The DVIB is a technological open-source network that aims to provide information, orientation, and secure access regarding educational programs and offers in Germany and Europe on a digital basis (DVIB, 2023). The network is not meant to create another learning environment, but to smartly connect existing educational platforms and offerings in a way that creates user-centric, functionally powerful, if desired personalized, and self-sovereign access to lifelong and cross-sector educational offerings ranging from school to vocational training and continuing education, while simultaneously enabling and supporting collaborative processes. For this purpose, common standards, formats and interoperable structures are being developed, used and tested. In the current pilot stage, more than 30 sample applications are connected to the infrastructure prototype through the use of five core components (DVIB, 2023). First, a personal (1) digital identity enables users to access content self-organized, to participate in collaborative teaching or learning scenarios, and to share information as well as results by authenticating through a single sign-on. A both central and linked (2) wallet is used to store and share data and documents like learning outcomes and certificates obtained throughout individuals' entire educational biography via multiple educational providers and platforms. The trustworthiness of such (3) digital credentials is ensured by deploying a public key infrastructure with certification authorities and central registration authorities as an established standard. An integrated information portal featuring a dashboard and learning path assistant acts as a (4) ‘storefront’ for users to access connected education providers based on their customized search, competency, and needs profile. The processing of requests and deposits within the DVIB is based on a (5) data space (metadata repository) that is used to store non-personal and non-transactional data as a basis for linking learning information and content to learners' individual educational journeys or teachers' customized queries. For an overview of the infrastructure components and their technological relations see Figure 1.

Figure 1: Technological architecture of the DVIB. AAI = authentication and authorization infrastructure; PDD = user's personal data and documents (adapted and translated from DVIB, 2023)

These explanations, taken from DVIB's official website, highlight the platform's focus on not only providing broad, customized, and secure access to educational offerings but also on enabling and supporting collaborative education. In particular, DVIB facilitates productive collaboration and a lively exchange of experiences and between teachers, users and providers of education through the interconnection of different educational institutions. This includes communication and collaboration of the scientific community as the generating entity of knowledge and evidence for the education of individuals. Within this scope, "AVILAB2" uses the technical possibilities of the DVIB to exemplarily connect the social VR software "TriCAT Spaces" and to investigate its suitability for the virtual realization of scientific academic conferences.

Since the expansion of the world wide web and the increasing equipment of companies and citizens with sufficient bandwidth internet connections, video-based communication has become the means of choice for remote collaboration. Attention in video-based meeting solutions increased in the late 1990s (Finn et al., 1997) when they were also empirically shown to provide a quality of communication that can be considered equal to face-to-face meetings (Olson et al., 1997). Although video calls provide users with multiple advantages, for instance the opportunity to see meeting participants or share their screens with their fellows, they also yield limitations like restricted possibilities for social interaction between participants (Bonfert et al., 2022). Furthermore, in recent years and in particular since the massive increase in the use of videoconferencing software during the COVID19 pandemic, a phenomenon called "Zoom fatigue" occurred more and more frequently. This state, which can be considered a form of computer-mediated communication fatigue, refers to user's exhaustion after participating in many or long video conference meetings (Bonfert et al., 2022). Besides experiencing increased cognitive load or mirror anxiety while permanently seeing one’s self-image, and hyper-gazing from feeling watched by many faces (Bailenson, 2021; Fauville et al., 2023), Zoom fatigue may also result from reduced mobility during video calls (Bailenson, 2021), consequential feelings of being physically trapped (Faulville et al., 2023), and the spatial reduction of conversation partners and their background to a 2D projection (Nadler, 2020). In this regard, social VR may provide a promising alternative for online team meetings, although potential drawbacks like concentration problems, cognitive overload, or VR sickness that may arise over time of use (e. g., Fischer et al., 2021) should not be left aside from this consideration.

