Simulation-based learning (SBL) is defined by Koppenberg et al. (2014) as an experiential form of knowledge gain through which concrete tasks can be presented in different realities and tested by learners. Pan et al. (2006) emphasized that virtual environments can lead to better understanding and motivate the learner. The authors describe VR as a combination of computer graphics systems with various display devices that enable immersion in an interactive 3D environment. Two main characteristics shape and influence learning with VR—immersion and presence (Lei et al., 2022). A sense of immersion is conveyed in VR systems by engaging visual, auditory, and kinesthetic senses, thus reducing awareness of information in the external world (Huang et al., 2023). Presence, on the other hand, can be characterized as the response to a certain degree of immersion (Slater, 2003). Various behaviors in the VR environment are signs of presence, which arises from a combination of human perception and immersion (Slater, 2003).

A distinction can be made between non-immersive, semi-immersive, or fully immersive VR systems. While non-immersive VR systems use ordinary devices, such as a desktop monitor, on which the user experiences the virtual environment and is not shielded from the external environment, the user of a fully immersive environment is surrounded by it (Huang et al., 2023; Jensen and Konradsen, 2018). This visual blanking of the external world is achieved through a head-mounted display (HMD) (Jensen and Konradsen, 2018; Rauschnabel et al., 2022). To create a higher level of immersion, wall-mounted screens and motion tracking devices are required for semi-immersive VR systems (Huang et al., 2023). For student teachers’ video-based learning, Dehne and Gröschner (2023) characterized immersion as a key affective-motivational characteristic but also emphasized its limitations. For example, reflecting on own—potentially unpleasant —classroom sequences might evoke feelings of shame (Chang et al., 2018).

Research on the motivation for using VR in education highlights the potential of VR in making various situations experiential and tangible which would, for example, not be feasible in ordinary classroom instruction (Freina and Ott, 2015). In terms of learning success, studies from the medical research field have shown that the learning gain for medical students is higher in fully immersive VR environments than in semi-immersive VR systems (Gutierrez et al., 2007) or in video and text-based learning formats (Sattar et al., 2019). In a study on self-regulated learning (SRL) behavior, various behaviors of learners could also be observed when passing through a VR environment, which contributed to the completion of the learning task (Ader et al., 2025). In the context of the current study, a systematic review by Cevikbas et al. (2023) reveals positive effects of implementing VR in mathematics teaching on cognitive and metacognitive aspects of learning. As the review found that geometry is the most studied area in connection with VR (Cevikbas et al., 2023), the present study deals with exponential growth and thus contributes to the research field with another topic. So far, insights into learning gains provide evidence that VR is relevant to the school setting and enriches the classroom. Yet, the question is how (student) teachers assess the benefits of VR and to what extent do they feel prepared to use these new technologies in their own teaching? This study contributes to this lack of knowledge.

Research on VR in teacher education shows that VR-based training programs start with student teachers and recreate various typical teaching scenarios in which the teacher is supposed to carry out instructions (Huang et al., 2023). The interactivity of VR environments is often “still a weak point,” as virtual students are usually controlled by pre-programmed scripts and thus can hardly respond to teacher behavior (Huang et al., 2023).

In addition, most applications are one-time experiences with program durations ranging from 10 min to 2 h. Furthermore, Huang et al. (2023) found that VR is mostly used to train procedural knowledge. The studies considered largely reported positive outcomes captured by teachers’ self-reports (Huang et al., 2023). For instance, Huang et al. (2022) showed higher increases in interest and self-efficacy related to classroom management using a VR simulation compared to an instructional video. An advantage of 360-degree videos viewed through a VR headset over traditional videos was also found around noticing (Kosko et al., 2021). Here, student teachers paid attention to more student actions than student teachers in the control group. In addition, an investigation using a virtual classroom found an improvement in the visual attention of student teachers in connection with adaptive feedback (Huang et al., 2025). The state-of-the-art in research shows that using VR in teacher education mainly aims to promote teacher knowledge and behavior (Stavroulia and Lanitis, 2017). So far, the use of VR for future practice in the profession has hardly been addressed in empirical research.

According to the Theory of Reasoned Action (Ajzen and Fishbein, 1980), a person’s beliefs and attitudes influence behavioral intention. The Technology Acceptance Model (Davis et al., 1989, TAM) is based on this assumption and generally addresses the factors influencing the acceptance of certain technologies (Figure 1). More specifically, perceived usefulness and perceived ease of use influence attitudes toward the use of technology, which in turn affect behavioral intentions to use the technology in question (Davis et al., 1989; Teo, 2009). If (student) teachers perceive a novel technology, such as VR, as useful and easy to use, they will develop a positive attitude toward its use (Davis, 1989; Teo et al., 2007). Thus, in terms of perceived usefulness, it is assumed that teachers need first to be convinced that the use of VR will improve their own teaching. Ferdinand et al. (2024) showed that even brief priming about the usefulness of VR can trigger cognitive processing and have a positive effect on learning achievement.

