Kenneth Frank, College of Education and Fisheries and Wildlife

1

• Kenneth Frank, College of Education and Fisheries
  and Wildlife
• Kyle Fahrbach
• Frank, K.A., and Fahrbach, K. 1999.
  Organizational Culture as a Complex System
  Balance and Information in Models of Influence
  and Selection. Special issue of Organization
  Science on Chaos and Complexity in Organization.
  Organization Science, 10(3) 253-277.
• All papers available at http//www.msu.edu/kenfr
  ank/

2
What Are Social Networks?

• A set of actors and the ties or relations among
  them.
• Close colleagues (relation) among teachers
  (actors)
• Biomass exchange (relation) among predators and
  prey (actors)
• Economic Exchange (relation) between
  multinational corporations or countries (actors)
• referrals (tie) among social service agencies
  (actors)

3
Definition of Globalization

• An increase in the rate of exchange of resources
  and information across geographic regions and
  cultures
• Not a new phenomenon
• We are increasingly globalized

4
The Human network mediates

• Globalization based on network of economic and
  resource exchange
• Human network of social relations and resource
  allocations
• Information as resource
• Ecological network of predator prey
• Frank, K.A., Mueller, K., Krause, A. and Taylor,
  W. (2005). "The Intersection of Global Trade,
  Social Networks, and Fisheries." To be published
  in a book by Cambridge Press, edited by William
  Taylor and Micheal Schecter editors. Cambridge
  University Press.

5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
Complexity of human network

• Dual Models
• Influence how actors are influenced via
  interaction with others
• Diffusion of innovation
• Spread of toxin through food chain
• Selection how actors choose with whom to
  interact
• Who to give information to?
• Which species to consume?
• Dual Mechanisms in Human Systems
• Information
• Influence Persuaded by information
• Selection seek new information
• Balance
• Influence conform to norm
• Selection seek others like yourself homophily
  (i.e., birds of a feather flock together)

12
Models x Mechanisms
13
Basic Influence Model
kii indicates extent of relation between i and
i, as perceived by i. yit represents an
attitude or behavior or sentiment of actor i at
time t ?i yit-1, the sum of the attributes of
others to whom actor i is related at t-1. Errors
are assumed iid normal, with mean zero and
variance (s2). Thus ? represents the extent of
influence of others on an actor And ? represents
the extent to which actors retain their own
beliefs or behaviors from one time to the next.
where
14
Influence Graphical Representation
15
Information as a Unique Resource

• No conservation of resource
• When A tells B, both have it, it doubles
• Unlike biomass which is consumed (with caveats)
• Uniquely human implications for system

16
Systemic Implications of Influence Model When all
paths are Operative Exploding Beliefs
17
Phase Portrait Convergence of System to y1 y2
18
Source of Explosions

• Explosions occur because information recirculates
• Must separate out informational paths from
  others redundancy

19
Redundant Effects through A Network
20
Influence Model with All Paths
21
Influence Model Differentiating maximal Paths
from All Others
mpiit-v?t represents the maximal path from
experience of i at time t-v to I
(eit-v) qiit-v?t-mpiit-v?t represents all
paths other than the maximum the redundant
paths d governs influence through redundant
paths not informational. d is KEY
22
Systemic Implications of Influence Model When
only Maximal Paths are Operative
23
Systemic Properties

• Can have exploding or stable equilibrium
• But very stable and non realistic
• No feedback loops
• Feedback through selection as actors choose with
  whom to interact

24
Basic Selection Model
In words Interaction between t-1 and t
(kiit-t?t ) is a function of absolute value of
difference in previous sentiments (a(k)yi t-1
yi t-1 ) prior tendency to interact
(?(k)kiit-t?t )
25
Model of Selection with Pursuit of new information
a(k) effect of pursuing similar others ?(k)
effect of continuity from previous interaction ?
effect of pursuing maximal paths to new
information
26
Oscillation Produced by Internal Dynamics of
System Including Selection and Influence, Balance
and Information
27
Technical Interpretation of systemic
implications complex

• Bounded, but never completely defined
• Sensitive to initial conditions
• Multiple equilibria depending on parameters
• Feedback loops
• Positive Normative influence
• Negative seeking new information
• Internal Dynamics produced complex behavior

28
Social Implications of Models

• Humans influenced by both information and
  conformity
• Humans might seek information
• Inherent feedback loops
• Do not assume static
• What do you think?

