|

Social Information Processing in MUDs:
The Development of Friendships in Virtual Worlds
by Sonja Utz, Ph.D.
Utz, S. (2000). Social information processing in MUDs: The development of friendships in virtual worlds. Journal of Online Behavior, 1 (1). Retrieved <date> from the World Wide Web: /JOB/v1n1/utz.html
Abstract
With the rapid growth of the Internet, new communication forms have
emerged. This study examines how friendships are developed in a special kind of virtual
world: multi-user-dungeons (MUDs). According to the Social Information Processing
perspective (Walther, 1992) people learn to verbalize online that which is nonverbal
offline, with increasing time. The use of verbal paralanguage should be an
important factor in the development of impressions. Sociability, as a general trait, and
skepticism towards computer-mediated communication (CMC), as a situation-specific
attitude, could also influence this process. One hundred and three MUD users completed a
questionnaire concerning their online friendships, MUD use, attitude about MUDding, use of
paralanguage, sociability, and skepticism toward CMC. Seventy-seven percent of the MUDders
reported relationships with others. Results supported the Social Information Processing
perspective: Sociability had little influence, whereas skepticism towards CMC was an
important predictor. Only participants who scored low in skepticism use the
paralanguage-features provided in MUDs and develop friendships. To examine why the
skeptical respondents play MUDs, a cluster analysis was computed. It revealed four types
of MUDders, who differ in their attitude toward MUDs. The consequences of different
motivations for playing MUDs on the development of friendships are discussed.
Introduction
The Internet has become
more and more widespread. World wide, 275.5 million people are online (Nua Ltd., 2000).
This medium not only provides access to information and immediate communication over long
distances; it is also changing society. New social venues have evolved, and virtual
communities have emerged. Enthusiasts consider virtual communities a liberation from
traditional constraints and boundaries like time and place. People from all over the world
can find others with the same interests. Friends have no longer to be chosen mainly from
the persons living in the same city. Similarity in interests and attitudes are the basis
of virtual relationships. Virtual worlds create new opportunities for the development of
friendships (Parks & Floyd, 1996; Parks & Roberts, 1998) but little is known about
the processes taking place in virtual worlds. People cannot see each other in text-based
virtual worlds, and many useful signals like gestures or intonation are missing in
computer-mediated communication (CMC). How is it possible to form relationships in such a
seemingly cold medium?
The Social Information
Processing theory (SIP; Walther, 1992) explains how people meet and develop online
relations, but it is troubled by two issues. First, there is a lack of evidence for the
specific communication processes taking place in these online relations, as researchers
have focused on media inputs, and relationship outputs, to the general exclusion of actual
processes (see Straus, 1997). For instance, while Walther (1992) proposes that CMC users
adapt nonverbally-expressed social information into online behavior, both verbal and
textual, research supporting his contention has not examined the micro-behaviors specified
in this proposition. Second, while the theory originally assumed that all communicators
are motivated to develop significant relations with CMC partners, subsequent tests have
shown a variety of factors that dampen this motivation and affect (e.g. Walther, 1994,
1997). While other approaches to CMC, such as those related to media selection, offer
alternative perspectives on how perceptions of media can influence their uses, these
approaches tend to focus on raw usage patterns, though, and not on the character or
content of that usage.
The present research
addresses both of these concerns. First, it examines the process specified in the SIP
perspective. According to Walther (1992, see also Hiltz & Turoff, 1981; Rice &
Love, 1987), people get used to CMC and learn how to verbalize relational content. Do
people really use more such verbalizations with more CMC experience? Does such
verbalization lead to friendships? Additionally, a second aspect is taken into account
regarding verbalizations and friendship formation. People differ in traits, attitudes, and
other relevant motivations. Do people who are very sociable in real life develop more
friendships via the Internet as well? Does their attitude toward CMC influence their
behavior, and therefore the formation of friendships? Can different types of users be
distinguished who differ in the use of relational strategies?
This article addresses these questions by
examining the process of making friends in a specific form of virtual communities:
multi-user-dungeons (MUDs). MUDs are adventure role-playing environments on the Internet.
MUDs were chosen as a research site because they may be considered as well-established
virtual communities. In contrast to new, commercial and graphical virtual worlds, they
have existed in form for over 20 years. MUDs are more social than newsgroups, yet more
centered around a common interest (role-play) than chats, combining characteristics of
different forms of virtual communities. Although most MUDs are simply text-based, they
feature a highly developed social system. It is also known that people form friendships in
MUDs (Parks & Roberts 1998). The purpose of this study is to examine some of the
underlying processes of making friends online.
Multi-User Dungeons
MUDs (for further
information see http://www.godlike.com/muds/ or
http://www.mudconnect.com/)--multi-user-dungeons
or multi-user-domains--are text-based virtual realities, adventure role-plays on the
Internet.Note 1 They can be compared
to various forms of CMC along the following lines. While CMC can happen synchronously or
asynchronously, MUDs are synchronous. In asynchronous CMC the communication
partners do not have to be using the system at the same time. E-mail and mailing lists, or
Usenet newsgroups, for instance, can be read immediately, or hours, or even days later. Synchronous
CMC, however, requires the participants to be using the system simultaneously. The text
typed is almost immediately displayed on the other participants screens. Synchronous
CMC also includes chats. Chats offer an opportunity for real-time discussions, and are
primarily used for socializing. In contrast, MUDs often involve competitive role-playing
games. The settings are fantasy worlds, in which players have to create their own
"characters" by choosing a name, a gender, and a race (e.g. demon, elf, gnome).
To join a MUD the player has to log into the server hosting the MUD via telnet Note 2 (see Figure 1), or special MUD
software.
 MUDs are virtual worlds
with rooms, shops, pubs, towns, streets, forests, seas, ships and so on, in
which each object is described by text only. In most cases, no graphical elements are
provided. MUDs are similar to MOOs (multi-user-dimensions, object-oriented) in these
aspects, but they differ in their main purpose. MOOs are primarily social; people meet
there to communicate, socialize, and build houses. This article focuses on adventure-style
LP-MUDs (after Lars Pensjoe, who developed the program code). Participants mainly take
part in role-playing, the aim of which is to solve quests and kill monsters in order to
get points, which can be used to advance the level of the character. At a
certain level it is possible to become a so-called wizard. Wizards are responsible
for the further development of the MUD. They code new areas and quests. MUDs consistently
change. They are dynamic worlds developed by the players.
A special feature of MUDs are feelings
and emotes. Feelings are small pre-programmed scripts to express actions and
emotions by simply typing an abbreviation (In MOOs these are called Feature Objects and
Social Verbs, but they function similarily). There are several hundreds of verbs and
adverbs which can be combined at pleasure. For example, "smi iro" results
in "smile ironically"; it is possible to smile sadly, happily, knowingly,
innocently, and so on. Other feelings are hug, laugh, cry, poke, kick, kiss, or sigh. If
the predefined feelings are not sufficient, players can easily create their own emotes by
appending text to their name. MUDders also use "smileys," little faces formed by
ASCII symbols. Smileys, feelings, and emotes are often collectively referred to as emoticons,
some examples of which appear in Table 1.
