Productive confusions: learning from simulations of pandemic virus outbreaks in Second Life

Productive Confusions, Learning from Simulations of Pandemic Virus Outbreaks in Second Life

Micha Cárdenas1a, Laura S. Greci c,d; Samantha Hurst c, Karen Garman c, Helene Hoffman c, Ricky Huang b, Michael Gates b, Kristen Kho a, Elle Mehrmand a, Todd Porteous c, Alan Calvitti c,d, Erin

Higginbotham d, Zia Agha c,d

a CRCA, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA, USA 92093-0037; b Calit2, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA, USA 92093-0436; c School of Medicine, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA, USA 92093-0436; d Health Services Research & Development, Veterans Administration San Diego

Healthcare System, 3350 La Jolla Village Drive, CA, USA 92161

ABSTRACT

Users of immersive virtual reality environments have reported a wide variety of side and after effects including the confusion of characteristics of the real and virtual worlds. [11] Perhaps this side effect of confusing the virtual and real can be turned around to explore the possibilities for immersion with minimal technological support in virtual world group training simulations. This paper will describe observations from my time working as an artist/researcher with the UCSD School of Medicine (SoM) and Veterans Administration San Diego Healthcare System (VASDHS) to develop trainings for nurses, doctors and Hospital Incident Command staff that simulate pandemic virus outbreaks. By examining moments of slippage between realities, both into and out of the virtual environment, moments of the confusion of boundaries between real and virtual, we can better understand methods for creating immersion. I will use the mixing of realities as a transversal line of inquiry, borrowing from virtual reality studies, game studies, and anthropological studies to better understand the mechanisms of immersion in virtual worlds. Focusing on drills conducted in Second Life, I will examine moments of training to learn the software interface, moments within the drill and interviews after the drill. Keywords: Second Life, medical, simulation, collaborative, immersion, education, training, pandemic virus

1. INTRODUCTION The Virtual Immersive Platform for Education and Research (VIPER) is collaboration between the Center for Research in Computing and the Arts (CRCA), the (VASDHS) and the UCSD SoM.

1.1 Goals

The main goals for VIPER are to create a collaborative learning environment using a 3-dimensional virtual world that allows participants to learn to work together in medical scenarios. The team’s work to date has focused on drills for emergency preparedness in the case of pandemic virus outbreak scenarios. The drills consist of collaborative experiences in which participants and members of the VIPER team meet in the same physical room, learn to use Second Life and then enact a scenario of a pandemic virus outbreak in Second Life. The goals of the project are multiple, including technological and pedagogical goals. One of the pedagogical goals attempted has been to familiarize users with the Hospital Incident Command System (HICS), which is comprised of a number of roles that must be assigned and performed in the case of an emergency situation. Another goal was to train Emergency Department (ED) nurses on the usage of both Emergency Severity Index (ESI) and START (simple triage and rapid transport) triage on simulated virtual patients. Finally, a primary reason for using a virtual environment for

* [email protected]; phone 1 858.534.4383; transreal.org

Three-Dimensional Imaging, Interaction, and Measurement, edited by J. Angelo Beraldin, et al.,Proc. of SPIE-IS&T Electronic Imaging, SPIE Vol. 7864, 78641C · © 2011 SPIE-IS&T

CCC code: 0277-786X/11/$18 · doi: 10.1117/12.878420

SPIE-IS&T/ Vol. 7864 78641C-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 11/10/2014 Terms of Use: http://spiedl.org/terms

