1 Introduction

The perceived self-efficacy of secondary school teachers is often low. Around 50% of teachers across OECD member countries reported that their training did not include this competence (Organisation for Economic Co-operation and Development [OECD], 2020). Furthermore, recent studies underscore the dissatisfaction among pre-service teachers, highlighting the insufficiency of purely theoretical training and the deficiency in practical components (AUTHORS 2023; Sarceda-Gorgoso et al. 2020; Shank et al. 2022).

In response to these challenges, Immersive Virtual Reality systems (IVR) have emerged as transformative tools in teacher training (Garland and Garland 2020; Mouw et al. 2020). A systematic review by Billingsley et al. (2019) confirms the efficacy of IVR systems in acquiring and refining the skills of proficient teachers. Despite this promising trajectory, there remains a conspicuous research gap —the limited exploration into the acceptance and adoption of VR systems specifically tailored for training pre-service teachers.

Most of the studies consulted primarily focus on assessing teachers' willingness to adopt such technologies in classroom settings, with positive success rates reported (Abd and Mohd 2019; Barrett 2023). However, the crucial facet of adequately training pre-service teachers through immersive virtual experiences has not been extensively explored, highlighting the need for a more comprehensive and in-depth investigation.

This study addresses this gap by delving into the usability and acceptance of an IVR system—Didascalia Virtual-ClassRoom (Didascalia VC hereafter). Positioned as an innovative platform, Didascalia VC is designed to allow pre-service teachers the opportunity to not only experience but also record and reflect upon affective and attitudinal dimensions essential for effective conflict management in secondary school classrooms. The scenarios simulated by Didascalia VC accurately replicate the most prevalent conflicts reported by secondary teachers in real classroom settings.

Our primary objective was to comprehensively address usability issues by identifying them through user feedback, proposing potential solutions, and examining teachers' behavioural intentions related to usage. This multifaceted approach aimed not only to enhance the tool's functionality but also to facilitate its seamless integration into the broader framework of teacher training.

1.1 Theoretical background

1.1.1 Classroom management from an ecological perspective

Based on the systematic review conducted by Wang et al. (2020), in this paper we understand classroom climate as a dynamic and multidimensional phenomenon that refers to the interactions between teachers and students at three levels: academic-instructional (characteristics of teaching–learning methodologies), socio-emotional (emotional exchanges between teachers and students) and organisational (practices for managing disruptive behaviour).

From an ecological perspective, in keeping with the above definition, effective classroom climate management requires foresight, forward planning of the strategic actions that will support students' social, emotional and cognitive learning (Doyle 2006; Evertson and Poole 2012). This approach therefore suggests that teacher performance—going beyond the transmission of knowledge – will depend on the teacher's ability to perceive and interpret classroom events that are relevant to students' learning and to decide how to adjust his/her performance to manage them strategically. Proactive strategies are essential for maintaining a positive and conducive learning atmosphere, and they align with the ecological perspective by addressing the root causes of behaviour within the classroom ecosystem (Alasmari and Althaqafi 2021).

Therefore, our conceptual framework for teaching classroom management competence (CM hereafter) encompasses cognitive as well as emotional and attitudinal variables. We consider the latter to be crucial for effective classroom management (Keller and Becker 2020; Sutton et al. 2009) and, since they are difficult to model, they require a practical-experiential learning scenario such as the one provided by IVR.

1.1.2 Virtual reality learning environment for teacher training

Traditionally, videos, role-plays, case examples, or even images are used in pre-service teacher training to simulate or discuss classroom situations. However, their effectiveness in terms of level of realism and complexity can be limited. A multitude of different technologies have been tested in this area. For example, the application of TLE TeachLivE (Dawson and Lignugaris/Kraft 2017) in American universities for teacher training or similar activities (Ade-Ojo et al. 2022). However, these activities using screens offer limited immersive capabilities. The interactive and immersive nature of IVR technology (Burdea and Coiffet 2003) overcomes the limitations of role-playing and video viewing, providing pre-service teachers with the opportunity to practice and observe themselves teaching in a safe and similar environment to that of a typical classroom (Richter et al. 2022; Shernoff et al. 2020).

A significant number of research has delved into the utilization of various VR simulations for pre-service teacher training, as evidenced by studies such as those conducted by Lugrin et al. (2016), Mouw et al. (2020), and Ferguson and Sutphin (2022). Their findings underscore the imperative need to enhance realism and presence in VR simulations. Specifically, the shortcomings identified include the lack of interaction between agents and participants, the need for more intricate interaction patterns, and the call for simulations to reflect real-life situations. Notably, participants in these studies expressed concerns about the time-consuming nature of learning and deploying VR technology.

On the topic of scenario-based learning (SBL) research, Kugurakova et al. (2023) explored the incorporation of both verbal signals and non-verbal reactions, highlighting their potential to create a more emotionally resonant background within simulated situations.

Although the use of technology to support teachers' professional development holds great promise for bridging the gap between theory and practice, few training models that use IVR to teach CM have been rigorously evaluated, and more studies are needed to determine which educational approach is most appropriate (McGarr 2021; Won et al. 2023).

Along these lines, in order to foster pre-service teachers’ classroom management competence from an ecological approach, the Didascalia VC platform —described in detail in the following section— has been conceived as a situated experiential education environment (SEEE) focused on the performance of a learner that will be exposed to various experiences (Schott and Marshall 2018). This virtual reality model focuses on four features: (i) sense of immersion, (ii) authenticity and complexity of the learning task, (iii) interaction and (iv) timely and effective feedback.

Firstly, our aim is to provide an environment that allows pre-service teachers to perform their teaching role, feeling as though they are in a real classroom. Immersion in virtual reality generates a sense of presence, defined as "the subjective experience of being in one place or environment, even when one is physically situated in another" (Witmer and Singer 1998, p. 225).

Secondly, the authenticity of the learning task is another feature of the SEEE and refers to the deliberate exposure of the student to experiences that incorporate 'troublesome knowledge' (Meyer and Land 2003).

In Didascalia VC, authenticity and complexity are reflected in the (sudden and unpredictable) disruptive behaviours that are simulated and their potential escalation depending on the more or less strategic decisions made by teachers and pre-service teachers (users) to manage them, both in terms of teaching methodology (cognitive variables) and emotional regulation during interactions with students (affective and attitudinal variables).

