Hierarchical Control and Management in a CAI Visual Authoring Environment

Abstract

A review of the historical development of CAI from the creative breakthrough in 1958, through the replication period [1959-66], the empirical period [1967-74], the theoretical period [1975-82], and the automation period [1983-90] was carried out. The evolution of the sequence control and courseware management aspects of CAI languages and authoring systems was examined. These two aspects are the focus of this thesis. It was found that almost all CAI languages and authoring systems tend to provide a two level system of management and control: a within-file system and a separate and distinct between-file system. It was concluded that a more unified multi-level system of management and control features in a CAI system would enhance courseware organization, design and development. Ideas for the design of a large scale CAI system were also contributed from the following areas of computer science: the concept of abstraction, visual programming, human-computer interaction, and graphical user interfaces. The design of a large scale, multi-user CAI system was proposed based on a modular CAI language, ERAS, which has six sub-languages: CONTROL, CONTENT, DISPLAY, INPUT, ANSWER, and MENU. The system supported a hierarchical courseware data base and a visual authoring environment. It was designed to have a unified look and feel for all classes of users, to incorporate features that support user and courseware registration, and to assist authors at the design stage of courseware development. Two single user prototypes were developed to test some of the design features and the user interface, one on a Digital Equipment VAX using Elf (Educational Language Facility) and the other on a Commodore AMIGA using Intuition and GFABasic with Extend. The features tested included a multibranch tree structured data base, the nodes of which held all courseware modules. The authoring environment was maintained on an execution stack as the author navigated the tree structure. The visual authoring environment consisted of a file card metaphor of the data in the tree structure as well as a tree structure editor which presented contextual information on the tree node currently being edited.