Learning Outcomes of Simulation Games

Learning Outcomes of Simulation Games

Many published studies focused on comparing simulation games to conventional instruction where the primary outcome measurement is content retention. Predictably, these studies have achieved mixed results. Some indicated no difference in content retention between the two types of instruction (Keeffe et al, 1993; Randel et al, 1992; Bredemeier and Greenblatt, 1981; Brenestuhl, 1975; Greenlaw and Wyman, 1973). Others claimed conventional instruction allowed more content retention (Boseman and Schellenberger, 1974). Others said simulation/games encouraged better learning (Page & Roberts, 1992; Stembler, 1975; Braskamp and Hodgetts, 1971). Other longitudinal studies claimed that remembering content over a period of several months was better for simulation/games (Sprecht & Sandlin, 1991). This line of inquiry suffers from the inherent complexity of simulation gaming. As a learning outcome, content retention is primarily a feat of sensory perception and memory. Simulation games also engage elements of personality (e.g., extroversion, introversion, ambition, aggressiveness, reticence, etc.), social ability (i.e. ability to assess social relationships, persuade, charm, tolerate ambiguity, etc), as well as higher forms of cognitive thinking (i.e. analysis, synthesis) (Keegan, 1994). Simulation games also invite a reflective contemplation of personal values and belief systems. While conventional instruction may intend to engage many of these same elements, it does so in an indirect manner, and ultimately changes in these elements are not measured or evaluated. Simulation games directly engage the elements of personality, social ability and value/belief systems. The debriefing process provides the opportunity to re-evaluate changes in these elements. In comparing simulation/games to conventional instruction, measuring content retention fails to acknowledge the complexity of learning outcomes for simulation/games.

Theoretical Perspectives for Teaching, Learning and Instructional Design

Only a few resources are available to those who might be interested in designing simulation/games, and even fewer are available to describe the design of computer simulation/games. The design of modest computer simulation/games for a high school setting has been described (Keegan, 1994). This resource offers a thorough review of the relevant learning theories and psychological research with regard to instructional design of simulation games. While it is an adequate how-to primer for small games, it offers little advice on how to develop the larger, complex games that would be more appropriate for adult education. Like most of the constructivist literature, it seeks to validate the value of constructivist theory but fails to provide step-by-step guidelines for designing software. However, Keegan (1994) developed a software product which accompanies the book and can be used by high school teachers (and students) to design relatively simple simulation/games.

The apprenticeship method is also a time-honoured way of providing learning opportunities. The expert is continuously present to provide guidance as needed to the apprentice, and act as mentor initiating the apprentice into a community of practice. But in the modern world, the matching of apprentices to an expert in a resource rich environment is often viewed as too expensive. Many computer technologies have attempted to recreate this relationship with the computer as ‘expert’ and the student as ‘apprentice’—so far none has succeeded (Woolf, 1990). Most versions of computer-assisted-instruction are little more than drill and practice (Suppes, 1990). The limitations of expert systems and artificial intelligence have been described (Illovsky, 1994). Useful applications of artificial intelligence and neural networks have yet to emerge from the research laboratory. But combinations of these technologies are producing useful tools. The architecture of a computer simulation/game is usually framed around an expert system of rules. Artificial intelligence is used to animate and activate the virtual competitors who play against the end-user. The design team uses a combination of these technologies to give the ‘appearance’ of a knowledge master. If professor/instructors were proficient with these same tools and processes, they might succeed in creating a virtual apprenticeship in decision making—a cognitive apprenticeship.

Learning through apprenticeship has been an expensive, labour-intensive process. Computer simulation/games distributed through online education may represent an opportunity to realize some efficiencies. If computer simulation/games succeed in recreating the master/apprentice relationship, the professor/instructors may need to retreat from ‘transmission of content’ and embrace the teaching perspective of apprenticeship.

Apprenticeship learning in a work environment offers the opportunity to create personally relevant meaning. Work environments also create a context for learning that is associated with high levels of learner motivation. One author reports the dysfunctionalities of conventional classroom instruction (Raizen, 1994). This is an overview of theory and practice as it relates to the informal learning of apprenticeship programs. Conventional instruction emphasizes declarative knowledge, while apprenticeship learning usually emphasizes tacit procedural knowledge. Theories of situated learning emphasize the importance of the environmental context that surround the learning opportunity. The ideas of ‘socially constructed knowledge’ and ‘learning from error’ also emphasize the relevance of environmental context. The theory about transition from novice to expert offers a rich model for considering learning through simulation/games (Milech et al, 1993). When learning something new, a novice accesses fragments from a variety of models, and does so incompletely and inadequately. After a process of experimentation, trial and error, the novice begins to synthesize the fragments into an entirely new model. Experts can be characterized as those who have synthesized many different models into new more efficient and robust models. Simulation/games may offer the learner a rich resource of models from which they can pick and choose a variety of fragments. Simulation/games also offer a meaningful context for experimentation, as well as trial and error.

Traditional perspectives of instructional design are grounded in behaviourist and cognitive epistemologies (Gagne, 1992). These perspectives are not appropriate for the simulation games because the focus is on systematically designed instruction that focuses on the attainment and measurement of fixed objectives. Simulation games immerse the learner in complexity, and the learner chooses how to and in what sequence to manage that complexity. Other authors have tried to review the variety of instructional design perspectives that go beyond the behaviourist perspective (Romiszowski, 1981, 1981a). None of the well known models of instructional design has proven to be applicable in every situation. Eight different types of instructional design were described, from reception learning (conventional rote instruction) to impromptu discovery (unplanned learning). On this continuum, guided-discovery most closely resembles simulation/games, where objectives are fixed and the learner is guided to appropriate methods. However, in a simulation/game the student/user often chooses the objective. A less than optimal choice of strategy may be just as instructive as choosing an optimal strategy. These tactics of traditional instructional design assume the learner will always need to approach the content in the following ways: (a) from simple to complex, (b) from known to unknown, (c) from particular to general, and (d) from concrete to abstract. The immersive environments of simulation/games challenge these tactics—often the learner must reverse tactics (i.e. from the complex to the simple, unknown to known, etc.)

One theoretical perspective offers insight into the difficulties of categorizing the kind of learning offered by simulation/games. The instructional design of simulation games could be more aptly described by a constructivist epistemology (Jonasson et al, 1995):

The constructivist sense of active learning is not listening and then mirroring the correct view of reality, but rather participating in and interacting with the surrounding environment in order to create a personal view of the world (p. 11).

While constructivism makes claims for the value of this orientation, it does little to guide the designer. Two main features of constructivist instructional design require more how-to guidance: interactivity and problem solving. Designing for interactivity should include the following considerations (a) immediacy of response, (b) non-sequential access of information, (c) adaptive communication, and (d) bi-directional communication (Borsook & Higginbotham-Wheat, 1991). Designing for problem solving would focus on cognitive operations that transform the mental representation of objects, both images and concepts (Dijkstra, 1991). Declarative knowledge is developed by solving problems, and the operations must allow the learner to discover relationships between objects. The author then claims that “there is no clear instructional design model for the teaching of relational concepts.” (p. 23). Managing the complexity of a simulation/game is dependent on analyzing the relationships between many different variables. Therefore, with respect to the design of simulation/games, the current state of instructional design theory is impoverished and may not provide a clear theoretical focus to analyze data yielded by this thesis.