Thirdly, we highlight interaction as a distinctive aspect of experiential education (Dewey 2004). The efficient use of technology in a SEEE must enable the learner to establish a strategic and immediate link between information and his or her actions in searching and processing information (Engel and Coll 2022), in this case, to make immediate decisions to deal with the critical incidents caused by disruptive behaviours in the classroom. Some usability studies have found this to be a decisive factor in technology acceptance (Dieker et al. 2017; Huang and Lee 2022).

Timely and effective feedback is the fourth element of note. According to Lugrin et al. (2016), effective CM training in an IVR environment requires precise control of training stimuli and scenarios and detailed performance feedback.

In the systems reviewed for this study, we identified that the decision-making process during execution is constrained by the lack of interaction between participants and agents (e.g., Ferguson and Sutphin, 2022; Kugurakova et al. 2023; Lugrin et al. 2016; Mouw et al. 2020) and the participant's performance is controlled by an external instructor, i.e., the teacher or researcher conducting the study (e.g., Dieker et al. 2017; Huang et al. 2016; Seufert et al. 2022). Other studies have shown that effective feedback systems should reinforce aspects of training that could be gamified (Honey and Hilton 2011), such as performance measurement, acknowledgement, skill level, character level, progression, points and stats (Lugrin et al. 2016; Toda et al. 2019), especially for engagement, motivation and evaluation purposes.

However, how best to provide feedback within an immersive CM training system remains an open question. Thus, beyond the biometric records that the system may capture, we pay attention to students' own experiential explanations, as they may help identify and characterise emotionally relevant "courses of action" for classroom climate management that they have practiced during the simulations (Evans et al. 2019; Seufert et al. 2022).

Finally, it is important for the tasks to be designed in a manner that would foster attitudes conducive to learning, particularly curiosity, considering its potential impact on reflective and creative thinking (Dewey 2004). Cheng's (2023) study on the usability of an epistemic curiosity-evoking model in an IVR to promote narrative reading in secondary school corroborated that the design of learning tasks based on this approach can provide students with a favourable user experience. Curiosity refers to the intention to probe for novel information to enjoy the learning experience, which can be triggered externally through task complexity, novelty, uncertainty and conflict (Berlyne 1960, cited by Cheng 2023).

Despite their acknowledged advantages, research on the pedagogical application of IVR has focused, primarily on two main streams: the effects of the unique system features of an immersive technology on user experience and how the use of immersive technologies enhances user performance (Suh and Prophet 2018). According to Bower et al. (2020), educators and researchers lack the empirical basis needed to determine 'when', 'why' and 'how' to use IVR in learning and teaching.

To optimise the effectiveness of deploying learning content, Sander (2010) underscores the criticality of usability testing. This practice enables developers, trainers, and end users to grasp interaction dynamics and identify usability issues (Holden and Rada 2011; Kari and Kosa 2023). Given the contextual dependency of 'magic number' rules for sample sizes in usability testing (Lewis 2014), it is imperative to complement quantitative measures with qualitative methods for assessing instructional design quality (Shernoff et al. 2018).

In 1998, the International Organization for Standardization (ISO) defined usability, highlighting user effectiveness, efficiency, and satisfaction within specific environments when utilising products or systems. Revisiting this definition, Bevan et al. (2015) emphasise that usability results from interaction rather than being an inherent product trait. This differentiation between usability and perceived usability is vital, with the latter heavily reliant on user subjective perception and experiences. User satisfaction, or perceived usability, significantly influences learning outcomes (Vlachogianni and Tselio 2023). In alignment with the study's objective, it is crucial to evaluate usability from all stakeholders' perspectives, including teachers and future teachers.

Given these considerations and acknowledging the ongoing development of the Didascalia prototype, we adopt a formative strategy concerning usability principles. Our ultimate goal is to identify obstacles and gather recommendations for potential improvements in the educational interface design, ensuring a secure, effective, and enjoyable experience for users.

As previously mentioned, comparable tools are designed to enhance teacher competence in CM by focusing on disruptive scenarios. For instance, Breaking Bad Behaviors (BBB) (Lugrin et al. 2016) revolutionizes teacher training through immersive VR simulations. The system provides a dynamic virtual classroom environment complete with essential features like classroom rules and an interactive whiteboard. These rules serve as foundational guidelines for preservice teachers to effectively address misbehaviors. Through a user-friendly session administration interface, instructors (lecturers) maintain precise control over the virtual classroom and its semi-autonomous virtual students. This control allows instructors to initiate scenarios, adjust student behaviors, and select disturbances from three different levels (low, medium, high), aligning with established classifications of disturbances in school settings.

Similarly, Didascalia VC is an innovative VR system focused on teachers training crafted to autonomously address classroom conflict situations. However, unlike other systems, Didascalia VC presents two important innovations.

Firstly, it is designed to address conflict situations. To that aim, it draws on an analysis of 1411 disciplinary reports from schools in Madrid and Barcelona (Masó 2022), Didascalia VC replicates the three most prevalent challenging incidents in secondary school classrooms, encompassing distinct communicative scenarios that are often fraught with conflict. These scenarios include a student challenging the teacher, two students disrupting the class with unacceptable behaviors (such as talking or disregarding rules), and group conflicts stemming from interpersonal difficulties among members. Within each of these communicative scenarios, a range of conflictive behaviors highlighted in the aforementioned study can be accurately simulated.

And secondly, Didascalia VC is an stand-alone system (users do not need anyone else to use it). It empowers users to autonomously navigate scenarios as teachers and practise their teaching skills safely within the virtual environment. The system provides the users with tailored feedback only based on their interactions.

These differences with existing applications make Didascalias a system worth exploring. It could play an important role in the professional development of future teachers, serving as practical training before preservice teachers encounter real classrooms.

The primary aim of this research is to assess user acceptance and discern the potentialities and challenges perceived by both teachers and pre-service teachers regarding the Didascalia VC platform as a facilitative tool for promoting classroom management competence.