Table 1. Smileys, Feelings and Emotes
Emoticon |
Description |
Example |
Smileys |
Faces formed by ASCII characters |
:-)
happy smiley :-(
sad smiley |
Feelings (called social verbs
in MOOs) |
Preprogrammed
scripts, combinations of verbs and adverbs
possible, results always in the preprogrammed reaction
Shortcuts
|
command: laugh ¿ displays: You fall down laughing
command: cuddle Minna ¿
displays: You cuddle Minna.
command: smi Aron und ¿
displays: You smile understandingly at Aron |
Emotes (posing in MOOs) |
Self-defined expressions |
command: me is so happy that he could embrace the whole world ¿ displays:
<Name> is so happy that he could embrace the whole
world. |
Due to
these features, MUDs can be regarded as a highly developed form of CMC. They offer
specific devices with which to overcome the lack of nonverbal cues. While Walther's (1992)
SIP perspective focuses explicitly on the natural language verbalization of relational,
traditionally nonverbal symbols, the stylized emoticons in MUDs offer a particularly
interesting and unique instance of verbalization. Before outlining SIP as the theoretical
framework of this study, a short overview of prior theories in CMC research is given.
Research on Computer-Mediated Communication
The Social Presence
model developed by Short, Williams, and Christie (1976) is the oldest theoretical
framework for analyzing CMC. Originally intended for comparing communication via
telephone, audio, and video conferences, it was often applied for explaining effects in
CMC (DeSanctis & Gallupe, 1987; Hiltz, Johnson & Turoff, 1986; Rice & Case,
1983). Media vary in the number of channels. In audio conferences, for example, only sound
is given, whereas in video conferences additional visual information is provided.
According to Short et al. (1976), the fewer channels a medium has, the lower its social
presence. Consequently, the perception of the communication partner should be very
impersonal in media with low social presence. Communication should be primarily
task-oriented, and social influence should be difficult to establish. Social presence was
regarded as a "quality of the medium itself" (Short et al., 1976, p. 66),
although it was measured as a subjective impression in some studies. Social presence is
seen "as a single factor that comprises a number of dimensions relating to degree of
interpersonal contact" (Spears & Lea, 1992, p. 32) and is often measured using
semantic differentials on dimensions like unsociable-sociable or cold-warm (Short et al.,
1976). Consistent confirmation of the Social Presence hypotheses was not attained, though.
For example, more attitude change towards the source was found in sound only conditions
than in face-to-face (FTF) communication (Short, 1972, 1973).
According to the Social
Presence model, it should be more difficult to build relationships via CMC than via
FTF-communication. If the communication is perceived as cold and impersonal rather than
warm and sociable, the communication partner would not be seen as a person who could be a
friend.
Another very
influential model is the Reduced Social Cues approach (RSC; Kiesler, Siegel & McGuire,
1984; Kiesler, 1986; McGuire, Kiesler & Siegel, 1987; Sproull & Kiesler, 1986).
The absence of social and contextual cues is the central factor in the RSC approach. The
lack of cues undermines the perception of leadership, status, and power, and leads to
reduced impact of social norms and therefore to deregulated, antinormative behavior.
Moreover, people become depersonalized because the attention focuses on the written text,
not the social context. Other consequences are difficulties in coordination,
deindividuation, and equality of participation (see for review Spears & Lea, 1992).
Deindividuation means a loss of identity, reduced self-regulation, and self-awareness.
Social norms and constraints are less present. The resulting disinhibited behavior-- for
example, flaming--causes a hostile impersonal atmosphere and impedes the development of
friendships.
Although several
studies (Kiesler, Siegel & McGuire, 1984; Siegel, Dubrovsky, Kiesler & McGuire,
1986) describe CMC as impersonal, hostile, and task-oriented, these results do not
generalize to other CMC situations. Field studies have shown that CMC can be very rich in
socio-emotional content (Rice & Love, 1987) and that many users develop friendships in
virtual communities (Parks & Floyd, 1996; Parks & Roberts, 1998).
Walther (1992) offers
an explanation for these discrepancies between laboratory and field studies. The crucial
factor is time. It takes longer to type than to speak, and people have to get used to the
new medium. Thus, there are rate differences between the media, rather than fixed
differences in impression development (Walther, 1993). The impression development process
takes longer in CMC; if there is sufficient time, the differences between CMC and FTF
diminish.
Walther (1992),
therefore, suggests an alternative perspective: the SIP model, which consists of the
following assumptions and processes. Relational motivators exist a priori, for
example, the affiliation motive. Other drives are impression management or dominance
drives. These drives motivate people to exchange social information with others. The next
step is impression formation. In CMC, this happens by decoding the verbal messages of the
communication partner. Despite the assumptions made by RSC, it is possible to get an
impression of the other via CMC. Language variations, for example, lead to assumptions
about education, social, and professional status. CMC users develop an interpersonal
epistemology, i.e. a distinctive representation of the communication partner (see also
Miller & Steinberg, 1975). They use "knowledge-generation strategies such as
interrogation, self-disclosure, deception detection, environmental structuring, and
deviation testing to gather psychological knowledge-level information about other
persons" (Walther, 1992, p. 71). Finally, the encoding of relational messages takes
place. People learn how to transmit relational content in text. According to Walther
(1992), relationships in CMC are possible, but the user has to learn to verbalize
relational messages. Time plays an important role as well: It takes longer to get used to
the medium, to get to know each other and to build up trust and friendships via CMC.
SIP has been confirmed in several studies in which geographically dispersed people
cooperated in work-groups for several weeks (Walther, 1993, 1997). It has not been tested
in more open-ended, recreational settings like MUDs or in long-term interactions which
take place in virtual communities.
The general hypothesis
that the development of relationships takes more time in CMC can be applied to MUDs. While
many language-based relational cues can be applied in CMC, the most obvious might be
emoticons. There are well-known language-based strategies in CMC such as capitalization or
acronyms. Capitalization is considered symbolic of shouting, and there are many well-known
acronyms like ROTFL ("roll on the floor laughing") among Internet users. As
reported above, MUDs offer several features for the encoding of relational cues. Feelings
and emotes as well as smileys can be used to express emotions. The first hypothesis of
this study is therefore:
H1: Emoticon
use (i.e. smileys, feelings, and emotes) increases with time.
Walther later developed
the Hyperpersonal Model of CMC which states that CMC is sometimes even more friendly and
social than FTF-communication (Walther, 1996). In CMC, users have the opportunity for
selective self-presentation. They have time to think about how to present themselves and
can choose the positive aspects. On the other hand, the reduced social cues in CMC lead to
an idealized perception by the perceiver. S/he has only the positive information, and
inflates the impression of the partner by generalizing these positive cues on other
unknown personality aspects. CMC can, therefore, be more social and intimate, or
"hyperpersonal" relative to FTF communication. The ability to express emotions
in text and self-presentation are very important for a social and friendly atmosphere,
leading to the development of friendships.
Roberts, Smith, and
Pollock (1996) demonstrated that the Hyperpersonal model also holds true for MOOs. They
showed how MOO-users form relationships by overcoming the lack of nonverbal cues using
feelings and emoticons. Since the main purpose of adventure-oriented LPMUDs is playing a
game, and socializing is only a side-effect, the ability to express emotions via text in
LPMUDs should be even more important for the development of relationships. The second
hypothesis is therefore:
H2: The more
emoticons (i.e. smileys, feelings, and emotes) a person uses, the more friendships s/he
builds.
Hypotheses 1 and 2 test
the predictions of SIP. As stated above, some studies in this domain show inconsistent
results. Not all CMC-users seem to be motivated to affiliate (Walther, 1994). Which other
factors influence the formation of relationships in MUDs?
