| dc.description.abstract | Augmented reality (AR) is the combination of virtual content with the real world through overlaying virtual objects in 3D space. An inherent challenge with AR is building frameworks for and evaluating the quality of users’ experiences. One dominant framework, presence, has led to multiple models on how and why users perceive virtual content as real. In this psychophysical study, we compared the relative importance of different factors in AR by manipulating interaction levels, physics, and shadows. We built a custom AR environment that allows participants to adjust these settings and analyzed their choices during a transition activity where participants were told to upgrade certain qualities based on their perception of the virtual content. Participants also performed budgeting tasks where they were given a limited number of upgrades to the environment in order to help rank which factors were most important. Lastly, we collected questionnaire responses on how participants ranked different factors and categories along with short responses to provide qualitative evidence. From the collected data, we drew multiple conclusions about presence in AR. We found that realistic user interaction was the most important factor for increasing plausibility. The real-world reference setting of the environment also had a strong impact on participants’ choices regarding plausibility. Lastly, we showed that high levels of realistic physical behavior was more important than graphical appearance. These conclusions support bottom-up processing in AR because users built their model of plausibility by first setting their reference frame to the real-world, followed by turning on a baseline perception of the objects by being able to grab them and, lastly, improving higher-order cognitive queues like matching the physics of the virtual objects to those of the real-world. | |