2 System description: Didascalia Virtual Classroom platform

2.1 System overview

As anticipated in the previous section, Didascalia VC is an immersive virtual reality learning environment (VRLE) that allows pre-service teachers to experience, record and reflect on affective and attitudinal issues that are crucial for effective conflict management in the secondary school classroom. To that aim, the platform transports the users into a classic secondary school classroom through a virtual reality headset (Oculus Quest 1 and 2) (Bocos-Corredor et al. 2020). In this section, we will describe in detail the specific characteristics of this platform.

2.2 System architecture

The system consists of four modules. As seen in Fig. 1, each module controls:

Fig. 1
figure 1

Didascalia’s α-prototype preliminary architecture

Classroom environment manager This module allows the user to configure the virtual classroom at the beginning of the game. It also takes care of the creation of the virtual students, manages the launching of critical incidents in each of the scenarios, and controls the different game menus.

Users interactions manager Manages all the input devices. This is an interface that acts as an intermediary between the capture device (e.g. motion capture suit or Oculus controllers) and the Users interactions analyzer module. It aims to isolate the capture devices from rest of the system.

Users interactions analyzer This module receives the information from the previous module and analyzes the parameters to construct complex variables by comparing them before and after the critical incident. This module is composed of the following sub-modules:

  • Tone of voice analysis sub-module. This module analyzes users’ voice, enabling the averaging of volume levels before and after critical incidents, as well as calculating the percentage change in volume.

  • Discourse content analysis sub-module. This module uses the wit.ai tool (https://wit.ai/) to evaluate the user's intention according to the content of his message. For example, all phrases containing the words 'out', 'shut up' or 'leave' are translated as an authoritarian response.

  • Non-verbal language analysis sub-module. This module captures metrics of the user's movement (in Oculus, exclusively hands and head). It allows us to average the speed of hand and head movement before and after the critical incidents.. From this velocity we infer whether the movement has accelerated or slowed down.

  • Proxemics analysis sub-module. This module calculates the distance of the user to each of the student avatars. From here we can get three different states: the user approaches, moves away or remains at the same distance of the student.

  • User gaze analysis sub-module. This submodule measures the time the user looks at each student. It calculates the percentage of time looking at each student and checks if it changes after the critical incident.

Students behaviour controller This behavioural decision tree receives as input the five outputs from the Users interactions analyzer, and represents the user state (US) in a 5-dimensional space. The behavioral decision tree processes the five outputs from the user interactions analyzer to define the user state (US) within a 5-dimensional space. It computes the distance between the US point and each possible path's representation (see Table 1), selecting the closest path. This selected path—proactive, reactive, or avoidant—is then transmitted to the animations controller.

Table 1 Different paths representation

Animations controller This module receives the output from the Student Behavior Controller and launches the animations based on the disruptive behaviour (see Sect. 1.1.2. Virtual reality learning environment for teacher training). This is the feedback that the user receives for their actions. Table 2 provides a summary of the various animations.

Table 2 Animations launched based on user response to critical incidents

2.2.1 Typical gameplay

A typical gameplay —once the pre-service teacher puts on the VR headset— works as follows: firstly, a virtual classroom with students appears. The user starts developing a brief teaching activity (i.e., explaining new content, carrying out a cooperative activity…). Unexpectedly, a disruptive behaviour occurs (e.g., a student openly expresses his or her displeasure with the content of the class; someone does not want to work in the team to which the teacher has assigned him or her, etc.). The system evaluates the user's actions by capturing the available parameters before and after a critical incident. The system responds to these inputs (with a decision tree) by triggering different reactions (proactive, reactive or avoidant) in the students. Figure 1 shows the full flow of information.

Thus, through user testing within an ecological approach to CM competence, it is hypothesized that the Didascalia Virtual-ClassRoom (Didascalia VC) platform, designed as a Situated Experiential Education Environment (SEEE) and focusing on immersion, authenticity, interaction, and timely feedback, will positively influence users' perception of hands-on experimentation.

Specifically, users, including both in-service teachers and pre-service teachers, are expected to perceive the platform as effectively simulating real classroom scenarios and facilitating reflective practices essential for enhancing CM competence, in secondary school settings. This hypothesis encompasses both quantitative indicators aligned with the Technology Acceptance Model and qualitative assessments of users' experiences and perceptions during the testing process.

3 Method

We followed an explanatory mixed methods design with a quantitative–qualitative sequence and data integration in interpretation (Creswell 2002). Established quantitative usability measures supplemented qualitative methods to assess the global usability of the system.

The study is contextualised in the Master's Degree in Secondary Teacher Training. In total, there were 278 participants: 116 in Madrid and 162 in Barcelona. Of these, 28 (33.3%) were teachers (eight Master's Degree teachers and 20 secondary school teachers and educational counsellors) and 56 (66.7%) were Master's Degree students from two universities: 34 (60.7%) from Barcelona and 22 (39.3%) from Madrid. The majority were women (56%), average age 25.6. The experience was carried out in a group format, with one participant performing the practice in the VRE while his or her peers observed. The data analyzed for this study corresponds to the 84 participants who carried out the experience on the platform (28 teachers and 56 pre-service teachers). The teachers evaluated the system as experts, and the pre-service teachers provided input based on their training needs. Participation was voluntary, subject to prior informed consent.

3.1 Experimental procedure

The experiment took 25 minutes to complete and consisted of three stages:

(1) Briefing. Welcome, informed consent signing and virtual reality gear equipping.

(2) Experience. First, the system offered a brief description of the environment. Next, the user started the teaching activity. After a short period of time, a disruptive situation occurred. The user (teacher) had to react to manage the conflict by implementing the strategy he/she considered most appropriate in each case. The system game recognised the path taken and showed the final feedback.

The emulated situations that occur during the gameplay stem from the analysis of 1583 critical incident reports from secondary schools in Madrid and Barcelona (Masó 2022). This analysis was supported by a series of semi-structured interviews conducted by both occupational guidance specialists and experienced and reputable teachers with over five years of teaching experience. The results of this first stage allowed for the creation of three virtual scenarios that simulate disruptive classroom behaviours, which could potentially escalate into conflicts frequently encountered in secondary schools. These are (1) Overt academic disinterest (e.g., getting up without permission, not working in class), (2) Disruptions among classmates (e.g., chatting, or distracting classmates, confrontations between classmates in a workgroup) and (3) Disruptions affecting the teacher-student relationship (e.g., talking to the teacher inappropriately, challenging authority, among others).

