Simulating Health Networks: Decisions and Performance

Tal Ben-Zvi, Donald N. Lombardi

Stevens Institute oI Technology, Hoboken, NJ 07030, USA

!"#$%&'$--This study uses a simulation as a vehicle for social
networks research application in healthcare. More than 300
graduating MBA students participated in the creation of
Accountable Care Organizations (ACOs), simulating a
healthcare industry. Our results suggest that organizations
positioning themselves at pivotal points within the ACO`s
network outperform organizations that do not. The findings
show the applicability of network theory and the use of
simulations in the study of ACOs.

I. INTRODUCTION

In the past three decades health costs have been on the rise
worldwide |32|. Expenditures in the United States, Ior
example, surpassed $2.3 trillion in 2008, more than three
times the $714 billion spent in 1990, and over eight times the
$253 billion spent in 1980 |11|.
Accountable care organizations (ACOs) have been
proposed as a novel way to slow the Iast-rising health care
costs, to improve the quality oI health provided, and to create
a new healthcare industry structure (|16|, |20|, |21|, |39|,
|40|). One oI the key questions in strategy research is where
an organization should position itselI within its industry (e.g.,
see |2|, |27|, |37|). This question has been given extra
impetus in the healthcare industry by the creation oI
integrated ACOs, where organizations (healthcare providers)
couple together to provide care Ior patients. Our hypothesis is
that entity perIormance in the healthcare industry can be
more Iully understood by examining the organization`s
relationships and ties within the network in which it is
embedded (the ACO). Such a network encompasses the
organization`s set oI relationships with other organizations in
the industry, e.g., suppliers, distributors, customers, and
competitors.
One way to deepen networks understanding Irom this
perspective is to investigate this area using a simulation.
Simulations are considered important motivational and
learning tools |23|, a link between abstract concepts and real-
world problems, a 'learning by doing or 'hands-on
approach (|29|, |33|). We use a simulation as the means by
which to establish a realistic environment Ior laboratory
research on ACOs and as the means by which to Ioster a
heightened awareness oI network attributes in order to gain
insights regarding organizations conduct and perIormance in
the healthcare industry.
Our investigation begins with a section reviewing ACOs,
recent network literature and simulations. Then, we introduce
the study`s hypotheses and present the study`s methodology.
Next, we discuss the value oI using a network approach in a
simulation design, Iollowed by an analysis oI perIormance,
according to network characteristics. Finally, we discuss the
applicability oI this study and propose some Iuture research
directions.

II. LITERATURE REVIEW

A. Network Theorv
Increasing interest in networks research in recent years
has resulted in an exponential growth oI studies across
several disciplines in this area (see |6| Ior a comprehensive
literature review). Network theory is an interdisciplinary Iield
that searches Ior a common Iormalism Ior networks Iound in
real-liIe. The goal oI network theory research is to gain a
greater understanding oI the structure and Ilow patterns
within networks.
Networks exist in all aspects oI liIe (see |34| and
reIerences therein). Some illustrations are as Iollows: (a)
social networks are sets oI people with interaction patterns
between them; (b) citation networks and the World Wide
Web (WWW) are examples oI inIormation networks; (c)
technological networks are man-made networks designed
typically Ior the distribution oI commodities or resources,
such as the electrical power grid and the Internet; and (d)
biological networks, where substrates and products are
connected with metabolic processes between them.
Each network consists oI basic atomic units, called
vertices (e.g., people, web pages, power plants or substrates)
and means by which they are connected, called edges (e.g.,
relationships, hyperlinks, power lines or metabolic
processes).
In this study we Iocus on the practical aspect oI networks
and measure their inIluence on entity perIormance. In
general, networks can operate on diIIerent levels and the
relationships between the actors play an important role in
how problems are modeled and solved. An extensive
literature review oI networks research (and its application in a
social context) may be Iound in |10|, |22|, |28|, |41|, |43|
and |45|. There is also a growing body oI research that is
coming to terms with the economic consequences oI
organizations participating in social or strategic networks
(e.g., |26|, |27|). This underlines the importance oI
understanding network theory, and highlights the need Ior
Iocusing research on this area. We address this notion using
the platIorm oIIered by simulations.

