Goal. Given a set of conditions, the student is able to predict the consequence of an event. Or given a consequence (expected or unexpected) the student is able to identify the conditions which were present in order for this consequence to occur. When an unexpected consequence occurs (an error) then finding the precipitating conditions is called trouble shooting.
Knowledge Structure. A set of process rules consisting of conditions and consequences. A goal to be accomplished (value for some property or set of properties) and a set of specific problems (initial conditions). Some of the problems may contain faults. A fault is an incorrect condition. That is, when a faulted process executes an unexpected consequence occurs because of the incorrect condition. An incorrect condition can represent a damaged component, a missing component, or a control set to an inappropriate value.
The same PEAnet knowledge structure described for the explore interaction is required for an INTERPRET transaction. It may be necessary to include additional conditions representing faults. Selecting a problem is an action which triggers a process which changes the initial conditions of the system to some predetermined values.
Presentation. The student is allowed to play with the learning environment to "see what happens if ƒ." An explanation is provided which indicates the consequence of each action ("what happened") and the conditions which were satisfied or unsatisfied ("why"). If the process applies to a number of different situations, then a number of different scenarios (variations on the learning environment) are provided to the student and the student is allowed to "play with" these variations. In some situations the student is provided a "control panel" which allows them to set the values of some of the properties of the system. In this way the student can perform "experiments" by observing the consequences of different conditions (property values) and receiving the explanation for these consequences. One advantage of a learning environment is that there need not be a distinction between conditions that the student can change as a result of their actions (action triggers process which changes a condition or consequence) and conditions which cannot be changed by student actions. We can always give a student some control action (an action not available in the real world) that allows them to experiment with the system. (An example is a gravity control which increases or decreases gravity.)
Practice. The student is presented a specific problem in the learning environment. The student is asked to observe the conditions of the device or system and to predict one or more consequences. In the system described the student is provided a list of properties and allowed to select the value for each property that constitutes the consequence of a given action. One or more of the conditions given may represent a faulted condition. A variation is to provide the student a "control panel" and direct them to set the conditions (property values) which will lead to a specified consequence. The consequence may be one that would result from a faulted condition. The prediction is confirmed by allowing the system to execute and show the consequence of the execution. In complex systems the student may have to trace the execution back through several events to find the condition(s) which caused the observed consequence. Or the student may need to execute several events of the system to see the consequence that results from the conditions.
Learner Guidance. During the presentation the conditions necessary for a consequence for given event are made clear to the student. Often the best guidance is to allow the student to ask for an explanation of an unexpected consequence during exploration. This explanation identifies the conditions which caused the consequence. During practice the student's predictions or trouble shooting are confirmed by executing the system with explanations of what occurred during each step of the process and why.
Space limitations makes it necessary to describe only a few of the transactions which can be implemented by knowledge object architecture. We believe that any instructional interaction can be implemented by the appropriate slots added to knowledge objects and the appropriate algorithm specified for manipulating these elements of knowledge objects. In subsequent papers we will specify some of these additional instructional algorithms in terms of knowledge objects and instructional transaction algorithms.
Perhaps one of the most exciting possibilities for the use of knowledge object architecture for instructional design is the possibility for truly adaptive instruction. An instructional strategy is an algorithm An algorithm can be pre-specified and pre-programmed. In addition an instructional algorithm can include a number of parameters, the different values of which control the way the strategy promotes interaction with the student. Changing a strategy parameter value changes the nature of the interaction. An adaptive system would include a set of expert system rules relating student parameters to instructional strategy parameters. When a student parameter changes the expert system would then change the parameter of the strategies and the subsequent interaction with the student would change. An adaptive system would include a system for monitoring student parameters (level of motivation, level of interest, performance, and other parameters). When the values on these student parameters change then the system dynamically changes the values of the strategy parameters thus adapting the system to the individual student.
The knowledge object architecture together with its simulation engine and inference engine makes it possible to build a learning environment in which the student him/herself is an object in the learning environment. Like other objects in the learning environment the "student entity" has properties and property values. These property values are changed by processes triggered by actions or other processes just like all other entities in the system. Thus, when the properties of the "student entity" change the actions that are available to the student change. The learning environment thus adapts to the student or interacts with the student in a dynamic way.
There is not space here to elaborate adaptive systems. A subsequent paper will describe a theoretical model for adaptive instruction.
In this chapter we have concentrated on the component methods of ITT. We have suggested that a more precise representation of the knowledge to be taught in the form of knowledge objects increases the precision with which instructional strategies can be described. We have indicated that instructional strategies can be described as methods for manipulating the elements of knowledge objects. This architecture enables the specification of executable knowledge making possible tutorial and experiential instruction from the same knowledge representation.
In this chapter we have described knowledge objects. We have described how knowledge objects are used to define a number of instructional strategies including: learning environments (exploration), identify transactions, execute transactions, and interpret transactions. We have illustrated one implementation of these strategies in an instructional learning environment.
In our previous presentations of CDT we have provided prescriptions for best case strategies for different kinds of instructional outcomes. In this chapter we could have recast these prescriptions in terms of strategy algorithms based on knowledge objects but we chose instead to present the architecture of the system rather than the prescriptions. This chapter is a presentation of the foundation for instructional design based on knowledge objects but is not a complete presentation of the theory and does not include all of the items that were suggested by the editor of this volume.
Gagné, Robert M. (1965). The conditions of learning. New York: Holt, Rinehart and Winston.
Gagné, Robert M. (1985). The conditions of learning and theory of instruction. Fourth Edition. New York: Holt, Rinehart and Winston.
Jones, M.K., Li, Z. & Merrill, M.D. (1990). Domain knowledge representation for instructional analysis. Educational Technology, 30(10), 7-32.
Li, Z. & Merrill, M. D. (1990). Transaction shells: A new approach to courseware authoring. Journal of Research on Computing in Education, 23(1), 72-86.
Merrill, M. David (1983). Chapter 9, Component display theory. In Charles M. Reigeluth (Ed.) Instructional-design theories and models. Hillsdale, NJ: Lawrence Erlbaum Associates.
Merrill, M. David (1987). Chapter 7, A lesson based on component display theory. In Charles M. Reigeluth (Ed.) Instructional design theories in action. Hillsdale, NJ: Lawrence Erlbaum Associates.
Merrill, M. David. (1997). Learning-oriented instructional development tools. Performance Improvement, 36(3), 51-55.
Merrill, M. D., Li, Z. & Jones, M.K. (1991) Instructional transaction theory: An introduction. Educational Technology, 31(6), 7-12.
Merrill, M. David, Jones, Mark K., & Li, Zhongmin. (1992). Instructional transaction theory: Classes of transactions. Educational Technology, 32(6), 12-26.
Merrill, M. David, Li, Zhongmin, & Jones, Mark K. (1992) Instructional transaction shells: Responsibilities, methods, and parameters. Educational Technology, 32(2), 5-27.
Merrill, M. David & ID2 Research Team (1993). Instructional Transaction Theory: Knowledge relationships among processes, entities, and activities. Educational Technology, 33(4), 5-16.
Merrill, M. David (with David G. Twitchell, Ed.), (1994). Instructional design theory. Englewood Cliffs, NJ: Educational Technology Publications.
Merrill, M. David & ID2 Research Team (1996). Instructional transaction theory: Instructional design based on knowledge objects. Educational Technology, 36(3), 30-37.
The buttons that appear below will be found at the top and bottom of each page of the discussion. The first button will take you back to the previous page (in this case, to the beginning of the second section of paper 22). The middle button will take you to the ITForum home page. The last button takes you forward into the discussion.