As an example, the third scenario simulates a conflict between classmates in a group-work class. A student, Pablo, leaves his work group, apparently offended, shouting "Here you are, you bunch of morons! The teacher has a wide range of options for reacting to the disruptive scenario. However, we reduce the possibilities to three lines according to the strategy used (proactive, reactive or avoidant). For example: a) tell Pablo to return to his seat and continue working with his classmates, (b) order Pablo to go to the back of the classroom and wait until the class is over, and (c) approach Pablo to discuss the reasons for his current behaviour. If the teacher approaches the student (the most appropriate strategy in this case), Pablo calms down, and the teacher can resume his/her class without further interruptions. Otherwise, Pablo remains angry, standing at the side of the class, and the classroom noise does not stop.

Once the three scenes were completed, the devices were removed, and the experience recording stopped. The VR experience lasted between 5 and 7 minutes. The group peers observing the practice could see, through the computer screen, what the participant was experiencing on the VR device. They had a script with questions and key topics to guide the discussion.

(3) Final feedback. The system provided feedback depending on the user’s behaviour during the incidents. Feedback is based solely on the reaction of individual agents (avatars) to the class as a whole. These reactions range from laughter in the class to agents leaving the classroom. This tentative feedback is later interpreted, explained and expanded by an educational advisor in an individual interview (for teachers) or in a pre-service teachers’ focus group, with the collaboration of the other students who participated as observers.

In the joint reflection on practice within the Didascalia VC system, participants emphasised the importance of organisational strategies, assertiveness, and empathy in managing classroom dynamics. They highlighted proactive engagement with students to address disruptive behaviors and acknowledged the influence of psychosocial needs on classroom interactions. However, some participants also perceived limitations, such as challenges in implementing certain strategies and navigating complex classroom dynamics. Despite these challenges, the reflective process fostered deeper insights into effective classroom management, promoting a sense of agency among pre-service teachers for future practice. Due to space constraints, we will not elaborate on these findings in this article. For an in-depth analysis of this discussion, please consult a previously published study (Álvarez et al. 2023)

3.2 Data collection

At the end of the user experience, we administered an adapted questionnaire based on the Technology Acceptance Model (TAM) (Davis 1989). The TAM is a widely used framework for understanding users' acceptance and adoption of technology. This model is grounded in the Theory of Reasoned Action (TRA) (Fishbein and Azjen 1975) and the theory of planned behaviour (Ajzen 1985). The two primary factors in TAM are Perceived Ease of Use (PEU) and Perceived Usefulness (PU). Overall, PEU and PU refer to the degrees to which a person believes that using technology would be free from effort (PEU) and that using technology would enhance their job or task performance (PU). These two perceptions, PEU and PU, directly relate to another TAM-core variable, attitudes toward technology (ATT). The questionnaire was drafted based on questions from questionnaires used in previous studies related to IVR (Huang et al. 2016; Huang and Lee 2022). The items were tailored to align with both the unique attributes of our platform and the study's objectives. Eighteen items were presented (see Appendix 1), grouped into four dimensions aligned with the TAM: 1) PU (eight items), 2) PEU (eight items, 3) ATT (three items, and 4) BI to use the system for learning purposes (three items). The correlations between all dimensions were statistically significant. Internal consistency reliability was assessed using Cronbach's alpha, with values of 0.809 (PU), 0.76 (PEU), 0.871 (ATT), and 0.931 (BI) corresponding to the four dimensions, respectively. The mean values of variance extracted from all constructs were 0.707, above the recommended value of 0.50, thus showing that convergent validity was met.

In addition, participants were invited to a semi-structured interview to reflect on their user experience and to provide suggestions for possible improvements. The interview protocol was modelled on the aspects assessed in the adapted questionnaire based on the TAM, followed by open questions for each specific dimension and starting with general questions regarding the PU (e.g., "How was the experience with the Didascalia VC system?" and "What did you like the most/least about the system?"); PEU (e.g., "Did you feel any soreness or physical discomfort?, "How realistic were the behaviours, actions and dialogue with the virtual learners?"); ATT (e.g., "Did you feel any emotional activation during the test?"), and BI (e.g., “Would you recommend including this experience in teacher training? Why? How?"). The average duration of the interviews was 28 min (SD = 0.26; range: 23–37 min).

Descriptive analyses of the quantitative data from the questionnaires and a complementary analysis using the ANOVA test were carried out to explore the possible impact of socio-demographic variables (gender, teaching experience, prior experience with the technology, and university affiliation). Assumptions of homogeneity of variances were assessed and met prior to conducting the ANOVA. Levene's test results did not reach statistical significance, indicating that the variances across groups are homogeneous.

The analysis of the qualitative data was based on the literal transcription of the interview audio recordings. A recursive procedure was followed in four stages: inductive exploration, coding, description and interpretation of the data. First, the transcripts were independently analysed with a representative sample of the primary data, and the units of analysis (segments of content of interest to answer the objective) were delimited. Then, through successive approximations of the data to the study objectives (Mills et al. 2006), each researcher developed a preliminary system of themes and categories in which descriptions, interpretations and examples of each code were tentatively proposed. At the end of this exploratory procedure, it was established that listening to new interviews did not provide additional categories. In a third phase, the researchers identified convergences, resolved disagreements and agreed on the final version of the codebook, which included four main themes, coinciding with the four variables defined by the TAM. For each of the variables, categories and subcategories were defined to allow a better reading of the participants' experience (see supplementary material). Based on this, the data were manually coded with the Atlas.ti software (v.23.0) using the "smart coding tool", which allows iterative comparison of the coded segments and marking of the most illustrative segments of each category to generate the final results report.

4 Results

Below are the results obtained in the experiment. We start with the quantitative results extracted from the Questionnaire. For a deeper analysis, we then provide some quotes extracted from the post-test interviews.

4.1 User experience as assessed through the TAM questionnaire

In order to perform a user-centred evaluation of the system, we collected the experience evaluation from both the pre-service teachers and the teachers. The results in Table 3 show high scores on all dimensions (min = 3.75, max = 6.05). Except for three items of the first dimension (Perceived usefulness), the means of each item were higher than 4 (see Appendix 1). The mean of the dimensions was above 5 points (M = 5.63, SD = 0.23) on a Likert-7 scale (from ‘strongly disagree’ to ‘strongly agree’).