B. Simulations
A simulation is, by deIinition, a highly complex man-
made environment. A simulation oIIers participants the
978-1-890843-23-6/11/$26.00 2011 IEEE
opportunity to learn by doing in as authentic a management
situation as possible and to engage them in a simulated
experience oI the real world (e.g., |23|, |33|). This approach
to simulation design enhances its characteristics to mirror
real-liIe and the observed participant behavior may be
generalized to reality |30|.
Over the years, researchers have reported the extent oI
usage oI simulations in both academe and business (e.g., |1|,
|3|, |12|, |15|, |17|, |19|, |42|). IS literature also suggests the
application oI simulations as a learning tool. For example,
|35| suggested a system development simulation in which
Iailure and escalation are introduced to InIormation System
students; researchers, such as |18| and |31|, used a simulation
game to teach Enterprise Resource Planning concepts; |36|
explored an Internet-mediated setting to simulate an
electronic commerce environment; |4| used a simulation to
teach decision technology and decision support systems.
Overall, the simulation method enables participants to
'learn by doing |23|. A simulation provides participants the
opportunity to take on the roles and responsibilities oI
executives, to become deeply involved in decisions Iaced by
real people in real organizations, to Ieel the pressure and to
recognize and to assume the risks. Moreover, this method is
an excellent tool to test the understanding oI theory, to
connect theory with application, and to develop theoretical
insights in a laboratory environment. The participants are
provided the opportunity to develop some useIul practical
skills and to practice the tools, techniques and theories they
have previously acquired.

III. HYPOTHESES

In this study, we Iocus on the practical aspect oI ACOs
and examine how this type oI collaboration between
organizations in the healthcare industry impacts their
perIormance. Numerous studies examined the structure oI
networks and the characteristics oI their vertices Irom
diIIerent perspectives (e.g., |6|, |10|, |34|); however, the way
through which the network characteristics aIIect perIormance
is still largely unknown.
Studies investigating the economic consequences oI social
or strategic networks (as ACO is a type oI a network) show
that organizations enter alliances to improve their competitive
position (e.g., |7|, |24|, |27|). It seems clear that iI healthcare
is to become more cost eIIective, better strategies Ior
disseminating inIormation and diIIusing innovations through
communities using social inIluence processes need to be
devised |44|. In addition, healthcare social networks have
been used beIore to yield meaningIul measures oI social
integration, and to investigate the social dynamics underlying
community Iunction and population health |13|. In this study
we examine how ACOs can proIit Irom collaboration.
ThereIore, we hypothesize:
)*+,$-.#/# )0. Organi:ations collaborating with other
organi:ations outperform organi:ations that do not.

Moreover, previous studies showed that the larger the
number oI collaborators, the better the state oI an
organization (see, Ior example, |5|, |14|).
ThereIore, the second hypothesis examines the
relationship between the number oI collaborators in an ACO
and entity perIormance, as Iollows:
)*+,$-.#/# )1. The larger the number of an organi:ations
collaborators, the better its performance.

Studies show that the lack oI strong links between groups
or individuals generates holes in the structure oI the network
(|9|, |25|). These structural holes create a competitive
advantage Ior those who span them |7|. Structural holes are
also related to network resilience. Network resilience is
deIined as a network`s ability to Iunction, or continue its Ilow
Irom one vertex to another, aIter some vertices and their
connections are removed |34|. The existence oI strong links
within the network strengthens its resilience. Structural holes
do the exact opposite, as network resiliency becomes
dependent on a Iew vertices that span those holes.
Researchers conIirm a positive correlation between proIits
and entities spanning over structural holes (e.g., |8|, |38|).
ThereIore, we hypothesize:
)*+,$-.#/# )2. Organi:ations having the biggest impact on
network resilience outperform the average
organi:ation.