Social VR platforms represent “applications that enable geographically remote users to interact with each other in shared virtual environments through VR technology, i.e., immersive head-mounted displays” (Sykownik et al., 2021, p. 537). Today, a multitude of online VR applications with a social component exist, including various characteristics and functionalities (for an overview see Liu & Steed, 2021). Due to its possibility of simulating any communication process in a highly imaginable way, social VR is credited with not only surpassing the experience and effectiveness of communication compared to videoconferencing, but even making it comparable to real-world environments (Moustafa & Steed, 2018; Torro et al., 2021). Recent research indicates that social VR meetings resemble in-person meetings more than video conferences with regard to group dynamics in gatherings before and after the main meeting, or for the impression of being together in the same space (Bonfert et al., 2022). Furthermore, social VR holds huge potential to foster socialization activities, like meeting colleagues in virtuality, as well as playful and creative activities that increase participants' relatedness, enjoyment, and self-expansion (Barreda-Angeles & Hartmann, 2022). Consequently, the use of social VR for online team gatherings may also help to overcome Zoom fatigue.

In the field of education, social VR platforms are recently piloted for various purposes. For instance, they provide a spatio-social environment for lectures, courses, and workgroup meetings but also academic conferences. Their potential as a more environmentally sustainable replacement to both costly and time-consuming conventional formats has been discussed for more than a decade (e.g., Welch et al., 2010). Compared to academic gatherings or lectures held in videoconference, the use of social VR creates a more authentic feeling of a social co-experience with other scientists, lecturers or learners in a common space. In particular, scientific conferences that inherently thrive on the shared event experience, exchange, and networking may greatly benefit from this. Early attempts using the online multimedia platform “Second Life” were considered reasonably successful while mainly suffering from technical issues (Erickson et al., 2011). With advancing technological development, and at least since the outbreak of the COVID19 pandemic, there have been growing attempts to conduct scientific conferences in social VR with increasing success. Accompanying empirical evaluations indicate high overall satisfaction among participants (Ahn et al., 2021) while perceiving lively experiences (Kirchner & Nordin Forsberg, 2021) and performing proxemic interactions between attendees comparable to physical settings also leading to dynamic group formations (Williamson et al., 2021). Given the DVIB’s subgoal of encouraging participation in collaborative teaching and learning scenarios, connecting a social VR environment to the digital infrastructure seems both appropriate and purposeful.

In the pilot project "AVILAB2" we exemplarily connect the social VR environment "TriCAT Spaces" to the DVIB. We selected this particular platform because it offers three key advantages: (1) It is hosted by the operating company on servers that are located in Germany and meet DVIB's requirements in terms of data protection and security. (2) It provides a very user-friendly interface and is highly customizable. Predefined room structures for common use cases can be adapted by users based on their individual requirements. In addition, an integrated 3D editor enables the creation of completely new worlds and educational settings. Furthermore, (3) users can access the VR environment in desktop mode while seated, which significantly reduces the risk of suffering from VR sickness.

From a technical view, the social VR software has already been and will continue to be comprehensively connected to the DVIB. Currently, the single sign-on of the DVIB is already available as a prototype, so that users do not need an extra account to access the virtual environment. The interface to the BIRD Wallet app of the DVIB has also already been prototyped, so that information and files (as well as certificates) from educational social VR events can be stored there. To establish a connection, one party must generate a QR code and mark all attributes in optional and mandatory entries that they (1) want to send themselves and (2) require the other party to send. On this basis, we plan to add the ability to generate, store, and transfer certificates that participants have earned at events or learners have acquired in courses (such as credits) in the further development process. Additionally, we intend to automate the booking of virtual rooms within the environment by the use of the DVIB.

The social VR platform is suitable for addressing several key use cases in formal and informal education in Germany. One major use case concerns scientific conferences. Just for the year 2023, conference directories list more than 500 events with a scientific reference to be hosted in Germany (COMS, 2023; IDW, 2023). Since many smaller events in particular are often not registered here, the actual number is estimated to be much higher. With usually at least double-digit and often even three- or four-digit numbers of participants traveling to these events all across the country, the use of social VR enables high resource savings while at the same time providing a realistic event experience - in contrast to mere video conferencing events. Another large target group is formed by the lecturers and students of distance learning universities. In 2021, approximately 190,000 students were enrolled in distance learning programs (IUBH, 2021), with numbers growing steadily. The social VR platform also offers them an authentic and social educational experience. Finally, cross-university teaching where students from two or more public or private on-campus or online universities collaborate with each other remotely (e. g., Breitenstein et al., 2018) is increasingly in demand. This scenario provides a good basis for increasingly cross-regional and globalized business processes in the later professional life of graduates. Consequently, social VR has the potential to increasingly reach these target groups as well.