Table 3 Statistics summary

The highest value of the mean corresponding to the ATT dimension (M = 6.01, SD = 0.07) stands out. This result is interesting for our study insofar as it informs about the potential of the platform to activate the teachers' emotions during the execution of the practice (item 15) and about the overall satisfaction with the experience (items 13 and 14). In terms of PEU, the results report on the perceived ease of performing the practice on the platform. The responses refer to interference from technological devices and understanding of the environment. Regarding the PU, as noted above, participants expressed some dissatisfaction in their responses. The aspects that received the worst rating were those associated with interactivity. Despite the limitations reflected in this assessment, participants favourably rated the formative potential of the platform to promote classroom management competence, as reflected in the mean scores of the items corresponding to the BI dimension.

Figure 2 shows a breakdown of the perceptions of both groups, which had higher than the mean scores on the scale. However, although no statistically significant differences were found, the teacher group had a higher acceptance of the technology than the pre-service teacher group. In general, most users (teachers and pre-service teachers) reported favourable Attitudes during the system test (Mteachers = 6.04, SDteachers = 1.1; Mpre-service teachers = 5.67, SDpre-service teachers = 1.9); they felt that the system is Easy to Use (Mteachers = 5.28, SDteachers = 1.36, Mpre-service teachers = 5.13; SDpre-service = 1.71), and felt that the Didascalia VC platform has potential as a tool to promote classroom management competence in teacher training (Mteachers = 6.21; SDteachers = 0.94; Mpre-service teachers = 5.62, SDpre-service teachers = 1.6). However, the lowest scores correspond to Perceived Usefulness (Mteachers = 4.63; SDteachers = 1.37, Mpre-service teachers = 4.2; SDpre-service teachers = 1.67), which indicates that there are perceived obstacles in the techno-pedagogical design that are affecting the system’s potential. The main difficulties perceived by the participants referred to the virtual learners’ lack of responsiveness during the interactions in the different scenarios (items 4, 5, and 6 of the TAM Questionnaire, see details of the scores in Appendix 1). Below, we elaborate on these results through qualitative data.

Fig. 2
figure 2

Results from the Technology Acceptance Questionnaire. The figure contrasts the mean values for each dimension on a scale from 1 to 7, corresponding to the response interval of the Questionnaire

Note: The figure contrasts the mean values for each dimension on a scale from 1 to 7, corresponding to the response interval of the Questionnaire.

4.2 User experience as reported in the interview

For the analysis of the participant interviews we grouped the data into the four main themes, which are aligned with the dimensions of the Technology Acceptance Model (TAM).

We believe that the categories defined in each dimension can help to better identify the strengths and weaknesses that participants recognised in experimenting with the Didascalia VC system. The categories corresponding to each dimension are highlighted in italics. In the quotations selected to illustrate the results, we contrast the perceptions of teachers and pre-service teachers.

4.2.1 Perceived Ease of Use (PEU)

The PEU refers to the degree to which a user perceives that using Didascalia VC will be effortless. As in the questionnaire responses, in the interviews most participants reported that the system is easy to use.

"As you progress, you do learn indeed, and very fast, because it doesn't take long to do so, really." [Teacher (T)]

"It was my first time using this virtual reality tool. I didn't know how I was going to feel with it. But I think the fact that the tool is so intuitive and that I just need to operate a controller with the joystick made me understand it very quickly and I felt comfortable with it." [Pre-service teacher (PT)]

And, in general, they felt comfortable with the technology:

"I didn’t feel self-conscious about expressing myself either verbally or with gestures." [T]

"I thought the interface was super easy to use and very user-friendly." [PT]

Some participants reported feeling some physical discomfort at the beginning of the test, mainly dizziness and blurred vision:

"At first I felt a bit dizzy because I was seeing blurry, but the glasses were adjusted and I started to see perfectly fine." [T]

"At first, when I put on the glasses, I felt a bit dizzy, but nothing serious. I was able to continue perfectly well, but at that moment I did feel a little bit off." [PT]

Technological difficulties were to do with using the Joystick and controlling the virtual space.

"What I found most difficult was moving around with the buttons. At some point I lost sight of the students. Then I pressed a button too hard and the pupils came at me." [T]

"Maybe I should have gone to the pupil, if I approach her, I feel it is more natural for me to talk to the [disruptive] pupil. I tried this in the third situation, but I didn't manage to press the joystick." [PT]

It also seems necessary to improve the Initial Instructions in terms of describing the context and allowing more time for the user to situate themselves in the scenario they will encounter. The following comments illustrate the discomfort caused by this issue:

"At the beginning I couldn't process it all. I would have liked it to be more auditory and not have to read [the instructions]. You want to see the whole environment first to be able to find your feet. And then, at the same time as you have to find your feet, you have to read and all in a very short time." [T]

"When I started the experience, I didn't know what I was going to find. I knew it was about facing a situation in the classroom, but I didn't know it was going to be about something conflictive. So, when the first character has this [disruptive] reaction, I froze. I thought, where is this coming from? I didn't know how to react and I started to feel a cold sweat, how do I tackle this?" [PT].

Similarly, some participants suggested giving them the chance to prepare the lesson they are expected to do during the simulation.

"(...) we were thrown right into the test and didn't understand what we had to do. They told us to start a class, as we would normally do. OK, but what class? What should I say? [PT].

4.2.2 Perceived usefulness (PU)

The PU indicates the degree to which the user perceives that the use of the Didascalia VC platform can foster the classroom management competence. The responses to the questionnaire indicate several perceived difficulties in this dimension. In order to explore this dimension in more depth in the analysis of the interviews, three key categories were identified: Presence, Realism and Safety.

The analysis of the participants' comments allowed us to identify two interesting issues associated with Presence: "being in the classroom" and "playing the role of a teacher".

"I had the feeling that I had to do something as a teacher, like keeping order in the classroom." [T]

"Yes, it does help you to put yourself in that situation of being in front of the students, who are silently expecting your response, and deciding what you need to say. You have to fill that silence and address those situations all the time, that feeling is very well replicated. That is what I liked the most". [PT]

Realism was an aspect that was frequently mentioned by the interviewees, and it was also the variable with the lowest score in the Questionnaire. On the positive side, and linked to the feeling of Presence, participants highlighted that the system offers the possibility of "real" practice, in experimental conditions.

