Playing with the model or modeling the game?
A comparison of Group Model-Building and Game-Based Learning
A master thesis project by Emmy van de Pasch
Supervised by Prof. Dr. R. Van der Heijden and Prof. Pål Davidsen In cooperation with T-Xchange, supervised by Dr. Johan de Heer
A
BSTRACTIn this study, an extensive literature study and exploratory qualitative data analysis are used to gain insights in two methods of organizational intervention: Group Model-Building (GMB) and Game-Based Learning (GBL). GMB is a process aimed at providing insights in strategic debates by jointly creating a model of the system that may support decision-making. Through sharing of Individual Mental Models, a shared perspective is created. GBL is a popular tool to support individual and organizational learning. It is argued that the methods have quite some overlapping elements, and an integration may aid in overcoming some shortcomings associated with the methods. Grounded in literature, an integrated tool Group Model-Based Game Learning was hypothesized. Through qualitative semi-structured interviews with experts from both the field of GMB and GBL, this hypothesized tool was further articulated. A case study of GO4IT, a serious game developed for military training, was drawn up and analyzed. Results of this research show potential for Group Model-Based Game Learning, although further research is needed to see how this tool may work in practice. For experts of both GMB and GBL, some recommendations are given, for both research and practice.
A
CKNOWLEDGEMENTSThis research project marks the end of the personal and academic roller coaster that the last few years have been for me. A journey through multiple countries and universities, that have not only taught me a great deal about the field of System Dynamics, but also about myself. I would like to take a moment to thank a few very special people that got me here.
First and foremost, I would like to thank my supervisors, Rob van der Heijden of Radboud University and Johan de Heer of T-Xchange, as well as my second supervisor Pål Davidson of the Universitet I Bergen for all the time and energy they have put in giving me feedback and pointing me in the right direction. Your engagement with this work has definitely helped to create something that I can be proud of as a final university project.
Secondly, thank you to all interviewees and all those that have given me fruitful ideas to prolong and expand this project. A big thank you to T-Xchange for the time and hours you have spent on my thesis. I hope the results will be as valuable to you as the researching process was for me. Thirdly, thanks, heel erg bedankt, tusen takk, grazie mille to all of my lecturers at Radboud University, Universitet I Bergen, and Università degli Studi di Palermo, as well as to the supporting staff of the EMSD program. Also a warm thank you to my fellow students, who became great friends or even family along the journey.
Finally, an acknowledgement to all my family and friends for supporting me through this thesis, in particular my parents for all their advice and support, and Ralph for all his encouragements and for being there in the most difficult moments of graduating.
1 C
ONTENTSAbstract ... 2
Acknowledgements ... 2
2 Introduction ... 6
2.1 System Dynamics as a tool to understand complexity ... 6
2.2 Group Model-building to enhance implementation of SD models ... 7
2.3 Overcoming barriers to implementation with games ... 8
2.4 Integrating GMB and GBL ... 9
2.5 Research Strategy, Aim and Questions ... 11
3 The wider research context: Organizational change, interventions and organizational Learning . 13 3.1 Organizational Change ... 13 3.2 Organizational Intervention ... 15 3.3 Organizational Learning ... 16 3.3.1 Individual learning... 16 3.3.2 Organizational learning ... 18 4 Group Model-Building ... 21
4.1 Defining Group Model-Building ... 21
4.2 System Dynamics ... 22
4.2.1 System Dynamics Model-Building ... 22
4.2.2 Implementation problems in System Dynamics ... 24
4.3 Outcomes and effectiveness of GMB ... 26
4.4 Underlying philosophy and assumptions of GMB ... 27
4.5 The GMB process ... 27
4.5.1 The role of facilitation ... 28
4.5.2 Set-up of a GMB session ... 28
4.6 Shortcomings of Group Model-Building ... 29
5 Game-Based Learning ... 31
5.1 Defining Game-Based Learning ... 31
5.2 Outcomes and effectiveness of game-based learning ... 32
5.3 Underlying philosophy and assumptions of GBL ... 34
5.4 The GBL process ... 35
5.4.1 Game Design and Development ... 37
5.4.2 Set-up of a GBL session ... 41
5.5 Evaluation and assessment ... 42
5.6 Shortcomings of Game-Based Learning... 44
6 Reflections on the literature review: Integrating GMB with GBL? ... 46
6.1.1 GMB and GBL in Organizational Change and Intervention ... 46
6.1.2 GMB and GBL in Organizational Learning ... 48
6.2 Group Model-Building versus Game-Based Learning ... 49
6.2.1 What GMB could learn from GBL ... 52
6.2.2 What GBL could learn from GMB ... 54
6.3 Premature Theory-Building: Integrating GMB and GBL ... 55
7 Methods ... 56
7.1 Qualitative Interviews ... 57
7.2 Case Study Research ... 58
7.3 Research limitations and research ethics ... 59
8 Results ... 60
8.1 Interview Results ... 60
8.1.1 Defining elements ... 60
8.1.2 Aspired outcomes ... 60
8.1.3 Philosophy and assumptions ... 61
8.1.4 Development and intervention process... 62
8.1.5 Shortcomings ... 64
8.1.6 Integration of GBL and GMB ... 65
8.2 Case Study results ... 66
8.2.1 Case description... 66
8.2.2 Case analysis and assessment ... 67
9 Conclusion and discussion ... 78
9.1 Conclusions ... 78
9.1.1 How are the methods of GBL and GMB comparable in terms of key concepts, goals and philosophy behind the methods? ... 78
9.1.2 How do GMB-based change processes differ compared to GBL-based change processes? ... 79
9.1.3 How can an integration of the methods of Game-Based Learning and Group Model-Building enrich each of the individual methods? ... 80
9.2 Research Limitations ... 82
9.3 Recommendations for future research ... 83
9.4 Recommendations for practice ... 84
References ... 85
Appendices ... 91
Appendix A: Topic Guide Interviews ... 91
Appendix B: Interview Transcripts ... 92
Interview I: General GBL expert and codeveloper of GO4IT ... 92
Interview III: Game designers and co-developers of GO4IT ... 102
Interview IV: Game-Based Learning Trainer ... 110
Interview V: Group-Model Building Expert ... 114
Interview VI: Group-Model Building Expert ... 121
Appendix C: Codes derived from interview data ... 126
2 I
NTRODUCTIONToday’s society is characterized by accelerating change and increasing complexity, resulting from an ever-increasing interconnectedness of the systems in society. Governments, organizations and all members of society need to be aware of their environment and understand the dynamics and feedbacks between and within their systems. A lack of awareness and understanding of these dynamic processes and systems in decision and policy-making may result in the creation of new, even more complex problems, or the unintended amplification of current problems (Sterman, 2000). To develop an awareness and understanding of dynamic processes and systems is not easy, as systems are interdisciplinary and learners themselves are often part of the dynamic complexity of a system.
