A Lattice Representational Definition of a Hierarchy of Instructional Processors Usable in Educational Courseware

The basic “recognize-act-recognize-end” cycle can be recognized in elementary as well as in more advanced forms of CAL This article attempts to offer a unifying formal framework in which different elaborations of this cycle (embodied in a “processor”) can be placed. Three different levels of elaboration are distinguished which can be considered to be situated into the nodes of a lattice of models of the instructional process. A formal definition of such a framework can serve at least two functions. In the first place a uniform and precise definition of various elaborations can be given and new elaborations can be created in a logically funded way. Secondly, such a framework can support the modelling of instructional processes and the stimulation of student behavior. Thus, pre-testing of courseware could become feasible. Aspects of the framework have been used to implement two prototypes of support systems for the development of CA1 courseware. INTRODUCTION AND OBJECTIVES In Computer Assisted Instruction, as far as the execution of the instructional process is concerned, one can speak of a basic “recognize-act-recognize-end” cycle in which, based upon the instructional situation of the learner, some information is presented, his or her reaction is analyzed and some new information is given. This basic model can already be found in early literature about CAI, for instance Stolurow[l,2] and Merril[3]. This elementary cycle can also be recognized in more complicated forms of CAI, such as those encountered in Intelligent Computer Assisted Instruction, for instance in O’Shea and Sleeman[4] and in O’Shea et aL[5], albeit in a more complicated form. The embodiments of this basic cycle functions as an instructional processor, by means of which the instructional process is brought from one process stage into the next. In this article three levels of elaboration of the basic cycle are given. These levels are based upon the extent of sophistication of the decision making process involved in the instructional strategy. A distinction will be made between “instrumental decision making”, “normative decision making” and “adaptive normative decision making”. In our opinion there are at least two incentives to try to describe this cycle more formally and to place various elaborations of it into common definitional framework. In the first place, a precise and uniform description of this basic cycle that underlies most CA1 forms could be given. Such a description can elucidate which (types of) relations exist between some less and some more advanced forms of CA1 and how they can be transformed into each other. Furthermore, these relations could be transformed into each other. Furthermore, these relations could be formalized as will be shown in this article. Secondly, such a formal description makes it possible to model various elaborations of this cycle. When this modelling can take place in an environment for the developing of courseware, the door to student-simulation, i.e. simulation of how students with various characteristics would behave in various elaborations of such a cycle, could be opened. This result makes pre-testing of educational courseware feasible; getting information about the quality of courseware while the product is still in the design stage. Presently, we are developing two lines of research based upon the formal description of the basic instructional cycle. The first one is of a theoretical nature and serves mainly to discern between elaborations of the cycle and new functions therein. This article is a contribution in this first line. The order of the names of the authors is alphabetical; they are equally responsible for the content.