Transcript Slide - VW - Indiana University

Computer-Mediated
Social Networks
Networks and Complex Systems
Talk Series, Indiana University
November 2004
Caroline Haythornthwaite
[email protected]
Graduate School of Library and Information Science
University of Illinois at Urbana-Champaign
Computer-Mediated Social Networks
►
Networks created, maintained, and dissolved by communication via
computer-mediated communication
 Email, listservs, webboards, online chat, instant messaging, text
messaging, MUD & MOOs, the Internet, etc.
►
Communication
 Who uses computer-mediated communication (CMC), for what, why, and
with whom?
 What media are used to communicate what kinds of information? And to
whom?
►
Life online
 Are online ties as “real” as offline ones?
 How do we work together, learn, create new knowledge at a distance and
via CMC, with people we rarely or never meet face-to-face?
 How do we design systems or recommend practices to support work,
learning, and community online, and at a distance?
►
Social and technical impacts
 How does being online change how we work and learn together?
 How will wide-spread online communication change how we interact, and
with whom we create and maintain relationships?
 What happens if media are removed and/or access denied?
Recognizing complexities
► Multiple
simultaneous influences
 Context, social construction, norms, practices,
distributed practices
 Theoretical perspectives:
► social
► New
construction, structuration, and adaptive structuration;
socio-technical systems perspective; information ecology; social
worlds; social and organizational informatics; social networks
forms of organizing
 made possible by the development of CMC and ICT
technologies
► Far
reaching impact of “the Internet”
 the connectedness created by the juxtaposition of
technical infrastructures and social uses
 What happens as a result of laying media and Internet
connections?
Traditional Approaches
►
Media and the people we communicate with are primarily
dealt with as singular
 Examining use of Email OR listservs OR chat; work-only groups, or
play-only groups
 Face-to-face is still treated as the ideal standard
► Media
declared as better or worse by the number of cues the transmit
rather than how they are used
 Message-medium fit
► Lean
messages should be sent via lean media
 Little recognition of context and variation in use across contexts
►
But, we …




Use multiple means of communication to effect relationships
Prefer rich media for all communications
Use media differently from our preferences
Communicate with people in many walks of life, in many roles, in
many different relationships with us
Complexity of CMC
► Media
 Multiple media available for communication and
maintenance of relations
► People
 Pairs of actors in relationships differing by type
(friendship; work), strength, duration, etc.
 Multiple relations tieing actors, with the set of relations
varying by relationship
► Contexts
 Distribution, across time, space, knowledge base, facility
with technology, interests, backgrounds, etc.
 External constraints on schedules, tasks to be
completed, media available and/or mandated for use
 Multiple calls on time and attention
My research
►
Adds media into the network equation & networks into the media
equation
 How can we support work, learning, knowledge, and community formation
at a distance, via the Internet and through computer-mediated
conversations?
►
Who talks to whom, about what, and via which media?
 What kinds of relations make up work and learning ties?
 How are these supported via all available means of communication?
 What patterns of connectivity emerge among group members because of
what they talk about and what media they use?
►
Connectivity effects
 What do these patterns of connectivity reveal about the group and/or
about its media use?
 What happens as a result of laying media and Internet connections?
►
Social and technological interactions
 How do communication media structure the way we interact with others?
Studies of who talks to whom,
about what, and via which media ?
►Cerise
 Co-located, academic researchers
►LEEP
 Distributed, distance learners
► Interdisciplinary
teams
 Distributed collaborative research teams (in
progress)
Cerise: Co-located, academic researchers
►
Media
 Unscheduled hallway meetings, scheduled classes and research
meetings, email (home and office), phone, fax,
videoconferencing system under development
►
Relationships
 Formal or informal working relations, acquaintance to close
friend
►
Questionnaire
 25 respondents (of 35 member group) answered 24 questions
about a variety of work and social interactions with 10-20 others
(data for a total of 378 pairs)
 Six dimensions of work and social interaction:
► Receiving
work (engaged in by 57% of pairs)
Giving work (57%), Collaborative Writing (32%)
Computer Programming (56%), Sociability (86%)
Major Emotional Support (7% of pairs)
Example of question format
TYPE OF INTERACTION
Group Members: 1
2
3
...