2.1 SYSTEM DYNAMICS AS A TOOL TO UNDERSTAND COMPLEXITY
The philosophy and modeling tools of System Dynamics (SD) are among the tools considered useful in generating an understanding of the system structure. SD, by analyzing the interconnectedness of a system and focusing on the inherent delays, feedbacks and nonlinearities in a system, helps decision-makers in understanding dynamic complexity (Sterman, 2000). This analysis is done by developing models, both qualitative to enhance systems understanding and quantitative models to simulate behavior of complex systems. By means of model structure and simulations, a better understanding of the decision-making system and environment can be generated. In essence, SD can be seen as a tool to enhance learning in and about systems characterized by complexity, and can aid in more effective decision-making (Sterman, 2000).
The System Dynamics methodology is used when high-stakes decisions need to be made, and when an integrated view of the system is needed to understand both current and future outcomes of the system, given the multitude of consequences and their differing time scales (Homer, 2012). Successful applications of SD may be found in corporate strategy, health care policy-making, and environmental policy-making. Here, SD models have provided a safe environment for testing the effects and effectiveness of decisions and policies before they have affected real-life (Forrester, 1971; Sterman, 2000). Additionally, they are tools for learning about the system. SD model simulations can be regarded as virtual micro-worlds in which decision-making can be tested and adapted (Sterman, 2001).
The aspired benefits of SD both for public and private organizational purposes are omnipresent, and the application of the method for organizational change processes and
consultancy has been widely researched and reported (Größler, 2007; Lane, 1992; Ledet, Monus, Cardella, & Burgess, 2005; Liehr, Größler, Klein, & Milling, 2001; Rouwette & Vennix, 2006; Snabe & Größler, 2006; Sterman, 2000; Vennix, 1996). Many of these reports, however, discuss instances where, although a valid model was developed, the change process in the organization was unsuccessful. Meadows and Robinson (2002) discuss, as one of the main factors, the implementation gap: supply and demand of models have different inherent goals, and modelers are often too concerned with academic robustness and publications. This assumption is also discussed by Coyle and Exelby (2000), who discuss that modeling for consultancy and modeling for academics are different, as there are differences in Customers, Actors, Transformation, Weltanschauung, Owner, Validation, Environment and Risk. For example, Commercial models start from a problem in an competitive environment – Academic models start from a gap in knowledge, and aim to openly contribute to a body of scholarship.
Another potential explanation for a lack of implementation can be found in organizational change literature: for effective implementation of change, next to managerial support, commitment to these changes in the organization at large is essential (Paton & McCalman, 2008). A way to achieve this commitment, is participation of the so-called change recipients in the change process (Armenakis & Harris, 2009). The idea of client involvement in the modeling process has been strongly articulated in the SD literature (Andersen & Richardson, 1997; Lane, 1992; Sterman, 2000; Vennix, 1996). Arguing that effective learning of dynamic insights comes from modelling the system problem, Vennix (1996) introduces an approach called Group Model-Building (GMB), in which communication, mental model elicitation, and commitment of participants to the model are central elements. The focus of the current study is here, as this approach reflects the highest level of participant-involvement (Stave, 2010). In a GMB project, “the client and facilitator embark on an inquiry process into a problem, a team learning process, aimed at increasing insight and, hopefully, finding a consensual decision to solve the problem” (Vennix, 1996, p. 99).
2.2 GROUP MODEL-BUILDING TO ENHANCE IMPLEMENTATION OF SD MODELS
In Group Model-building, a diagram, or model, is created based on the mental models of the participants. By eliciting those mental models and connecting individuals’ input, a shared mental model is created. GMB has proven successful in creating consensus and commitment amongst participant groups, as well as involvement of the participants in the overall process (Rouwette, Vennix, & Mullekom, 2002). Another benefit of GMB as a participatory modelling process comes from the effectiveness of the method to gain dynamic insights. Greenberger,
Crenson, and Crissey (1976), as well as several authors after them (Richardson, 1996; Voinov & Bousquet, 2010), report that modeling mainly educates the policy modeler, rather than the policy maker. Thus, those who model learn. Effective learning of dynamic insights thus occurs when people model system problems themselves (Sterman, 1994).
Still, although some barriers to organizational implementation can potentially be overcome with GMB, implementation problems persist (Vennix, 1996). Größler (2007) argues that the implementation issues with SD models can be of two types: (1) No changes in policies or structures; or (2) Immediate changes in policies and structures but no sustainable use of the new dynamic insights. A potential barrier to implementation can be found in organizational learning theory, as for true organizational learning to occur, the interests and mental models of individuals should be aligned with those of the organization, not just the interests and mental models from the decision-makers (Argyris & Schön, 1978). This argument aligns with that of Andersen, Chung, Richardson, and Stewart (1990), who argue that even if one gained dynamic insights, the model would not automatically be suitable for transferring and explaining these dynamic insights to others. In order to increase the amount of interaction individuals have with the dynamic model developed, an additional tool may be necessary. By means of Gaming, Andersen et al. (1990) argue, individuals learn about the system under investigation, as well as how the dynamic problem came about.
2.3 OVERCOMING BARRIERS TO IMPLEMENTATION WITH GAMES
Game-based learning (GBL) as a tool to move from GMB-based models to organization-wide implementation programs appears suitable. Games sustain the principles of system dynamics, such as interconnectedness, feedbacks and delays (Meadows, 1989). Many so-called ‘Management Flight Simulators’ have been developed over the years in the form of physical models, ‘Serious’ board games, or computer simulations (Sterman, 2001). The learning experience from gaming results from interactions of individuals with the system at hand, and from trial-and-error of potential policies. Following an ancient saying:
“When I hear, I forget. When I see, I remember. When I do, I understand” Meadows (1989, p. 636) argues that traditional SD practice to convey dynamic insights is based on presentations of model results, whereas by embedding SD models in a game and providing the opportunity for experimental learning, true systems understanding can be gained.
Game-Based Learning, or, Serious Gaming, is increasingly used as a deliberate intervention, aimed at examining or changing an organization (Mayer, Warmelink, & Zhou,
2014). The origins of gaming in (military) decision-making goes back to ancient times, and gaming in business and management has been on the rise since the 1950s (Shubik, 1975). The effectiveness of GBL for learning has found a large body of empirical support (Boocock & Schild, 1968; Klasson, 1964; Mayer & De Jong, 2004; Norris, 1986; Van Der Hulst & Ruijsendaal, 2012). Serious Games can thus be considered as a strong tool to support organizational (and individual) learning, and thus may be a good tool to support and follow-up a GMB modelling process to increase the likelihood of implementation.
Game-based Learning is not the only organizational change or training tool aimed at learning, but the effectiveness for learning about complex systems is not the only reason why GBL may be a solution to solve the implementation problems associated with GMB change processes. Games, or gaming, and the System Dynamics methodology have quite some things in common: both focus on dynamic models and feedback loops (Hunicke et al., 2004; Schreiber, 2009) and both fields are interdisciplinary (Crookall, 2010). Like SD, gaming “proves an appropriate process for dealing with the increasing complexity of organizational environments” (Geurts, Duke, & Vermeulen, 2007, p. 535). Following Geurts et al. (2007), gaming aims to create communication, consensus and commitment, supplemented by an understanding of complexity as well as creativity among participants. These aims are closely related to those of GMB, as described above (Vennix, 1996).
