase in the engagement and improvement in the learning experience (Motejlek and Alpay, 2023). Besides, as stated by Xiang et al., VR can also help in solving the problem of inaccurate and non-objective evaluation in skill-based practical training (Xiang et al., 2022).

Finally, after a study carried by Vergara et al. to assess the use of VR in public and private universities, the results revealed differences of a sociological nature between the teaching staff of both types of universities, where male private university professors make better assessment of VR, while in public universities the women were the ones making a better assessment of VR. In addition, they concluded the need to introduce specific training plans in higher technical education, both in private and public universities (Vergara et al., 2022).

Impact of VR in education in science

In a research about innovative approaches in development of educational materials, teachers that took part in the study pointed out the high value of using VR for complex subjects, like Astronomy. In addition, they also suggested the need to increase the trainings given by experts and applied professional development activities (Kapucu et al., 2021). In this regard, another study considering pre-service teachers concluded that in order to enhance technological pedagogical content knowledge, pre-service teachers should be involved in the discussions about pedagogical alignment with technology and content knowledge from the early stages of lesson design (Kim and Yoon, 2025).

Besides, a quasi-experimental study with 105 high school students compared three learning materials in a physics laboratory course: paper textbooks, wearable AR material and wearable hybrid AR/VR material. The situational interest and the learning performance of the hybrid AR/VR group were significantly higher than those of the traditional learning group, and its engagement was significantly higher than that of the other two groups, which supports the use of wearable hybrid materials to enhance science laboratory learning environments (Sun et al., 2023).

On the other hand, Kersting et al. performed a study where education and public outreach professionals from Australia reflected on design principles to apply VR in astronomy education. In this research, the design principles focused on the immersion and sense of presence as a mean to promote inquiry learning; on the visualization as a unique opportunity for learners to explore abstract concepts; collaboration to build collaborative environments within and across virtual spaces; and facilitation as a core aspect of learning consisting in the expertise and guidance of astronomers and astronomy educators to maintain and enhance learners interests in the matter (Kersting et al., 2024).

Furthermore, a research on the joint attention dynamics in an immersive collaborative astronomy simulation revealed the potential benefits of initial attention divergence for knowledge convergence in later stages and emphasizes the need to identify how and when joint attention promotes collaboration (Kang et al., 2024).

Finally, a study carried by Montalbo et al. highlighted the importance of adopting innovative technologies in the educational landscape to foster a deeper understanding of complex subjects and inspire the next-generation of learners (Montalbo et al., 2025).

Limitations of VR

The results from studies related to VR both in education and astronomy research showed certain limitations of this technology.

In relation to human health, some users noted dizziness, headaches, motion sickness, nausea, eyestrain and tired eyes when using VR (Predescu et al., 2023; Motejlek and Alpay, 2023). Nevertheless, these symptoms affected a minority of participants, around 24% in the studies that reported them. In this regard, Predescu et al. commented on the need of high details in multimedia elements or materials to reduce the mentioned symptoms; while Soliman et al. concluded that a proper design of the VR application will solve the drawbacks of it (Soliman et al., 2021).

Other studies also underscored the software and hardware limitations of VR, such as having to use hand controllers for VR interaction, a slightly out of focus VR view due to not enough GPU power and bandwidth available, or the engine for developing VR only being able to render relatively small cubes one at a time (Jarrett et al., 2021; Piñal and Arguelles, 2024).

Additionally, professors participating in the study from Kapucu et al., expressed their concerns about the potential negative effects of VR in student’s attention and technology dependency (Kapucu et al., 2021).

Finally, Kwon reflected that if the learner’s sensation in a negative situation is excessive, the learning effect might be hampered. So, in this concern, ethical considerations are necessary when developing VR (Kwon, 2020).