IV. METHODOLOGY

A. The Simulation Emploved
In order to gain signiIicant insights Irom applying
network theory to ACOs using a simulation, the simulation
must IulIill three Iundamental requirements: First, it must
hold numerous basic atomic units, or vertices, that interact
between themselves. Second, this interaction should be
properly deIined and measurable. Third, each vertex must
have a properly deIined perIormance measure.
We used a simulation developed in the United States,
commonly known as the International Operations Simulation
- INTOPIA B2B (http://www.intopiainc.com), hereaIter
INTOPIA. The simulation is designed to yield substantial
payoIIs in practical training. It involves the participants in the
executive process, motivates their need Ior decision-making
aids and Iorces them to adopt a managerial viewpoint.
The simulation is highly realistic, meant to simulate the
total environment. Participants immerse themselves in an
artiIicially created healthcare industry. Incoming participants,
working in groups (organization`) take part in six or more
simulated periods. The task oI the organizations is to make
decisions which will guide operations (simulated by a
relatively easy computer interIace) in the current period and
which will aIIect operations in subsequent periods.
Decisions were made once a week and were e-mailed to
the simulation administrator to be Ied to the computer
program. AIter the program ran the data, it generated outputs
that included Iinancial reports (e.g., a balance sheet, an
income statement) and market reports. These outputs were
then e-mailed to the groups and were used Ior their decision
making in sequential periods. Dozens oI decisions, covering
the entire range oI a typical healthcare enterprise, were
required oI the groups in each simulated period. Each group
(organization) assumed one (or more) oI the Iollowing
organizational roles: innovative research and development
(R&D) organization, developing diIIerent patents, healthcare
provider, distributor or pharmaceutical wholesaler. The
decision-making process was based on an analysis oI the
organization`s history as presented to the participants at the
beginning oI the simulation, interaction with other
organizations and the constraints stated in the simulation
manual. The perIormance oI an organization in each period
was aIIected by its past decisions and perIormance, the
current decisions, simulated customer behavior, and the
competition the other organizations in the industry.

B. Participants and Procedures
This study was conducted in a university accredited by the
Association to Advance Collegiate Schools oI Business
(AACSB). The participants were senior MBA candidates. We
conducted Iive (independent) runs oI the simulation, each
with diIIerent participants. Table 1 details the number oI
simulated healthcare organizations created in each run.
At the beginning oI each run, the students were asked to
Iorm competing teams. The Iormation oI the teams and
allocation oI executive roles within teams proceeded without
any external intervention or manipulation, and were reported
to the instructors beIore the simulation itselI began. Our
experience shows that executive roles are usually allocated
according to the participants` expertise in certain Iunctional
areas (e.g., accountants and bankers are usually assigned the
role oI chieI Iinancial oIIicers). In each run, we recorded the
decisions made by all the teams. We also kept track oI the
teams` perIormance. For this research, we aggregated all the
results and statistically analyzed them, as presented later.

V. RESULTS

A. Network Analvsis
This study proposes analyzing the INTOPIA simulation as
a network, with all oI the associated implications being
acknowledged. In Table 1 we detail the number oI
organizations the students operated in each run. As can be
observed, the number oI entities in the industry varied Irom
16 to 20 organizations, with an average oI 17 organizations.

TABLE 1. THE NUMBER OF HEALTH ORGANIZATIONS IN EACH
RUN.
Semester Run I Run II Run III Run IV Run V
No. oI Organizations 20 17 16 16 16

We consider INTOPIA as another kind oI an inIormation
network, where each organization serves as a vertex and its
relations or interactions with other organizations (licensing,
inter-organizational sales, etc.) are considered as edges.
Figure 1 illustrates the network structure at the end oI Run I.
The industry was made oI 20 organizations. Figure 1
demonstrates the complexity oI the network structure in the
simulation. Note that in that particular example, 19
organizations had a least one collaborator (entity 11, Ior
example, had 5 collaborators). One organization, entity 18,
did not collaborate with any other organization.



Figure 1. Network structure at the end oI Run I. The industry consists oI 20
organizations and exhibits a complex network structure.

Table 2 presents the average number oI edges oI each
organization in each run and the standard deviation. On
average, in all Iive runs, each organization had 2.51 edges on
average with a standard deviation oI 1.39. The correlation
between the number oI organizations and the number oI
edges is 0.48, indicating that the larger the number oI
organizations participating in the simulation, the larger the
number oI interactions between them.