"What surprised me most is that there is an environment for experimentation, a laboratory for teaching. You are in a real situation. Until now that was unimaginable for me." [PT]

However, respondents pointed out some issues that they felt detracted from the realism of the environment, such as the number of students in the classroom and their physical appearance. In this regard, the participants made some interesting suggestions, such as the following:

"Maybe it should have more students. Classes with 32 students are already daunting, you feel more confident when there are few students." [T]

"Another thing that could be improved is to make it more real in terms of movement and expression (of the pupils)." [PT]

"Maybe because of the issue of diversity, it would be good if there was a student identified with a label in red or another colour, which indicates that they have attention deficit disorder, for example, so that you are able to decide ‘well, I'm going to put this person in front’, or something like that." [PT]

In addition to the realism perceived in the environment and in relation to the simulated conflicts, it is worth noting that most of the participants reported feeling challenged to make decisions in the different scenarios of the test.

"I appreciated the chance to learn how I would act in certain situations. Watching a video is not the same. In a video you don't have any say, but here, depending on what you do, the situation develops." [PT]

However, the difficulty in interacting with virtual students was the biggest barrier encountered by both teachers and pre-service teachers.

"I might have ignored (the disruptive student). And of course, I would have liked to see what would happen after my actions, but it's all over very quickly. There is no more mention of it, but then you keep thinking about how the action may evolve. I think that a specific action in real life would have had more consequences." [T]

"In general, my way of approaching things is to read the situation all the time, what the student is saying and what impression the group is giving you. And because I was not able to read the room, I was a bit lost." [PT]

In terms of the PU, the feeling of Safety was a prominent issue. For a more detailed analysis, we defined three subcategories in this category: Controlled environment, Enjoyment, and Confidence.

Most participants positively valued the feeling of practising in an experimental environment—Controlled environment—that simulates reality without compromising the consequences of their performance if they were to make a mistake, as the following quotes show:

"I appreciate the value of rehearsing in a safe environment where there are no real consequences for the young people (students). I think that's interesting, so that we are prepared before we face a real classroom." [T]

"It was more like 'getting rid of the fear', which you can get out of your system after a couple of classes. I see it more as a first approach to the classroom." [PT]

Despite the uncertainty that most participants reported experiencing, they reported having enjoyed the experience. They highlighted the playfulness and the swiftness of the actions that took place, as the following quotes illustrate:

"I liked the quick reactions, that everything happened fast. That I didn't have to teach much, that things happened immediately, that I didn't have to concentrate much on the class because I was already playing, let's say." [T]

"In fact, as it's a game we can pause it. Obviously, a pause to think. This would be an ideal exercise to do in a psychology class. Imagine yourself doing this activity and then stopping to discuss with the whole class: What would you do? What could you do? What would be a possible action-reaction?" [PT]

The chance to practice in a controlled and playful experimental environment is linked to the perception of Confidence to act in a natural way, without fear of making mistakes.

"I liked the experience itself, that is, being able to face that situation with the certainty that you can make a mistake, you can fail. Nothing happens. It's a virtual environment, you are in a friendly environment, you're with your classmates, with your teachers." [PT]

"In the end, when you are in the classroom environment, you are in front of the students and in other work experiences I have had when I had my boss around, it conditioned how I handled myself, in this case with the children, because I felt the pressure of "let's see how I do it". [PT]

Finally, in relation to the PU, the pre-service teachers expressed the need to receive more feedback during the experience.

"Obviously, it is complicated in terms of technology, but I think it is very important that the reactions, the feedback that the programme gives you, are as realistic as possible, because it is not about your handling of the situation itself, but about the reaction to your handling of the disruption that the students create. We are talking about to what extent we can expect a reaction to your tone of voice, to the words you are saying (...)" [PT]

4.2.3 Attitude (ATT). Perceived emotions

The ATT refers to the emotions (positive or negative arousal) experienced by the user during the Didascalia VC experience. The questionnaire responses generally conveyed a positive assessment of this dimension. To enhance the evaluation, participants were interviewed about the emotional states evoked during the experiment. This article specifically reports emotions related to the overall experience as users of the Didascalia VC system, with numerous emotions documented across each of the three scenarios.

The analysis of the interviews showed that the majority of the participants experienced more positive emotions (57.5%) than negative ones (42.5%). Enjoyment (30.2%) and Curiosity (27.1%) were the most prominent positive emotional states. The correlation analysis was conducted using Pearson's correlation coefficient and showed a statistically significant positive association (r = 0.267, p < 0.05) between the variables.

Among the negative states, Nervousness (44.9%) and Uncertainty (24.6%) were reported more frequently in both groups. The correlation between these variables also showed a positive and significant association. (r = 0.319, p < 0.01). Comparing both groups, the ANOVA test showed significant differences in relation to two variables: Nervousness [F(1, 84) = 5.78, p = 0.04] and Uncertainty [F(1, 84) = 3.58, p = 0.032]. The mean values were higher in the pre-service teacher group. When considering the gender variable, significant differences were observed regarding Enjoyment [F(3,80) = 3.60, p = 0.017]. The mean values among men was higher (Mmale = 0.39; SDmale = 0.49, Mfemale = 0.14; SDfemale = 0.35).

The following quotes evidence the experiences related to the feeling of Enjoyment.

“I felt very comfortable. I found the activity itself was enjoyable” [T]

“I really believed that I was teaching a class and to be honest, I enjoyed it” [PT].

These other quotes illustrate perceived Curiosity:

“I liked it because it makes you aware of how you react to situations, so it awakens sensations in you. It felt curious indeed." [T]

“Surprisingly, I thought it was quite real in the sense that you are dealing with a tricky situation, in front of the class, and you need to act.” [PT]

Many participants perceived the test scenarios as Challenging (Challenges).

