A text-based networked virtual environment represents to a user a system of rooms joined by exits and entrances. When navigating this system of rooms, a user can communicate in real time with other connected users occupying the same room. Hence, these virtual environments are aptly suited for networked conferencing and teaching. Anecdotal information suggested that some people feel a sense of "being there" or presence when connected to one of these environments. To determine how many people feel this sense of presence, we surveyed 207 people from 6 different groups of users of text-based networked virtual environments. The results indicated that 69% of these subjects felt a sense of presence. Experiments with people in text-based networked virtual environments may be helpful in understanding the contribution to presence by social interaction in other virtual environments.
It has been suggested that degree of presence in a communication medium is related to two factors, vividness of the environment and interactivity, the degree to which users may influence the form or content of the mediated environment (Steuer, 1992). Vividness of the environment is a crucial factor in immersive environments where communication involves a variety of sensory stimuli. Interactivity includes such things as social presence, or the degree to which the environment contains other people who are reacting to you (Heeter, 1992). Our experiences (Towell, Hansen, Mercer, Leach, Rubin, Prilusky, & Glusman., 1995; Towell & Towell, 1995; Hardy, Robinson, Doughty, Findsen, Towell, Towell, & Wilson, 1996), those of Schiano (1997), and other anecdotal expositions suggest that many people feel a sense of presence when experiencing the primarily social environments of text-based networked virtual environments (TNVEs), which are also commonly referred to as MUDs (multi-user dungeons/dimensions). However, we are unaware of any objective attempts to measure how many people experience this sense of presence. To assess the potential of TNVEs as a medium for studying the interactive or social aspects of presence, we have studied six different groups of TNVE users. This manuscript reports survey results from 207 TNVE users and indicates that a sense of presence was reported by 69% of those connected to these virtual environments.
2.1 Text-Based Networked Virtual Environments
Emergence of the Internet has facilitated the use of TNVEs as conferencing tools (Anderson, 1994; Evard, 1993; Towell, Hansen, Mercer, Leach, Rubin, Prilusky, & Glusman, 1995; Towell & Towell, 1995; Hardy, Robinson, Doughty, Findsen, Towell, Towell, & Wilson, 1996). A TNVE consists of a server that provides information to clients about a database that it also maintains. The database contains information about a collection of objects that can be characterized as rooms, things and other connected persons. The information in the TNVE database is organized in such a manner that the person using the client is presented with a textual representation of a room in which there could be other people with whom he or she may "talk." Talking in this context means using the keyboard to communicate with the other person(s) who share the same room with you; likewise, "hearing" means to see textual communication on a monitor. Typically, TNVEs are compartmentalized into rooms joined by entrances and exits. Hence, there is a topography which can be navigated with simple commands such as: "go south." When people move from one room to another they lose the communication they had with the person(s) in the previous room and can "hear" and "talk" only with those in the new room. This presentation of a virtual space to the TNVE user has been referred to as "imposing a spatial metaphor."(Evard, 1993). In addition to the communication and navigation properties, a TNVE is extensible and dynamic. Specific commands enable users to create their own rooms and objects in such a manner that the environment becomes a medium of self-expression. The objects respond to commands from the users (or other objects) and hence provide a dynamic environment that interacts with its inhabitants. Examples of these interactive objects are recording machines, blackboards, and slide projectors. The type of TNVE that was used in the studies reported here is referred to as a MOO (Curtis, 1992) which stands for Multi-user dimension Object Oriented. MOO not only indicates the type of TNVE but also the programming language used to extend the virtual space from within it.
The subjects were from six different groups of TNVE users. Group I (N = 53) consisted of scientists who had recently attended an Internet-based scientific conference (Hardy, Robinson, Doughty, Findsen, Towell, Towell, & Wilson, 1996). This conference utilized a TNVE (Towell & Towell, 1995) for scheduled real-time discussions of papers that were concurrently available via the World Wide Web. Over a four week period, 80 people used the TNVE to participate in discussion sessions. While connected to the TNVE, each person was identified by their real name and only other scientists registered for the conference were present. Groups II (N = 16) and III (N = 18) consisted of students at Northern Illinois University who had participated in a one hour laboratory exercise which consisted of a brief introductory lecture followed by a "hands-on" experience in a TNVE. Only students with no previous TNVE experience responded to a survey which was sent via e-mail following the laboratory exercise. All students connected to the TNVE were identified by their real names and they shared the virtual space with no one other than their classmates and the instructor. Groups IV (N = 40), V (N = 40), and VI (N = 40), consisted of subjects registered at an Internet-accessible public TNVE, called LambdaMOO. Groups IV-VI were sampled because they represented people who were motivated to use the medium for reasons other than those in Group I, the scientists who were using it for scientific discussions, or those in Groups II and III, the students who were required to use it. Group IV consisted of subjects selected randomly from a list of all the people who were connected to the TNVE during the sampling period. Groups V and VI consisted of subjects who were also selected from those currently connected but who had not been registered for longer than a week. The real identities of the users from Groups IV, V, and VI were unknown.