training of multiple participants simultaneously is to allow participants to experience and reflect on their ability to communicate and work collaboratively. The VIPER project builds on theories of collaborative learning that are student-centered and encourage a learning environment that facilitates engagement, group work and the construction of knowledge through interactions between peers and instructors. [7] On the technological level, the goals are to create a platform for collaborative learning that new users can begin to participate in with only a few hours of initial training. Another main goal is to create a system with some degree of immersion, albeit a low degree of immersion. The project seeks to present participants the possibility of acting out their actual roles in emergency situations through virtual avatars in Second Life instead of simply reading about their roles and responsibilities in a table top exercise. Research into immersive learning situations has shown that “immersion in a digital environment can enhance education in at least three ways: by enabling multiple perspectives, situated learning and transfer.” [4] Claims such as these motivate the VIPER group’s research into collaborative training in virtual worlds and are also the focus of this paper. Researchers such as Harrell S. and Harrell D. [8] have proposed complex scales for measuring degrees of immersion in virtual environments with regard to participants’ identification with their avatars. Similarly, other characteristics can be considered for their impact on immersion and the sensation of presence, such as the fidelity and complexity of the virtual environment and spatialized audio, where sounds and voices are displayed based on their 3-dimensional position. [1] Due to the preliminary stage of development of our work, having been only tested with a handful of users during the platform’s development, this paper will look primarily at individual remarks made by participants during the drills in Second Life and afterwards in debriefing interview sessions. These remarks will be used to consider the potential of low cost systems to still provide a degree of immersion in a virtual world and increase a participant’s ability to learn.

1.2 Constraints

The primary constraints for the project are constraints on the usability of the system and constraints created by Second Life.

A main constraint has been the need to create a system that is usable to participants who have never used virtual world environments before or other gaming systems in general. Many of the VA hospital staff we are seeking to train are this kind of user, with little or no experience with either Second Life or 3D gaming. Given the level of user experience, as well as the need to conduct the training and the drill in only a few hours, we have attempted to streamline our usage of Second Life, only relying on a limited set of the functionality of Second Life. Many areas of the virtual environment have also been scripted or modified to work as simply as possible. Lastly, we face a number of constraints presented by the limitations and protocols of Second Life itself. One such limitation is the 15,000 primitive object (prim) limit per region in Second Life. Another is the inability to make a voice chat call directly to another person without cutting out all other sound. Another is the general slowness of activities such as downloading textures or downloading objects as you move from one region to another. The groups’ more recent research has been directed towards moving the platform to OpenSim, an open source derivative of the Second Life server software that works with the same viewer software, in order to overcome these limitations. Yet for the purposes of this paper, I will focus on drills that took place in Second Life.

2. DESIGN OF DRILLS IN SECOND LIFE 2.1 Technological Overview

The VIPER training platform consists of two regions in Second Life. One of these regions contains a virtual copy of the VA hospital on the UCSD campus. This region is frequently at its limit of 15,000 prims, including an ED, multiple stories, rooms for treating patients, a lobby, and a patient education room, just to name a few elements of the hospital. Within the model of the physical architecture are many features to make the simulation realistic, including detailed textures of the floors, plants, vending machines and medical instruments such as heart monitors, ventilators and

SPIE-IS&T/ Vol. 7864 78641C-2


wheelchairs. The second region contains roughly 9,000 prims representing the parking lot for the VA as well as areas for training students to use Second Life.

Figure 1. The VA hospital in Second Life, viewed from the front entrance.

For drills, participants meet in the physical space of the School of Medicine’s medical education computer lab at the University of California, San Diego. (UCSD) The drills begin with an instruction session in which the facilitator teaches the participants how to use Second Life. The training of how to use Second Life then moves “in world” or into Second Life, in a space for learning basics of avatar movement and camera movement. Initial training includes an obstacle course and a “scavenger hunt” where participants find medical equipment throughout the hospital and bring it back to a chosen location, complete with a timer and scoreboard. The scavenger hunt allows participants to orient themselves to the virtual environment using a familiar setting, allowing them to feel empowered by their previous knowledge to begin practicing communication and collaboration.

The participants wear headphones with microphone attachments whenever they are in world, and communicate using the voice chat feature of Second Life. Drills to date have consisted of groups between the sizes of 5 participants and 25. In these group settings, the spatialized 3-dimensional audio becomes very important for managing multiple conversations in world during the simulation of emergency situations.