“You have to think quickly about how you are going to react. And maybe you decide not to react, but you think quickly, you stay still for a moment and you refrain from entering into an argument with him (disruptive student).” [T]

“What I liked the most was the challenge of, okay, how do I deal with this situation now? [PT]

Regarding the reported Nervousness, having to improvise was found to be an influential factor, as the following comment shows:

“I think this experience forces you to improvise. And of course, that makes you nervous, it can make you feel overwhelmed.” [PT]

The previously mentioned limitations in interacting with pupils also caused stress, as this teacher relates:

“I got a little nervous because I didn't quite know how to act based on the boys' reactions. And I felt a little lost, blocked.” [T]

Some participants linked the stress they experienced during the test to a lack of practical knowledge. For example:

“I realized that despite the many theoretical classes that I have had and the many scenarios that I have been able to imagine, during this practice I actually went blank now and then and thought: what am I supposed to do now?”. [PT]

Uncertainty is linked to lack of prior information on how the scenario developed:

“I was waiting for something to happen because of the type of activity it was and that created a lot of uncertainty for me.” [PT]

“The whole time I was in this pose, with a passive-aggressive attitude, or rather an aggressive one, I think, waiting to see what would happen.” [PT]

4.2.4 Behavioural intention (BI)

BI refers to the probability reported by users of adopting the Didascalia VC system as a teaching and learning environment for the development of CM competence. As pointed out earlier, in the questionnaire, the majority of the participants showed willingness to use the Platform in teacher training. In the analysis of the interviews, we identified interesting suggestions that we will comment on below.

First, the participants suggested an observer experimentation design, similar to the one used in the user test. Students highly value the contributions that their classmates can make:

"They (my classmates) were super encouraging: ‘You've done very well!’ They told me that they had seen an evolution in my attitude. For example, at first, they noticed that I was more tense, and then I became more authentic, even being able to make some jokes with those fictitious students." [PT]

In this format, both teachers and pre-service teachers highlighted the benefits of having an opportunity for reflection, as a group, after the experience.

"It is an important experience, not only at the time of doing it and on an individual level, but it would also be useful to comment on it with colleagues. For example, we could analyze each scenario and ask yourself: What did you say and why did you do this? And what if you had done this? [T]

"Yes, I would like it to offer immediate feedback, like that data that appears now on the screen about the emotions you felt or how you moved. But a shared reflection after the VR test could be more interesting. A chance for us to see what happened and for you (the trainer) to tell me that I did not realize that my gaze was not directed at the student, or that I did not deal with something that was happening. It would be a chance to become aware of yourself, reflecting on how you have acted" [PT]

Secondly, the participants suggested using the platform for progressive and personalized training, throughout a training module.

"I find it interesting for teacher training, but it would have to be gradual, to go from simpler to more complex activities, like a proper training." [T]

"I would conceive of it more as a final project, one that included prior preparation, not having to go straight into facing the conflict. Perhaps it could be done before the work placements, or after the first observation period to be able to contrast information." [PT]

Participants found it advantageous to align this training with the placement period. They recommended a two-phase structure: sessions before placements to acquaint themselves with the classroom environment and delve into emotion regulation, and sessions afterward to provide additional opportunities for hands-on application of scenarios, thereby reinforcing theoretical understanding and improving the CM competence. This approach aligns with the understanding that hands-on experience is pivotal for refining competence in classroom management.

Beyond its possible uses as part of the Master’s, the students also thought it possible to use the Didascalia VC platform as a long-life training tool, as the following comment suggests:

"I would recommend using the system in workshops to explain how it works and then we can use it ourselves at our own pace, or in projects at the school, when we are working as teachers." [PT]

5 Discussion

Based on the results obtained, we believe that Didascalia VC could provide a safe environment to implement much needed practical training, in which pre-service teachers can experiment and reflect on their classroom management competence. With high mean scores in usability dimensions and variables, we confirm that the Didascalia VC platform enables pre-service teachers to explore cognitive, affective, and attitudinal aspects crucial for effective conflict management in secondary classrooms. Our hypothesis posits that Didascalia Virtual-ClassRoom (Didascalia VC), with its focus on immersion, authenticity, interaction, and timely feedback, positively influences users' perception of hands-on experimentation, thereby enhancing classroom management skills in secondary education.

Firstly, Attitude towards its use achieved a remarkable score (6.01/7). The IVR designed with the experiential and situated learning approach (Schott and Marshall 2018) was perceived as challenging and it aroused curiosity, two issues associated with technology acceptance (Cheng 2023; Holden and Rada 2011; Kari and Kosa 2023). In terms of teacher training, these attitudes may foster critical thinking during reflection on practice, thus reinforcing professional learning processes, as it allows pre-service teachers to re-signify the experience and gain greater understanding of their actions (Clarà et al. 2019; Seufert et al. 2022).

The majority of participants felt that the system provided an authentic, enjoyable and fun classroom experience, which allowed participants to perform the teaching role and face the challenges of managing the classroom climate. As previous studies have corroborated (Dieker et al. 2017; Garland and Garland 2020; Huang et al. 2016; Kugurakova et al. 2023; Mouw et al. 2020; Richter et al. 2022; Shernoff et al. 2020), IVR offers pre-service teachers the opportunity to practice and observe themselves teaching in a school-like setting.

Secondly, the feeling of being in a real classroom was highlighted as a key feature of the system. The simulated scenarios place the teacher in an authentic classroom situation in which It is essential to interact with a group of students (represented by avatars in this paper referred as agents) and make multiple decisions immediately in order to achieve a classroom climate that is conducive to the achievement of learning goals (Doyle 2006). Moreover, as demonstrated in previous studies (Ke et al. 2021; Stavroulia et al. 2019), this potential provided by IVR allows users' feelings and fears to emerge in a similar way as they do in the real classroom. In the user experience analysed, participants reported stress and uncertainty as emotional states that interfered with a more strategic performance during the test, thus evidencing the importance of paying attention to the socio-affective variables that influence effective classroom management (Keller and Becker 2020; Sutton et al. 2009).

Specifically for the students (pre-service teachers), uncertainty and nervousness affected satisfaction with the learning experience and these feelings were attributed to being forced to improvise, their lack of theoretical knowledge and the limitations in interacting with the virtual students. In order to address some of these issues and provide a more comfortable environment, introducing students to conceptual knowledge before engaging in the IVR environment, structured upon SEEE instructional principles (Schott and Marshall 2018), would be beneficial.

