Collective Intelligence – problems and possibilities 15/03/2012 Michael Hogan

advertisement
15/03/2012
Collective Intelligence –
problems and possibilities
Michael Hogan
National University of Ireland, Galway
Table 1. Heuristic Maneuvers and their Reasoning Disadvantages
Developmental automaticity, heuristics, and intelligence
Simulation
Availability
Cognitive Maneuver
Given an option that is good
enough decide in favor of that
enough,
option
Take an initial stance in support
of or in opposition to a given
choice consistent with one’s
initial affective response to that
choice
Estimate the likelihood of a
given outcome based on one’s
ease in imagining that outcome
Base the estimate of the
likelihood of a future event on
the vividness or ease of
recalling a similar past event
Disadvantage/Risk
Good enough may not be
best
Feelings may mislead
y=1181.04+-191.576*log10(x)+eps
2400
2000
Over-estimation of one’s
chance of success or
likelihood of failure
Mistaken estimations of
the chances of events
turning out in the future
as they are remembered
Distincts Targets
Name
Satisficing
and
temporizing
Affect
1600
1200
800
400
1
15/03/2012
Developmental automaticity
IT, Information System,
and Semantic Web supports
Lanier argues:
1)The algorithm-driven direction of Web 2.0 is
disempowering individuals and reducing the creativity
of people online
2)The social-semantic web is distorting human
relationships and distancing people from true intimacy.
Lanier does not address social problem solving, or the
pragmatic web. The pragmatic web uses the knowledge
within the social network to facilitate problem solving.
However,
• Collective intelligence within the pragmatic web can never
be an exclusively algorithm-driven process; cultivating critical
thinking, systems thinking, and computational thinking skills
within individuals is important.
p
• Technology can support the development of these thinking
skills and facilitate collective intelligence and collective action
• The social psychology of collective action presents other
real challenges (or problems) that require higher-order socialemotional intelligence
2
15/03/2012
John N. Warfield (1925-2009)
Warfield, past president of the society
for systems science, developed
Interactive Management
•
•
•
•
•
•
Relevant Books
1976. Societal Systems: Planning, Policy, and Complexity. New York: Wiley Interscience.
1990. A Science
S
off Generic
G
Design: Managing Complexity
C
through Systems
S
Design. Ames,
IA: Iowa State University Press 1994.
1994. A Handbook of Interactive Management. With Roxana Cárdenas, Ames, IA: Iowa
State University Press 1994.
2002. Understanding Complexity: Thought and Behavior. AJAR Publishing Company, Palm
Harbor, FL.
2003. The Mathematics of Structure. AJAR Publishing Company, Palm Harbor, FL.
2006. An introduction to systems science. Singapore: World Scientific.
People use different mental models to
describe the same problematic situation.
Hofstadter’s integration of Gödel's
Incompleteness Theorem, the music
of Bach, and the art of Escher
Kurt Fischer (Harvard):
People integrate at different
levels of complexity
3
15/03/2012
Mathematical models can capture significant complexity
but less often facilitate the resolution of social problems.
Jay Forrester’s
World Model.
Forrester’s original World Model (1973) included 58 elements, 81 pair relations, and complex mathematical interdependencies.
1. The model is quantitative, but includes many unstated qualitative assumptions.
2. The model is not the product of consensus and it is not presented in a way that can be readily understood by the public.
3. Many solutions can be generated depending on what assumptions are entered in a particular computer run.
4. Any decisions and solutions anticipated as a consequence of belief in the validity of the model would involve many individuals,
groups, and organizations to implement solutions.
When a group works to resolve a shared problem they need to
build consensus models that facilitate collective action
Tools
Words
Graphics
Mathematics of
logic and structure:
eg., formal logic,
graph theory, matrices
Mathematics
Mathematics of content:
e.g., differential equations,
integral equations
used to describe
phenomena in physics,
chemistry, biology,
psychology, sociology.
4
15/03/2012
What is Interactive Management?
• Integrated design process
• Based
B
d iin th
the th
theory off systems
t
d
design
i
• Specifically developed for dealing with complex
issues
• Goals are to help participants:
– Develop understanding of the issues they face
– Establish collective basis for thinking and working
cooperatively
– Produce framework for effective action.