2.4 INTEGRATING GMB AND GBL
The potential benefits of integrating GMB and GBL is not only considered beneficial for Group-Model-building, but may also help the GBL method overcome some challenges, both on a methodological and a practical level. The field of game-based learning, although increasingly popular, with a $ 2.16 billion revenue in 2016 (Adkins, 2016), still lacks consensus on how games can be defined. Further, evidence on effectiveness is mixed (Garris et al., 2002). Mayer (2012) argues that the field of GBL lacks an overarching methodology, and is in need of theories to formulate and test hypotheses and generic tools for unobtrusive data-gathering for Serious Games, amongst others. Thus, where GMB, and SD as a whole, face issues in gaining acceptance to the larger audience, GBL lacks an agreed-upon definition of its aims, and lacks the tools to collect data (Crookall, 2010; Mayer, 2012). It is here where the integration of GMB and GBL may provide most benefit for GBL. Participatory modeling and other mental-model elicitation methods used by the SD society may help GBL to gain a stronger methodological basis.
The interest of the System Dynamics community in gaming is not new, nor is the interest of the field of Gaming and Simulation in System Dynamics. Over the years, some widely used and valued games, such as the “Beer Game” and the “Fish Banks Ltd.” game have been developed in the SD field. Although mainly used in an educational environment, the games have been successfully played with corporate managers to support organizational learning (Sterman, 2000). Meadows (1989, p. 640), e.g. predicted that “gaming will grow quickly over the next decade to become an extremely important tool in implementing system dynamics models”, and Lane (1995) reported on a growing interest in ‘management games’ within the field of SD. Analysis of the SD literature suggests that this interest, at the very least, has cooled down. Figure 1 below portrays the development of interest of SD/GMB in GBL and vice versa. The dark, dotted line represents the amount of articles published in Simulation and Gaming that have ‘System Dynamics’ in their titles; the light, solid line represents the amount of articles published in System Dynamics Review that have ‘Gam*’ in their titles. The article count indicates that, especially in recent years, the interest of Game-Based Learning in Dynamic modeling methods has grown quite a lot stronger than vice versa.
Based on the mutual interest of both fields in one another, although the interests fluctuate, an observation can be made that there are some overlapping elements in both fields and in these commonalities, enrichments of the methods can be found. So, it can be questioned, how one can put this mutual interest and overlapping elements between GBL and GMB to use.
0 10 20 30 40 1985 1990 1995 2000 2005 2010 2015 NU M BER OF A RT ICL ES YEAR I N T E R E S T O F S D I N G A M I N G A N D V I C E V E R S A
SDR Articles "Game" S&G Articles "System Dynamics"
Figure 1: Number of articles published in the System Dynamics Review related to “game” or “gaming” and number of articles published in Simulation & Gaming related to System Dynamics
2.5 RESEARCH STRATEGY,AIM AND QUESTIONS
This research will compare the roots of the two tools or methods themselves by elaborating on the key concepts, goals, processes and philosophy behind the methods, as well as their shortcomings. This study does not elaborate on a comparison of the effectiveness of the methods in organizational learning, as this cannot be done rigorously: too many contextual factors may impact the learning process, and the two methods are expected to differ on more than one critical aspect (Größler, 2001). The comparison of GMB and GBL will, alongside an extensive literature study of both methods, be grounded in qualitative data, gathered through interviews with experts in both fields. Additionally, a case study of a Serious Game developed for the Civil-Military Cooperation Centre of Excellence will be used. This game, situated in a complex and dynamic setting, aims to train military, which gives it a strong validity for case studies, as the roots of GBL can be found here (Shubik, 1975). By these methods, the current research aims to meet the following research objective:
To analyze the applicability and usefulness of game-based learning as an integrated follow-up to a Group Model-building change process, as well as the enrichment GMB can give to the field of GBL. By integration, this study aims to boost the development and maturing of both methods, as well as their impact on the organizational change process.
The objective can be translated to the conceptual model presented in Figure 2., that will provide the basis for the theoretical analysis, data analysis and theory-building process that will follow in the subsequent chapters. The diagram portrays the theoretical assumption underlying this study: when a GMB change process is used to analyze the problem at hand and is followed-up by a Serious Game, it is more likely that the mental models of participants change and implementation takes place, thus the impact of the intervention will be larger. This is portrayed in Figure 2 by distinguishing two paths: the red arrow, portraying the GMB process currently used, where organizational change is expected to follow after sessions. The red arrow is partially crossed out, to emphasize that this process is often inefficient. Alternatively, the green path goes through game-building and playing, to truly change organization-wide mental models and create the planned changes.
In order to meet the research objective and validate the conceptual model, this research aims to answer the questions:
RQ1: How are the methods of GBL and GMB comparable in terms of key concepts, goals and philosophy behind the methods?
RQ2: How do GMB-based change processes differ compared to GBL-based change processes?
RQ3: How can an integration of the methods of Game-Based Learning and Group Model-building enrich each of the individual methods?
In the subsequent chapters, an extensive literature study is presented. First, the wider research context of organizational change and organizational learning is discussed, followed by a thorough discussion of the philosophy, definitions and processes of Group Model-Building and Game-Based Learning. The two methods are subsequently compared, to allow for some premature theory-building. Next, the empirical research methods will be discussed, as well as the limitations and ethics of this study. Following the methodological discussion, the results of the qualitative data gathering methods will be presented. The discussion of results will be done within the theoretical frame developed before. Finally, based on the presented results and in light of the theory and preliminary conclusions following this, the research questions are answered and discussed, followed by a discussion of research limitations. This thesis concludes with some recommendations for both future research and for practice, that are derived from the current study.
Figure 2: Conceptual model of one of the main assumptions to be investigated. A GMB intervention process that is followed-up by a Game leads to stronger implementation and organizational change.
3 T
HE WIDER RESEARCH CONTEXT:
O
RGANIZATIONAL CHANGE,
INTERVENTIONS AND ORGANIZATIONAL
L
EARNINGTo understand the potential for integrating Group-Model-building and Game-Based Learning, it is not only necessary to explore and analyze each of the methods, but also to put them into their real-life context. Hence, to fully comprehend the procedures, concepts and philosophy of GMB and GBL, the wider context of organizational change needs to be explored. The field of organizational change is becoming increasingly important in today’s highly dynamic society. Organizational change generally occurs due to some failure in an organization, and aims to improve the performance of individuals within the firm (Weick & Quinn, 1999). That being said, organizational change takes many forms. It can be episodic or continuous, deliberate or emergent, but regardless, organizational change is dependent on the organizational context, and within that context, there can be many barriers to change (Weick & Quinn, 1999).