Cross-study synthesis

Reading the twenty studies together, and taking into account the quality appraisal, three patterns cut across the individual results; in this reading, the fifteen appraised empirical studies carry the effectiveness evidence, while the five technical contributions inform the capability and limitation themes. First, the studies converge on the visualization capability of VR as its main contribution, both for learners exploring abstract astronomical concepts (Kersting et al., 2024; Montalbo et al., 2025) and for researchers interrogating three-dimensional datasets (Baracaglia and Vogt, 2020; Jarrett et al., 2021). Second, the evidence on learning effectiveness is positive but not uniform. Several of the appraised studies reported measurable gains in performance, grades or efficiency (Kwon, 2020; Predescu et al., 2023; Sun et al., 2023; Vahdatikhaki et al., 2024), including a quasi-experimental comparison in which the group using wearable hybrid AR/VR material significantly outperformed the traditional learning group (Sun et al., 2023) and a controlled comparison in which VR users obtained significantly higher quiz scores (Predescu et al., 2023), a pattern that the argumentative review of Soliman et al. (2021) had anticipated from the previous literature. In contrast, the randomised non-inferiority trial of Motejlek and Alpay (2023) found retention comparable, though not superior, to a multimedia video, while still observing logistic and engagement advantages. This range of outcomes suggests that the effect of VR depends less on the technology itself than on the instructional design that accompanies it, an interpretation consistent with the insistence of Ng and Chu (2021) and Predescu et al. (2023) on linking VR to pedagogy. Third, the studies that involve teachers (Kapucu et al., 2021; Vergara et al., 2022; Kim and Yoon, 2025) coincide in identifying teacher training as the main condition for successful adoption, above the technical characteristics of the systems. The distribution of these themes across the corpus is detailed in Supplementary Table S2.

The findings obtained from the systematic review point to VR as a technology with substantial potential for Space-related education and for astronomy research, although the strength of the evidence differs markedly between the two dimensions: the educational findings rest on fifteen empirical studies with heterogeneous designs, interventions and outcome measures, synthesized narratively and without meta-analysis, while the research dimension rests on two technical contributions. Across both dimensions, the recurring driver is the visualization capability that VR provides to users, enabling them to understand and perceive complex phenomena. Besides, the analysis brings to light that VR can improve students’ visualization and spatial thinking, and several of the appraised studies associate it with higher engagement, collaboration and learning effectiveness; de Moraes Rossetto et al. (2023) also reported that it can help to prevent abandonment and the lack of motivation to go to school.

The mechanisms behind these benefits can be explained through current theories of immersive learning. The Cognitive Affective Model of Immersive Learning (CAMIL) proposes that presence and agency, the two affordances that distinguish immersive VR from other media, act on learning through affective and cognitive factors such as situational interest, motivation, embodiment and cognitive load (Makransky and Petersen, 2021). This model accounts for the pattern observed in the corpus. When the immersive experience is aligned with the learning goals, presence increases situational interest and engagement, which is precisely what Sun et al. (2023) measured in their three-group comparison and what de Moraes Rossetto et al. (2023) also observed; when it is not, the added sensory richness can raise extraneous cognitive load without improving retention, which offers a plausible explanation for the outcome of Motejlek and Alpay (2023). The overlapping brain activation between simulated and real experiences reported in one study (Piñal and Arguelles, 2024) is also coherent with this framework, although a single observation cannot ground the sense of presence on an established neurophysiological basis.

In this regard, the review underlines how important it is to link VR to pedagogy, and three pedagogical frameworks emerged from the corpus, each one anchored in specific results. Csikszentmihalyi’s flow theory underpins the work of Kwon (2020), whose experiment indicated that reproducing the negative situation with higher fidelity helped the learner’s anxiety translate into flow rather than hinder learning, suggesting that, in that setting, the learning effect depended on the fidelity of the experience and on keeping the learner’s sensations within the band that sustains flow. The Constructivist Learning Theory, combined with the Variation Learning Approach, frames the results of Ng and Chu (2021) and the argument of Soliman et al. (2021), where students consolidate science concepts and develop problem-solving skills by actively contrasting varied representations of the same phenomenon within the simulation. The Design-Based Research paradigm structures the iterative dialogue between researchers and practitioners reported by Kersting et al. (2024) and the co-development of materials with pre-service teachers described by Kim and Yoon (2025). The three frameworks converge on the same practical implication: the educational value of VR is unlocked by the design of the learning experience, not by the technology alone.