20
How often have you received instructions (i.e., exact
directions on what work to do) from this person?
in unscheduled face-to-face meetings
in scheduled face-to-face meetings
by telephone
by fax
by electronic mail
by videoconferencing
How often: D for daily W for Weekly M for Monthly Y for Yearly 0 for never
For in between amounts use e.g., 2D for twice a day, 6Y for six times a year
LEEP: Distributed, distance learners
►
Media
 Internet Relay Chat (IRC) for classes, ‘live’ office hours, subgroup discussions, and whispering; Real-Audio for instructor
delivery of lectures; Webboards for classes and the program;
 Email, Phone, Face-to-face once a semester
►
Questionnaire
 4 classes of 14-23 (2 classes studied over time)
 4 questions: Collaboration on Class Work, Exchanging
Information or Advice about Class Work, Socializing, Exchanging
Emotional Support
►
Interviews
 Students from across the program
 Exploring characteristics of online community, learning to be part
of an online program and community, social support, coping with
technology, impacts of being “at school” at home
Major Results
► Media
use is associated with the strength of
the tie
► Media use is not associated with the content
of the message
► Media choice is not arbitrary
 follows a unidimensional scale
► Patterns
of ties and media describe tiers of
media use supporting networks of different
tie strength
Media use is associated with the
strength of the tie in number
► The
intensity of the work tie and closeness of the
friendship are associated with
 Higher frequency of interaction
* expected
► Higher
frequency overall, for each relation, and via each
medium
 Maintenance of more relations
(relational multiplexity)
* expected
 Use of more of the available media
(media multiplexity)
* unexpected
Found for Cerise and Leep; also found by Koku, Nazer & Wellman
(2001) for distributed scholars
Media use is associated with the
strength of the tie in adoption
►
Within a group, media appear in pairs’ communication
repertoires in a similar order
 Those who use only one medium, use the same one medium
Those who use two, tend to add the same second medium, etc.
►
Cerise
 (1) Face-to-face Unscheduled meetings (2) Scheduled meetings
(3) Email then (4) ‘Other’ media (phone +/or fax +/or
videoconference) (Guttman scaling: CR=.92)
►
LEEP (two classes)
 IRC, Webboard, Email, then Phone (CR=.99)
 IRC, Email, then Phone (CR=.94)
CR=coefficient of reproducibility;
10% cutoff (CR=.90) accepted as indication of a fit
Media use is *not* associated with the
content of the communication
► What
pairs communicate about *does*
differ by the type of tie (self-reported)
 Work-only pairs communicate about work relations
 Pairs who combine work and friendship communicate
about both work and social relations
 Friends include more frequent communication and
more emotional and social communication than nonfriends
► BUT,
they do not allocate communications
of different types to different media
 i.e., no support for message-medium fit
Two Patterns of Media Use Emerge
► The
ordered use of media by tie strength
leads to media-based group-wide
networks that are also tie-strength related
► Patterns emerge over time, and vary with
local use and purpose
► Illustrated
in the following sociograms
 Note the differences across media, and across
time (approx. one month apart)
F97: Collaborative work via IRC and Email by Time
Internet Relay Chat
Time 1
Time 2
Time 3
Email
Group projects; Webboard also used for discussion, connected all to all
Communication Networks x Medium
LEEP: Distance students,
end of semester
Internet Relay Chat
Guttman scaling, overall
communication all term
(CR=.99): IRC, Webboard,
Email, then Phone
[nb: sociograms show end of
term]
Webboards
Email
F98: All communications, IRC and Email by Time
Internet Relay Chat
Time 1
Time 2
Time 3
Email
No group project; Rotating pairs for presentations; Webboard use started
but abandoned in this class; it connected very few after abandonment
Communication Networks x Medium
Cerise: co-located
computer scientists,
socializing
Network densities:
.32, .13, .06
Unscheduled Meetings
Guttman scaling, overall
communication (CR=.92):
face-to-face Unscheduled
meetings; Scheduled meetings;
Email; Other [nb: sociograms
show the socializing networks]
Scheduled Meetings
Email
Cerise: Overall Work, and Socializing by Medium
Overall Work
Unscheduled
Scheduled
Email
Socializing
Unscheduled
Scheduled
Email
Two Patterns of Media Use
► Wide
connectivity
► Selective connectivity,
with low frequency of
with higher frequency of
communication
communication
 Webboard +/or IRC
 Email, Phone
 class-mandated media
 optional media
 class-wide, public
 person-to-person, private
 communicate with the
class as a whole
 communicate with friends
and project work mates
Theory
► Does
this pattern of media use suggest
some simplification – a general rule – that
can explain the apparently complex
behaviors?
Theory: A new medium …
► Creates
Latent Ties
 A connection that is available technically even if not yet
activated socially
► Recasts
Weak Ties and Weak Tie Networks
 A new medium connects formerly disconnected others
thereby creating new weak ties
 A change in the medium supporting weak ties breaks
existing weak ties, disrupting existing weak tie networks
► Causes
minimal change to Strong Ties and Strong
Tie Networks
 Strong ties carry on through other media
 Strongly tied pairs adopt, adapt or jointly resist new
media
Creates Latent Ties
► By
providing a technical means of connectivity,
social connection is now possible
 e.g., a physical location, a group meeting, a
webboard, a chat room, an email listserv, an online
environment
► But
… ties are not activated, i.e., converted from
latent to weak, until some sort of social
interaction occurs
 e.g., by attending a group wide meeting, by reading
the webboard, by posting to a bulletin board
► NB:
This latent tie structure is likely to be set up
by an authority beyond the individuals affected.