Parks and Roberts
(1998) studied several forms of virtual communities and showed how friendships in MOOs,
newsgroups and real-life differ in dimensions such as interdependence, breadth, or depth.
MOO friendships were more advanced than newsgroups relationships but less developed than
off-line relationships. Reasons for these differences were not specified. Parks and
Roberts (1998) did not find that demographic variables such as sex or age differentiated
people with online-relationships from those without. If such gross demographic factors do
not influence the formation of online-relationships, more specific personality factors may
play an important part.
Sociability may be such
a factor. People differ in the general trait sociability, and it could be that individuals
with difficulties making friends offline may experience the same difficulties in CMC.
Becker and Mark (1998) observed three different virtual environments (text-based,
graphical and with audio) and reported that most individuals experienced the same degree
of shyness as FTF. Thus, shyness or sociability may be a factor that predicts the
development of friendships, reflecting the same influence in CMC as in FTF communication.
On the other hand, it
can be argued that shy and inconspicuous persons profit by the anonymity of the Internet.
They cannot be judged primarily by their appearance, they do not have to fear any
consequences offline, and may therefore feel encouraged to approach other people. Roberts,
Smith, and Pollock (1997) studied MOO- and IRC-users and found that "individuals who
self-identified as shy reported that they were less inhibited and less conservative in
on-line environments" (p. 2). If a safe environment provided by virtual communities
compensates for shyness, different degrees of sociability may also be less relevant
for the development of friendships. Because of these different predictions and the lack of
consistent empirical results, an open research question is formulated:
RQ1: Does
sociability affect the development of online-friendships?
While sociability is a
general trait, situation-specific attitudes may also play an important role. Evidence for
the influence of attitudes and perceived media attributes derives from research on media
selection (e.g. Daft & Lengel, 1984; Fulk, Schmitz,& Ryu, 1995), which has
ultimately demonstrated that attitudes about CMC affects its use.
As reported above, CMC
was rated as quite impersonal, cold, and unsociable in early studies following Social
Presence theory (Short et al., 1976). A similar proposal is made by the Media Richness
model (Daft & Lengel, 1984; Daft, Lengel, & Trevino, 1987; Trevino, Daft, &
Lengel, 1990): Media vary in their capability of reducing equivocality and ambiguity. A
medium is regarded as rich if it facilitates feedback, communicates multiple cues,
presents individually tailored messages and uses natural language. Trevino, Daft, and
Lengel (1990) concluded that media can be ranked according to their richness, with FTF
communication highest, formal numeric text at the end, and e-mail somewhere in between.
Studies on media
selection came to inconsistent results. While some confirmed Media Richness, others, e.g.
Markus (1994a), found that managers used electronic mail successfully for tasks ostensibly
requiring higher degrees of media richness.
A more recent
contribution is the Social Influence model of technology use (Fulk, Schmitz, & Ryu,
1995; Fulk, Schmitz,& Steinfield, 1990). It suggests that individuals' media
perceptions, which are socially constructed, vary considerably. As Schmitz and Fulk state,
"the perceptions of objective features may differ across individuals" (1991, p.
490). Their results showed that the perceived media richness is a significant predictor of
actual media use. And in a more recent study, perceived media usefulness was a better
predictor of media use than the perceived richness (Fulk, Schmitz,& Ryu, 1995). From
these studies it appears that despite the objective characteristics of media, users
perceptions of CMC lead to particular uses, and desired outcomes.
While most of the
studies on media selection concentrated on which medium is used, the question of how
a medium is used was not as often addressed (see for exception Markus, 1994b). This study
assumes that perceived media characteristics also influence the actual content of the
selected medium. While MUDders already have chosen synchronous CMC in virtual worlds, it
is not certain that all MUDders regard CMC, and especially MUDs, as appropriate for
developing friendships. If individuals judge CMC as impersonal and not appropriate for
making friends, this attitude will affect their behavior. As a result, their CMC will be
impersonal even if they are quite sociable in other situations.To distinguish this concept
from media richness, it is called skepticism towards CMC and towards the features
of MUDs. A skeptical person will neither believe in the possibility of developing
friendships via CMC in general nor the specific possibilities provided by MUDs. Therefore,
skepticism was predicted to be another determinant of the development of online
relationships.
H3:
Skepticism towards the usefulness of MUDs for developing friendships impedes the
development of online friendships.
Beside skepticism,
there may still be other factors influencing the process of making friends, and
motivational factors should also play a role. One variable affecting sociability in CMC is
anticipation of future interaction. Walther (1994) compared CMC in two conditions to find
this effect. In one condition the CMC groups expected future interaction over several
weeks, while in the other condition, group members expected to have different partners for
their next task. Persons who thought that they would work together for several weeks
showed slightly higher values in immediacy/affection, similarity/depth, and
receptivity/trust. This result implies that expectations and goals influence medium use.
Anticipation of future interaction may be a potent factor in MUDs as well. MUDs imply
further interaction, as the character is saved for future use at the end of each session.
The level structure in LPMUDs is an incentive for further playing. The more points a
player has, the better the skills and abilities of his/her character become.
Further evidence for
the role of expectations and goals is given in a study by Kayani, Wotring, and Forrest
(1996). Participants were confronted with several communication situations differing in
content (information exchange, persuasion) and relational dimension (friend, colleague,
stranger). The dependent variable was media selection. Results showed clearly that the
situation specific goals influenced media choice. Markus (1994b) showed also that socially
positive and negative e-mail use was not only determined by perceived media
characteristics, but also by users' intentions. She concludes that "research on the
social effects of electronic communication technology should consider not only the
technological characteristics of various media, but also those purposes and goals that
users atttempt to accomplish through the media" (1994b, p. 145).
These studies show the
importance of goals and motivations, but again they focus on media selection. It is
supposed that goals also influence how a medium is used. What does this mean for
the development of friendships in MUDs? What goals could be important? People may play
MUDs just for the fun of the game or for role-playing, and if their reason for playing
MUDs is not to meet other people, this should have an impact on in forming friendships. To
them, other MUDders are considered part of the game, fighting the same dangers, not as
persons with whom interpersonal bonds could be formed. If the goal is playing a game and
not joining the virtual community, less friendships should be developed. The second goal
is role-play. Strict role-players act always in terms of the role. The feelings and emotes
these players express are the emotions of their characters. They fit in the story. For
example, a thief caught for stealing is angry. The verbalizing of nonverbal context cues
is only a successful strategy for making friends when the actors express their true
feelings. Although this argument sounds quite plausible, these ideas have not been tested
in studies so far. In this study, the motivation for MUD-playing is considered as another
goal influencing the formation of online-relationships. Interest in playing a game and in
role-playing are regarded as two possible motivations. H4 is therefore:
H4:
The motivation for playing MUDs (i.e. game, role-play) affects relationship building, such
that role-play oriented users form more friendships than game-oriented users.
Method
Procedure
The study was announced
in the online news of three German MUDs (Adamant, First Light, and TappMud), requesting
interested people to send an email with their real life postal address to receive a
questionnaire.The use of hard-copy postal mail for the questionnaire avoided technical
problems like script programming. It also provided a check for multiple participation and
real life gender (using first names which, in German,are quite unequivocal). A hundred and
twenty-five questionnaires were sent out and 103 questionnaires were returned to the
author, resulting in a response rate of 82%.Note
3
Measures
A hundred and sixty
items tapped the following aspects: data regarding MUD-use, attitude towards MUDs,
development of friendships, use of paralanguage, and personality aspects.