TABLE 2. THE NUMBER OF EDGES PER ORGANIZATION IN EACH SEMESTER.
Semester Run I Run II Run III Run IV Run V
No. oI Organizations 20 17 16 16 16
No. oI Edges per Organization 2.70 2.59 2.56 2.63 2.06
Standard Deviation 1.63 1.42 1.31 1.36 1.24
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
B. Network Resilience
The concept oI network resilience reveals the Iollowing
characteristics oI a network: (1) organization dependency on
other organizations; (2) the notion oI centered or pivot
organizations; and (3) ineIIectual or weak organizations. The
removal or collapse oI centered or pivot organizations may
lead to a network breakdown, whereas the collapse oI
ineIIectual or weak organizations does not signiIicantly aIIect
the Ilow oI inIormation or goods within the network.
Network resilience is a measure oI the number oI centered
organizations within the simulated network. For example, in
Run I, the (artiIicial) removal oI only two organizations
(entities 5 and 15) results in a large dysIunction oI the
network, as shown in Figure 2: the large component oI 19



Figure 2. The network structure oI Figure 1 aIter the (artiIicial) removal oI
two organizations (entities 5 and 15).

organizations breaks up to 5 smaller ones. On the other hand,
a removal oI an entity placed on the edge oI the large
component, connected to only Iew other organizations (Ior
example, entity 10, which is connected to only one
organization), would have little eIIect on the 'Ilow oI
inIormation and goods within the network, as this entity
serves as an insigniIicant satellite oI the large component.

C. Investigating the Hvpotheses Performance Analvsis
This section examines the research hypotheses and tests
entity perIormance versus network characteristics. In all runs,
entity perIormance was measured by its accumulated retained
earnings (i.e., the accumulated proIits). For example, Table 3
exhibits the perIormance oI organizations in Run IV in
absolute values and in percentage, relative to the average
organization in that run. The average organization in Run IV
achieved accumulated retained earnings oI about 3.1 million
dollars. Entity 6, Ior example, achieved accumulated retained
earnings oI more than 10 million dollars, which is 238
more than the average organization in that run. Note that
organizations that achieved negative proIits may present
perIormance worse than -100. To avoid biases, we do not
measure entity perIormance in absolute values, but in
percentage, relative to the average organization oI the
associated run. For example, the perIormance oI entity 6,
described above, would be 238 (which represents 238 more
than the average organization), while the perIormance oI
entity 9 would be -59. We emphasize that the results in this
section are aggregated Ior all Iive runs.


TABLE 3. PERFORMANCE IN ABSOLUTE VALUES AND IN PERCENTAGE RELATIVE TO THE
AVERAGE ORGANIZATION IN RUN IV.
Entity No.
PerIormance in Absolute
Values (in K$)
PerIormance (in ) Relative to the Average
Organization
1 1,267 -59
2 (456) -115
3 1,358 -57
4 6,248 100
5 (2,354) -175
6 10,564 238
7 562 -82
8 (3,214) -203
9 1,267 -59
10 16,234 419
11 (235) -108
12 23 -99
13 (5,248) -268
14 3,624 16
15 7,562 142
16 12,834 310
Average 3,127 0

1
2
3
4
6
7
8
9
10
11
12
13
14
16
17
18
19
20
In all runs, 85 or more oI all organizations collaborated
with at least one other entity. Table 4 shows the average
perIormance oI the collaborating organizations and the
independent` organizations (those organizations that decided
not to collaborate) in each run, relative to the average
organization.
The results reveal that organizations that did not
participate in alliances with other organizations had below-
average results. We cannot determine that all results are
signiIicant due to the relatively small number oI
organizations. We also note that some oI the collaborating
organizations perIormed much worse than the independent`
organizations in the same run, but overall, on average,
collaboration prevailed.
Previously, we assumed connectivity by the number oI
edges. Figure 3 exhibits the relationship between the number
oI collaborators the organizations maintained and their
perIormance, relative to the average organization oI all runs.
The number oI collaborators ranged Irom zero and seven. As
can be seen, the larger the number oI collaborators, the better
the perIormance oI the entity (with R
2
0.7798). While this
result can be explained by several Iactors, it is mainly
ascribed to the reduction oI risk when increasing the number
oI health partners, leading to a greater competition among
them and, thus, an increase in the negotiation power oI the
organization.
In the previous section, we showed that by removing
entities Irom the network, the large component, consists oI
most organizations, may break up into smaller components.
Using a computer program, we analyzed the perIormance oI
the organizations whose (artiIicial) removal would result in
the greatest Iragmentation oI the network. The Iindings reveal
that when only one organization was removed, it
outperIormed the average organization by 82.8. When two
organizations were targeted, those organizations
outperIormed the average organization by 52.2. Those
results were statistically signiIicant. The Iindings show that
organizations positioned at the heart oI the connection
between network components were those that beneIited most
and outperIormed the average organization. They simply
exploited their centrality and signiIicance to their own beneIit
and thus enhanced their perIormance.