In turn, perceived usefulness can be influenced by factors such as interactive learning environments, self-efficacy, and perceived satisfaction (Liaw and Huang, 2013). In addition to variables such as perceived usefulness, attitudes toward use, and perceived ease of use, previous research has considered further influential factors that impact teachers’ technology acceptance. Teo (2009) reported a direct effect of computer self-efficacy on technology acceptance, as well as indirect effects of technological complexity and enabling conditions. Enabling conditions, such as good technical or personal support for better use, have a direct effect on perceived ease of use (Teo, 2009). Technological complexity circumscribes the degree of difficulty in understanding a particular application (Teo, 2015; Teo et al., 2016; Thompson et al., 1991).

According to the TAM model, behavioral intention directly determines actual usage behavior. As previous studies with student teachers showed, behavioral intention is directly influenced by the perceived usefulness of computer technology (Ma et al., 2005). Moreover, attitude exerts a direct effect on intention (Davis et al., 1989). With regard to attitude, Liaw and Huang (2003) demonstrated a relationship between the perception of a technology and the attitude toward its usefulness. A study conducted with teachers in the school context also confirmed these relationships when examining attitudes toward the use of new technologies as tools (Nair and Das, 2012). The results indicated an influence of perceived ease of use on perceived usefulness and on the attitude toward usefulness. More specifically, the better teachers were trained in the professional use of new tools, the more useful they perceived these tools to be, and the more positive their attitude became toward integrating them into classroom instruction (Nair and Das, 2012).

This study contributes to the field of research on teacher professionalization by investigating the extent to which student teachers perceive VR as useful for teaching. In reference to the TAM (Davis et al., 1989), the aim is to examine the extent to which the development of perceived usefulness through their own experience of VR changes their willingness to use this technology in future professional practice. Studies such as those by Yi and Hwang (2003) have shown that the intention to use has a direct influence on the actual usage of a web-based environment. Since the TAM model has both a direct and indirect influence on perceived usefulness and intention to use, the current study decided to focus on these two variables in order to investigate the direct influence on willingness to use in one’s own professional practice. Interventions in the field of immersive media have demonstrated that utility values are constructs sensitive to change (Dehne and Gröschner, 2023; Nickl et al., 2023).

As part of an international collaboration in the Learning To Teach Lab: Science (LTL:S) (Gröschner et al., 2024), a total of N = 57 student teachers from four countries (Finland, Germany, Israel, and Turkey) participated in the VR experience during the summer 2021 (age: M = 26.18 years, Mdn = 24.50 years, SD = 4.65 years; gender: 36.8% female, 57.9% male, 5.3% not specified). About 90% of students indicated that English was not their first language. Nevertheless, 61.4% rated their level of English as good and another 22.8% as excellent. English language proficiency was relevant because the language of the VR environment was English. The student teachers who agreed to participate in the experiment were studying mathematics (29.8%), English (15.8%), biology (10.5%), other languages (7.1%), or other subjects (36.8%) as their primary subject. As teacher education programs at different universities are structured differently, participants of our study pursue different university degrees: Bachelor (28.1%), Master (17.5%), state examination (26.3%) or other degrees (28.1%). More importantly, 38.6% had never experienced a VR environment before.

Student teachers were asked to complete an online questionnaire before and after completing the VR environment individually to assess the perceived usefulness and the intention to use. The processing of the individual questionnaires took approximately 10 min on average.

Student teachers generally took about 60–90 min to complete the VR environment. The VR environment could be navigated using the Meta Oculus Quest 2 (Oculus VR). Interaction with the VR was possible through movement in a predefined space and using two controllers held in the hands. To accompany the student teachers during their navigation and provide support if necessary, the VR was transferred to a connected laptop and streamed (Figure 2).

The VR environment generally deals with the spread of a virus (Figure 2). Using this frame story, the participant’s role in the VR is to conduct investigations to prevent further spread. Within these investigations, mathematical problems related to exponential growth are to be solved. The VR is divided into two sections. The first section serves as an introduction to the topic. In the second section, the participant is in a laboratory. There, the participant is confronted with tasks such as, for example, determining the growth factor or the new cases of disease per week by calculation and entering them in a table. In addition, a function equation is to be created with the aid of a coordinate system and individual multiple-choice questions are to be answered in the meantime to check understanding. Before the individual subtasks, individual explanatory videos appear in each case, which are to prepare the solution of the task accordingly. Also, the model representation of the spread per week right at the beginning of the section is meant to contribute to the understanding about exponential growth via a practical reference.