Demographic data were also collected in order to characterize the sample with regard to
age, sex, and major subject of study or occupation.
Data on MUD use were
collected to get information about MUD behavior and to test whether a broad range of
MUDders had taken part in the study. This part of the questionnaire included questions
about the hours per week spent in the MUD, MUD-time in months, number of characters and
MUDs played, etc.. While these data were self-reported, and therefore subjective, other
research shows substantial agreement between self-reported time per hour and objectively
measured data (Zielke, Schildmann, & Wirausky, 1995).
Three scales assessed
the attitude towards MUDding. Initially, 14 items were developed. Responses were made on
six-point scales ranging from "not true" to "true," with higher scores
indicating higher levels of agreement.
To test whether the
newly developed items for game, role-play and skepticism formed three
different factors, a principal components factor analysis using varimax rotation
specifying three factors was conducted. The three resulting factors explained 52.8% of the
total variance. The first factor, game, explained 19.0%, the second, role-play,
18.3%, and the third, skepticism, 15.5% of the total variance. One item not loading
highly (less than .35) on any factor was excluded. Each of the remaining items loaded
clearly on its appropriate factor (see Appendix 1). The items on
each subscale were averaged for further analysis.
The scale skepticism
(Cronbachs a = .68) was
intended as a situation-specific measure of the attitude towards CMC and MUDs. The items
focus on the social potential of MUDs such as the belief in the possibility of making
friends or expressing emotions adequately in MUDs. The four items were "I think there
are often misunderstandings in the MUD because I don't see or hear the others,"
"I think it is impossible to express feelings adequately in CMC," "I think
getting to know each other via CMC is impossible," and "I think getting to know
each other via CMC is possible only under the condition that people are willing to show
their true attitudes and emotions". The last item was reverse scored so that higher
scores meant higher skepticism.
As MUDs are not only a
form of CMC, but also adventure-style role-playing games, different motivational
reasons for playing MUDs are possible. Two scales assessed these possible motivations in
order to characterize the group of MUDders further. Game (a = .70) focused on the game-character of MUDs.
People can play MUDs primarily to solve quests, kill monsters, train their skills, win
points, and to get a position in the highscore list. Getting to know the other
participants is less central. Game consisted of the items "For me the MUD is
only a game" or the reverse, "The MUD is not only a game because relationships
with other persons are developed;" "MUD is connected with real-life, because I
know many MUDders in rl" (real life), "The boundary between MUD and rl fades
away sometimes," "When my MUD-character dies, I am also sad in rl". Items
were coded such that higher scores indicated seeing the MUD more as game than as virtual
community.
Role-play (a = .78) taps a different aspect. Especially
well-developed MUDs offer a number of different races (e.g. human, elf, demon, gnome) and
guilds (e.g. fighter, cleric, bard, thief, sorcerer). Role-play is therefore a central
component of MUDding. Role-play differentiates between playing a role in the MUD or
behavior as seen in real-life. Role-play measurement consisted of these four items:
"With my characters I try to play different roles which have not much in common with
my true personality," "My behavior in the MUD is different from that in
real-life," and reverse coded "I identify myself with my MUD-character/s,"
and "Behavior and character of my MUD-figures are very similar to me".
In order to control the
involvement, social identification (a = .89) with the group of the MUDders was measured by a modified scale from
Simon and Massau (1991), based on the Social Identity Theory (Tajfel, 1978).
Sociability (a = .73) was assessed with a subscale of the
Multidimensional Measurement of Self-Description (Mummendey, Riemann & Schiebel,
1983). This instrument consists of 56, 7-interval bipolar adjective pairs (e.g.
social/unsocial) that assess six different dimensions of the self-concept.
Verbalization of
relational contents was operationalized by self-reported frequency ratings of the usage of
smileys, feelings, and emotes. While smileys are quite common in CMC, feelings and emotes
are the MUD-specific features for expressing emotions through text. Ratings on this paralanguage
scale ranged from 1 (very seldom) to 5 (very often); a = .72.
The scale friends (a = .72) was intended to measure the development
of online friendships. Four items addressed the quality of online-relationships: "I
have friend(s) in the MUD, with whom I can talk about private topics," "The
people in the MUD are just casual acquaintances to me" (reversed), "Some of my
MUD-acquaintances resulted in real life friendships"; "Some of my MUD
acquaintances resulted in romantic real life relationships." The categories
acquaintance, friend and romantic friend differ in the extent of depth and intimacy of the
relationship. The exact criteria of judgement for another person as acquaintance, friend,
or romantic friend was left to the respondent.Note 4
Romantic relationships led this ranking.
The questionnaire also
included some additional scales which are not relevant to the present research (see Utz,
in press).
Subjects
A hundred and three
players (85.4% male, 14.6% female) of adventure-style MUDs took part in this research. The
average age was 23.5 years (SD 2.7 years). Of these participants, 101 came from Germany,
and 2 from France; 83% were university/college students, 1% were "going to
school" (connoting high school in the German phrasing), and the others were employed
outside the home. The high education level of the respondents probably is due to the free
Internet access at universities in Europe, where commercial dial-ins and local telephone
charges are costly relative to other regions.
Participants had been
playing MUDs an average of 19.6 months (SD = 14.8). They spent an average of 12.3 hours
per week in these virtual worlds (SD = 10.7). While 52.4% had only one character, 35%
played two characters, and the remainder had three and more. The average number of MUDs in
which each user participated was 1.85 (SD = 1.0). Most (76.7%) of the MUDders reported
that their MUD-friendships developed into offline friendships, and 24.5% formed romantic
relationships. The mean of the scale friends was 3.85 (SD = 1.17), indicating
moderate agreement.
Results
Social Information Processing in MUDs
According to SIP,
people use more paralanguage with increasing time (H1) and this should increase
the development of friendships (H2). To test these hypotheses, correlational
analyses were conducted. The correlation between use of paralanguage and MUD-time in
months was marginally significant, although weak, r (92) = .17, p
(one-tailed) = .05. The correlation between MUD hours per week and paralanguage was also
significant, r (103) = .32, p < .01. H1 is therefore
supported: The expression of otherwise nonverbal contents through text increases with
time.
Paralanguage
correlates with development of friendships, r (98) = .49, p < .01
(one-tailed), confirming H2. The partial correlation between paralanguage
and friends, controlling for MUD-time and hours per week, was r (95) =
.38, p < .01, showing that the substantial link between the use of paralanguage
and the making of friends is not only due to time spent in virtual worlds, but also to the
expression of relational contents in words. These results support the SIP perspective:
People learn how to verbalize nonverbal contents, and this behavior predicts the
development of friendships, explaining, beyond the time factor, 14% of the variance.
Effects of Sociability and Skepticism
RQ1 asked whether the general trait, sociability,
influences the development of relationships via CMC as well. There were significant but
only moderate correlations between sociability and friends, r (100) =
.23, p (two-tailed) < .05, and between sociability and paralanguage,
r (98) = .24, p (two-tailed) < .05. Trait sociability evidently explains
significantly less variance than the behavioral use of paralanguage in the development of
friendships (z z' paralanguage,friends vs. z' sociability,friends
= 2.075, p < . 019).