3.1 ORGANIZATIONAL CHANGE
In the current study, focus is on deliberate organizational change, or, organizational development. Organizational development is a planned and systemic change process, which aims to improve an organization’s well-being and effectiveness by adapting to the environment, improving internal relations and increasing learning and problem-solving capabilities, by changing attitudes of employees through team-building, and interventions (see, e.g. (Daft, Kendrick, & Vershinina, 2010)).
Organizational change processes can be characterized by three phases. In the first phase people are made aware of problems and the need for change, creating the motivation to change attitudes and behaviors. Subsequently, in the changing phase, individuals experiment and learn new skills, which may occur through interventions. Here, individuals can experiment with and discuss on the various options and potential impact. Finally, when a change is made and individuals have acquired new attitudes or values, the impact of new behavior needs to be evaluated and reinforced (Daft et al., 2010). After the change process, the implementation of the changes needs to managed properly. All too often, resistance to change leads to unsatisfying results (Loewe & Dominiquini, 2006).
Resistance to change has been widely researched in both past and recent research (Dent & Goldberg, 1999; Ford, Ford, & D'Amelio, 2008; Jones, Jimmieson, & Griffiths, 2005; Pardo del Val & Martínez Fuentes, 2003). Resistance to change leads to additional costs and delays in the change process, but is also a source of learning and “can have value for the existence,
engagement, and strength of change, serving as an asset and a resource in its implementation and successful accomplishment” (Ford et al., 2008, p. 368). For these reasons, it is of utmost importance to manage change and resistance to change thoroughly. Change agents should take into account the self-interest of employees, create a perceived need for change and gain understanding and trust. Various tools to deal with change resistance have proven successful: (1) communication and education to provide solid information; (2) participation to allow potential resisters to design the change, leading to commitment and understanding; (3) negotiation to gain formal acceptance and approval; (4) coercion to force employees to change; and (5) top management support to symbolize that the change is important (Nutt, 1986).
Implementation of any organizational intervention thus needs to be managed carefully. It is of utmost importance to give the right people the right responsibilities to build a critical support mass of the change, and to integrate implementation in the entire organizational change process (Nadler, Tushman, & Nadler, 1997). Thus, in planning for organizational change, in analyzing the problem to be dealt with, in learning new skills, and in adopting the learned behaviors, implementation of the changes should be considered. These steps of the organizational change process are elegantly summarized by Größler (2007), as portrayed in Figure 3
.
The model, based on the work of Schein (1999) portrays two cycles of organizational problem-solving, where the inner cycle captures the analysis of the problem and planning solutions, and the outer cycle describes the action steps taken to implement changes. The model can serve as a basis for designing an organizational intervention process (Schein, 1999)The two cycles described by Größler (2007) and Schein (1999) can aid in distinguishing the methods of GMB and GBL as well. It appears that Group-Model-building with its focus on
problem formulation and generating consensus, commitment and new insights for participants, by creating a framework or model for analysis, fits best in the inner cycle (Van Daalen et al., 2014; Vennix, 1996). GBL, on the other hand, appears to be more focused on the outer cycle as one uses a proposed model of a system, reflected in a game, to initiate action planning and experimenting with actions in a safe environment (Van Daalen et al., 2014).
3.2 ORGANIZATIONAL INTERVENTION
As mentioned, one way to achieve organizational change is by means of interventions. Organizational intervention processes such as GMB and GBL are generally supported by an intervener, who enters “into an ongoing system of relationships, to come between or among persons, groups or objects for the purpose of helping them. The intervener exists independently of the system” (Argyris, 1970, p. 15). The intervener, or facilitator, values the system as self-responsible: it is obliged to control its own destiny. The facilitator merely assists to create more effective problem-solving, decision-making and decision- implementation systems, in a way that decreases the need for the facilitator. In assisting, the facilitator, can adopt one of three roles, as described by Schein (1990) and portrayed in Table 1. In an expert role, one provides the client organization with relevant information; as a doctor, one investigates and diagnoses a problem and suggests a cure; or, as a process consultant, one guides the organization in identifying the problem it faces and aids in finding a sustainable solution.
The role of process consultant appears most appropriate for GMB interventions. Here, the facilitator aids the client in generating an understanding of the dynamic complexity surrounding the problem (Rouwette & Vennix, 2006; Snabe & Größler, 2006; Sterman, 2000;
Vennix, 1996). This role is the most appropriate for GBL interventions as well, as they are aimed at providing clients with those skills needed for further organizational decision-making, thereby educating the client within a safe learning environment (Crookall, 2010; Sauvé, Renaud, Kaufman, & Marquis, 2007; Van Daalen et al., 2014)
Regardless of the facilitator style chosen, interventions should be designed in such a way that they do not only change the mental models and behavior of those who participate in the intervention, but also those who are the targets of change and those who will be affected by the change (Schein, 1990). Changing mental models of all affected will help overcome barriers to change (Jones et al., 2005). Interventions that do not successfully take away those barriers and only show incremental structural changes to solve the problem, appear to reflect a sense of organizational adaptation. No understanding of causal relationships is generated, as opposed to organizational learning, where understanding and knowledge is generated beyond a single event (Fiol & Lyles, 1985).
3.3 ORGANIZATIONAL LEARNING
Many definitions surround the concept of organizational learning, referring to new insights or knowledge, new structures, new systems, or improved actions (Kim, 1998). The role of knowledge and understanding is central to move from mere organizational adaptation to true organizational learning (Fiol & Lyles, 1985). Intervention tools enable two types of learning: learning about the issue, and learning about other participants (Lenferink, Arciniegas, Samsura, & Carton, 2016). Focus here is on learning about the problem at hand, though the impact of learning about other participants, especially in complex problems, is not to be ignored.
Essential for organizational learning is that the interests and mental models of individuals should be aligned with those of the organization (Argyris & Schön, 1978). Learning thus needs to occur both on an individual and a group level. In a way, individual learning can be considered as a necessary condition or even a metaphor for organizational learning, as organizations in fact learn via their members (Kim, 1998). By formulating and implementing an intervention strategy, organizational learning can be enhanced, as this enables all individuals within an organization to learn and establish desired behavioral changes.
3.3.1 Individual learning
Learning, as discussed by Argyris (1976), is heavily influenced by the perception that individuals have of the system, as individuals determine what the system will look like in the future by conscious or unconscious actions and behavior. Individuals learn from the greater
system based on the feedback they get following their actions, however, merely adapting solutions within the already existing practices and values will not likely change the individuals’ mental model and likewise not change the problem at hand. So-called Double Loop Learning is required to sustainably solve problems and learn about the system (Argyris, 1976). Though difficult to internalize, the double-loop learning mode, emerging from valid information, informed choice, commitment to the choice and constant monitoring, has shown an increased effectiveness (Argyris, 2000).
