17.5 SYMBOLIC SIMULATIONS

In contrast to experiential simulations, a symbolic simulation is a dynamic representation of the functioning or behavior of some universe, system, set of processes, or phenomena by another system (in this case, a computer). In other words, symbolic simulations are populations of events or sets of interacting processes. The role of the learner in relation to a symbolic simulation is typically that of a researcher or investigator. 'Mat is, the learner manipulates different variables in order to discover scientific relationships, explain or predict events, or confront misconceptions.

Symbolic simulations may be classified according to the nature of the variables and the nature of the interactions among them. Four types of symbolic simulations are currently in use that differ in these characteristics. They are data-universe simulations system simulations process simulations. and laboratory-research simulations.

17.5.1 Data Universe Simulations

A data universe simulation represents the behavior of sets of related elements that compose a population of continuing events. The simulation expresses the relationships among the variables through the use of mathematical equations. An example is the population ecology simulation described earlier. The simulation illustrates the effects of the 75 variables on population, capital investment, food production, pollution, and quality of life (Forester, 1971; Hinze, 1984). The output is a graph that illustrates the effects of continued turn-of-the-century trends on the five characteristics of civilizations. Trends also may be altered by the user and the effects observed.

The situation typically posed for the student in a data universe simulation is to test student-generated hypotheses about a large population of interrelated variables and outcomes. The goal is to discover relationships or trends among the variables. The purpose of a data universe simulation typically is to provide students with opportunities to discover scientific laws and principles, such as the laws of genetics (see 24.9 for a discussion of databases and cognitive tools).

Note that data universe simulations differ from other simulations that involve the manipulation of variables. First, students are functioning as researchers by testing their hypotheses, reviewing the outcomes, and testing new hypotheses or continuing their research strategy. In other interactive exercises, students are often attempting to solve a problem that has been posed for them and/or they are working with a smaller database. For example, in a data management simulation, the student is executing specific role-related responsibilities in which the goal for the student or the team is to enhance the economic status of an institution or enterprise.

17.5.2 System Simulations

A system simulation demonstrates the functional relationships between the components of a physical or biological system (such as a small ecosystem) or a constructed system (such as complex equipment systems). Students learn about the particular system or solve problems involving the system by manipulating the components.

One important role for the interactive graphics and videodisc capability of current computer technology is to provide functional representations of complex systems that students can operate. An example is the steam plant system and subsystems developed for the U.S. Navy known as STEAMER. The exercise also includes a quantitative component so that the student can open and shut valves, turn components on and off, adjust throttles, and observe the effects on indicators, such as dials, thermometers, and digital readouts (Stevens & Roberts, 1983).

System simulations are often used to teach the operational principles of complex equipment composed of subsystems. They also are used to teach procedures and may, depending on the design of the simulation, develop students' cognitive strategies. The use of a simulation to teach maintenance procedures, for example, is the procedural simulation referred to by Riegeluth and Schwartz (1989).

Examples that develop students' cognitive strategies are the low-cost plywood MI tank simulators and M2/3 fighting vehicles, each with its own microprocessor database of the terrain, graphics, and sound system developed in project SIMNET. Each "armored vehicle" is a system that generates the battle engagement environment required for the combat mission training of its crew. Each crew member sees a part of the virtual world defined by his line of sight (e.g., forward for the driver) (Alluisi, 1991, p. 350).

17.5.3 Process Simulations

The focus of a process simulation is a naturally occurring phenomenon in the physical, chemical, or biological realm (Riegeluth & Schwartz, 1987). Interactive graphics images can illustrate processes that are unobservable and/or are not easily experimented with in the classroom. Students can manipulate variables and attempt different tasks in order (1) to discover the relationships among the variables or (2) to confront their misconceptions.

Confronting student misconceptions about Newtonian mechanics is the goal of several process simulations developed in physical science (Flick, 1990; White, 1984). DiSessa (1982, 1985) and others note that students' intuitive knowledge about force, motion, and velocity derived from experience in a gravity-bound world often prevents Students' construction of accurate mental models of physics principles. White (1984, 1995), for example, has designed several progressively more difficult gamelike tasks that require the student to perform several actions on a "spaceship" in a frictionless environment (space). Force, velocity, and speed are illustrated in the interactive exercises.

DiSessa (1982) identifies three important contributions of process simulations that represent physics principles. First, they provide students an opportunity to interact with phenomena at a qualitative level. Often, students only interact with quantitative problems in which getting the right answer typically becomes their goal. Second, students' fragmented and often naive knowledge of phenomena is challenged. Third, simulations can change the time scale of exercises from the 20 minutes or so per type to problems that can engage students in investigations that can span days or weeks.

17.5.4 Laboratory Research Simulations

Laboratory-research simulations are specific to courses that include laboratory sessions as part of the course work. Among them are biology, chemistry, physics, and, occasionally, physical science. These exercises provide visual and graphic components for students to manipulate, and they illustrate the results. Early examples of chemistry experiments used color microfiche images projected onto the back of a plasma panel with a PLATO I-V system (Smith & Sherwood, 1976). Currently, computer laboratory simulations are making use of videodisc technology to expand the range and complexity of the experiments conducted by students.

These simulations differ from data-universe and process simulations in that they are a series of discrete problems. Because laboratory research exercises are a series of discrete experiments instead of a complex evolving problem, they are categorized by some theorists as problem-solving exercises in a simulated context (Gredler, 1992a). Nevertheless, the computer videodisc simulations provide realistic experimental reactions. Further, students can conduct experiments that involve hazardous or costly materials. Also, slow reactions that students may not ordinarily be able to observe may be sped up (and others may be slowed down). Moreover, experiments can be repeated (Smith & Jones, 1989).

17.5.5 Discussion: Symbolic Simulation

Symbolic simulations may be developed at any of several different levels of complexity. Data universe simulations are the most complex, in which a large population of events is represented and the causal models are quantitative. System simulations are less broad and may involve either quantitative or qualitative models of causality. Process simulations, in contrast, typically address specific processes in the physical world that are often poorly understood by students. In addition to biological processes, the interactions of variables such as force, speed, and velocity are typical examples. Causal models for process simulations also may be quantitative or qualitative. Laboratory research simulations, in contrast, involve a series of discrete activities that are directed by students. Again, the causal models for the specific experiments may be quantitative or qualitative (see Table 17-3).