VI. DISCUSSION AND CONCLUSIONS

This research used network theory concepts to better
understand how ACOs are Iormed and how healthcare
organizations should position themselves within the
healthcare network. For that, simulated organizations were
Iormed. Although the general environment was mutual to all
participants, the organizations became diIIerentiated: each
assumed considerably a diIIerent strategy, diIIerent operating
decisions, and a diIIerent approach to collaboration with
other organizations. Leaving the decision on network strategy
to the groups resulted in a variety oI behaviors toward other
organizations in the industry: Iully integrated organizations
that conducted all the activities along the supply chain
themselves, wholesalers that developed dependency in other
entities, innovating organizations that sold their R&D
products, etc. It appears that these organizations reIlect most
real-liIe approaches in the healthcare industry.

TABLE 4. ENTITY PERFORMANCE HYPOTHESIS H1.
Run Run I Run II Run III Run IV Run V
oI collaborating organizations 95 88 94 94 87
PerIormance oI single organizations -42.68 -8.98 -31.68 -59.00 -20.54
PerIormance oI collaborating organizations 2.24 1.20 2.11 3.93 2.93


R = 0.7798
-50
-40
-30
-20
-10
0
10
20
30
40
50
0 1 2 3 4 5 6 7
P
e
r
f
o
r
m
a
n
c
e

(
i
n

%
)
No. of CoIIaborators


Figure 3. Correlation between Entity PerIormance and the Number oI its Collaborators.


Beyond the creation oI simulated organizations and
industries, this study tested three hypotheses relating ACO
structure, network characteristics and entity perIormance. All
three hypotheses were conIirmed. These results agree with
those oI previous similar Iield studies in other industries (e.g.,
|24|, |46|). Furthermore, our Iindings complement and extend
traditional strategy and social Irameworks and perspectives in
healthcare. They shed light on our main question oI where a
healthcare organization or provider should position itselI with
regard to other organizations in the industry. The answer is
complex and has two main aspects: (a) work with numerous
health organizations in a large component (an ACO); or (b)
position the organization in the junction between two
components (two ACOs). Combining these aspects, we come
to the Iollowing answer: "position the organization at the
pivotal point oI the network."
Nevertheless, although simulations today present
suIIicient complexity to provide realistic network Ieatures
and characteristics, no simulation can seize all aspects oI real-
liIe networks. As more data Irom real ACOs become
available, it will be easier to determine the extent to which
simulation situations resemble reality. ThereIore, the
applicability oI the simulation Iindings to the real-world must
be examined with caution. Also, there is a need to determine
how simulations can be applied in studying various aspects oI
ACOs. For example, we showed that generally, perIormance
is improved with the number oI collaborators. This begs the
question oI why this phenomenon is not so Irequently Iound
in real liIe. A deeper investigation may provide important
insights to better comprehend these collaboration
relationships and address the notion that some organizations
succeed in coalescing into collaborative components while
others suIIer Irom conIlict.