29
The Role of the Change agent Administering
Exogenous Shocks

• Consensus inducing
• Try to bring certain actors together
• Conflict management
• Factionalization
• Spread out actors
• Divide and conquer

30
Consensus Inducing and Factionalizing Shocks
31
Implications of systemic shocks to beliefs

• Equilibrium reemerges regardless of consensus
  inducing or factionalization

32
The Timing of the Shock

• Induce Consensus when it is most likely to emerge?

33
Consensus Inducing Shock when Consensus is
Emerging
34
Implications of systemic shocks to beliefs

• Timing matters shock induced when system is most
  consensual brings actor 3 close enough to 4 and 5
  to create a subgroup ? factionalization
• Counter intuitive

35
Implications for administering exogenous shocks

• Shocks to a single actors beliefs or control of
  resources
• Diffusion of Innovation
• Where to cultivate or plant the new idea?

36
(No Transcript)
37
Implications of shocks to individual actors

• Most sensitive to actor 3
• Actor 3 bridges between others
• Actor 3 does not have highest centrality in terms
  of common static measures
• Actor 3 has neutral predisposition to belief
• Dont recruit just the people most likely to
  appreciate your innovation
• Consider social location
• Dynamic conceptualization of centrality,
• account for influence, selection, systemic
  dynamics

38
Conclusion

• Agent based modeling people are agents
• Influence
• Conformity
• Information
• People pursue (through selection)
• Balance (homophily birds of a feather)
• Information
• ? Changes in interaction as a result of influence
• ? unintended consequences
• Consider subgroups
• Concetration of interaction within subgroups
• Similar beliefs within subgroups
• Mediators, or bridgers (actor 3) between
  subgroups key to system
• Target innovations for mediators
• Likely moderate in their beliefs
• Cultivate mediators or bridgers
• What makes a complex system an organization?

39
Guiding Pragmatic Questions

• How does an innovation diffuse throughout a
  school?
• Frank, K. A. and Zhao, Y. (2004). "Subgroups as a
  Meso-Level Entity in the Social Organization of
  Schools." Chapter 10, pages 279-318. Book
  honoring Charles Bidwell's retirement, edited by
  Larry Hedges and Barbara Schneider. New York
  Sage publications

40
(No Transcript)
41
Access to Resources and Expertise
Ripple around A indicates increase in use between
time 1 and time 2
B
provides help to
C
provides help to
A
provides help to
D
Change in A is a function of interaction with
people with expertise the greater the mean
expertise of Helpers B, C, and D, the greater the
change in A
42
Ripple Plot

• Overlay talk about technology on geography of
  crystallized sociogram
• Lines indicate talk about technology
• Size of dot indicates teachers use of technology
  at time 1
• Ripples indicate increase in use from time 1 to
  time 2

43
(No Transcript)
44
Crystalized Sociograms with Subgroups Integrate
Organization Theories

• Openness of schools to external institutions
• Some teachers respond more directly to external
  changes and institutions than others.
• Loose Coupling
• Not all teachers respond directly to the
  administrative change.
• Not all teachers respond directly to each other.
• Multilevels
• Teachers make sense of change through their
  subgroup members. Subgroup members have the most
  direct influence on each other.
• School Decision-making
• Not by edict or policy, but through slow change
  of teachers behaviors as influenced by each
  other.
• Transfer of institutions external to school into
  school
• Actors are sandwiched between external and
  internal (e.g., actor 2) (lemke)

45
Implications

• Talk guided by enduring social structure (grades,
  subgroups, close colleagues)
• True for traditional (curriculum) and innovation
  (technology)
• Suggests most coordination structured by same
  components,
• although different specific conversations
• Innovation may change social structure

46
Krause, A., Frank, K.A., Mason, D.M., Ulanowicz,
R.E. and Taylor, W.M. (2003). "Compartments
exposed in food-web structure." Nature
426282-285
47
Conclusion

• Agent based modeling people are agents
• Influence
• Conformity
• Information
• People pursue (through selection)
• Balance (homophily birds of a feather)
• Information
• ? Changes in interaction as a result of influence
• ? unintended consequences
• Consider subgroups
• Concetration of interaction within subgroups
• Similar beliefs within subgroups
• Mediators, or bridgers (actor 3) between
  subgroups key to system
• Target innovations for mediators
• Likely moderate in their beliefs
• Cultivate mediators or bridgers
• What makes a complex system an organization?