Not all the findings point in the same direction, and the divergences are informative. The retention results of Motejlek and Alpay (2023), comparable but not superior to a multimedia video, contrast with the performance gains measured by Soliman et al. (2021), Sun et al. (2023) and Vahdatikhaki et al. (2024); it should be noted, however, that a non-inferiority design tests precisely whether VR matches a traditional medium, so its result supports the logistic case for VR rather than contradicting the corpus. The remaining discrepancies are plausibly explained by the different outcome measures employed (retention of information versus grades, task performance or situational interest), the different degree of alignment between the virtual activity and the assessed content, and the novelty effect, which tends to inflate short-term engagement measures. Something similar occurs with the teachers’ perspective: while the teachers in Kapucu et al. (2021) feared distraction and technological dependency, the empirical results of the corpus associate VR with higher engagement, a tension that the sociological differences found by Vergara et al. (2022) between public and private universities help to nuance. These contrasts reinforce the need, identified below, for standardized assessment instruments and longitudinal designs.

Another suggestive observation, reported in a single study of the corpus, is the overlapping activation of the same brain areas whether the experience is real or simulated (Piñal and Arguelles, 2024). Coming from one system-validation context, this observation cannot by itself establish a neurophysiological mechanism linking realism to learning effectiveness; what it does support is the value of incorporating physiological indicators into future evaluations. Nevertheless, it is telling that there is a lack of studies regarding the brainwaves when using VR.

Additionally, it is worth noting that VR not only impact students, but teachers and professors too. The preconceived idea of the teachers related to the use VR in the classroom was that student will get distracted (Kapucu et al., 2021), but nothing further from reality, because it was shown that VR enhanced student’s engagement and effectiveness. For so, it is necessary to introduce specific training plans on the didactic use of VR for teachers and professors, as well as to involve them from the early stages of lesson design in VR to enhance their technological pedagogical content knowledge (Kim and Yoon, 2025).

Finally, as expected, VR has also certain drawbacks related to software and hardware, as well as to users’ health, like cybersickness. In this context, Soliman et al. proposed to have a proper design in mind before developing the VR application, like simplifying the virtual environment (Soliman et al., 2021). This solution could solve both the software/hardware issue and the health matter. However, in this systematic review there is no assessment of solutions for such problems. In this regard, a new systematic review focused on cybersickness, for example, should be carried out for further awareness.

Theoretical and methodological gaps

Although there is a vast body of literature related to VR applications, several gaps were identified regarding the theoretical and methodological approaches: Despite the numerous studies about VR in education, most of them are regarding the health sector rather than the Space related field, even though VR is a powerful tool for visualization that can help understanding complex astronomical phenomena.

Quantitative studies related to measuring brain behaviour (like brainwaves) when using VR in education or Astronomy research are lacking.

VR allows users to easily interact in the virtual environment and promotes collaborative activities. However, there are limited studies on the benefits of VR for joint attention and students’ interactions when learning science.

There is a lack of studies related to the use of VR in working environments for Space related field (e.g. Astronomy research).

Regarding the negative effects of VR, there are insufficient studies evaluating the student’s dependency on technology after using VR in class.

Research on the uses of VR for developing creative competence and critical thinking is also missing, as pointed out by Yoo (2024).

There is a shortage of longitudinal studies following the same participants over time, so the durability of the learning gains and of the engagement associated with VR remains unknown.

The outcome measures employed across the corpus are heterogeneous, and no standardized assessment instruments exist to compare the effectiveness of VR interventions between studies.

Cross-cultural research is scarce; although 40% of the studies were international collaborations, none of them analysed whether the effects of VR vary between cultural or educational contexts.

The integration of Artificial Intelligence with VR, for example through adaptive environments or intelligent tutoring inside the simulation, is barely explored in the Space related field, a gap that Yoo (2024) also identified in the metaverse literature.

There are no cost-effectiveness analyses comparing VR with traditional laboratories or teaching methods, despite cost reduction being one of the claimed advantages of the technology (Soliman et al., 2021).