Recasts Weak Ties and Weak Tie Networks
► Integrative
Effects
 Connecting disparate others creates weak ties
 Change can be
► Technical,
e.g., creation of a community network;
Administrative, e.g., creation of an electronic list for
organization wide discussion; Social, e.g., creating a web
environment for interest-based discussions
► Disintegrative
Effects
 Removing passively accepted connectivity breaks weak
ties
 Change can be
► Technical,
e.g., removal of network connections; Administrative,
e.g., removal of access to lists and or removal from a list;
Social, e.g., from a ftf meeting to a listserv for discussion, from
asynchronous discussion mode to synchronous
Integrative effects: A new medium creates
connections among weak ties
F97: Internet Relay Chat
R2
R13
R15
R3
R4
R9
R5
R8
R10
R7
R2
R3
R12
R11
R4
R9
R10
R5
R9
R3
R2
R10
R5
R12
R13
R12
R6
R14
R14
R6
R4
R15
R14
R8
R11
R6
R11
R13
R15
R7
R7
End of Month 1
End of Month 2
R8
End of Month 3
Increase in connectivity via IRC over time; IRC supported weak ties
as determined by to where IRC fit in the unidimensional scale, and
by the frequency of communication typically found for IRC
Disintegrative effects: A change in the medium
supporting weak-ties breaks existing weak ties
F98: Webboard (non-)Use
End of Month 1
End of Month 2
End of Month 3
Rapid decline in Webboard after official class use was terminated.
Note the loss of connectivity among the weakly tied pairs.
Causes minimal change to
Strong Ties and Strong Tie Networks
► If
there is a change in a medium used by strong
ties, such ties can carry on through other media
 e.g., strongly tied work pairs can carry on through
Email if IRC is removed, or vice-versa
► Because
of their need or desire to communicate
 They are likely to adopt an extra medium if it is useful
for maintaining relations important to the tie
► Because
of their influence on each other
 They can jointly adapt the use of a new medium to
be useful for the tie, or resist its use if it does not suit
their needs or ways of working
Implications
► Adding
media use to characteristics of ties
► Revisiting conflicting results about CMC as
arguments about maintaining ties
► Resolving conflicting results about the integrative
and disintegrative effects of CMC
► Support of static and dynamic needs of groups
► Choices about media implementation
► Wider impacts of media connectivity
Add Media Use to Tie Strength
Characteristics
Weak Ties …
► Media Use
… Strong Ties
► Media Use
 Use multiple means of
communication
 Use organizationallyestablished media as a base
on which they add other
media
 Use private, person-to-person
communication (e.g., use
email very frequently)
 Use few means of
communication
 Use organizationallyestablished media
 Communicate infrequently
via the one to two media
they use
►
Whole network support
 Weak tie networks were
created and sustained via
the mandated, organization
or group-wide media
►
Whole network Support
 Strongly tie networks are
supported through both
mandated media and other
optional, more private, means
of communication
Revisiting conflicting results about CMC
Arguments against CMC
Lean
► Text-based, reduced cues
► Ill-suited to emotional,
expressive, complex
communications
Disintegrative
► Anti-social flaming
► Decreased social involvement
(Nie, 2001)
► Abandonment of local
relationships (Kraut, et al,
1998)
Arguments for CMC
Rich
Emoticons and acronyms (McLaughlin,
et al, 1995)
► Group defined genres and rules of
conduct (Orlikowski & Yates, 1994;
Bregman & Haythornthwaite, 2003)
► Interpersonal self-disclosure, emotional
support; online communities (e.g.,
Haythornthwaite et al, 2000; Baym
2001)
►
Integrative
Connecting disparate others: Bringing
in peripheral players, spanning time and
space (Sproull & Kiesler, 1991)
► Maintaining connections even when
distributed (LaRose, Eastin & Gregg,
2001; Hampton & Wellman, 2002)
►
Revisiting conflicting results about CMC
as arguments about maintaining ties
Arguments against CMC
Lean
► Inadequate for the
emotional and intellectual
interactions that support
strong ties
Disintegrative
► new media are disrupting
existing networks of
communication
Arguments for CMC
Rich
► Strong tie needs
reintroduced by those
with strong desire to
communicate
Integrative
► new media are creating
networks of
communication
Implications for Support Groups :
Static Demands
► Information
exchange for innovation, learning,
exposure to new ideas
 The strength of weak ties
 The strength of media group-wide, mandated, public
media
► Task
completion related to working together,
joint problem solving, projects, etc.
