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 ACOs network outperform organizations that do not.
Descrição original:
Título original
Simulating Health Networks- Decisions and Performance
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 ACOs network outperform organizations that do not.
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 ACOs network outperform organizations that do not.
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.