Figure 4 portrays a simple model of individual learning. Kim (1998) discusses that individual learning consists of two parts: the Observe-Assess-Design-Implement (OADI) cycle and the Individual Mental Model (IMM). According to the OADI cycle, individuals cycle through a process of having a concrete experience they observe, which they assess by reflections and observations. Based on the reflections, individuals design new ideas and courses for action. By implementing these ideas in new situations, new concrete experiences emerge (Kim, 1998). Though helpful in understanding learning, the OADI cycle is incomplete, as it merely focusses on acquisition of new ideas. The role of memory, tightly interconnected with learning, needs to be considered to understand how individuals retain that what has been learned. Therefore, attention is given to the role of Individual Mental Models, described by Senge (1990) as internal images of how the world works that are deeply held and powerful, because they affect what we do as well as what we see. IMMs are dynamic concepts that provide the context in which individuals view and interpret new information, and thereby not only help us to understand the world, but also restrict our understanding (Kim, 1998).
Linking the IMM to the OADI cycle of learning helps to understand the dynamic nature of the mental model. Generally, individuals learn within the realm of their mental model, but this may be updated with new information. Constant feedback on one’s IMM results in more
effective decision-making and learning, as it transforms the narrow, static view of the world in a holistic, dynamic view, which enables one to change the system itself (Sterman, 2000). Changing an organizational system, however, cannot be done by one individual, as an “[o]rganization is an artifact of individual ways of representing organization” (Argyris & Schön, 1978, p. 16). Paradoxically, as organizational learning is dependent on individual learning, organizations can only learn through the experience and actions of individuals. Hence, to understand the learning process, the link between organizational and individual learning needs to be further explored
3.3.2 Organizational learning
As discussed, organizations learn through their individual members, who take actions based on their IMMs. Within the framework of organizational learning, it can be assumed that within an organization, individuals have a set of shared mental models (Argyris & Schön, 1978). Shared Mental Models (SMM) are both necessary and dangerous, as they may be so institutionalized that double-loop organizational learning is inhibited.
Cyert and March (1963) discuss a balanced perspective, where the behavior of the organization is changed in response to short-term feedback from its environment according to well-defined rules, and adapts to long-term feedback on the basis of more general rules. Their behavioral perspective gave input to the model of organizational learning as portrayed in Figure 5. The model portrays how individual actions are rooted in individual beliefs. Individual actions lead to organizational action, which produces an environmental response, which, in turn, affect individual beliefs. Additionally, March and Olsen (1975) portrayed the broken links that can inhibit learning: (1) Role-constrained learning when individual learning cannot cause individual action as the individual’s role constraints this; (2) Audience learning when individuals affect organizational action in an ambiguous way; (3) Superstitious learning, when learning takes place without a real basis for connections between organizational actions and environmental
response; and (4) Learning under ambiguity, when the causal connections among events are not clear, and only operational learning occurs. These barriers to learning may, in an organizational intervention process, act as barriers to change, and are therefore important to consider when designing the change process and interventions.
To fully understand organizational learning, an integrated model of organizational learning is presented in Figure 6, which links the OADI Cycle to both Individual- and Shared Mental Models. Both the individual and the organization can engage in Single- and Double-Loop learning, depending on whether environmental feedback leads to updating of IMM as well as SMM. The model shows that an organization can only learn through its members, but is not dependent on specific members. Individuals, however, can learn without the organization, and not all individual learning has organizational consequences. Again, learning cycles may be incomplete. The integrated model has identified three additional types of incomplete learning cycles that need to be considered in organizational interventions: (5) Situational learning when individuals improvise a solution but does not learn from it; (6) Fragmented learning when individuals learn but the organization as a whole does not; and (7) Opportunistic learning when organizations purposely try to bypass established ways of doing business, because it impedes a particular task (Kim, 1998).
Like the four barriers to organizational learning introduced by March and Olsen (1975), the three additional barriers to change identified by Kim (1998) need to be considered in designing organizational change programs. Individuals should be given the opportunity to
change their behavior, the freedom to make mistakes and correct them, and the time to understand their actions. Change requires reflection and communication with other members in the organization. Additionally, change-related decisions should be thoroughly analyzed and reflected upon by the organization and its members, to increase the likelihood that learning occurs with an understanding of the connections in the system and thus for the right reasons.
Organizational change, organizational interventions and organizational learning are widely researched concepts, that provide the context for the current study. Both GMB and GBL are intervention tools, that can support learning and change at both an individual and an organizational learning level, by changing individual and organizational mental models.. Learning, communicating and developing a shared organizational perspective will unfreeze the status quo within the organization, and allow for organizational change to take place (Weick & Quinn, 1999), providing that the barriers to learning and change are overcome. Though both GMB and GBL may lead to organizational change, a deeper analysis of the methods is needed to understand how they lead to change and what they have to contribute to overcome barriers to learning and change.
4 G
ROUPM
ODEL-B
UILDINGIn the previous chapter, the learning and organizational change context of interventions was discussed. The following two chapters will zoom in on two intervention methods: Group Model-Building (GMB) and Game-based Learning (GMB), thoroughly discussing the literature surrounding the methods, focusing on the roots and philosophy of the methods, the aspired outcomes and effectiveness, the process of the intervention, and possible shortcomings. By analyzing each of the methods thoroughly, a grounded comparison of the two methods can be made. The current chapter will focus on Group-Model-Building, an intervention method in which a facilitator, together with a group of stakeholders, develops an SD model.
4.1 DEFINING GROUP MODEL-BUILDING
Though briefly discussed in the introductory chapter of this research, a more elaborate theoretical discussion of the method of Group Model-Building will follow. The method was created in the mid-1990s, and its origins can be found in both the work of Vennix (1996) and the work of Richardson and Andersen (1995). Following their work, two different but comparable definitions of GMB can be found:
“Group model-building is a process in which team members exchange their perceptions of a problem […]. An important characteristic of group model-building (and system dynamics in general) is that ‘fact’ is separated from ‘value’. The primary focus is descriptive and diagnostic: the way team members think a system works is separated from the question how they would like a system to work. And this separation in itself often proves helpful to clarify the strategic debate” (Vennix, 1996, p. 3)
“Group model building […] signals the intent to involve a relatively large client group in the business of model formulation, not just conceptualization. The goals are a wider resource base for insightful model structure, extended group ownership of the formal model and its implications, and acceleration of the process of model building for group decision support” (Richardson & Andersen, 1995, p. 113).
Both the definition employed by Vennix (1996) and that of Richardson and Andersen (1995) are process definitions. Where the first focusses on the differences in perceptions that individuals may have and how a debate of their values may lead to additional insights, the latter focusses on the wider goals of GMB: increasing information availability and model ownership to accelerate the decision support system. The definitions are compatible enough to be integrated: GMB is a process aimed to provide insights in the strategic debate, which, by jointly creating a model of the system, supports strategic decision-making. To better understand the
types of decision-support models created, the underlying methodology System Dynamics (SD) needs to be understood. It is important to note, before the SD methodology is discussed, that an understanding of SD is not sufficient to understand the assumptions underlying group model-building, which is grounded in human information processing and behavioral decision-making (Vennix, 1996).
