Linking continuous behavior to aesthetic enjoyment in a walkable virtual-reality museum tour: effects of agency and a painting-level analysis framework
This study investigates the impact of user agency and continuous behavioral metrics on aesthetic enjoyment in a virtual reality museum tour, finding that while agency significantly alters viewing patterns, it does not reliably increase self-reported enjoyment, and gaze engagement during audio guides serves as the strongest predictor of liking among behavioral measures.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
For decades, museums have operated on a simple, unspoken contract: the curator decides the path, and the visitor follows it. You enter a gallery, and a pre-set route guides your feet from one masterpiece to the next, with the assumption that this curated journey is the best way to appreciate art. Yet, a long-held belief in psychology suggests that having a choice—whether to turn left or right, to linger or move on—makes us feel more engaged and satisfied. As museums increasingly move into digital spaces, a new question arises: does giving people the freedom to choose their own path in a virtual world actually make the experience more enjoyable? This question sits at the intersection of how we perceive beauty and how we move through space. Researchers have long known that where we look and how we walk can tell us what we find interesting, but measuring these subtle behaviors in a real museum is difficult. People move differently when they know they are being watched, and standard computer screens cannot capture the full body movement of a person walking through a room. Virtual reality offers a unique solution, creating a digital space that feels real enough to walk through, yet precise enough to track every step and glance with scientific accuracy.
A team of researchers at Tel Aviv University and Panthéon-Sorbonne University decided to test these ideas by building a life-size virtual museum. They invited forty-eight adults to walk through an eight-meter by four-meter digital gallery containing seven famous paintings, each accompanied by a synchronized audio guide. The goal was to see if the way people moved and looked at the art could predict how much they enjoyed it, and whether giving them the power to choose their route changed that enjoyment. To do this, the researchers created three different versions of the tour. In one version, visitors had to make a choice before every single painting, selecting a feature they wanted to see next. In another, they made choices for only the first three paintings before the route became fixed. In the third version, the visitors simply followed a pre-set path with no choices at all. Crucially, the actual paintings and the order in which they appeared were identical for everyone; the only difference was the feeling of having a choice.
The researchers equipped the visitors with headsets that tracked their eye movements and head position fifty times every second. This allowed them to reconstruct exactly where each person stood and what they were looking at for every moment of the tour. They broke the experience down into specific viewing episodes for each painting, measuring how long people stood in front of the art, how often their eyes jumped from one spot to another, and how much of the time they kept their gaze fixed on the canvas while listening to the audio guide. After the tour, the participants rated how much they liked the experience as a whole and how much they liked each individual painting.
The results revealed a surprising disconnect between what people felt and how they behaved. The researchers found that the level of agency—the amount of choice a person had—did not change how much they reported enjoying the tour. Whether a visitor made choices for every painting or none at all, their satisfaction ratings were statistically the same. The feeling of freedom did not translate into a more enjoyable experience in this setting. Furthermore, the researchers discovered that the way people moved and looked at the art was a poor predictor of how much they liked it. The distance they walked, the speed at which they moved, and the total time they spent standing in front of a painting did not reliably tell the researchers whether the visitor liked that specific artwork. The only behavioral signal that showed a faint connection to liking was the percentage of time a visitor kept their eyes on the painting while the audio guide was playing. Even this link was modest, suggesting that while looking at the art while listening might help, it is not the sole driver of enjoyment.
What the study did find, however, was that the act of choosing did change how people behaved, even if it didn't change how they felt. Visitors who had to make choices spent more time in the virtual gallery, largely because they had to stop and answer questions. More importantly, their patterns of looking and standing were different. Those with more choice tended to stand in front of the paintings for different amounts of time and looked at the art in different ways compared to those on the fixed route. This suggests that while giving people control alters their physical engagement with the space, it does not necessarily make the experience more pleasurable. The study concludes that in a virtual museum tour, the micro-structure of how we look and move is distinct from our subjective enjoyment. We can change the way people walk and look by giving them choices, but that does not guarantee they will like the art more. The research establishes a new method for studying these behaviors, proving that virtual reality can capture the complex, continuous dance of a museum visit without the need for a physical gallery, offering a powerful tool for understanding how we experience culture in the future.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.