Sociability is
not related to hours per week MUDding, r (96) = -.02, n.s.; or MUD-time history, r
(89) = -.11, n.s. A selection effect can therefore be excluded: It is not the case that
only sociable persons continue MUDding. In answer to RQ1, it can be said that sociability
influences the formation of online relationships to a moderate degree.
H3 suggested
that the attitude toward MUDding, especially skepticism, would be another
important variable. A significant correlation, r (99) = -.50, p < .01
(one-tailed), between skepticism and friends confirmed this hypothesis. The
more skeptical about CMCs friendly capacities a person is, the less s/he develops
friendships.
Regression Analysis
To examine the effects
of all mediating variables simultaneously, a regression analysis was calculated (using a
forced entry of all predictors simultaneously). Development of friendships was the
criterion variable. The significant predictors were paralanguage, sociability,
and skepticism. Results are presented in table 2.
Table 2. Regression Analysis on the Variable
"Friends," Multiple adj. R2 = .32, F (3, 88) = 15.1, p < .01
|
Beta |
t |
p |
|
Paralanguage |
.322 |
3.250 |
p < .01 |
Sociability |
.084 |
.933 |
n.s. |
Skepticism |
-.330 |
-3.371 |
p < .01 |
The
results indicate that skepticism towards CMC and the use of paralanguage predict the
development of relationships in MUDs. In this analysis sociability did not reach
significance. This result qualifies the result of the correlational analysis: Compared to skepticism
and paralanguage, sociability is not a significant predictor. H2
and H4 were also supported through this analysis. The absolute Beta weights of skepticism
and paralanguage do not differ. The SIP perspective, therefore, should be
extended, and skepticism towards CMC taken into account as one of several influences that
affects motivation to develop relationships online. Individuals who believe it possible to
build up relationships in virtual worlds learn how to use smileys, feelings, and emotes
and thus make friends in MUDs, but those who are more skeptical of CMC do not.
The results so far
extend our knowledge about relationships in virtual worlds and the processes by which
social information processing takes place in at least one form of CMC. The question
remains why people very skeptical toward CMC play MUDs. A possible explanation lies in the
assumption that they do not want to socialize primarily but are more interested in
role-play, in killing monsters, or in solving quests. The number of players developing
friendships speaks for this hypothesis. The 76.6% proportion of respondents in this study
may seem relatively high, but in the study by Parks and Roberts (1997), 93.6% of the users
of the primarily social MOOs reported online-relationships. According to SIP (Walther,
1992), motivational factors are an assumed precursor to social information processing.
Different types of MUD players may be possibly distinguished by their motivations to play
the game. To test this hypothesis, an additional analysis was conducted.
Types of MUD Players
A hierarchical cluster
analysis (Ward, 1963) was conducted using the z-standardized values in the game,
role-play, and skepticism scales. The appropriate number of clusters was
determined by visual examination of the amalgamation coefficient schedule (Aldenderfer
& Blashfield, 1984), a procedure analogous to the Scree test in factor analysis. Four
types of MUDders emerged. Ninety-nine participants could be classified, while four cases
had to be sorted out because of missing data. The fit of the classification was tested by
discriminant analysis; 94.9% of the originally-classified cases and 91.9% of the
cross-validated cases were correctly classified. Therefore, the cluster analysis resulted
in an adequate classification. The group means on the three attitude scales are presented
in table 3. Multivariate analysis of variance (MANOVA) confirmed the differences between
classifications on these scales (see table 3).
Table 3. Types of MUDders: Means and Standard
Deviations (in parentheses) for Skepticism, Game and Role-play
|
Group
1
( n = 24) |
Group
2
(n = 44) |
Group
3
(n = 15) |
Group
4
(n = 16) |
F (3,95)
|
p
|
|
Skepticism |
2.27
(.57) |
3.61
(.86) |
2.22
(.56) |
4.33
(.96) |
34.76 |
p
< .001 |
Game |
3.93
(.82) |
4.02
(.77) |
2.39
(.53) |
5.00
(.62) |
33.88 |
p
< .001 |
Role-play |
4.11
(.63) |
2.68
(.70) |
2.30
(.62) |
4.69
(.66) |
60.44 |
p
< .001 |
The
participants in group 1 (n = 24) can be considered real role-players. They
are low in skepticism, the MUD is a game for them, and they are first and foremost playing
different roles. Role-play is a central aspect in adventure-style MUDs. For example, there
are different guilds, and a bard must behave differently than a cleric or a thief. The
largest group (n = 44) consists of persons for whom the MUD is primarily a game.
They do not adopt roles, but they want to play, i.e., to kill monsters and to solve
quests. They are significantly higher in skepticism, but only just above the midpoint of
the scale. Group 3 consisted of 15 players who were very low in skepticism, but who
neither want to game-play nor to role-play. They seem to be very involved in the virtual
world per se. They may be wizards whose main purpose is to develop the MUD further; for
them, the MUD is more a software project than a game. Alternatively, they may be
"chatters," who visit virtual worlds to meet other people. In the latter case,
MUDs are seen primarily as virtual communities. Group 4 includes 16 participants who can
be called "extremely skeptical." They score very high on skepticism, game, and
role-play. The question remains, why they are MUDding, if they are so skeptical and do not
take it seriously at all.
To further characterize
the groups with regard to involvement, another MANOVA was conducted, with the variables social
identification, hours per week MUDding, and MUD-time involved. The last
two variables were regarded as implicit measures of engagement in the MUD, whereas social
identification with the group of the MUDders is an explicit criterion for involvement. The
group means are presented in table 4.
Table 4. Types of MUDders: Means, and Standard
Deviations (in parentheses) for Implicit and Explicit Involvement Measures
|
Group 1 |
Group 2 |
Group 3 |
Group 4 |
F (3,83) |
p |
|
MUD-time |
20.41 (15.00) |
17.92 (13.22) |
23.86 (17.91) |
17.80 (8.94) |
.28 |
p = .84 |
Hours per week |
14.04 (10.81) |
12.74 (8.31) |
17.63 (16.87) |
4.20 (3.94) |
5.34 |
p < .01 |
Identification |
3.93 (.87) |
3.59 (1.02) |
4.66 (.78) |
2.60 (.94) |
13.16 |
p < .01 |
The only
significant differences between groups are for MUD hours per week and social
identification. These results are consistent with the descriptions above. Members of group
4 are very skeptical, high on role-play and gaming, and therefore not very
involved in the MUD. They spend clearly fewer hours per week, and they don't identify with
the group of MUDders (although it may well be that they were once more engaged in the
game). Respondents in group 3 were most identified with the game. They also spend
the most time in the MUD, confirming the assumption that these are the respondents most
involved. The role-players in group 1 identify to a high degree with the other MUDders,
too. Low skepticism therefore corresponds to high identification with the group. The
players in group 2, for whom the MUD is primarily a game, identify only to a moderate
degree.
Do these different
motivations and attitudes carry consequences for the use of paralanguage and the
development of friendships? Another MANOVA with paralanguage and friends as dependent
variables was computed. Both effects were significant. The means are presented in table 5.
Table 5: Types of MUDders: Means and Standard
Deviations (in parentheses) in Use of Paralanguage and Friends
|
Group 1 |
Group 2 |
Group 3 |
Group 4 |
F (3,91) |
p |
|
Paralanguage |
4.21 (.47) |
3.70 (.75) |
4.19 (.77) |
3.33 (.98) |
5.44 |
p < .01 |
Friends |
4.14 (.85) |
3.73 (1.10) |
4.82 (.89) |
2.73 (.98) |
11.20 |
p < .01 |
The
effect on the use of smileys, emotes, and feelings follows the same direction. The low
skeptical participants in group 1 (role-play) and 3 (involved) use the most paralanguage.