As discussed in Methods, Groups I-VI had experienced a TNVE that was of the MOO type. Hence, being connected to such an TNVE was referred to as being "connected to a MOO." Instrument I consisted of the following question:
Please indicate the strength of your agreement with the statement below where a "1" indicates you strongly agree and a "5" indicates you strongly disagree.
I feel a sense of actually being in the same room with others when I am connected to a MOO.
_____1. strongly agree _____2. moderately agree _____3. neither agree or disagree _____4. moderately disagree _____5. strongly disagree
Instrument II was designed to pose the same question to the user but in a manner whereby the subject would have to disagree with the statement to yield the same results as Instrument I. Hence, Instrument II was identical to Instrument I except for the statement which read:
I *do not* feel a sense of actually being in the same room with others when I am connected to a MOO.
The purpose of Instrument II was to investigate potential bias resulting from a tendency to agree with positive assertions.
Subjects in Groups IV, V, and VI were queried by direct communication while they were connected to the TNVE. They were sent a message which stated:
Could I trouble you to answer one survey-type question about MOO's?
If they responded affirmatively, they were sent either Instrument I (Groups IV & VI) or Instrument II (Group V) in a manner such that it appeared on their screen. Over 97% of the responses were subsequently returned via a similar messaging technique. Subjects in Group I were e-mailed the survey question (Instrument I) at the conclusion of the scientific conference; all subjects responded to the survey via e-mail. Similarly, subjects in Groups II and III responded to Instrument I, which was e-mailed, or Instrument II, which was presented directly to them,respectively, following their TNVE experience. Only subjects who had not been in a TNVE prior to the class were asked to respond to the survey.
A nonparametric analysis (median test) of the ordinal data (Table I) was performed to test the hypothesis that data from Groups I-VI had identical distributions of ratings. For this analysis the data from the groups receiving Instrument II was altered to represent its mirror image. Hence, in Groups III and V, all 5's were converted to 1's and all 4's were converted to 2's. The six sample groups were combined into a single distribution and the grand median, Md, for the sample was obtained. Thus, if the above hypothesis was true, then it could be assumed that any bias due to the positive assertion of the survey statement in Instrument I could not be considered significant in this study.
In the following discussion the term "agreement" will be used to indicate either a result of 1 (strongly agree) or 2 (moderately agree) in groups receiving Instrument I, or 5 (strongly disagree) or 4 (moderately disagree) in groups receiving Instrument II. The responses for all the groups are shown in Table I along with that percentage of the group that indicated "agreement." The responses from Group I (scientists) indicated "agreement" in 74% (the percentage of those answering with either '1' or '2') of the subjects. Of the 53 who responded, 36 included comments and 44% of these comments (16) made some reference to the distress caused by Internet lag which hindered the conference. Group I was the only group with more '1' (strongly agree) responses than '2'(moderately agree). Group II (students) indicated "agreement" in 88% of the subjects but 86% of these responses were of the '2' rating. The responses from Group IV indicated that 60% answered with either a '1' or a '2' but, as in Groups II & VI, most (79%) were at the '2' level. Responses from Groups III and V (marked with asterisks in Table 1) which were administered Instrument II, indicated that the reversal of the Instrument I statement resulted in reversal of the ratings and indicated "agreement" in 56% and 58% of the subjects, respectively. Several comments from Groups IV, V, & VI indicated that the feeling of presence was not static but dependent on whom they were with.
Analysis of the data in Table I indicated an Md of 1.75 and when each rating was compared to the value of Md a chi-squared = 8.53 resulted. Thus, the distributions from Groups I-VI were indistinguishable and hence the results were combined and depicted in the form of a histogram (Figure 1).