REFERENCES

|1| Asakawa, T. and N., Gilbert 'Synthesizing Experiences: Lessons to be
Learned Irom Internet-mediated Simulation Games, Simulation &
Gaming. An Interdisciplinarv Journal, (34:1), 2003, pp. 10-22.
|2| Barney, J. 'Firm Resources and Sustained Competitive Advantage,
Journal of Management, (17:1), 1991, pp. 99-120.
|3| Ben-Zvi T., 'The Changing Shape oI Networks: Lessons Ior the Auto
Industry, Proceeding of the Portland International Center for
Management of Engineering Technologv (PICMET) Conference,
Portland, Oregon, 2009.
|4| Ben-Zvi, T. 'Using Business Games in Teaching DSS, Journal of
Information Svstems Education (18:1), 2007, pp. 113-124.
|5| Ben-Zvi T. and Gordon, G., 'Corporate Positioning: A Business Game
Perspective, Developments in Business Simulation & Experiential
Exercises, (34), 2007, pp. 101-110.
|6| Borgatti, S. P., and P.C., Foster, 'The Network Paradigm in
Organizational Research: A Review and Typology, Journal of
Management, (29:6), 2003, pp. 9911013.
|7| Burt R. Structural Holes. The Social Structure of Competition, Harvard
University Press: Cambridge, MA, 1992.
|8| Burt, R., Guilarte, M., Raider, H.J., and Yasuda, Y. 'Competition
Contingency and the External Structure oI Markets, In Advances in
Strategic Management, Ingram, P. and Silverman, B. (Eds.),
Greenwich, CT: JAI Press, 2002.
|9| Burt, R., Cook, K. and Lin, N. Social Capital. Theorv and
Research, New York: Aldine de Gruyter, 2001.
|10| Carrington, P.J., J., Scott, and Wasserman, S., Models and Methods in
Social Network Analvsis, Cambridge: Cambridge University Press,
2005.
|11| Centers Ior Medicare and Medicaid Services, OIIice oI the Actuary,
National Health Statistics Group, National Health Care Expenditures
Data, 2010.
|12| Chen, L. and Lin, C. 'An Experiment in DSS EIIectiveness,
Proceedings of the Joint SIGDSS & TUN Users Group Pre-ICIS
Workshop and Congress, Phoenix, Arizona, 2009.
|13| Christakis, N.A., and Fowler, J.H. Connected. The Surprising Power of
Our Social Networks and How Thev Shape Our Lives, Little, Brown
and Company, 2009.
|14| Cool, K. O., and D.E., Schendel ,PerIormance DiIIerences between
Strategic Group Members, Strategic Management Journal, (9), 1988,
pp. 207-233.
|15| Dasgupta, S. 'The Role oI Controlled and Dynamic Process
Environments in Group Decision Making: An Exploratory Study,
Simulation & Gaming. An Interdisciplinarv Journal, (34:1), 2003, pp.
5468.
|16| Devers, K., and Berenson, R. 'Can Accountable Care Organizations
Improve the Value oI Health Care by Solving the Cost and Quality
Quandaries?, Timelv Analvsis of Immediate Health Policv, (13), 2009.
|17| Dickinson, J. R., J.W.A., Gentry, and A.C., Burns, 'A Seminal
Inventory oI Basic Research Using Business Simulation Games,
Proceedings of ABSEL Conference, Las Vegas, 2004.
|18| Draijer, C., and Schenk, D-J. 'Best Practices oI Business Simulation
with SAP R/3, Journal of Information Svstems Education (15:3),
2004, pp. 244-261.
|19| Eldredge, D. L., and H.J., Watson, 'An Ongoing Study oI the Practice
oI Simulation in Industry, Simulation & Gaming. An International
Journal, (27:3), 1996, pp. 375386.
|20| Fisher, E.S., McClellan, M.B., Bertko, J., Lieberman, S.M., Lee, J.J.,
Lewis, J.L., et al. 'Fostering accountable health care: moving Iorward
in Medicare, Health Affairs, (28:2), 2009, pp. 219231.
|21| Fisher, E.S., Bynum, J.P., Skinner, J.S. 'Slowing the Growth oI Health
Care Costs Lessons Irom Regional Variation, The New England
Journal of Medicine, (360), 2009, pp. 849-852.
|22| Freeman, L.C., The Development of Social Network Analvsis. A Studv
in the Sociologv of Science, Vancouver: Empirical Press, 2004.
|23| Garris, R., R., Ahlers, and J.E., Driskell, 'Games, Motivation and
Learning: A Research and Practice Model, Simulation & Gaming. An
Interdisciplinarv Journal, (33:4), 2002, pp. 441-467.