4.2 SYSTEM DYNAMICS
System Dynamics is a method to enhance learning in systems characterized by complexity (Sterman, 2000). The method, developed in the second half of the 1950s, takes a closed-loop approach to decision making, where decisions are seen as a means to affect the environment of a system, which, in turn, provides input to new decisions (Forrester, 1968). By connecting the field of engineering to management, social decision-making systems are studied as information feedback control systems. In this, the essence of SD can be found: “social systems can productively be studied as […] systems in which a decision affects the environment which in turn affects the decision” (Vennix, 1996, p. 43). Experimental simulation by means of computer models is considered a technique to reveal the behavior of the system, given the nonlinearities and complexity. SD as a tool is widely applicable and multidisciplinary. Reports of successful application are found in supply chain management; eco-systems; health care; project management; and climate change, amongst others (Sterman, 2000; Vennix, 1996).
On a very basic level, system dynamics theory thus describes and discusses the structure and behavior of complex systems, given the closed-loop approach. Systems’ behavior is determined by characteristics of the whole rather than the individual parts (Forrester, 1968). Through feedback-loops of interconnected variables, time delays, nonlinearities and accumulations, the system’s behavior is determined. Insights into the dynamic complexity of the system, or, systems thinking, can be generated through qualitative analysis of interconnected elements, or from a more quantitative simulation-focused analysis. Regardless of the approach, models are developed in the context of the real world, to solve a problem and gain insights. 4.2.1 System Dynamics Model-Building
Before going deeper into the assumptions and processes of GMB, some understanding of SD modelling is necessary. Modeling, as portrayed in Figure 7 is embedded in the larger cycle of learning and organizational change (Sterman, 2000). Although inherently creative, the overall modeling process includes five steps that are taken iteratively: “(1) articulating the problem to be addressed, (2) formulating a […] theory about the causes of the problem, (3)
formulating a simulation model to test the dynamic hypothesis, (4) testing the model until you are satisfied it is suitable for your purpose, and (5) designing and evaluating policies for improvement” (Sterman, 2000, p. 87). The first step, Problem Articulation, is perceived to be most important in the modelling process.
SD does not aim to provide models of entire systems. All models, in the end, are mere representations of the system, and are therefore always wrong in representing reality. Models can be useful, however, if they aim to reflect the complexity of a specific problem (Sterman, 2000). Usefulness of models is found in the simplification of reality, thereby creating a representation that one can understand. The effectiveness of model use is not only based on the problem definition, however, but also on how it uses the information base of the represented system (Forrester, 1991). System Dynamics models are grounded in a broader range of information sources than traditional social sciences, drawing upon a mental, written and numerical data base (Forrester, 1991). The three sources of data have a very different level of significance when it comes to SD modelling, as portrayed in Figure 8. Mental data contains more information than written data bases, which contains more information than numerical data bases. Additionally, mental data as an information category contains significantly more information that is needed for constructing a dynamic model (Forrester, 1991).
SD has proven its usefulness in analyzing complex problems and SD models are considered a good environment for designing policies and for learning. Learning from SD models, Forrester (1991) argues, is about an increased understanding of the general Figure 7: The System Dynamics modelling process and context. Retrieved from Sterman (2000, p. 88)
characteristics of a system, as well as surprising discoveries. SD has been widely applied and researched for organizational intervention and consultancy, but implementation of suggested policies remains troublesome (Größler, 2007; Lane, 1992; Ledet et al., 2005; Liehr et al., 2001; Rouwette & Vennix, 2006; Snabe & Größler, 2006; Sterman, 2000; Vennix, 1996).
4.2.2 Implementation problems in System Dynamics
One of the biggest challenges in System Dynamics modelling is implementing the model results. Organizational impact resulting SD projects is often rather low, even though the projects can be considered successful from an SD perspective (Größler, 2007). Success, from an SD point of view, is rooted in a clear identification of a problem. The model resulting from a systems’ analysis “should give people a more effective understanding about an important system that has previously exhibited puzzling or controversial behavior” (Forrester, 1991, p. 15). An influential project changes the way people think about a system, and it can only have this impact if it couples the concerns of the target audience, thereby altering relevant mental models. But, to achieve success, Forrester (1994) argues, implementation should be integrated in the SD-modelling process from the beginning.
Martinez‐Moyano and Richardson (2013) explicitly included implementation in the system dynamics modelling process, as portrayed in Figure 9. In contrast to Figure 7portrayed and described above, Figure 9 includes the use, implementation, and distribution of the model results. Implementation results from both an understanding of the model and an understanding of the system and the problem at hand. Größler (2007) argues as well that SD projects do not end with a model and proposed strategies, but rather should aid to implement the strategies in the organizational structure. But this gives rise to new challenges, which are only amplified by the nature of SD models, which are made of, built in, and built for social systems (Vriens & Achterbergh, 2006).
Figure 8: Decreasing information content in moving from mental to written to numerical data bases. Retrieved from (Forrester, 1991, p. 23)
Vriens and Achterbergh (2006) argue that SD modelers need an explicit understanding of the social dimensions in which they operate, in order to understand the correctness of the model and to aid in implementation. Implementation problems, according to them, result from a lack of understanding among SD modelers that the problem owner and the social system are not identical: each problem owner involved is part of various other social systems, leading to different values, perspectives, and agendas regarding the adoption of results. Additionally, SD based solutions are subject to communication and coordination problems. It could be argued that the ‘learning experience’ of the modelling process, be it participatory or not, remains at the individual level, or fades out in daily practices (Vriens & Achterbergh, 2006). This argument strongly aligns with two of the barriers to organizational learning identified before: fragmented learning and opportunistic learning. In addition, it captures some barriers to change, being a lack of consensus between members, or even managerial support.
To serve implementation of model results, Martinez‐Moyano and Richardson (2013) discuss some best practices, suggested by experts in the field of SD, the most important being: ❖ “Make sure the entire modeling exercise revolves around the problems of concern to the
client and audience” (Martinez‐Moyano & Richardson, 2013, p. 113);
❖ “Tell clear, coherent stories of model behavior using simple language and pictures of system structure” (Martinez‐Moyano & Richardson, 2013, p. 113);
❖ “Involve clients in telling model-based system stories and illustrate the stories appropriately using simplified causal diagrams” (Martinez‐Moyano & Richardson, 2013, p. 113);
❖ “Provide clear guidance for the use of flight simulators and model-based games, carefully debriefing participants after all learning exercises to ensure that they understand model-based insights” (Martinez‐Moyano & Richardson, 2013, p. 113); ❖ “Explain counterintuitive model behavior to contribute real system insights” (Martinez‐
Moyano & Richardson, 2013, p. 113).
These best practices support the argument by Vriens and Achterbergh (2006) that communication of System Dynamics models to both the direct client and the audience is difficult, but essential for implementation. A ‘simplified diagram’ implies something different for the modeler and for the client. In line with Meadows and Robinson (2002), it appears that SD modelers have different goals compared to the client organization, and as many SD practitioners are academics, this effect is only amplified (Coyle & Exelby, 2000). To overcome some of the problems associated with implementation, some tools emerged, like GMB. Communication, mental model elicitation, commitment, consensus and the generation of insights are central elements. By involving the client, or, to use the terminology of Vriens and Achterbergh (2006), by building the model in the social system, as a collective activity, the modelling process actively deals with some barriers to learning and change.