In preparation for a sustainable connection of social VR environments (like "TriCAT Spaces") with DVIB, it is crucial to first investigate their suitability for the use cases mentioned above and their acceptance by the users. As a first step, we aim to examine these success factors with regard to the case of academic scientific conferences. Beyond the previously mentioned initial empirical glimpses, we however still know very little about the appropriateness of social VR for scientific conferences and its acceptance by participants. Subsequently, we first describe a sound theoretical framework as a basis to examine such factors.

As one of the very few recent pilot studies to assess technology appropriateness of social VR for academic conferences, the evaluation of a discussion-driven academic conference on embedding augmented reality and VR technologies in vocational education and training using “AltSpaceVR” showed a high to very high perceived suitability among participants (Zender & Mulders, 2022). Going more into detail, a further study evaluating a virtual scientific conference held via “MozillaHubs” investigated the appropriateness of social VR with regard to typical conference tasks and in comparison to different forms of multimedia platforms (Ahn et al., 2021). Participants reported a low to medium appropriateness for asking questions and learning new skills while perceiving a medium to high suitability of social VR for the activities of listening to talks, discussing research, socializing, and networking, thereby outperforming any further provided kind of tools (such as social network systems and text-based chat platforms). Unfortunately, the exciting results of both studies are limited by the circumstance that they each only measured the suitability of social VR for scientific conferences on the basis of one ad hoc item, which, moreover, is not theoretically substantiated in depth within the articles. The present study tackles this issue by using an integrated approach of the task-technology fit theory and the technology acceptance model as a theoretical foundation as well as the accompanying, validated scales for measurement.

In a nutshell, task-technology-fit theory proposes that tasks and technologies interact to “the degree to which a technology assists an individual in performing his or her portfolio of tasks” (Goodhue & Thompson, 1995, p. 216). Such degree is called task-technology fit and, in turn, predicts users’ reactions and performance moderated by utilization. Although the task-technology theory was never tested entirely, the framework yet received multiple empirical support with regard to its partial links, containing studies in the fields of education and profession (e.g., Kerr & Murthy, 2004; Maity & Dass, 2014). The technology acceptance model describes how users' perceptions of technologies ultimately lead to use (Davis, 1985). With regard to technologies, it proposes that perceived ease of use influences perceived usefulness while both predict users' attitudes. Subsequently, both perceived usefulness and attitudes towards the technology predict users' behavioral intentions to use it, which, in turn, predicts the actual technological use. Similar to the task-technology theory, the technology acceptance model is well supported not only with regard to education and work life (e.g., King & He, 2006; Šumak et al., 2011).

Both the task-technology fit theory and the technology acceptance model share the common goal of predicting how users perceive a technology and to what extent they (will) use it, by taking specific characteristics into account. In consequence, more recent work attempted to integrate both models by substituting the rather general construct of external factors by the more specific characteristic of task-technology fit to create a hybrid theory (Dishaw & Strong, 1999; Howard & Hair, 2023). Notwithstanding minor peculiarities among previous integration approaches, we aim to use their joint core (see Figure 2) as a sound theoretical basis for the present study.

Figure 2: Integrated task fit and acceptance model of technology

On this basis, we investigate whether social VR technologies are to be considered suitable for educational collaboration purposes within the frame of the DVIB with particular focus on scientific academic conferences as a use case. While technology acceptance of general VR in education and profession has at least received some study (e.g., Cabero-Almenara et al., 2023; Li et al, 2022; Pletz & Zinn, 2018), to the best of our knowledge there is neither evidence on the task-technology fit and technology acceptance of social VR nor empirical data on integrated approaches. To address the aforementioned research question and issue, we carried out a methodical rigorous study by the use of valid scales rather than ad-hoc items to measure the to be examined characteristics.

Since to date only few scientific academic conferences have been held in social VR, it can be assumed that the majority of scientists do not yet possess experience with this technology and are accordingly not able to submit assessments regarding its acceptance and task fit. Therefore, and in order to provide an authentic field of investigation, a virtual scientific conference was organized and carried out by the "AVILAB2" staff in association with the Center for Open Digital Innovation and Participation and the coordinator of the structured graduate program "Education & Technology" of the Technische Universität Dresden in Germany. The conference was titled "Empirical Status Conference on Digitization in Education and Society" and, consequently, focused the presentation of recent empirical findings regarding the role and use of digital media in education and beyond.