Practical and academic relevance

From a practical perspective, the findings of this systematic review emphasize the benefits of VR for visualization, building collaborative environments and its effectiveness to learn. So, with these aspects in mind, the development of VR applications for astronomical research can ease the work of scientists and engineers, while at the same time VR can be the perfect tool for education and for disseminating Space related science to people of all ages.

On the other hand, from an academic perspective, this systematic review offers an integrative theoretical framework that can guide empirical research on VR advantages and disadvantages for Space exploration and education, incorporating brain data and behavioural responses.

Limitations of the review

The conclusions of this review must be read against its limitations. The search was restricted to three databases and two broad Boolean strings. As argued in the Methodology, the document types that most distinguish IEEE Xplore, the ACM Digital Library and ERIC fall outside the eligibility criteria of this review, and the journal literature they index is very largely covered by Scopus and WoS (Gusenbauer and Haddaway, 2020; Singh et al., 2021); a residual risk of selection bias nevertheless remains for journal articles published in venues indexed only by those sources. If such articles exist, they most plausibly belong to the engineering and human-computer interaction side of the field, so their absence would tend to underrepresent technical implementations rather than to overturn the educational findings, which rest on the convergence of studies published across a wide range of venues and disciplines. The dimension where a single missed study could alter the picture most is VR as a working instrument in astronomy, which rests on two studies only, and the conclusions of that dimension are qualified as preliminary for this very reason. In particular, the finding that only two studies examine VR as a working research instrument must be read as referring to astronomy and astrophysics, the domain the second search string covered, and not as evidence about the prevalence of VR in the wider space sector, since activities such as aerospace engineering, astronaut training, planetary science or mission operations were not searched for specifically. PubMed/MEDLINE was deliberately excluded because the health field falls outside the scope. Only studies in English or Spanish were considered, so relevant work in other languages may have been missed. The protocol was not registered in PROSPERO, since that registry is limited to reviews with health-related outcomes, although the protocol, the extraction sheet and the appraisal records are available from the corresponding author upon reasonable request. The corpus itself imposes limits: only two studies address VR as a working instrument in astronomy, so the answer to RQ1 in that dimension is preliminary; five studies are technical or descriptive contributions that could not be appraised with the MMAT and that were consequently not used to support effectiveness claims; and the heterogeneity of designs and outcome measures prevented any quantitative meta-analysis, which is why the synthesis is narrative and thematic. Finally, although the MMAT appraisal informed the interpretation of the findings, no study was excluded for quality reasons, and the appraisal of qualitative studies retains an unavoidable degree of judgement.

In answer to RQ1, the reviewed evidence indicates that VR can influence knowledge acquisition mainly through its visualization capability, with several of the appraised studies reporting higher engagement, performance and understanding of complex phenomena, especially those that can hardly be reproduced in a laboratory. Fully immersive virtual laboratories can reduce cost and infrastructure while improving efficiency and the effectiveness of the learning, provided that the technology is linked to pedagogy; under those conditions VR can help to consolidate science concepts and to develop problem-solving skills, with controlled comparisons reporting higher quiz scores, higher situational interest and less time needed in the physical laboratory. Regarding its effectiveness as a working instrument in astronomy research, the two available studies indicate a clear potential for the exploration and measurement of three-dimensional astronomical datasets, but the evidence base is too small for this part of the answer to be considered consolidated.

In answer to RQ2, the corpus divides into fifteen empirical studies and five technical or descriptive contributions; among the empirical studies, mixed approaches are the most frequent (six studies), followed by quantitative (five) and qualitative designs (four), and the influence and effectiveness of VR have been evaluated mostly through surveys and users’ experience, with very few objective measures. The methodological approaches that allow a comprehensive and empirically validated evaluation are those that combine behavioural performance measures with physiological indicators, as suggested by the studies measuring brain activity, yet such combinations are still exceptional in the field.

Beyond these evidence-based conclusions, two recommendations follow. VR applications for Space education and research should be designed taking into account the negative effects reported for a minority of users, as well as the hardware constraints of the current systems. And future evaluations should adopt standardized instruments and longitudinal designs, so that the effectiveness of VR can be compared across studies and its persistence verified over time.