Note: See Part 1 for an intro to virtual worlds
Avatar design is one small piece of the bigger puzzle that is the design of the environment as a whole. Bente’s study of 150 undergraduates which compared avatar, text, audio, and video communication showed that while different communication platforms had an effect on users’ experience and behavior, the virtual world environment showed no significant difference from the video chat environment in terms of reported levels of social presence, trust, or user satisfaction. Bente also concludes that the visual channel in the avatar and video environments was “mainly used for first impression checks instead of non-verbal interaction regulation.” (310) Other studies, though, report higher levels of satisfaction and co-presence presence with virtual world communication than with video chat. Out of these, one by Tapsis also makes the case that virtual world communication is a “more natural communication medium” than video chat (2013, p.439). What might account for these contradictory results? One factor is surely the design of the virtual world.
Looking back at Bente’s study, it seems that one of its limitations is that the virtual environment under consideration in the study is quite bare bones, with little in terms of a compelling, memorable graphical space in which the avatars interact. In other words, it’s not surprising that users of this MUVE didn’t look at the screen much, as there wasn’t much to look at. Given other research, it is likely that the study would have borne different results had the MUVE been designed differently. A recent paper marshalls significant evidence, for instance, that it it is precisely the “eye-catching”, “entertaining”, and “realistic” visual features of an environment that foster presence and motivation (Yilmaz, Topu, Goktas, 2013, p. 824). Clouding the issue, though, is an article by Lee (2009) that suggest that the “features and intricacies” of the 3D environment “may work to the educator’s disadvantage, distracting or discouraging students from attending to the key conceptual tasks in a collaborative learning activity” (p.156). That word “realistic” and also the word “realism” crop up frequently in the research, and deserves further attention.
As one example, Bulu et. al. (2012) examine the relationship between different kinds of presence and their influence on user satisfaction and immersion. The suggestion offered by this small study is that there is a positive correlation between presence and co-presence, and that the design of a virtual world, since it can lead to feelings of immersion, plays a key role in fostering feelings of presence. One of their assumptions is that the “realism” of the virtual world, as depicted by its design, makes immersion possible. Again, an elaboration of what the authors mean by “realism” would be helpful, though, since the term within the visual arts and other media, has a long, variegated history. For example, realism is a style of painting in which figures and scenes are depicted as they are experienced or might be experienced in everyday life, with an emphasis on the everyday life of regular people and a de-emphasis of the supernatural and fantastical. Under this definition, a virtual world with “realism” could plausibly be envisioned existing in “real-world”, physical space. The same term, though, is also used to describe the rich and detailed representation in art of visual appearances. This can evidently include appearances of things that the artists haven’t actually seen in physical space, as paintings of winged angels are classified as examples of Realism, presumably due to their convincing portrayal of what those things would look like, were we to see them. In this way, a dragon could be described as “realistic” if the viewer though it was depicted in such a way that made it seem real, despite the non-existence of dragons.
Guy Merchant (2009) reflects on this issue in a paper about a MUVE which he helped to develop called Barnsborough, a 3D virtual world for literacy learning. Teachers and educational consultants from Barnsley began this project in 2006. In Barnsborough, users occupy avatars and co-create narratives with their peers by reading, discussing, and writing about the cryptic clues scattered around the environment. Merchant makes the important point that the design process is an act of simultaneous creation and control. For the design of Barnsborough, the team decided on a contemporary urban environment that felt “slightly run down”, but with glimpses of affluence. One of the justifications for this is that it would “partly reflect pupil’s everyday reality” (43). These decisions are based on what the adults imagined the children would respond to best. Rather than make a virtual world that corresponds visually to conventionally designed learning spaces, this team asks a different question: what world would children most like to inhabit (43)? Admittedly, I wonder how this team came to the conclusion that children would most like to inhabit a world modeled after their regular daily world.
Bulu et. al. mention the use of details like three dimensions, texture, light, sound, and motion, which “can bring realism” to the virtual world environment” (156). These same details, though, can be used to show fantastical worlds and which bear little relation to physical reality or everyday life, but which are shown in a convincing way (e.g. “that really looks like a table resting on the ocean surface!”). Some designers of learning spaces choose not to embrace any kind of realism, opting instead for highly abstract environments. A paper by Prasolova draws attention to a growing amount of virtual places, some of which are highly abstract constructions, that “only to a limited degree resemble any real ones”.
A paper by Minocha and Reeves, one of the few studies that takes the design of 3D virtual worlds for learning purposes as its primary subject, tackles this subject as well. The study looks at qualitative responses from educators, students, and designers and compiles some suggestions about MUVE design. Since a MUVE can be designed like a replica of actual physical space or a highly abstract, imaginative fantasy world, what are the benefits or downsides of “visual realism” in these environments (113)? Her responses show that many teachers opt for “realism” under the impression that virtual learning spaces looking like their familiar physical counterparts will help relate to users’ expectations and models from the physical world, helping new users adjust. One designer finds that in his experience only a slim minority are actually excited about the possibility of abstract, more fantastical environments.
To shift away from this issue of realism for a moment, note again that Bente’s study also describes successful MUVE designs as “entertaining” and “eye-catching” alongside “realistic”. When someone designs a learning environment based on what will be “best” for the learner, they could be thinking in terms of what the teacher wants learners to know, or in terms of what will capture learners’ attention and interest. Given that there are far fewer design limitations than in physical space, it is possible to make quite captivating and detailed 3D learning environments. Possibility is not a reason, though, and more research needs to be done on whether the richness of an environment can detract from the quantity or quality of conversation within it.
An example of a successfully eye-catching, entertaining visual space is found in a MUVE hosted by Appalachia State University called the AET Zone. Bronach et al. (2008) draw attention to one of the more popular spaces within the AET environment, the So What? Saloon, which is modelled after an old-fashioned saloon (certainly not something from everyday life), replete with wood furnishings, a player piano, and bottles. If a user hovers his or her mouse cursor over a bottle, a contested question in the field appears, along with a form through which the user can attempt an answer. Interacting with the player piano reveals a navigable database of user-submitted answers. The environment is a popular conversational hub for users. One assumption here seems to be that the visual attractiveness of this simulated saloon is a crucial factor that contributes to its success. If the same functionality existed in a 3-D space that looked more conventional or bland (an uncolorful office, for example), and instead of bottles and a piano the user had to click on different file cabinets, for instance, would this other space be as popular? What about the studies that demonstrate the distracting aspects of entertaining visuals?