In particular, the model-based representation of the spread of the virus provides a much more tangible visualization than conventional methods of teaching exponential growth. Interacting with the model and transferring it into formulas and coordinate systems makes it easier for learners to relate theory and practice. The playful component, which comes from embedding it in a storyline and giving learners an overarching task to save humanity from the virus, can also have a positive effect on their motivation to solve the task (Ader et al., 2025).

After going through the VR and completing the post-survey, the student teachers participated in a guided and structured interview. Each interview took about 10–15 min on average. The interview data of N = 41 student teachers from Germany, Israel and Turkey were used for the qualitative analysis in this reported study. The focus of the study with Finnish students teachers were rather on think-aloud data (Sobocinski et al., 2024), so the interviews did not cover aspects in relation to the present study. In addition to questions about general perceptions and how they handled the tasks in the VR environment, the focus was on what contribution VR could make to teaching. In addition, the student teachers were asked to express their thoughts on what hurdles and limitations they see in the use of VR in schools and classrooms.

The VR tool was in English for all participants. In addition, the Israeli research team conducted the questionnaires and interviews in English. The other teams switched to their native languages for this purpose. This subsequent quantitative-qualitative mixed-methods approach enables the collected questionnaire data to be linked and enriched with arguments from the interviews to obtain explanatory approaches for the quantitative results.

To assess Perceived Usefulness the scale from Teo (2009) was adapted (Table 1). The three items, which had to be rated on a 5-point Likert scale (ranging from 1 = “Strongly disagree” to 5 = “Strongly agree”), reflect the prospective teachers’ perceptions of the extent to which VR can contribute to improving their own work (α = 0.80). This scale was also used in the posttest to detect change (α = 0.82). In addition, the Intention to Use scale by Teo (2009) was added to the posttest (α = 0.76). Intention to Use consists of two items that had to be rated on a 5-point Likert scale (1 = “Strongly disagree”; 5 = “Strongly agree”).

For the interviews, a structured guideline was developed with two main sections. The first section included questions on the perception of the VR environment and the handling of the task. The second section dealt with more general questions about the use of VR in schools and lessons. This part included specific questions on the current relevance for teaching, the contribution that VR can make to pupils’ learning and specific possible ways of using the technology in lessons. In addition, the student teachers were asked about the potential hurdles and limitations of VR for use in the classroom. In more details, the relevant questions for the present study were:

In your opinion, how important is the use of innovative learning technologies in the classroom?

In order to examine the change in students’ usefulness beliefs from the pretest to the posttest, a Wilcoxon signed-rank test was calculated (Siegel, 1988). This test can be used to investigate whether the median difference between the pretest and posttest is significantly different from zero. The requirements for this non-parametric procedure are that the measurements are dependent, the independent variable has two expressions, and the dependent variable is at least ordinally scaled (McCrum-Gardner, 2008). In addition to these basic requirements, the distribution of the differences should be approximately symmetrical. For the Wilcoxon signed-rank test, the correlation coefficient r is calculated as an effect size measure. According to Cohen (1988), a weak correlation exists for an r = 0.1, a moderate correlation exists for an r = 0.3, and a strong correlation exists for an r = 0.5.

To investigate the second research question, a multiple regression model was calculated in which Intention to Use was the criteria and the change in Perceived Usefulness was the predictor. In addition, age and gender were included as covariates. Multiple determination coefficient R² is given to determine model fit. With an R² = 0.02, weak variance clarification is present, with R² = 0.13, moderate variance clarification is present, and with R² = 0.26, strong variance clarification is present (Cohen, 1988). As prerequisites for the procedure, it is first necessary to check that there is a linear relationship between the variables, no outliers, and no multicollinearity. Second, it is necessary that the residuals are independent and normally distributed and that they exhibit homoscedasticity (equality of variance). The calculations were performed using the statistical software SPSS 27 (IBM SPSS Statistics, 2020).

The interview data were transcribed and translated into English. For the development of a common coding scheme, first an open coding of three interviews from different countries was conducted, based on which suggestions for codes were generated. This preliminary coding scheme was developed in accordance to previous research by Gaspard et al. (2015). The definitions of the codes for the main categories and subcategories were agreed on in the international research team and steadily adjusted in three coding rounds. With the final scheme, one example interview was coded by three coders and an acceptable interrater reliability was reached between all pairs of coders (0.68 < Cohen’s κ < 0.77). For the analysis of the qualitative data, summed values were calculated from the codes of the individual participants for the respective main and subcategories. Regarding the second research question, the categories Value (N = 284 codings) and Costs and Limitations (N = 248 codings) (see Supplementary material) are examined in more detail in the results section in which example statements of the students can be found. The category Value encompassed participants’ ideas regarding the benefits, importance, and integration of VR into education and students’ lives. Value as category is intended to emphasize the aspects of usefulness within the current evaluation. The category Costs and Limitations, on the other hand, captures codes with perceived possible negative effects or negative associations with VR (Gaspard et al., 2015).

