When museums and archaeological sites adopt XR, the first question is often which device they should choose. Visitors may enter a reconstructed environment through a VR headset, use AR to place vanished structures over a physical site, or explore panoramic scenes through a 360-degree virtual tour.
But is the most immersive technology always the best option? Researchers from the University of Salerno and collaborating institutions implemented the Roman archaeological site of Villa Regina in three formats and compared motivation, presence and engagement. The results do not identify a single winner. They reveal three different experiential profiles.
The Same Site Was Recreated in Three Formats
Comparing XR technologies is difficult when the content changes along with the device. If one condition uses a detailed 3D reconstruction and another uses simple photographs, it becomes difficult to separate the effect of the medium from the effect of the content.
To reduce this problem, the researchers created VR, VR360 and AR experiences from the same digital assets, information and narrative structure. Architectural features, storage areas and artefacts from Villa Regina were reconstructed through photogrammetry and 3D modelling before being adapted to each platform.
The VR condition used a head-mounted display with positional tracking and handheld controllers. Participants navigated the reconstructed site from a first-person perspective. The VR360 condition presented panoramic scenes on a tablet. The AR condition used a tablet equipped with LiDAR or an RGB-D sensor to anchor digital reconstruction within the physical environment.
Ninety Participants Were Assigned to Three Conditions
The study included 90 adults. Participants were randomly assigned to VR, VR360 or AR, with 30 people in each condition. Immediately after the experience, they completed questionnaires measuring intrinsic motivation, presence and user engagement. The researchers also recorded exploration time.
A subsample of 20 participants completed semi-structured interviews lasting approximately 20 minutes. This group included seven VR participants, six VR360 participants and seven AR participants.
The experiment used a between-subjects design. Each participant experienced only one technology. The results therefore compare the averages of three groups rather than the preferences of the same individuals after trying all three formats.
VR Created the Strongest Sense of “Being There”
The clearest difference concerned spatial presence. On a seven-point scale, VR achieved a mean of 6.12, compared with 5.41 for AR and 5.02 for VR360. The overall group difference was statistically significant, and the adjusted post-hoc comparisons reported differences among all three conditions.
VR also received the highest focused-attention score. The means were 5.71 for VR, 5.33 for AR and 5.11 for VR360. VR scored significantly higher than both tablet-based conditions. It also received higher aesthetic-appeal ratings than AR and VR360.
The interviews reflected a similar pattern. VR participants frequently described curiosity, emotional involvement and a desire to continue exploring. Some, however, reported brief disorientation or a need to adapt at the beginning. Deeper immersion did not automatically mean easier use.
AR Led in Perceived Realism
VR did not outperform the other formats on every measure. AR received the highest realism score at 5.89, compared with 5.77 for VR and 5.21 for VR360. AR was rated significantly higher than both conditions.
The paper interprets this through contextual anchoring. VR replaces the physical environment with a reconstructed world. AR instead places vanished architecture and artefacts in relation to the site in front of the visitor, showing where digital reconstruction belongs within the existing space.
AR and VR360 were also rated as easier to use than VR. Mean perceived-usability scores were 5.91 for AR, 5.88 for VR360 and 5.02 for VR. The study did not establish a significant difference between AR and VR360, but both scored significantly higher than VR.
VR360 Was Less Immersive but Served a Different Role
VR360 received lower spatial-presence and involvement ratings. The researchers discuss restricted navigation and limited interaction as possible explanations. Users could rotate their view, but they had fewer opportunities to move through the space or manipulate content.
This does not make the 360° tour unnecessary. Interview participants associated it with accessibility, visual coherence and comfortable exploration. Because it can be used on a tablet without a headset, it may be suitable for remote visits or pre-visit orientation, where ease of access matters more than deep immersion.
Format | Strength identified in the study | Consideration |
|---|---|---|
VR | Spatial presence, focused attention, aesthetic appeal | Requires a headset and controllers; adaptation may be needed |
AR | Realism, connection between reconstruction and place, usability | Depends on the physical setting and device capability |
VR360 | Easy access, comfortable viewing and remote-use potential | Lower spatial presence and less active interaction |
Longer Mean Exploration Time Does Not Prove a Causal Effect
Mean exploration time was 1,142 seconds for VR, 983 seconds for VR360 and 867 seconds for AR—approximately 19, 16 and 14 minutes, respectively.
However, the paper did not report an inferential test for these time differences. It is therefore accurate to say that VR had the longest mean exploration time in this sample, not that VR statistically caused longer use.
Motivation, presence and engagement were positively correlated with exploration time. These correlations do not determine whether stronger presence led participants to stay longer or whether longer exposure increased presence.
The Study Does Not Show That VR Produces Better Learning
The study measured immediate self-reports of motivation, presence and engagement, along with exploration time. It did not test historical knowledge acquisition, long-term retention, later visits or revisiting behaviour. Its findings should not be translated into the claim that VR improves heritage learning.
The sample was predominantly highly educated and included students and researchers. The experiment focused on one Roman archaeological site. The results cannot automatically be generalized to children, older adults, other cultural settings, museums or natural-heritage sites.
The delivery conditions were not identical even though the researchers used the same digital assets. VR required a headset and controllers, while AR and VR360 used tablets. The authors acknowledge that rendering, interaction and visual fidelity had to be adapted to each platform, making complete equivalence impossible.
The publicly available PDF is an Article in Press awaiting final editing. Its methods section describes the intrinsic-motivation measure as a five-point scale, while the table reports means above five. One set of post-hoc comparisons for the user-engagement measure is also inconsistent between the narrative and the table. This review does not rely on those ambiguous values. The final Version of Record should be checked for corrections.
The Goal of the Experience Matters More Than the Device Ranking
The paper does not ask heritage institutions to select one universal winner. It shows that technology choices should follow the experience an institution wants to create.
VR360 may support remote access or help visitors understand a site before arrival. AR can place lost architecture in its original physical context. VR may be more suitable for intensive exploration of spaces that are inaccessible or no longer exist.
The most immersive system is not always the most appropriate system. The success of heritage XR depends less on device specifications than on where visitors will use it, what they need to see and how the technology supports interpretation. The paper’s most useful contribution is not a ranking of three technologies, but a clearer distinction among their roles.