Construction of the avatars for participants to use as well as patient avatars is also an important and detailed process. Each avatar has their own Animation Overrider to replace the default animations from Second Life with more realistic animations for movements like walking and running, for example. Additionally, the group has developed numerous participant Heads Up Display elements, including one which allows participants to activate particular gestures and another allowing facilitators to send out injects, content messages which add realism and difficulty to the drills.

Patients are represented in two ways, by patient avatars and by objects shaped to look like patients. Patient avatars are customized to present various stages of illness, including skin discoloration and bleeding. The object patients are designed for triage, due to their ease of duplication. The object patients include sound recordings describing their symptoms, notecards with vitals and interactive triage tags that can be changed by clicking on them. A recent training included a line of roughly 100 object patients waiting to be triaged outside of the VA hospital.

SPIE-IS&T/ Vol. 7864 78641C-3


Figure 2. Patient objects with triage tags in Second Life.

2.2 Drill Content

Once the initial training of how to use Second Life has completed, participants move on to informational content sessions before starting the drill. During the recent triage drill, participants watched a video in Second Life about the need for and proper usage of triage tags.

Drills to date have focused on simulations of pandemic flu patient surges and the procedures that should be followed in the event of such outbreaks. The drills begin with the announcement of the situation, and the group begins working. Initial tasks include assigning roles within the HICS structure, and avatars putting on vests indicating the role the have been given. The group then proceeds to enact procedures for dealing with a large-scale patient surge, such as shutting down particular hospital services. As the drill proceeds over the next few hours, injects of new information come in, in the form of text based HUD announcements or verbal exchanges simulating phone calls to the CDC or other agencies which provide information about the viral outbreak scenario.

A recent drill included two groups of participants, one group simulating the HICS command roles and the other, a group of ED nurses, performing the triage of a large surge of patients. The simulations include both essential small-scale procedures, such as wearing gloves, masks and other protective gear, and large-scale procedures as well. One benefit of the virtual training is that large scale actions, such as the evacuation of a hospital or the clearing of the hospital parking lot to allow tents to be set up, can be enacted by participants without the extreme impact such a simulation would have on the real hospital.

2.3 Challenges

Throughout the development of this platform, we have encountered a number of challenges, some of which seem technologically solvable in the short term and others, which do not. Based on my observations of an actual drill in the VA hospital, where staff that were learning and practicing HICS roles met face to face, not in Second Life, the main challenge to training in virtual worlds is the lack of emotional response. The technology of Second Life is not at the point where avatars mimic facial gestures. While voice chat does provide a good indicator of emotional tone, the real drill still contained an emotional intensity that is totally lacking in virtual world drills. The furrowed brow and intense gaze of participants in the real drill making decisions about hospital functioning during a pandemic flu outbreak has no virtual mirror in the simulations the VIPER group has created. Given the importance for this level of learning in response to emergency events, this is the main challenge today. While Linden Labs has demonstrated technology for tracking and mimicking facial gestures in Second Life, it remains to be seen how far this feature is from daily use. [10]

SPIE-IS&T/ Vol. 7864 78641C-4


Another primary challenge is the use of voice chat and its limitations. Again based on real life drills in the VA hospital, people engaging in drills of emergency situations often use multiple simultaneous levels of conversation, easily managing cell phone conversations, walkie-talkies and conversations at one’s immediate table and at the level of the entire room. Communication in Second Life is far more simplistic and difficult to manage. While Second Life has spatial audio in 3-dimensions, so that avatars behind one sound like they are in that position and farther avatars are quieter, the reduction of the sound into headphones seems to level out the sounds until they are all at similar volume levels. The VIPER group continues to try to find ways to enable participants to be able to use the kind of flexible conversation navigation techniques they do in real life settings.