Essential Components of
IM Process
1.
2.
3.
4.
5.
Participants
Facilitation Team
Group Methodologies
Software Support
Productive Workshop Space
2
1
3
4
5
5
15/03/2012
Group Methodologies
1.
1
2.
3.
4.
5
5.
Generating ideas
Categorizing
Structuring
Sequencing
Developing action agendas
IM Methodologies aim to:
1. Prevent under-conceptualization
2. Avoid cognitive overload
3. Allow informed choice-making
4 Enable clear communication
4.
5. Promote positive group dynamics
6
15/03/2012
What
are
barriers
to ….
Step 1: Idea Generation
Step 2: Recording and posting ideas on
display walls, and clarifying the meaning of
ideas
7
15/03/2012
Problem set and problem
categorisation
(3) Selection and ranking of sub-problems in the problem field
Statement
Number of
votes
Sum of ranks
Category
2. Lack of clear incentives to
4
16
8
23 Clashing
23.
i
personalities
i i
and
4
10
4
12. Challenge of identifying l
3
8
6
4. Lack of identity for the new
3
9
2
17. Uncertainty regarding new
2
7
2
25. Lack of reward systems to
2
6
8
9. Difficulty in defining clust
2
6
1
24. Unrecognized value of soci
2
7
2
5. Specialization (mitigates ag
2
6
5
7. Lack of clear language that
2
6
5
19. Overdependence on "bureauc
2
4
6
22. Some individuals want to w
2
2
4
3. Lack of motivation or intere
2
7
7
13. Lack of opportunity for fo
1
3
3
26. Turf issues: individuals w
1
5
4
32. Someone needs to commit si
1
4
6
20. Divergence in methods, pro
1
5
5
28. Not really an existing, re
1
4
3
33. Institute based on what we
1
2
6
14. Lack of information/certai
1
1
5
15. Lack of translation of res
1
2
8
-----------------------------------------------------------------------
8
15/03/2012
(4) Structuring
the interdependence between problems in the problem field
From voting, to matrix, to structure
Figure C-1 shows a Boolean matrix (i.e., a matrix, all of whose entries are either 0 or 1). The
matrix rows and columns are identically indexed, the index numbers ranging from 1 to 15. Such a
matrix might represent the information provided to the computer by a group of participants, in
which e.g., 15 problems are interrelated (or possibly 15 options, or 15 events, etc.). The entries in
the matrix could represent a mix of the answers to questions posed to the participants and the
answers inferred by the computer (using a property of relationships known as transitivity).
9
15/03/2012
(5) Group analysis of emergent influence structure
Lack of clear
incentives to
work together
and across
disciplines (2)
Turf issues:
individuals want
to maintain their
distinctivness or
be seen as the
experts in an
issue area (26)
Clashing personalities
and histories (23)
O d
Overdependence
d
on "bureaucratic
democracy" without
vision and collective
design (19)
Specialization
(mitigates against
collaboration) (5)
Lack of identity
for the new
institute (4)
Uncertainty regarding
new dynamics,
d
i
management, activities,
in moving beyond the
past (old unit) and
moving towards a new
institute (17)
Lack of reward
systems to work on
real social problems
- the challenge of
bridging basic and
applied research
goals (25)
Challenge of
identifying leaders
and mentors to
drive/facilitate
cluster, thematic
and
d iinstitute
tit t goals
l
(12)
Difficulty in defining
cluster objectives that
align with national and
EU priorities (9)
Lack of
motivation
or interest
in change
(3)
Divergence in
methods,
problem
definition,
outcomes,
between
disciplines
(in/between
clusters). (20)
Unrecognized value of
social sciences at NUI,
Galway (24)
Lack of clear language
that communicates with
outside world and across
disciplines (7)
Some individuals
want to work alone
rather than as part of
a team (22)
(6) Generate options as a response
to the structural problems
Multiple actors work in
multiple directions
10
15/03/2012
Sample Influence Structure:
Obstacles to Effective Communication in Group Problem Solving
Failure to
remain
focused on
the topic
(21){J}
Lack of ownership
of thee problem
o
p ob e or
o thee
objective (58)
Failure to
allocate
sufficient time
to the group
effort (24)
{E}
Nonconducive
physical
meeting
eenvironment
v o e
(32)
{D}
• Lack of
clearly
understood or
agreed purpose
(27)
Inadequate
process for
facilitation
(70)
{J}
• Failure to
utilize
appropriate
methodologies
(102)
{D, C, A}
The Arrow Should be Interpreted as:
“Significantly Aggravates”
• Makes it more difficult to resolve
• Increases the Severity of
• Makes Worse
• Exacerbates
Organizationa
l politics
inside and
outside the
group (101)
{F}
• Conventional
wisdom that groups
or meetings are a
waste of time (28)
Demand for
quick
resolution
and
d singular
i l
solution (95)
{F}
• Failure to
involve the
head, heart,
and body
(76)
•
"Groupthink"
(35)
• Lack of
trust among