Although the two groups do not differ in the extent of paralanguage use, the participants
in group 3 managed to develop significantly more friendships. This effect can be explained
by greater involvement and motivational factors. People in group 1 are primarily
interested in role-play. As MUD behavior differs from real-life, and verbalized emotions
in role-plays are not emotions of the actor, but the pretended emotions of the character
played, this dissimulation may make it more difficult to build up relationships. By
contrast, group 3 does not play roles and is highly identified with the virtual community.
Consequently its members develop the strongest bonds to other MUDders. Participants in
group 2 use paralanguage to a smaller extent and they also report fewer friendships. This
corresponds with their higher skepticism. They are primarily interested in the game, and
forming friendships seems to be a side-effect of playing MUDs. However, the results in
group 4 are striking: The skepticals deny the development of friendships, and score only
3.33 at paralanguage use--a scale on which a value of 3 meant
"occasionally"--and is very low for a MUD, due to the purely text-based
communication. A certain minimum of verbalization must be used. It seems that individuals
in this group could as well play a single-user game on their own. This analysis shows that
low use of paralanguage goes along with a low number of friendships formed via the
Internet. However, as the differences between group 1 and 3 show, high use of paralanguage
is not the only predictor of making friends. The motivation for playing MUDs also plays
also a role.
Discussion
The purpose of this
study was to examine how people develop friendships in virtual worlds, especially in MUDs.
Is it just a question of familiarity and time in CMC, or do other relevant factors exist?
The results of the correlational and regression analyses confirmed the SIP perspective
(Walther, 1992). Over time, people get used to CMC and come to compensate for the lack of
nonverbal cues by using linguistic cues. The more that MUDders used smileys and the
MUD-specific feelings and emotes, the more friendships they formed. However, while the SIP
perspective suggests that such behavior may be consistent over time, this specific
relationship between verbalizing relational contents and development of friendships
remained, even when the data were controlled for MUD-time in months and hours per week
spent in MUDs. These findings also contrast Parks and Floyd´s (1996) contention that the
development of friendships is a function of familiarity and experience with CMC. The time
spent in virtual worlds may be a necessary, but not a sufficient predictor for making
friends. It appears that while friendship development occurs among some long-term users,
it does not occur among all such users. It could be argued that time spent in MUDs is not
a good measure of experience or ability. People who continously run a window with the MUD
in the background of their computer report high numbers of MUD use per week, although
their actual usage may be lower. Therefore, only if experience is conceived as ability to
deal with the social possibilities of CMC, and the ability to verbalize emotional
contents, is it a significant predictor of the development of friendships.
Another important
finding was that the development of friendships in MUDs is only weakly correlated with the
general trait sociability. The Internet seems to provide an opportunity for shy
and socially inhibited people to overcome their difficulties. This is in line with the
contention of Parks and Roberts' (1998), among others, who suppose that MOOs and MUDs
provide users with "the perception of a safe environment for social interaction in
which individuals can explore all types of relationships without fear of repercussions in
their physical lives" (p. 531). Another study, by Roberts et al. (1997), examined shy
people and found that they were less inhibited in CMC, and formed intimate relationships
online. Zimmerman (1987) demonstrated the positive effects of CMC on schizophrenic
adolescents. The potential of virtual worlds for overcoming shyness has long been
considered and now appears to be somewhat conclusive.
The most important
result of this research is the influence of skepticism on development of friendships. If
individuals do not believe that it is possible to express feelings in CMC or to become
acquainted with others in virtual worlds, they refrain from doing so. They also do not use
the features offered by MUDs: feelings and emotes. Not trying to verbalize emotions and
other nonverbal contents leads to a confirmation of their skeptical attitude. Skepticism
towards CMC seems to be the missing link between familiarity with CMC and the development
of relationships. Because the results presented are based on correlational
analysis, the direction of causality cannot be proved. Nevertheless, it was shown that
time alone is not sufficient to form friendships in virtual worlds. People have to
compensate for the lack of nonverbal cues and replace nonverbal expressions with verbal
indicators. Beliefs about the medium are a presupposition to that. The attitude towards
the medium has been neglected in most studies on actual media use, and the
exception--perceived media richnesshas only been used to predict media selection and
not media content.
Not every individual
communicating online believes in the possibility of making friends in virtual worlds, nor
intends to. There are other reasons for communication via computer. Motivational factors
have to be taken into account. A first clue to their importance is the number of
friendships found in different settings: 73.6% of the respondents in this study reported
that they had made friends in the MUD. Parks and Floyd (1996) found that 60.7% of
Newsgroup users developed relationships, whereas Parks and Roberts (1998) found that 93.6%
of MOOers formed relationships. These figures reflect the expected motivation for making
friends. MOOs are built for socializing; MUDs have the additional role-play and
game-component, whereas newsgroups are primarily intended for discussing certain topics.
Individuals may participate in a newsgroup to gain information as well as to form
relationships; most MOOs and MUDs are less information oriented.
For a further
specification of the motivational aspects, a cluster analysis was computed in order to
divide the MUDders into subgroups. The resulting types differed not only in skepticism,
but also in their goals, motivations, and the development of friendships. Four types could
be distinguished: role-players, game players, involved ones and skeptical ones. Skeptical
MUDders are low in identification as well as in hours per week. Parks and Roberts (1998)
found that 46.2% of the MOOers without online friendships spent less than three hours per
week in the MOO, and interpreted this as low involvement. This assumption is supported by
the results in the social identification scale in this study, which showed
similar results. Another finding is that although the role-players and the involved ones
have similar values in skepticism, and use feelings, emotes, and smileys to a
similar extent, the involved MUDders report significantly higher values on the friends
scale. This result indicates that in addition to the ability to verbalize nonverbal
contents, motivation also plays an important role. If the reason for playing MUDs is the
interest in role-playing, development of friendships seems to be a secondary feature.
Although the design of
the study does not allow the testing of causal hypotheses, there is evidence for the
important role of motivations. First, this finding is in line with Walther's (1992, 1994)
assumption and verification of underlying motivational factors. In the present study, in
relationship development, if people are not primarily motivated to get to know other
individuals, they verbalize nonverbal communication to a lesser intent. Secondly, research
on media selection has proven the influence of motives and goals on media choice
(Kayani, Wotring,& Forrest, 1996), while other studies are needed to replicate and
further analyze the connection between motivations, goals, and CMC-use (e.g.
Markus, 1994b). Longitudinal studies are needed to specify temporal developments. An
initial motivation could be supplemented by another one which could result in different
medium use again. The present study can be seen as an early step in studying the influence
of goals and motivations on actual medium use.
A limitation of the
study can be seen in the highly selective sample. First, respondents were self-selected.
This problem is difficult to avoid in research in virtual communities. The MUDders' real
identities are not known; participants cannot be "forced" to fill out a
questionnaire. More important, though, is the question whether participants and
non-participants differ in important characteristics. The high SDs in MUD-time and hours
per week indicate that a broad range of MUDders participated. There were
"Newbies" as well as Wizards, players with only one character, and players with
more characters in different MUDs. The mean of hours played per week is in the same range
as the objectively measured time in another study on German MUDs (Zielke, Wirausky,&
Schildmann, 1995). The results are therefore assumed to be quite representative.