For answering research question 1, the requirements of a Wilcoxon signed-rank test showed an acceptable fit. Based on the visual examination of the histogram of the differences, a symmetrical distribution of the values could be demonstrated. There was a statistically significant change in perception of Usefulness (Mdn = 0.17) from the pretest (Mdn = 3.33) to the posttest (Mdn = 3.67), z = 2.14, p = 0.032, r = 0.29.

The second research question addressed the influence of the change in perceived usefulness on the intention or willingness to use VR in future professional practice. Requirements for the multiple regressions models were first controlled for in the quantitative data. A linear relationship between the variables was given and based on the studentized excluded residuals, there are no outliers in the data set. The Durbin-Watson statistic had a value of 1.90, according to which there was no autocorrelation in the residuals. No multicollinearity between the predictors was found. After visual inspection, the conditions of homoscedasticity and normal distribution of the residuals were also controlled. The model shows a (moderate to) high goodness of fit with an R² = 0.26 (corrected R² = 0.21) according to Cohen (1988). The predictors Change in Perceived Usefulness, age, and gender predict the criteria Intention to Use statistically significantly [F(3, 48) = 5.59, p = 0.002]. Interpretation of the regression coefficients shows that the predictor Change in Perceived Usefulness is significant, B = 0.60, p < 0.001 (Table 2). Thus, a linear relationship is assumed. The coefficients age and gender are not significant. Consequently, these variables have little associations on Intention to Use.

In the interviews from the three countries, the majority of the statements for Value (Table 3) was assigned to the subcategory Utility for education and teaching profession, which addresses the teaching-related aspects of usefulness (n = 210 coding). Herein, for example, interview participants stated that the use of VR in the classroom can make certain content tangible, create a change from conventional teaching formats, and ensure easy access to the teaching topic. In addition, the new technology is perceived to offer great potential to transcend certain boundaries of everyday teaching. One participant commented: “Especially with the topics or things that you cannot necessarily do with the students in the real laboratory, because they might be too dangerous, or because it would simply be too expensive or would not work out in terms of time if you were to do it with everyone there.” (EFG03S07, 90).

The subcategory Utility in general comprises the aspects of utility that are not related to learning and teaching (n = 41 coding). When considering the general usefulness of VR, an advantage may be the simulation and hands-on practice of theoretical content realizable by using technology. The statement “It’s just that some things, like abstract things, − and they are increasing, so in our society we have to think more and more abstractly [and] the students also have to be able to remember that, because the VR glasses are almost indispensable” (EFG03S03, 105) underlines the perceived growing importance today.

A positive “motivation for learning,” “fun in testing” and a certain “surprise effect” as affective components are aspects that were mentioned by the participants and have individual impact on the intrinsic value of VR (n = 25 codings). Related to this subcategory, one participant stated: “I do think that it triggers motivation, simply because it’s a new technology, gadget, that everyone thinks is totally great. I was also excited and thought it was exciting to take a look at it.” (EFG03S12, 106). A few statements were assigned to the subcategories Importance for achievement (n = 6 codings) and Personal importance (n = 2 codings), where the latter subcategory views mastering the VR task as part of one’s own identity. Importance for achievement refers to improving one’s own performance, as one student described: “And you [can] play through that over and over again and improve through that.” (EFG03S01, 108).

Within the Costs and Limitations category, the subcategory Limitations appeared most frequently with n = 123 codings in the interviews. Limitations in applicability and usefulness, which also have an influence on the intention to use it in later professional practice, are considered primarily in the different effect on the adolescents, the lack of realism and the currently still limited possibilities. “I would not give it to young students because I do not think the abstraction is there yet.” (EFG03S03, 142). This statement indicates that an age limit for use in the classroom is perceived as relevant, because the different effects of the technology on children should be considered when use in school is planned or discussed.

In terms of Costs, a distinction was made between personal costs (n = 68 codings) and opportunity costs (n = 57 codings). The subcategory of personal costs implies the associated energy investment and negatively associated emotions. Statements that can be assigned to the other subcategory often included financial and time aspects, such as: “So, there is also a bit of a lack of time for the teachers to simply try things out.” (EFG03S11, 229). In addition, the lack of adaptability of a VR environment to teaching is criticized. A student teacher said: “The potential is very high, but it has to fit in with what I would like to achieve as a teacher, and also with how my students are. If I now have students who are – I do not know – mathematically super talented, then I certainly need a different environment that also runs faster.” (EFG03S11, 132). This statement also shows a challenge that the contents or tasks of the VR environment are not individually adapted to the needs and the performance level of the individual students. It requires a great deal of content embedding and support from the teacher. “It could be that sometimes you get lost too much in the technical gadgetry, um, if it’s not very strongly guided” (EFG03S15, 61).