the members
(6)
• Lack of necessary
or appropriate
information (23)
Lack of
cultural
sensitivity
((59))
{B}
{H, J, H}
Lack of clarity
on who is the
client (63)
{C}
• To help build a country where everybody's needs
are everybody's concern (E-1)
• To try and create a proper climate where Turkish
Cypriots will ask for Greek Cypriots whatever they
want for themselves and where G/Cs will refuse
anything for T/Cs that they don't want for
themselves (D-10)
•
Dominatio
n of group
discussion
by strong
personaliti
es (14)
• Lack of
effective
communic
ation
norms (20)
{G, G}
• Failure to
suspend
assumptions
(48)
• (Refusal to
recognize
difference in
mental
models) (15)
{G, G}
• Hesitancy to
bring out sensitive,
undiscussable
issues (13)
• Failure to involve
correct
people/stakeholders
(37)
• Presence of
hidden agendas or
ulterior motives (9)
{F, D, H, E, I}
COMM 531
Spring 1999 Design
Sessions
Updated 2/9/99
• To promote the idea that the security problem
of both communities could be lessened by
demilitarization (K-26)
• To come up with a solution that will satisfy
the security needs of both communities (K-41)
• To minimize the effects of outside powers
(K-29)
• To empower Cypriots to envision a Cyprus where people live in peace and
to believe that this is possible now (D-2)
Collective Vision Statemen
C
nt
• To learn to accept differences in culture and start thinking of them as
positive, interesting, and enriching factors (E-3)
To promote in both
communities a better
understanding of
basic needs, fears, and
aspirations of each
community
concerning the
property issue (H-17)
• To work for an independent bicommunal bizonal federally united Cyprus
with full respect to human needs and rights and develop the appropriate
public mentality to support it (K-43)
• To make known and acknowledge the basic needs, fears and aspirations of
each community to the other community (H-21)
• To eliminate the "enemy image" (E-11)
• To stimulate the idea of freedom of movement throughout the island (K-19)
• To encourage youths of both communities to get to know each other (I-36)
• To accept and respect each others' identity as T/Cs and G/Cs and see this as
a richness rather than an obstacle to peace building (G-32)
• To develop public confidence in the peace building
process (A-4)
To minimize the effects of
national chauvinism and
promote the idea of a polyethnic
i l i federal
inclusive
f d l society
i t (G-14)
(G 14)
To try and find ways to bridge the
economic differences between the
two communities (K-37)
• To strengthen the peace movements on the citizen
level (B-12)
• To promote the idea in G/C community that the
existence problem of T/C community is vital for the
whole of Cyprus (H
(H-18)
18)
• To help the two communities create a common vision
(H-27)
• To build a climate in which the ethnic identities of the two
communities are not threatened (G-6)
• To support in both communities the idea of joining the
European Union (K-28)
• To provide opportunities for interaction between the two communities (I-8)
• To build intercommunal institutions and centers (I-9)
• To promote bi-communal commercial, health, educational and sports activities (I-15)
• To establish an effective bicommunal peace movement (B-30)
• To promote cultural and social relations between the two communities (C-16)
• To promote both communities working together on common projects (I-7)
To build
bridges of
mutual
empowerment
and understand
with Track I
people (C-42)
11
15/03/2012
12
15/03/2012
13
15/03/2012
Applications
IAS
Individual
as system
IAS
Family/
Small
group
Government planning
and policy and the
Wisdom of the crowd
IAS
Community
problem solving
Relational training
Cooperative learning In schools
Product design
Organizational
g
design
g
Business management
Functional social networking
Google experiments
(global research aided by google support
and google consultants)
Google G Earth
(representation of groups and avatar transport)
Universities
(basic and applied science;
Relational knowledge import;
Web of knowledge)
Google rewards for group
participation
Dynamic Groups
(dynamic update of problem/option structures,
group comparison, group merger)
14
15/03/2012
Facilitating Cross-Cluster
Connectivity
Political Innovation
Economics of Health and Ageing
Applied Social Science
Health, Ageing and Lifespan Development
Group Decision Making
Sustainability Groups
Agility
Performance Management
Systems Science Education
•
Problematiques generated in IM
sessions are like structural
equation models (SEMs) without
effect sizes included
•
However, effect sizes can be
imported for relations in IM
structures and thus allow for model
fit indices to be generated (i.e.,
computation science link).