The sample is highly
selective in another aspect as well. Only players of German LPMUDs were studied. Most of
the players are young well educated males. That raises questions about the generalization
of the results found. One argument in favour of the generalization is that similar
dynamics at the gross level have been found in the original SIP and hyperpersonal
experiments which were not based on role-playing. The process specified in SIP should be
the same for different kinds of virtual communities.
Another question
remaining is whether the results presented here can be also generalized to asynchronous
forms of CMC. Maybe individuals are primarily skeptical in synchronous CMC where only a
limited time for choosing the right words is given. Selective self-presentation as
described by Walther (1996) is made much more difficult in synchronous CMC. The messages
may be less thought about, and misunderstandings may therefore be more likely. For a
further testing of this hypothesis, studies in different settings are needed.
The fact that goals and
motivations influence media use should also be true for other environments. Goals will
differ, though; in a newsgroup, role-play is generally not an important goal.
(Information-seeking should be one of the dominant motivations in this case).
The special sample of LP-MUDders determines the percentage of people
who built up online-friendships as well. In more social MOOs the percentage is higher, and
in more task-oriented newsgroups it is lower, as reported above (Parks & Floyd, 1996;
Parks & Roberts, 1998). Thus, the goals studied here do not play the same role in
other virtual communities; attitudes (skepticism) may. Therefore, motivations and
expectations are important factors that should be taken into account in future studies.
The potential of
virtual communities is not yet examined enough. What consequences do friendships have that
are formed in virtual worlds and what consequences does the identification with virtual
communities have for traditional bonds and communities? Will virtual friendships and
communities replace, compensate or complement ftf-relationships? Distinguishing different
groups of users seems to be helpful for answering such questions. Skeptical participants
will not make friends in cyberspace. For people primarily interested in the non-social
aspects of the Internet (e.g. information, fun) friendships are just a side-effect.
Replacement or a substantial complement of traditional bonds is not expected for these
users. Sociables, on the other hand find fertile ground for relating in virtual
communities. For some, as Parks and Roberts (1998) note, these are places for people to
meet and initiate eventual FTF encounters. For others, virtual communities are a safe
place to develop relationships that they may not have in other domains. But, sociable or
not, an individual who goes on line with the expectation that relationships can form on
CMC, does so in this new environment.
References
Aldenderfer, M. S.
& Blashfield, R. K. (1984). Cluster analysis. Newbury Park, CA: Sage.
Becker, B. & Mark, G.
(1998, June). Social conventions in collaborative virtual environments. Paper
presented at the the Collaborative Virtual Environment Conference CVE98,
Manchester.
Daft, R. L. & Lengel,
R. H. (1984). Information richness: A new approach to managerial information processing
and organizational design. In L. L. Cummings & B. M. Staw (Eds.), Research in
organizational behavior (pp. 191-234). Greenwich, CT: JAI.
Daft, R. L. & Lengel,
R. H., & Trevino, L. K. (1987). Message equivocality, media selection, and manager
performance: Implication for information systems. MIS Quarterly, 11, 355-366.
DeSanctis, G. &
Gallupe, R. B. (1987). A foundation for the study of group decision support systems. Management
Science, 33, 589-609.
Fernback, J., &
Thompson, B. (May 1995). Computer-mediated communication and the American collectivity:
The dimensions of community within cyberspace. Paper presented at the annual
convention of the International Communication Association, Albuquerque, New Mexico,
Retrieved May 23, 1997 from the World Wide Web: http://www.well.com/user/hlr/texts/VCcivil.html
Fulk, J., Schmitz, J.,
& Steinfield, C. W. (1990). A social influence model of technology use. In J. Fulk
& C. Steinfield (Eds.), Organizations and communication technology (pp.
117-140). Newbury Park, CA: Sage.
Fulk, J., Schmitz, J.,
& Ryu, D. (1995). Cognitive elements in the social construction of technology. Management
Communication Quarterly, 8, 259-288.
Hiltz, S. R., Johnson, K.,
& Turoff, M. (1986). Experiments in group decision making. Human Communication
Research, 13, 225-252.
Hiltz, S. R., &
Turoff, M. (1981). The evolution of user behavior in a computerized conferencing system. Communications
of the ACM, 24, 739-762.
Kayani, J. M., Wotring, C.
E., & Forrest, E. J. (1996). Relational control and interactive media choice in
technology-mediated communication situations. Human Communication Research, 22, 399-421.
Kiesler, S. (Jan-Feb
1986). The hidden messages in computer-networks. Harvard Business Review, 1, pp.
46-58.
Kiesler, S., Siegel, J.,
& McGuire, T. W. (1984). Social psychological aspects of computer-mediated
communication. American Psychologist, 39, 1123-1134.
Markus, M. L. (1994a).
Electronic mail as the medium of managerial choice. Organization Science, 5, 502-527.
Markus, M. L. (1994b).
Finding a happy medium: Explaining the negative effects of electronic communication on
social life at work. ACM Transactions on Information Systems, 12, 119-149.
McGuire, T. W., Kiesler,
S., & Siegel, J. (1987). Group and computer-mediated discussion effects in risk
decision making. Journal of Personality and Social Psychology, 52, 917-930.
Miller, G. R. &
Steinberg, M. (1975). Between people: A new analysis of interpersonal communication. Palo
Alto,CA: Science Research Associates.
Mummendey, H. D., Riemann,
R., & Schiebel, B. (1983). Entwicklung eines mehrdimensionalen Verfahrens zur
Selbsteinschätzung [Development of a multidimensional method of self-assessment]. Zeitschrift
für personenzentrierte Psychologie und Psychotherapie, 2, 89-98.
Nua Internet Surveys
(2000). How many online? Retrieved February 11, 2000 from the World Wide Web: http://www.nua.ie/surveys/how_many_online/index.html
Parks, M.R. & Floyd, K. (1996). Making
friends in cyberspace. Journal of Communication, 46, 80-97.
Parks, M. R. &
Roberts, L. D. (1998). "Making MOOsic": The development of personal
relationships on-line and a comparison to their off-line counterparts. Journal of
Social and Personal Relationships, 15, 517-537.
Rice, R. E. & Case, D.
(1983). Electronic message systems in the university: a description of use and utility. Journal
of Communication, 33, 131-152.
Rice, R. E. & Love, G.
(1987). Electronic Emotion: Socioemotional Content in a Computer-Mediated Network. Communication
Research, 14, 85-108.
Roberts, L. D., Smith, L.
M. & Pollock, C. (1996). Social Interaction in MOOs: Constraints and Opportunities
of a Text-Based Virtual Environment for Interpersonal Communication. Paper presented
at Virtual Information Digital Workshop, The Centre for Research in Culture and
Communication, Murdoch University, Perth, Western Australia. Retrieved June 4, 1998 from
the World Wide Web: http://psych.curtin.edu.au/people/robertsl/vidtext.htm
Roberts, L. D., Smith, L.
M., & Pollock, C. (1997). "u r a lot bolder on the net": The social use
of text-based virtual environments by shy individuals. Paper presented at the
International Conference on Shyness and Self-consciousness, Cardiff, Wales. Retrieved June
4, 1998 from the World Wide Web: http://psych.curtin.edu.au/people/robertsl/shy.htm
Schmitz, J. & Fulk, J.
(1991). Organizational colleagues, media richness, and electronic mail: A test of the
social influence model. Communication Research, 18, 487-523.