3. RESULTS – PARTICIPANT RESPONSES 3.1 Methods for Studying Virtual Worlds

In order to look at the possibilities of immersion with simple desktop systems running Second Life, a number of moments in the drills will be examined. These moments are in a way anecdotal, not numerical averages, but the goal of this paper is not to prove in a scientific manner specific claims about virtual environments. In contrast, this paper attempts to look at the pedagogical, cultural and social implications of using virtual worlds as collaborative learning environments. Contemporary cultural theorists are attempting to find new methods for understanding emerging media environments such as virtual worlds. While feminist theorists such as Jane Gallop has upheld the value of the anecdote for understanding pedagogy, game studies scholars including Marinka Copier have questioned the application of scientific paradigms to media studies, proposing a “hybrid-hybrid” researcher-participant which calls for a blurring of boundaries between player and academic to better understand online role-playing games. [6, 3] Tom Boellstorff is an anthropologist who has conducted a participant-observer ethnography of Second Life, in which he questions basic assumptions about the notion of immersion and its application to virtual worlds. [2] He builds on the writing of Salen and Zimmerman who name the “Immersive Fallacy” as the notion that immersion happens based on the ability of technology to create a realistic illusion. They claim that instead, immersion “is an engagement that occurs through play itself.”[10] This claim, which lines up with collaborative learning theories well, is applied by Boellstorff to Second Life when he says “In virtual worlds, ‘virtuality’ refers to sociality, not the senses.”[2] Additionally, Boellstorff links presence and immersion and finds support in Žižek’s claim that “the ultimate lesson of the ‘virtual reality’ is the virtualization of the very ‘true’ reality: by the mirage of ‘virtual reality’, the ‘true’ reality itself is posited as a semblance of itself, as a pure symbolic construct”.[2] Given these two views that the effect commonly referred to as immersion is a result of a complex interplay of the real and the virtual, of real social interaction mediated through virtual worlds, then the importance of examining moments of the confusion of the two, mixing the real and the virtual, for understanding how virtual worlds work, becomes clear. While previous studies of immersion in virtual reality have focused on the physical effects of adaptation by muscular, ocular, neural and proprioceptive systems, often resulting in cybersickness and cyberadaptation syndrome, an understanding of the mechanisms that create a feeling of immersion or presence in a virtual world need other methods. [9,11]

3.2 Participant Confusion of Response to Drill

In the drills that the VIPER group has conducted, a number of participants have demonstrated behaviors that indicate the confusion of the boundaries between the real and the virtual in different ways. The prime example of this occurred during a HICS drill simulating a pandemic flu patient surge. A participant enacting one of the HICS roles needed to inform another participant in the drill of the number of remaining supplies. The numbers of supplies were being regularly updated on a board in the HICS command room in Second Life. Under the pressure of time to respond, instead of turning his avatar around in the virtual space to see the board, this participant turned his actual head around to look at the back of the room, expecting to see the number of supplies. Throughout the drill, this participant had remained very engaged in the simulation, at one point breathing heavily into his microphone when his avatar was running. And yet this moment of turning his head around, his level of immersion in the drill was so great that he confused his own muscular response, responding in reality in a way that would slow his ability to respond within the virtual emergency scenario. The social reality of the drill was so great, that his reactions became confused, his signals mixed. Conscious that the scenario was only a drill, on a pixelated screen, and not an actual viral outbreak, the participant was able to accept the fictional scenario through social interactions with a facilitator and his peers to the degree that the unrealistic avatar onscreen became confused with his actual physical body, even if for just a moment.