Before considering the implications of the studies reported here, it is necessary to briefly discuss the terminology of presence. Steuer (1992) has defined presence "...as the sense of being in an environment." Telepresence was then defined "...as the experience of presence in an environment by means of a communication medium." Thus, the sense of presence reported here by subjects connected to TNVEs could be referred to as "telepresence" according to Steuer (1992). Sheridan (1992) intended "telepresence" to apply to environments involving teleoperation and "virtual presence" as "...feeling like you are present in the environment generated by the computer." Thus, according to Sheridan (1992), the sense of presence reported here may be referred to as "virtual presence." To avoid this terminological conundrum, we refer to the feelings of "being there" experienced by the subjects studied here as simply presence in a TNVE. Also, TNVEs have been referred to as a type of virtual reality (Curtis, 1992). With the use of Steuer's (1992) definition of virtual reality, which is defined as "a real or simulated environment in which a perceiver experiences telepresence," it would be proper to refer to a TNVE as a type of virtual reality. In partial agreement with Ellis (1996) who considers the term virtual reality an "oxymoronic buzzword," we avoid its use here. However, we believe the term text-based virtual reality may be more readily comprehended (as opposed to text-based networked virtual environment) by a lay audience.
The instruments used in the studies reported here were intended to assess the degree to which the subjects felt presence in the TNVEs. It is important to consider the shortcomings which may have resulted from the way the instrument was worded. In the statement: "I feel a sense of actually being in the same room with others when I am connected to a MOO," the inclusion of "with others" in the question may have biased the subject towards the social aspects of the environment. Hence, one must consider the possibility that the instruments were assessing the feelings of "being with" instead of "being there." However, as noted in the introduction, interactive features and social presence may be contributory to presence and in this sense one may be justified in characterizing our results as having tested for feelings of presence. Nevertheless, it leaves unanswered the question of whether there is presence contributed by interaction with objects in the virtual environment and/or by the spatial metaphor of the TNVEs.
Recent studies have focused on the complexities of measuring presence in virtual environments. Many excellent expositions on presence are intended for immersive environments (Held and Durlach, 1992; Slater & Usoh, 1993; Slater, Usoh and Steed, 1994; Schloerb, 1995; Sheridan, 1996; Ellis, 1996) and are not relevant to TNVEs. This is because the focus is on perception of and interaction with the environment primarily through sensory experience and not on textual communication with other objects, including humans, within the environment. An example of this is found in Robinett's (1992) proposal of a taxonomy for the classification of technologically mediated experience which excluded verbal descriptions.
As introduced earlier, social presence refers to the degree to which the environment contains other people who are reacting to you (Heeter, 1992). Of the 207 subjects surveyed in the studies reported here, no one commented on the spatial metaphor of the TNVEs as contributory to presence. This is puzzling in that it has been shown that readers of narrative scenes of spatial environments construct a spatial mental model that "consists of extensions of the three model axes, and they associate objects to it." (Franklin, Tversky & Coon, 1992). One might hypothesize that this construction of spatial mental models would be contributory to feelings of presence. Some subjects indicated that the sense of presence was dependent on whom they were with in the virtual environment and, for the scientists, their interests and involvement in the topic of discussion. Social presence and interactivity are most likely related to the "dramatic presence" discussed by Kelso et al (1993) in their studies of interactive drama. People who interact with a virtual world inhabited by "dynamic and complex characters" are referred to as "interactors." An interactor who isn't required to keep to a script is called "highly interactive." In the TNVEs studied here, the subjects may be likened to the Kelso et al (1993) "interactors" wherein all are "highly interactive" by virtue of the absence of any scripts. Interactivity and its relationship to presence is also discussed in Fontaine's studies of the heightened absorption experienced by people in intercultural or international encounters (Fontaine, 1992). He considers the concept of "flow" as "...a state of consciousness occurring when we are challenged by tasks in which equilibrium exists between our high skill and the task's high difficulty." This "flow" is suggested as a source of "immediacy" of presence and is perhaps related to the sense of presence in TNVEs, particularly in the Group I (scientists) whose discussion sessions were intense, complex and energetic.
The purpose of this study was to objectively assess how many people in TNVEs feel some degree of "being there" or presence. Out of 207 subjects from six different groups, 69% indicated that they experienced a sense of presence in TNVEs. This presence may be relevant to studies of presence in immersive environments particularly in the area of social interaction within virtual environments. Further experimentation is needed to determine the relative contributory effects of social interaction, the spatial metaphor, and other factors (i.e.., psychological profile of the user), on the degree of presence experienced in TNVEs.