4.3 OUTCOMES AND EFFECTIVENESS OF GMB
A GMB process has three inherit goals: (1) it aims to create a climate for team learning, so that understanding of the problem at hand is enhanced; (2) it aims to create consensus among the members of the team. This consensus needs to be robust: premature consensus or groupthink may harm the GMB process and outcomes. Finally, (3) the GMB intervention should help to create acceptance of and commitment to decisions resulting from the GMB sessions.
It is difficult to assess to what extend the aspired outcomes of GMB are achieved. GMB processes do not follow a standard protocol, nor do the assessment studies following them (Rouwette et al., 2002). To aid the systemic analysis of GMB, the CICC model was developed, emphasizing Communication quality, Insight, Consensus and Commitment to conclusions. These constructs are measured by focusing on, e.g., the contributions participants made, the way in which discussions took place, the way information was treated, and how the process was perceived (Rouwette, 2011). The CICC model, supported by in-depth interviews, observations and content analysis, has shown positive results with regard to the GMB outcomes (Rouwette, 2011; Vennix, 2016). Definite conclusions on the effectiveness of GMB are very preliminary. Too little objective, experimental research is conducted, and subjective measurements are prone
to a success bias: only successful cases will make it to publications (Vennix, 2016). Generally, GMB sessions are received better than ‘meeting-as-usual’ (Vennix, 2016).
4.4 UNDERLYING PHILOSOPHY AND ASSUMPTIONS OF GMB
SD models, and thus models resulting from GMB processes, are based on real world processes and mental models. In complex problems, however, individuals may have different perspectives on the problem. People, by their very nature, have limited information processing capabilities, and thus make assumptions that may be irrational or unjust (Evans, 1984; Kahneman, 2011). Besides information distortion resulting from a limited ability to process information, different IMMs and selective perceptions lead to different interpretations of situations (Hogarth, 1987; Schein, 1990). By Group Model-Building, Vennix (1996) argues, mental models are systematically elicited and shared in teams to create a holistic problem view. Additionally, GMB is based on the assumption that individuals think in causal processes, (Weiner, 1985), but face difficulty identifying interconnections, seeing multiple causes of problems, and holistically understanding the relations between elements in the system (Dörner, 1980; Sterman, 2000; Vennix, 1996). By means of GMB, causal assumptions are elicited and integrated to get a complete and holistic picture of a problem.
GMB makes individuals learn about a messy problem, and by learning in the context of a group, a shared mental model should be created (Argyris & Schön, 1978; Phillips & Phillips, 1993; Vennix, 1996). By means of this SMM, commitment to potential solutions is created (Eden, 1992; Senge, 1990), thereby overcoming some barriers to learning and change. Thus, in addition to learning, consensus within the team is fostered throughout the GMB process, in order to create sufficient basis for implementation (Jones et al., 2005; Schein, 1990). The effectiveness of problem-solving is determined by both the quality of the decision, as the acceptance of the decision by those who have to implement it (Hart, 1985; Majone, 1984).
4.5 THE GMB PROCESS
The model-building process of Group-Model-Building is comparable to that of ‘normal’ System Dynamics modeling, as portrayed the diagram of Martinez‐Moyano and Richardson (2013) in Figure 9 discussed above. Iterations in the building process aim to validate the model, and implementation of the model is central to the building process, thereby increasing the likelihood of model use (Forrester, 1968; Sterman, 2000; Vennix, 1996). Model use is essential for tools like GMB, and System Dynamics as a whole, that is, to have an impact.
Group-Model-Building is not just a model: it is a tool that supports organizational change, by generating a better understanding of the situation at hand and of the policy options. Although models are useful here, the communicative and group aspects of the process are essential. GMB sessions are tailored to fit the needs of client organizations, and are heavily influenced by the organizations’ culture. The GMB facilitator, though adopting a neutral role, strongly influences the proceedings of a session (Vennix, 1996).
4.5.1 The role of facilitation
In a GMB session, three elements contribute to the outcomes: the content, the procedure used, and the process and ways of interacting in a meeting. GMB differentiates itself on the latter two: procedure and process (Vennix, 1996). It can be argued that facilitators do not need to be knowledgeable about the problem at hand. However, too little knowledge may hinder the process. The facilitation style needed is partially dependent on the type of group. It can range from a procedural to a more therapeutic role (Keltner, 1989). In GMB groups are guided in modeling in order to increasing insights. A role model style is most applicable (Vennix, 1996). The facilitator has a procedural role, assisting the group in the process of problem-solving.
Good facilitation is crucial to assure effective group model-building (Vennix, Scheper, & Willems, 1993). Both the problem and the GMB meetings are complex phenomena. The more complexity one aims to tackle in a meeting, the worse the group performance will be. All too often, groups move to irrelevant discussions, where problem-solving and critically evaluating alternatives are subordinate to finishing the agenda. It is the facilitator’s task to manage procedural difficulties and understand the group process (Jensen & Chilberg, 1991). Although there is no clear definition on how to facilitate, a set of attitudes was distinguished by Vennix (1996). These affect the way people will respond to certain situations. The way a facilitator reacts to situations will partially determine the group process and effectiveness of problem solving. A GMB facilitator needs a helping attitude, be authentic and integer, have an inquiring attitude and be neutral. The facilitator needs to be able to structure group processes, handle conflict on task and relational levels, and communicate effectively. By the attitudes and skills, consensus, commitment, insights and enhanced communication can be achieved and the relationships within a team strengthened (Vennix, 1996).
4.5.2 Set-up of a GMB session
The facilitator of a GMB process is not only involved during the session, but also responsible for designing it. There are a few choices to be made in session design. The facilitator needs to determine, often together with the client organization, who to involve. Additionally,
times and dates need to be set, rooms need to be set-up, and the session itself needs to be designed. Andersen and Richardson (1997) argue that modelers, to design GMB sessions, rely on fairly sophisticated pieces of small-group processes. These so-called scripts are used sequentially, to move the modeling process through various stages, such as problem definition, sketching the model structure, and determining the actions to be taken. A catalogue of scripts as tested and refined procedures is created, to support high quality client-centered GMB projects, and to make the method more transparent and accessible. The catalogue, Scriptapedia, can be freely accessed online (https://en.wikibooks.org/wiki/Scriptapedia). The scripts draw upon well-established SD wisdom, focusing on preceding’s of a workshop (e.g. planning, logistics and scheduling), scripts and techniques for various GMB tasks (e.g. concept models, graphs over time, and nominal group techniques), and closing and evaluation of the session (Andersen & Richardson, 1997). Luna-Reyes et al. (2006), through case-study research, demonstrated the effectiveness of scripts in designing a GMB session with successful outcomes. Generally, GMB sessions follow what Kaner (2014) describes as the Diamond of Participatory Decision-Making, as portrayed in Figure 10. The group, during the sessions, will move through three stages: divergence of ideas; the groan zone in which discussion leads to prioritization of ideas and concepts; and the convergence stages where the broad range of ideas and solutions are brought back to a point of closure. Facilitation is essential in all of these stages, to prevent groups from reaching premature consensus, or from futile conflict (Vennix, 1996).