SPIE-IS&T/ Vol. 7864 78641C-5


3.3 Strong Feelings Toward Avatars

Numerous interactions with avatars demonstrate moments of the confusion of realities, in ways surprising to the VIPER group because of their apparent insignificance given the gravity of the simulated situation. Since researchers have demonstrated that in only 90 seconds, users become affected by their avatars, these moments may come as no surprise, but still reveal the mechanism discussed here. [5] In numerous debrief sessions after drills, participants have made comments such as “It was nice to have hair again” or “It was nice to be skinny”. Looking at Harrell S. and Harrell D.’s dimensions of stances towards avatars, the usage of the language “to be” implies a strong leaning toward the ontological status of the avatar being a “Mirror of Real Self”. These participants in the drills do not describe their avatars as characters on a screen separate from their own identities, but as a mode of their own being. Perhaps the intensity of the scenario in a social context of peers who could believably inhabit that scenario produced more identification with avatars, in contrast to producing avatars that were simply instrumental to complete the task. Participants’ strong identification with their avatars is also expressed in other ways that may provide more insight. Numerous participants, male and female, have made statements expressing that they feel that the female avatars are excessively sexual. While the VIPER developers have attempted to find avatars and animations that are not excessively sexual, all of the animations and clothing purchased in Second Life are still seen by participants as over sexualized, and the only option is to create new animations from scratch. As an example, in the image below, numerous participants noted how the female animation overrider included a standing pose in which the avatar’s chest was thrust forward. In addition, one participant stated that she would not come to work to deal with an emergency situation wearing a mid-riff shirt. This discomfort, arising from a close identification with one’s avatar, may indicate one reason for this feeling of immersion. The participants in the VIPER group’s drills are in a work environment. For them, the appearance of their avatar becomes a question of agency and their ability to function as command decision makers. They are interacting with peers who for the period of time of the drill see them as their avatars, and this moment of being seen by their peers creates a feeling of immersion through a deep engagement with one’s avatar.

Figure 3. Avatars of drill participants in a pandemic flu surge triage drill with patient objects in background.

3.4 Immersion and Authenticity

In numerous cases, participants express a desire for authenticity, or fidelity to real world experiences. Users want to draw parallels between reality such as how one dresses, looks and acts in real life versus in Second Life. Authenticity of the simulation, based on participants’ everyday experiences from reality, seems to be an important element in creating immersion. Authenticity matters to participants across multiple domains including the authenticity of physical surroundings, the authenticity of scenarios, and the authenticity of roles. This was a main motivation for modeling the VASDHS hospital to scale with as much fidelity to the real VASDHS hospital as possible. For example, the triage

SPIE-IS&T/ Vol. 7864 78641C-6


medical scenario in one of the drills ended prematurely without closure for each patients care and the nurses were very upset. They wanted to complete the medical triage by treating patients, including giving medicine or IV fluids, as they would perform in real life, despite the fact that this was outside of the scope of the particular simulation. In another example, HICS trainees want to perform the real life HICS role they perform in the VA. Repeatedly, participants express a desire to enact the social situations they are familiar with from their everyday experiences, and elements perceived as lacking authenticity hinder feelings of immersion.

3.5 Willingness to Step Out In Order To Be More Immersed

Another participant behavior involving avatars leads to the final example of user feedback regarding feelings of immersion in Second Life. Due to time constraints, very little time is devoted to teaching participants modes of customizing their avatars. While participants can choose their outfits from a limited selection, that is all the customization that is explained. Still, some participants feel so strongly about their avatar appearance, that they decide to step out of the rules of the simulation to do so. The third mode of mixing realities that can indicate a level of immersion is the willingness to step out of the rules of the simulation in order to more effectively act within the simulation.

Figure 4. Avatars in editing appearance stance in the middle of a group listening to scenario description.

A number of different participants have demonstrated this level of commitment to the simulation, a willingness to step out in order to be more immersed upon stepping back in. In the image above, a participant is modifying her appearance, changing her hair and clothing, which is easily identified by the stance the software puts avatars in during this mode. While this method of editing one’s avatar is not taught during the lesson, this participant took the initiative to leave the simulation during the description of the scenario in order to make her own comfort level with her appearance acceptable to her. In another case, a psychologist was asked to deal with a patient avatar being performed by a volunteer who was displaying symptoms of mental illness. After attempting to have a conversation with the patient, and being unable to move her camera in such a way that she could look at the patient avatar in the face, she became so frustrated that she stopped the simulation, took herself out by taking off her headphones and asked for help. Only once she could look directly at the patient avatar was she satisfied that she could do her job. The social and emotional need the participant had to engage with her patient was so great that she refused to settle for the simulation to be lacking the visual element of her looking the expressionless patient avatar in the face. During a triage drill involving ED nurses, they collectively made a decision to break the rules of the simulation. Due to documentation needs of the study, and to protect the privacy of participants, avatars are given made up names that do not correspond to the participant’s real name, only matching their gender. After learning how to use Second Life’s interface, once the actual drill began and the handful of emergency staff were faced with triaging over 100 patients, they decided to eschew the fictional names and use their real first names