Short, J. A. (1972). Medium of
communication, opinion change, and the solution of a problem of priorities. Unpublished
paper. Communication Studies Group, no. E/72245/SH.
Short, J. A. (1973). The
effects of the medium of communication on persuasion, bargaining and perception of the
other. Unpublished paper. Communication Studies Group, no. E/73100/SH.
Short, J., Williams, E.,
& Christie, B. (1976). The social psychology of telecommunications. London:
Wiley.
Siegel, J., Dubrovsky, V.,
Kiesler, S., & McGuire, T. W. (1986). Group processes in computer-mediated
communication. Organizational behavior and human decision processes, 37, 157-187.
Simon, B. &
Massau, C. (1991). Soziale Identifikation, Eigengruppen-Favorisierung und
Selbst-Stereotypisierung: Der Fall Oskar Lafontaine und die Saarländer [Social
identification, ingroup favoring, and self-stereotyping: The case of Oskar Lafontaine and
the Saarländers]. Zeitschrift für Sozialpsychologie, 22, 193-207.
Spears, R. & Lea, M.
(1992). Social influence and the influence of the "social" in computer-mediated
communication. In M. Lea (Ed.), Contexts of computer-mediated communication (pp.
30-65). London: Harvester-Wheatsheaf.
Sproull, L. & Kiesler,
S. (1986). Reducing social context cues: Electronic mail in organizational communication. Management
Science, 32, 1492-1512.
Straus, S.G. (1997).
Technology, group process, and group outcomes: Testing the connections in
computer-mediated and face-to-face groups. Human-Computer Interaction, 12, 227-266.
Tajfel, H. (1978).
Differentiation between social groups. London: Academic Press.
Trevino, L. K., Daft, R.
L., & Lengel, R. H. (1990). Understanding managers' media choices: A symbolic
interactionist perspective. In J. Fulk & C. Steinfield (Eds.), Organizations and
communication technology (pp. 71-94). Newbury Park, CA: Sage.
Utz, S. (in press). MUDs--Identitätswerkstätten
oder nur Spiele? [MUDs - identity-workshops or only a game?] Soziale Wirklichkeit.
Utz, S. (1999). Soziale
Identifikation mit virtuellen Gemeinschaften Bedingungen und Konsequenzen
[Social identification with virtual communitiesCauses and consequences]. Unpublished
doctoral thesis, Katholische Universität, Eichstätt.
Walther, J. B. (1992).
Interpersonal effects in computer-mediated interaction: A relational perspective. Communication
Research, 19, 52-90.
Walther, J. B. (1993).
Impression development in computer-mediated interaction. Western Journal of
Communication, 57, 381-398.
Walther, J. B. (1994).
Anticipated ongoing interaction versus channel effects on relational communication in
computer-mediated interaction. Human Communication Research, 20, 473-501.
Walther, J. B. (1996).
Computer-mediated communication: Impersonal, interpersonal and hyperpersonal interaction. Communication
Research, 23 (1), 3-43.
Walther, J. B. (1997).
Group and interpersonal effects in international computer-mediated collaboration. Human
Communication Research, 23, 342-369.
Ward, J. H. (1963).
Hierarchical grouping to optimize an objective function. Journal of the American
Statistical Association, 58, 236-244.
Zielke, A., Schildmann,
S., & Wirausky, H. (1995). Spiel- und Sozialverhalten im Morgengrauen [Play and
social behavior in the MUD Morgengrauen]. Retrieved February 28, 2000 from the World Wide
Web: http://www.mud.de/Forschung/verhalten.html
Zimmerman, D. P. (1987).
Effects of computer conferencing on the language use of emotionally disturbed adolescents.
Behavior Research Methods, Instruments & Computers, 19, 224-230.
Appendix 1. Rotated Components Solution
(VARIMAX), With Factor Loadings Above .30
Item |
Component |
1
(Role-Play) |
2 (Game) |
3 (Skepticism) |
|
MUD is only a game for me. |
|
.821 |
|
MUD is connected with real-life, because I
know many MUDders in rl.(-) |
|
-.508 |
|
MUD is not only a game because relationships
to other people are developed.(-) |
|
-.706 |
|
The boundary between MUD and rl fades away
sometimes. (-) |
|
-.751 |
|
When my MUD-character dies, I am also sad in
rl. (-) |
|
-.398 |
|
Behavior and character of my MUD-figures are
very similar to me.(-) |
-.842 |
|
|
With my characters I try to play different
roles which have not much in common with my true personality. |
.802 |
|
|
My behavior in the MUD is different from that
in real-life. |
.814 |
|
|
I identify myself with my MUD-character/s.(-) |
-.587 |
-.400 |
|
I think there are often misunderstandings in
the MUD because I don't see or hear the others. |
|
|
.511 |
I think it is impossible to express feelings
adequately in CMC. |
|
|
.672 |
I think getting to know each other via CMC is
impossible. |
|
|
.801 |
I think getting to know each other via CMC is
possible only under the condition that people are willing to show their true attitudes and
emotions.(-) |
|
|
-.806 |
I am never sure whether the other person is
acting a role in front of me. |
|
.313 |
|
Notes
Note 1 Graphical MUDs exist as
well. At the time of the study there was no running graphical MUD in Germany where the
study took place, and (despite the language problem with a German questionnaire) accessing
foreign sites is often hampered by excessive transmission lag. For insights on graphical
MUDs, see http://www.mudconnector.com/mud_graphical.html
Note 2 telnet: A terminal
emulation program which enables users to log into another computer.
Note 3 Of course, it is
useful to know the size of the population of the three MUDs, but this is difficult to
ascertain. It is only possible to get the number of players. Altogether, there were about
five hundred and fifty active players in the three MUDs at the time of the study. But many
players have more than one character in one MUD, and many of them play more than one MUD.
In this study, 25% of the players reported to play only one character in only one MUD.
Another 25% played more than one character in one MUD, the other half reported playing one
character in several MUDs or several characters in several MUDs. In a more recent study
(Utz, 1999) even more players reported having more than one character. So, the number of
active players should be seen as limiting value, the actual number of MUDders is at best
half. However, the sample covers quite a lot of the population of the three MUDs.
Note 4 In the original German version of the questionnaire, these
categories are somewhat more clear.
Author's Note: Sonja Utz received her Ph.D. in Psychology from the
Catholic University of Eichstadt (Germany) in 1999, where her thesis dealt with
"Social Identification with Virtual Communities: Causes and Consequences."
She is currently a member of the faculty in the Department of Economic,
Organizational, and Social Psychology at Chemnitz University of Technology; after April,
2000, she will be affiliated with the Free University in Amsterdam. Her research
interests include computer-mediated communication, virtual communities, social identity,
cooperation and competition. Dr. Utz's homepage is http://www.tu-chemnitz.de/phil/psych/professuren/sozpsy/Mitarbeiter/Utz/utz.shtml.
The author wishes to thank Kai Sassenberg and two anonymous
reviewers for insightful and critical comments on earlier drafts of the paper; Ruth
Loemke, for language and stylistic assistance; and the 103 MUDders who participated in
this study.
DISCUSS THIS ARTICLE
In citing this document please use the following style:
Utz, S. (2000). Social information processing in MUDs: The development of friendships in virtual worlds. Journal of Online Behavior, 1 (1). Retrieved <date> from the World Wide Web: /JOB/v1n1/utz.html
|
|