|24| Goerzen, A., Networks and Locations. Organi:ing the Diversified
Multinational Corporation for Jalue Creation, Palgrave Macmillan,
2005.
|25| Greve, A. and SalaII, J.W., 'Social Networks and Entrepreneurship,
Entrepreneurship Theorv and Practice, (28:1), 2003, pp. 1-22.
|26| Gulati, R., 'Network Location and Learning: The InIluence oI Network
Resources and Firm Capabilities on Alliance Formation, Strategic
Management Journal, (20:5), 1999, pp. 397-420.
|27| Gulati, R., N., Nohria, and A., Zaheer, 'Strategic Networks, Strategic
Management. Journal, (21), 2000, pp. 203-215.
|28| Hu, L., and Pekin, O. 'Social Behavior and Network Analysis
Proceeding of the Pacific Asia Conference on Information Svstems
(PACIS), Taipei, Taiwan, 2010.
|29| Kolodner, J.L., Gray, J. and Fasse, B.B. 'Promoting TransIer through
Case-Based Reasoning: Rituals and practices in learning by design
classrooms, Cognitive Science Quarterlv (3:2), 2003, pp. 183-232.
|30| Lainema, T., and P., Makkonen, 'Applying Constructivist Approach to
Educational Business Games: Case REALGAME, Simulation &
Gaming. An Interdisciplinarv Journal, (34:1), 2003, pp. 131149.
|31| Leger, P-M. 'Using a Simulation Game Approach to Teach Enterprise
Resource Planning Concepts, Journal of Information Svstems
Education (17:4), 2006, pp. 441-447.
|32| Lombardi, D.N. 'Preparing Ior the Future: Leadership and
Management Strategies, Healthcare Executive, (11), 2005, pp. 8-14
|33| Martin, A., 'The Design and Evolution oI a Simulation/Game Ior
Teaching InIormation Systems Development, Simulation & Gaming.
An Interdisciplinarv Journal, (31:4), 2000, pp. 445-463.
|34| Newman, M. E. J., 'The Structure and Function oI Complex
Networks, Siam Reviews, (45:2), 2003, pp. 167256.
|35| Nulden, U., and Scheepers, H. 'Increasing Student Interaction in
Learning Activities: Using a Simulation to Learn about Project Failure
and Escalation, Journal of Information Svstem Education (12:4),
2001, pp. 223-232.
|36| Parker, C. M., and Swatman, P. M. C. 'An Internet-mediated
Electronic Commerce Business Simulation: Experiences Developing
and Using TRECS, Simulation & Gaming. An Interdisciplinarv
Journal (30:1), 1999, pp. 5169.
|37| Porter, M. E. 'Industry structure and competitive strategy: Keys to
proIitability, Financial Analvsts Journal, 1980.
|38| Reagans, R., and Zuckerman, E.W. 'Networks, Diversity and
PerIormance: the Social Capital oI Corporate R&D Units,
Organi:ation Science, (12), 2001.
|39| Shortell S.M., Casalino, L.P., and Fisher, E. Implementing Accountable
Care Organi:ations, Berkeley Center on Health, Economic and Family
Security, 2010.
|40| Shortell S.M., and Casalino, L.P., 'Health care reIorm requires
accountable care systems, The Journal of the American Medical
Association, (300:1), 2008, pp. 95-97.
|41| Smith, B. and Goldman, D. 'An Experiment in Social Networks,
Proceedings of the Joint SIGDSS & TUN Users Group Pre-ICIS
Workshop and Congress, Phoenix, Arizona, 2009.
|42| Smith, D. 'Distance Learning: A Game Application, Developments in
Business Simulation & Experiential Exercises, (37), 2010.
|43| Wasserman, S., and K., Faust, Social Network Analvsis. Methods and
Applications, Cambridge: Cambridge University Press, 1994.
|44| West, E., Barron, D.N., Dowsett, J. and Newton, J.N 'Hierarchies and
cliques in the social networks oI health care proIessionals: implications
Ior the design oI dissemination strategies, Social Science & Medicine,
(48), 1999, 633-646.
|45| Ye, L. and Kasemsarn, B. 'Analyzing Social Networks
Characteristics, Proceedings of the Portland International Center for
Management of Engineering Technologv (PICMET) Conference,
Phuket, Thailand, 2010.
|46| Zaheer, S., and A. Zaheer 'The Structure oI Global Competition: A
Network Approach, Paper presented at the Strategic Management
Journal Special Issue Conference on Strategic Networks, 1999.