4.6 SHORTCOMINGS OF GROUP MODEL-BUILDING
Although the outcomes and the reaction of participants to GMB sessions is generally positive (Vennix, 2016), the method may not have been able to solve all implementation problems. Größler (2007) reports on a few instances where the model appeared successful and was received well by the team, but no use of the results, being it immediately after or sustainably over the long term, was achieved. It was argued that low-impact models result from Figure 10: The diamond of participatory decision-making. Retrieved from Kaner (2014, p. 20)
practitioners neglecting the importance of implementation. An understanding of GMB projects as organizational interventions, in which stakeholder management and implementation are central elements is crucial to impact organizational results (Größler, 2007). Figure 11elegantly portrays this argument of Größler: after the discussion phase, in which the GMB process is embedded, a decision will follow. But the decision on its own will not change the organization: it needs to go through a phase of implementation. Thus, GMB projects may need to be more embedded in an organization-wide, comprehensive intervention process (Größler, 2007). This embedded process, or intervention architecture, leads to total organizational change (Zock & Rautenberg, 2004).
To summarize, the purpose of GMB is threefold: “[t]he first is to create a climate in which team learning can take place in order to enhance understanding of the problem. The second is to foster consensus. Finally, the intervention should help to create acceptance of the ensuing decision and commitment with the decision” (Vennix, 1996, p. 6). This purpose appears to be achieved, as the method leads to consensus, commitment, insights and communication. Grounded in System Dynamics, GMB models are built provide an understanding of complex problems. Models are participatorily built based on IMMs through facilitation. The method appears to overcome some of the implementation problems related to the wider field of SD, particularly by explaining and communicating insights to those involved. However, as only a rather small group of stakeholders is involved, some implementation problems remain. GMB may benefit from a combination of other methods, focusing on both organizational intervention and implementation (Größler, 2007). Martinez‐Moyano and Richardson (2013) suggest that model-based games may aid in learning about the model, and thus overcome barriers to change. A thorough understanding of Game-based Learning is needed, however, to effectively design, guide and debrief these model-based games.
5 G
AME-B
ASEDL
EARNINGFollowing the literature analysis of the philosophy and processes of GMB, the method of Game-Based Learning is analyzed. GBL is increasingly used in organizational change and interventions (Mayer, Warmelink, & Zhou, 2016). Simulation games have been a primary tool for learning and development for decades, and have increasingly gained attention in research and practice. Games are generally considered as an easily accessible, low-cost, yet effective tool for training, although some argue that games only lead to superficial learning (Wilson et al., 2009). Nevertheless, interest in GBL is growing as the field of learning shifts from a didactic model of instruction to a learner-centered model (Garris et al., 2002).
5.1 DEFINING GAME-BASED LEARNING
Despite the increasing attention, papers and practical projects involving GBL, there is still a lack of consensus on the terminology used (Sauvé, Renaud, Kaufman, et al., 2007). An important aspect of this discussion is the distinction between simulations and games. Based on a thorough analysis of various articles and reports, Sauvé, Renaud, Kaufman, et al. (2007) gave the following definitions for Simulations and Games:
“A game is a fictitious […] situation in which players are put in a position of conflict […]; at other times, they are together and are pitted against other forces. Games are governed by rules which structure their actions in view of an objective or a purpose which is to win, to be victorious or to overcome an obstacle. They are integrated into an educational context when the learning objectives are associated formally to the content and the game enhances learning in the cognitive, affective and/or psychomotor domains” (Sauvé, Renaud, Kaufman, et al., 2007, p. 253). “[A] simulation is a simplified, dynamic and precise representation of reality defined as a system. A simulation is a dynamic and simplified model or reality and it is judged by its realism, by its correspondence to the system which it represents. A game is created without any reference to reality, what is never the case for a simulation or a simulation game. Simulation is not necessarily a conflict, a competition, and the person who uses it is not looking to win, what is the case in a game” (Sauvé, Renaud, Kaufman, et al., 2007, p. 253)
The definitions imply that simulations, unlike games, can function without human interventions. In that regard, a System Dynamics model can be considered a simulation. It is a
simplified model of reality. Given the aim of this research, however, it is important to consider, within the simulation definition, the role of human interventions, in what Sauvé, Renaud, Kaufman, et al. (2007) describe as Simulation Games.
“A simplified and dynamic model of a real or hypothetical system in which players are in position of competition or cooperation, rules structure player actions, and the goal is to win” (Sauvé, Renaud, & Kaufman, 2010, p. 193).
All three concepts discussed above, simulations, games, and simulation games can be used in an educational context, where each of the concepts draws upon a different but comparable set of critical attributes, as portrayed in Table 2. Despite the comparable attributes of the three different game concepts, the outcomes associated with each of the concepts are different.
5.2 OUTCOMES AND EFFECTIVENESS OF GAME-BASED LEARNING
Simulations, games and simulation games can all be used for learning, as these game environments enable participants to deal with problems in authentic situations (Kriz, 2010). Games can include multiple contexts and generally draw upon multiple competencies, by exposing learners to complex situations (Kriz, 2010). Additionally, games played with multiple players emphasize team-based problem solving, team learning, and collaboration skills, such as communicating and negotiating (Hesse, Care, Buder, Sassenberg, & Griffin, 2015). Games, according to Kriz (2010), make the consequences of decisions transparent, by diagnosis, analysis and assessments of responses to critical situations.
Table 2: Critical attributes of Games, Simulations and Simulation Games in an educational context. Based on Sauvé et al. (2010)
Problems associated with the outcomes of game-based learning, however, can be traced back to the differences in definition described above. Games, simulations, and simulation games, although comparable in definitions, critical attributes are somewhat distinct in outcomes. Sauvé, Renaud, and Gauvin (2007) analyzed the three concepts in existing literature in terms of their outcomes. An analysis of 190 articles associated with ‘Games’ found that games are positively associated with knowledge structuring and assimilation, the development of problem-solving skills, motivation of learning, and communication skills, amongst others. ‘Simulations’ are, based on 58 articles, associated with knowledge structuring, self-confidence, transfer of learning, self-reflection and evaluation; ‘Simulation Games’ (64 articles) are associated with, again, knowledge structuring, problem-solving skills, communication, motivation, transfer of learning and development of critical thinking (Sauvé, Renaud, & Gauvin, 2007). In Table 3 all outcomes of the three concepts are summarized. Based on the outcomes, it can be concluded that a distinction of games, simulations and simulation games is necessary to define the learning outcomes associated with game-based learning as a whole.