Figure 3: Participants located in the area for virtual exchange and resting

The social VR software "TriCAT Spaces" served to technically create a virtual, immersive 3D conference environment for the conference. The tool provides users with multiple functions to perform activities like (1) customizing their own avatar, (2) exploring the environment in a self-directed manner, (3) interacting with other users (through different possibilities like voice talk, gestures, and text chat), (4) viewing the event plan including the agenda and virtual locations in order to freely choose a session to attend, and (5) importing and managing their own slides and further media (like images, audios, or videos) in the role of a speaker. The conference environment consisted of a virtual plenary hall, as well as an area intended for free exchange among the participants and resting (see Figure 3), and three break-out rooms to host subject-specific parallel sections designed in three different colors. The plenary hall had a large grandstand with seats and a stage with a lectern and giant monitors displaying the presentation and the webcam image of the speakers as well as the conference agenda. The breakout areas each included several tiers of seating for the audience, as well as a stage equipped with a standing desk and large monitor for the presenters, and sections with posters. The networking and resting area contained preview rooms for each section showing its agenda and posters, as well as a hall of fame to honor the highest quality contributions, and a resting room to enable participants to play integrated browser games and videos.

All staff members of the organizing institutions and programs were invited a few weeks prior to the opening of the conference via their usual channels for communication (in particular mailing lists). Participation was mandatory for the scientific staff of CODIP. All additional staff as well as the personnel of the two further hosting organizations were invited to participate voluntarily in the event. Further information and communication regarding the conference also took place through the established channels for communication. Participation in the conference was free of charge. However, in order to facilitate organization, prior registration was requested.

In advance to implementation, a call for papers was issued to gather appropriate submissions. The submitted proposals underwent a peer review process by the organizers based on an evaluation guideline to enable as comparable and transparent feedback as possible. The call was addressed by a total of 12 submissions, which could all be accepted due to sufficient quality and thematic suitability. Accepted speakers were asked to prepare both slides and overview posters to be presented in a digital format. Few days prior to the conference, two trial runs was held simulating the entire process in short in order to familiarize moderators and speakers with the social VR environment and its functions, to clarify technical issues, and to prevent technical problems as well as possible. In addition, a virtual room for individual try-outs and technology checks had been available to all participants one week in advance of the conference.

Figure 4: The virtual conference's keynote speech taking place in the plenary hall

The virtual conference took place on February 6 of 2023 and was opened by a keynote on the role and design of social cues in multimedia learning grounded in the cognitive-affective-social theory of learning in digital environments (CASTLE; Schneider et al., 2022) to take place in the virtual plenary hall (see Figure 4). Subsequent to the keynote lecture, a guided discussion took place, in which the audience had the opportunity to actively participate in the discussion and ask questions to the speaker. Afterwards, the audience split up to attend the talk sessions based on their individual interests. Any of the three sessions consisted of two time slots that each contained two short talks to present results of empirical studies related to the digitization of educational and societal processes. For instance, the presentations addressed competence acquisition and assessment in higher education, and the role of AI and VR support in education. In between, a poster session was held where participants were able to virtually interact with the speakers by asking them questions or telling their opinion and ideas regarding all presented studies. The event closed with a group discussion about the advantages and shortcomings of social VR for scientific gatherings in comparison to other digital tools (like video conferencing systems), carried out in the plenary hall.

Figure 5: A talk being held in the red colored breakout room

Subsequent to the event, a survey was carried out through an online questionnaire provided to all participants of the virtual scientific conference in order to gather data regarding the aforementioned research questions. Participation in the survey was voluntary for the conference attendees. The survey data was exported and evaluated with regard to descriptive statistics and inference in a semi-automated manner using analysis software.

A standardized online questionnaire containing a total of 13 bound questions served as the instrument for survey. It measured socio-demographic characteristics of the participants as well as their perceived task-technology fit and technology acceptance. With the exception of participants' technology commitment, ad hoc scales and items were used for sociodemographic data collection. Technology commitment, task-technology fit, and technology acceptance were assessed using well-established and to some extent validated scales (see below). Prior to its use the questionnaire was tested by two experts. The pretest results served as a basis for optimization.