SPIE-IS&T/ Vol. 7864 78641C-7


instead. The group, containing one participant who stated that they were an avid World of Warcraft player, decided that the rule of the simulation regarding names was in conflict with their ability to immerse themselves in the situation to the level they needed to succeed in the drill. In order to maintain the speed of communication necessary to accomplish the fictional task, these participants decided to change the rules. What is demonstrated here is a commitment to the pedagogical goals shared by the group and to the demands of the imagined scenario above the real world constraints placed on the participants by the study.

4. CONCLUSIONS This paper has described a number of instances in which users of the virtual world Second Life engage in mixing the real and virtual in ways that indicate a level of immersion. Further still, these instances not only indicate degrees of immersion in the traditional sense used to describe virtual reality environments, but also qualitatively different aspects of immersion which may be said to be characteristic of virtual worlds. Through observations gained my experience as an artist/researcher working with the VIPER group, I have described user feedback from training drills of pandemic flu outbreaks simulated in Second Life. In these training sessions, users demonstrate a number of ways of mixing the real and virtual, by confusing the two and responding with misplaced physical gestures, by strong emotional responses to the appearance of their avatars and by demonstrating a willingness to consciously navigate the boundaries of the simulated scenario in order to make their level of immersion deeper.

REFERENCES

[1] Brooks, Jr., F., “What’s Real About Virtual Reality?”, IEEE, 1999 [2] Boellstorff, T., [Coming of Age in Second Life: An Anthropologist Explores the Virtually Human], Princeton

University Press, 113-117 (2010). [3] Copier, M. “The Other Game Researcher” In: Copier M. and Raessens J. eds. Level Up! The proceedings of the

Digital Games Research Conference at Utrecht University (2003). [4] Dede, C., “Immersive Interfaces for Engagement and Learning,” Science Magazine, 66, (2008) [5] Dell, K., “How Second Life Affects Real Life”, Time, 2008,

http://www.time.com/time/health/article/0,8599,1739601,00.html [6] Gallop, J., Anecdotal Theory, Duke U. Press. Books (2002) [7] Greci, Laura et al., “Pandemic Flu Patient Surge Planning and Practice in a Virtual Environment,” (2010) [8] Harrell, S. and Harrell, D., “Exploring the Potential of Computational Self- Representations for Enabling Learning:

Examining At-risk Youths’ Development of Mathematical/Computational Agency”, Proc. Digital Arts and Culture Conference, (2010)

[9] Lackner, J. and DiZio, P., “Virtual environments and the cyberadaptaion syndrome”, Science Direct, 1-2 (2004). [10] Lowensohn, J., “Linden Lab demos hands-free interface for Second Life”, http://news.cnet.com/8301-10784_3-

9916946-7.html CNET news (2010). [11] Salen, K. and Zimmerman, E., [Rules of play: game design fundamentals] MIT Press, Cambridge, 450-451 (2004). [12] Stanney, K. and Gavriel, S. et al. “Aftereffects and Sense of Presence in Virtual Environments: Formulation of a

Research and Development Agenda”, Seventh International Conference on Human Computer Interaction, 8-11 (1997)

SPIE-IS&T/ Vol. 7864 78641C-8


Productive confusions: learning from simulations of pandemic virus outbreaks in Second Life

Documents