Despite these limitations, it appears that a majority of participants are inclined to utilize our tool prior to facing a real classroom situation. They expressed agreement that Didascalia VC provides a secure environment for practicing the necessary skills to handle conflict situations.

Specifically, with regard to the Intention to use the system (BI), participants highlighted the possibility of experimenting whilst being observed (by classmates). Post-action reflection can help identify emotional aspects that compromise the effective management of disruptive behaviour, an essential issue in CM learning (Evans et al. 2019; Sutton et al. 2009).

In addition, during the dialogue with their peers and/or the teacher, participants can view the activity log on the IVR, which allows them to revisit the scenarios and/or pause at crucial moments, thus adding a more practical focus to the training, as frequently demanded by teachers. As evidenced by several studies (Won et al. 2023), contextualisation of tasks, collaborative peer interactions and multiple learning opportunities are essential to provide and receive feedback on users’ performance in virtual environments, improve the quality of the experience, help develop knowledge and problem-solving skills.

As argued by Winstone et al. (2017), peer feedback can foster motivation to learn, engagement, agency (thinking about actions that can be taken with the knowledge acquired) and shared responsibility for ensuring that feedback is effective to improve future actions, which can result in improved perceived self-efficacy for classroom management as future secondary school teachers. Furthermore, it is worth incorporating in the system tools and/or algorithms that would allow providing real-time feedback which, as corroborated by several studies (Dieker et al. 2017; Huang et al. 2016; Kugurakova et al. 2023; Seufert et al. 2022), can help users to identify their weaknesses during their performance and to continuously adapt their behaviour to effectively achieve the training goals. In this regard, it would be worth harnessing the potential of learning analytics to map learners' temporal behaviour patterns using, for example, sequencing, which offers valuable insights into learners' learning strategies (Elmoazen et al. 2023).

Participants also highlighted the need to offer progressive training. As Honey and Hilton (2011) point out, progressive training in IVR can be achieved by reinforcing the gamification aspects, gradually increasing challenges and offering more perceptual support during execution (Lugrin et al. 2016; Toda et al. 2019). Task complexity can foster curiosity, an attitude that impacts reflective and creative thinking (Cheng 2023; Dewey 2004; Seufert et al. 2022), which is a desirable and rare approach in virtual simulations used to foster CM in initial teacher education (McGarr 2021).

In terms of usability issues, although the majority of participants perceived the system as easy to use, we observed that some participants reported problems with learning with the tool due to the lack of time to familiarise themselves with the environment before practicing, and problems mastering technical issues such as using the joystick and becoming space aware, which can spoil the feeling of ‘presence’ (Dieker et al. 2017). Prior training, such as a tutorial, would be helpful and would strengthen the sense of security to practice CM in a controlled and playful environment, which is an aspect that was highly rated by most participants in the retrospective analysis of their experience in the interview. Furthermore, as suggested by the participants, it would be worth offering them the chance to prepare in advance the lesson to be taught when entering the virtual classroom. In line with the ecological approach to classroom climate, anticipation, planning and advance consideration of strategic actions by the teacher contribute to more efficient management of disruptive behaviour (Doyle 2006; Evertson and Poole 2012).

Immersion and interaction in an educational and playful VR environment allow the user to engage more of their senses and learning styles and provide the user with a sense of presence (Han et al. 2023). However, our results show that the user experience would have been better with more interactions. Consistent with other studies (Ferguson and Sutphin 2022; Huang and Lee 2022; Lugrin et al. 2016; Mouw et al. 2020), limitations in interacting with virtual learners was a constraint highlighted by most participants. In line with Engel and Coll (2022), this constraint prevents the learner from connecting with the information provided by the technology in order to make strategic decisions in relation to the goals of the learning task, and this may affect motivation. In addition, from the ecological approach to classroom climate (Doyle 2006), constraints to interacting with virtual learners are a significant barrier, as teaching performance is determined by the teacher's ability to perceive and interpret classroom events that are relevant to students' learning, and to decide how to (re)act appropriately in the face of these events.

To improve this limitation that affects the sense of realism and immersion (Dieker et al. 2017), it is essential to improve the appearance of agents and, especially, their responsiveness to users' actions, as well as to make the virtual classroom environment more dynamic and complex (number of students, interactions between them, conflict escalation, etc.). As corroborated by Huang et al. (2021), the complexity of the virtual classroom can affect the behaviour of pre-service teachers during their practice in IVR.

6 Conclusions and future work

The main result we have obtained is that Didascalia-VC, as confirmed by the users, could be a very useful tool in teacher training, as it would allow pre-service teachers to improve their competence in solving complex situations in the classroom. Participants highlighted the importance of the immersiveness of the system and its realism.

All participants recognised that the tool is an ideal complement to pre-service teacher training. In addition, the experiences provided by IVR can foster reflective and critical learning about effective classroom management. This can be very enriching for pre-service teachers, who often resent the theoretical nature of their training and the ineffectiveness of practical approaches, where they have no opportunity for self-reflection or receiving constructive feedback on their performance.

In conclusion, this study not only highlights the success of Didascalia Virtual-ClassRoom in meeting usability standards, but it also opens avenues for continuous improvement. The proposed enhancements, rooted in user feedback, not only optimize the tool's functionality but also pave the way for its effective incorporation into the broader landscape of teacher training.

Following suggestions from participants, future work will include 1) improvements to the graphical aspects to make the virtual environment more realistic and immersive, 2) additional improvements to the menus and learner interactions, and 3) provision of a tutorial-like opening scene to overcome learning barriers for some pre-service teachers who may be unfamiliar with virtual reality.

Nonetheless, there are limitations to the findings of this study. Firstly, the tool is still in the development phase. Secondly, the methodology proposed for its implementation requires further trials, in successive years and in other universities, in order to continue incorporating the voices of teachers, students and managers from universities that offer Master's programmes in teacher training. Additionally, it would be worth incorporating the use of this IVR tool, with the suggested improvements, into a training module that would impart relevant knowledge both prior to practicing with the IVR tool and after it.

Future research should also aim to examine short- and long-term outcomes in a variety of conflict situations to which pre-service teachers are exposed during professional practice, thus exploring the extent and sustainability of CM learning in the IVR environment.