Sociodemographic. Participants’ age, perceived gender, qualification, and current scientific activities were assessed by the use of five questions with suitable items (e. g., “Which gender do you feel yourself to belong to?”, offering the options “female”, “male”, “diverse”, and “other”, and "What activity do you currently carry out in the field of science?" providing items like "student assistant", "research associate", and "professor"). Previous experience with VR technology and virtual scientific conferences in general, and with regard to social VR in particular, was evaluated by the use of four questions, for instance “How often did you use virtual reality (VR) technologies (like "Occulus Quest", "PlayStation VR" or at the "Time Ride Dresden") prior to the Empirical Status Conference?” and "How often did you participate in online scientific conferences hosted in social VR (e.g., via "AltspaceVR" or "TriCAT spaces") prior to the Empirical Status Conference?". Answers were to be reported on a 5-point Likert scale ranging from "very rare" to "very often". To assess the participants’ technology commitment (α = .80), we used the Technology Commitment Short Scales (Neyer et al., 2012). They encompass four items for each of the three subscales to measure the participants' technology acceptance (α = .73; e.g., "I am very curious about new technical developments."), technology literacy beliefs (α = .78; e.g., "For me, dealing with technical innovations is usually too much of a challenge."), and technology control beliefs (α = .72; e.g., "It mainly depends on me whether or not I am successful in using modern technology.") to be rated on a 5-point Likert scale (ranging from “Does not apply at all.” to “Does fully apply.”).

Task-technology fit. For the assessment of the task-technology-fit, we used the scales by Schrier et al. (2010). They originally consisted of seven items (e.g., "The available options had the appropriate level of detail for my use.") to be assessed on a 5-point Likert scale with the two poles of “Does not apply at all.” and “Does fully apply.”. However, since the performed analysis of the scales' internal consistency showed poor reliability (α = .577), two items ("The software was up-to-date and available when I needed it." and "The functionality of the software was comparable to those of other providers / tools.") were excluded from further analysis. Their exclusion led to an overall reliability that can at least be considered as questionable (α = .615).

Technology acceptance. In order to measure participants' technology acceptance (α = .93), the scales by Davis (1985) were used. They consist of three subscales for perceived ease of use (4 items; α = .89; e.g., “Learning to operate the software was easy for me.”), perceived usefulness (6 items; α = .92; e.g. “Using the software, I can participate more quickly in scientific conferences.”), and intention to use (4 items; α = .90; e.g. “I intend to continue using the software in the future if I have the opportunity to do so.”) that were to be assessed on a 5-point Likert scales ranging from “Does not apply at all.” to “Does fully apply.”.

Age, perceived gender, qualification, and current scientific activity. A total of 50 people attended the virtual academic conference, of which 31 (62 %) participated in the voluntary survey. The majority of participants (51,1 %) perceived themselves as female (38.7 %; 3,2 % miscellaneous). They varied in age from 26 and 60 years (M = 37.10 years, SD = 7.81). Of the participants, six persons (19.4 %) were younger than 20, 17 persons (54.8 %) were between 30 and 39, seven persons (22.6 %) were aged 40 to 49, and two persons (6.5 %) were more than 50 years old. Among the subjects, two persons (6.5 %) held a bachelor's degree, while 22 persons (71.0 %) graduated with a diploma, state examination, magister, or master's degree, and seven persons (22.6 %) with a doctorate or habilitation. At the time the conference was held, three participants (9,7 %) were employed as technical staff, 25 participants (80,6 %) were working as research associates, and two participants (6,5 %) held a chair as a professor, whereas one participating person was engaged as a doctoral student without an employment contract in academia.

Prior experience and technology commitment. Participants assessed their prior experience with VR technologies rather low (M = 2.0 out of 5; SD = 1.18) in general, and very low to low with particular regard to social VR software (M = 1.52 out of 5; SD = .89), While participants' previous experience with scientific online conferences is to be considered high (M = 4.03 out of 5; SD = 1.20), they only reported a very low to low experience with such conferences held in social VR (M = 1.23 out of 5; SD = .67). Regarding their technology commitment, participants reported high acceptance of technology (M = 3.98 out of 5; SD = .79), high to very high technology literacy beliefs (M = 4.47 out of 5; SD = .59), and medium to high technology control beliefs (M = 3.70 out of 5; SD = .78), resulting in a high overall commitment (M = 4.05 out of 5; SD = .54).