•
Also, each relation in an IM
structure is a claim and the logic of
argumentation linked to each claim
can be worked out in an argument
map (AM).
•
AM training facilitates development
of critical thinking skill. IM training
facilitates development of systems
thinking/action abilities. SEM
develops computational skills.
15
15/03/2012
The Social Psychology of Effective Teams:
The Big Five
1.
2.
3
3.
4.
5.
Team Orientation – a preference for working with others and also a tendency to enhance individual
performance through the coordination, evaluation, and utilization of task inputs from other members.
Mutual Performance Monitoring – Effective teams are comprised of members who maintain an
awareness of team functioning by monitoring fellow members work in an effort to catch slips/mistakes
Backup Behaviour Management – Ability to anticipate other team members needs through accurate
knowledge about their responsibilities. Includes ability to shift workload among members to achieve
balance during high periods of workload or pressure
Adaptability – Ability to adjust strategies based on information gathered from the environment and
through the use of backup behaviour and reallocation of intra-team resources.
Team Leadership – Ability to direct and coordinate team activities, assess performance, assign
tasks, develop team knowledge, skills, and abilities, motivate team members, plan and organize, and
establish a positive atmosphere
Supporting Coordinating Mechanisms
A. Mutual Trust – The shared belief that team members will perform their roles and protect the interests
of their teammates.
B. Shared Mental Models – An organizing knowledge structure of relationships among the task the
team in engaged in and how the team members will interact
C. Closed-loop communication – The exchange of information between a sender and a receiver
irrespective of the medium. Involves (a) the sender initiating a message, (b) the receiver receiving the
message, interpreting it, and acknowledging its receipt, and (c) the sender following up to insure the
intended message was received
16
15/03/2012
Small Group Research
Psychodynamics of ‘open’ versus ‘closed’ groups
• We asked two groups of student to map
interdependencies between the costs and benefits of
online social media using IM.
• Participants were divided into high and low dispositional
trust by means of a median split based on their prescores on dispositional trust.
• Participants high and low on trust were randomly
assigned to open or closed voting group.
Results
• Main effect of trust on perceived consensus, F(1,26) = 8.43, p =
.007, with perceived consensus higher in the high trust group (M
14.6,
6 SD = 2.75)
2 75) than in the low trust group (M =12
12.0,
0 SD = 2.27).
2 27)
=14
There was also a main effect of condition on perceived
consensus, F(1,26) = 7.50, p = .011, with perceived consensus
higher in the open group (M =14.53, SD = 2.53) than in the closed
group (M =12.07, SD = 2.58).
• Main effect of trust level on perceived efficacy of IM, F(1,26) =
4.34, p = .047, with higher perceived efficacy in the high trust group
(M = 23.27, SD = 4.40) than in the low trust group (M = 20.07, SD =
3 94) Also
3.94).
Also, higher perceived efficacy in the open group (M =
23.53, SD = 3.23) than in the closed group (M = 19.80, SD = 4.74).
• There was also a trend whereby trust scores tended be to lower in
the low trust group after closed voting compared with after open
voting.
17
15/03/2012
Finally
• We are developing new IM software that
will
ill b
be available
il bl as stand-alone
t d l
software
ft
and networked app.
• We aim to build capacity in IM use for
basic and applied research purposes
• How might you use IM in your
basic/applied research work?
18
Download