The authors thank Pavel Curtis and the MOO-cows community for a lot of help and a marvelous MOO-support network, and especially Gustavo Glusman, founder of BioMOO, for teaching and sharing his knowledge of MOO. This manuscript is dedicated to the memories of Vivian Towell (JT's mother) and Donald Middleton (ET's father), recently deceased.
Anderson, C. (1994, May). Cyberspace offers chance to do 'virtually' real science. Science 264, 900-901. Curtis, P. (1992) Mudding: Social phenomena in text-based virtual realities. Proceedings of the 1992 Conference on Directions and Implications of Advanced Computing, Berkeley, CA, May. (Also available as Xerox PARC technical report CSL- 92-4.). Ellis, S.R. (1996). Presence of mind: a reaction to Thomas Sheridan's "Further musings on the psychophysics of presence." Presence: Teleoperators and Virtual Environments, 5(2), 247-259. Evard, R. (1993, November). Collaborative networked communication: MUDS as systems tools. Proceedings of the Seventh Systems Administration Conference (LISA VII), pp. 1-8, Monterey, CA. Fontaine, G. (1992). The experience of a sense of presence in intercultural and international encounters. Presence: Teleoperators and Virtual Environments, 1(4), 482-490. Franklin, N., Tversky, B., & Coon, V. (1992). Switching points of view in spatial mental models. Memory & Cognition, 20(5), 507-518. Hardy, B.H., Robinson, A., Doughty, S., Findsen, L.A., Towell, E.R., Towell, J.F., & Wilson. I.B.H. (1996, January). A new direction in conferencing: the First Electronic Glycoscience Conference, Trends in Biochemical Sciences, 21(1), 31-33. Heeter, C. (1992). Being there: The subjective experience of presence. Presence: Teleoperators and Virtual Environments, 1(2), 262-271. Held, R.M., & Durlach, N.I. (1992). Telepresence. Presence: Teleoperators and Virtual Environments, 1(1), 109-112. Kelso, M.T., Weythrauch, P., & Bates, J. (1993). Dramatic presence. Presence: Teleoperators and Virtual Environments, 2(1), 1-15. Robinett, W. (1992). Synthetic experience: A proposed taxonomy. Presence: Teleoperators and Virtual Environments, 1(2), 229-247. Schiano, D.J. (1997). Convergent methodologies in cyber-psychology: A case study. Behavior Research Instruments, Methods and Computers, 29(2), Schloerb, D.W. (1995). A quantitative measure of telepresence. Presence: Teleoperators and Virtual Environments, 4(1), 64-80. Sheridan, T.B. (1992). Musings on telepresence and virtual presence. Presence: Teleoperators and Virtual Environments, 1(1), 120-126. Sheridan, T.B. (1996). Further musings on the psychophysics of presence. Presence: Teleoperators and Virtual Environments, 5(2), 241-246. Slater, M., & Usoh, M. (1993). Representations systems, perceptual position, and presence in immersive virtual environments. Presence: Teleoperators and Virtual Environments, 2(3), 221-233. Slater, M., Usoh, M., & Steed, A. (1994). Depth of presence in virtual environments. Presence: Teleoperators and Virtual Environments, 3(2), 130-144. Steuer, J. (1992). Defining virtual reality: Dimensions determining telepresence. Journal of Communication, 42(4), 73-93. Towell, J.F., Hansen, P., Mercer, E., Leach, M., Rubin, I., Prilusky, J., & Glusman, G. (1995, November). Networked virtual environments and electronic conferencing. In S.M. Bachrach, D.B. Boyd, S.K. Gray, W. Hase, and H. Rzepa (Ed.), Proceedings of the First Electronic Computational Chemistry Conference [CD-ROM]. ARInternet: Landower, MD. Towell, J.F., & Towell, E. R. (1995). Internet conferencing with networked virtual environments. Internet Research 5(3), 15-22.
FIGURE AND TABLE LEGENDS
Table 1. Survey Results from 207 TNVE Users. Agreement Defined as Indicating a '1' or '2' in Response to Instrument I or a '4' or '5' in Response to Instrument II (see section 3 Results).
Figure 1. Overall number of TNVE users versus ratings of presence. Data from Groups III & V, who were administered Instrument II (opposite of Instrument I), were reversed (see section 2.5 Analysis).