 The strength of strong ties
 The strength of person-to-person, optional, private
media
Implications for Support of Groups:
Dynamic Demands
►
Support first time connections
 Organizationally established, group-wide
media lay the “latent tie networks”
►
►
Need to provide the technical and social means
for initial contact
Support growth of tie strength
►
Need to provide social and technical interventions
such as public and private means of
communication, and opportunities for social and
emotional interactions that build strong ties
Implications for Choices about
Media Implementation
► Organizational,
administrative, governmental
choices have greatest impact on weak tie
networks
 Which medium is used across the group as a whole
depends on what is organizationally mandated
 Choices about group-wide means of communication
lay the groundwork for weak tie contact
 Which medium supports weak tie networks will vary
according to local conditions
► Choices
are both technical and social
 Interventions, mandates, support for use affect how
much the new media changes, supports and serves
the weak tie network
Implications for Wider Impacts
► Internet
as groundwork for latent ties
 Provides easy way for groups to adopt peerto-peer communication
 Weak ties emerge based on interest
►Activists,
scholars, support groups, Usenet
discussion groups, online courses and degree
programs
 Weak ties can grow into stronger ties
►Online
communities, community networks
►Adding new connections: face-to-face meetings,
synchronous online meetings, private email added
to public discussion
Implications for Wider Impacts
► Latent
ties connections get appropriated
 Hijacked
►Viruses
use your email lists to spread
 Borrowed/Mined
►Address
lists used to create business contact
networks
 Severed
►Viruses
that bring down systems
►Job changes that lead to loss of network
connection and contacts
Summary
► Ties
Matter . . .
 Linear tie strength has non-linear effect on CMCsupported social networks
 Different media effects reconcile by considering the
strength of tie between communicators
► Media
Matter . . .
 Organizationally established and sanctioned media
provide latent and weak tie connectivity, and a base
on which ties can grow
 Changes will impact weak ties more than strong
 Impact of changes will depend *not* on what
medium is added or removed, but on *what niche*
the medium fills (e.g., public or private)
Summary
► Time
and Timing Matter
 Ties grow over time, and their needs change over
time
 Media use differs over time as ties grow or change
► Ties
x Media
 Tiers of media support networks of different tie
strengths
 Changes in media will have a more permanent effect
of weak ties than strong ties
The End
►
Haythornthwaite, C. (2002). Strong, weak and latent ties and the impact of new
media.The Information Society,18(5), 1-17.
►
___ (2003). Supporting distributed relationships: Social networks of relations and
media use over time. Electronic Journal of Communication, 13(1).
http://www.cios.org/getfile/haythorn_v13n1
►
___ (2001). Exploring multiplexity: Social network structures in a computersupported distance learning class. The Information Society, 17(3), 211-226.
►
____(2000). Online personal networks: Size, composition and media use among
distance learners. New Media and Society, 2(2), 195-226.
____(2002). Building social networks via computer networks. In K.A. Renninger & W.
Shumar, Building Virtual Communities (pp.159-190). Cambridge, UK: Cambridge
University Press.
► ____ (forthcoming, 2005). Social networks and Internet connectivity effects.
Information, Communication and Society.
►
►
Haythornthwaite, C. & Hagar, C. (2004). The social worlds of the web. Annual Review
of Information Science and Technology, 39, 311-346.
►
Wellman, B. & Haythornthwaite (Eds.) (2002). The Internet in Everyday Life. Oxford
UK: Blackwell.
► EXTRAS
BEYOND HERE
The Meeting Room
Meet and Greet
Down to Business
After Hours
The Meeting Room
Meetings = a socio-technical means
of connecting people. Meeting called
by authority beyond meeting
members. Authority establishes who
will be there, the agenda, how
interactions will proceed.
Meet and Greet
Latent ties existing because
of joint membership in the
meeting are activated into
weak ties by introductions,
exchange of personal
history.
Down to Business
Weak ties grow stronger as
people learn to work together,
learning more about each other,
sharing information, completing
tasks.
After Hours
Strong ties are reinforced as
weak, working ties are
extended to include
experiences outside work, in
different settings, and more
personal exchanges.
The Online Chat Room
Chat Room = a socio-technical means
of connecting people. Chat technology
and Chat room orientation established
by authority beyond chat room
participants. Authorities dictate who
can join, what they will be allowed to
post, how interactions will proceed.
Meet and Greet
Latent ties existing because
of joint presence in the
chat room are activated
into weak ties by
introductions, exchange of
personal history.
Down to Business
Weak ties grow stronger as
people learn to work or play
together, learning more about
each other, sharing information,
completing tasks.
After Hours
Strong ties are reinforced as
weak work or play ties are
extended to include
experiences offline, in different
settings, and more personal
exchanges.