The design of a rule-based CAI tutorial

Intelligent Tutor, Environment and Manual for Introductory Programming

“…This reflects the state of things: in ITEM/IP the domain expert is replaced by the interpreter of the mini-language with the same abilities as the expert. The same technique was used successfully in BIP (Barr, Beard and Atkinson, 1976), QUADBASE (Hudson and Self, 1982), and ReGIS (Heines and O'Shea, 1985) systems. The structure and functions of ITEM/IP components are given below.…”

“…The ITEM/IP programming laboratory, like the BIP (Barr, Beard and Atkinson, 1976), and the ReGIS (Heines and O'Shea, 1985) laboratories, supports `the experiment style' of working: design V the program -experiment (interpret the program) -observe -change the program -experiment again, etc. This style of work enables the student to learn from experience the concepts and constructions of the Domain model, to `discover' their features, and to get an idea of programming technology.…”

“…According to McCalla and Greer (1987), intelligent computer-assisted instruction (ICAI) attempts to bridge the gap between traditional CAI systems and microworld environments. There exist a few intelligent tutoring systems in the field of programming that are able to work as tutoring or coaching systems as well as to create a friendly environment for novice programmers; namely: BIP (Barr, Beard and Atkinson, 1976), Lisp-Tutor (Anderson and Reiser,1985), ReGIS (Heines and O'Shea,1985), Bridge (Bonar and Cunningham, 1988), IPTS (Cheng, Hu and Yang, 1988), GEL (Reiser, Ranney, Lovett and Kimberg, 1989), and Ugo (Innocenti, Massucco, Persico and Sarti, 1991). One of the possible ways of constructing such an integrated intelligent system for programming is suggested in (Brusilovsky, 1987).…”

A method for sequencing instructional objectives which minimizes memory load

“…As other authors (Heines and O'Shea, 1985;Stelzer and Kingsley, 1975) have observed, a learning hierarchy can be formally represented as a directed graph (digraph). A digraph consists of a set of nodes and a set of ordered pairs (u,v), called arcs, where u and v are distinct members of the set of nodes.…”

“…One hypothesis accounting for the contradictory fail-pass response pattern (in which a prerequisite is failed but its parent is passed) which can thwart the validation of a learning hierarchy, is that a known prerequisite is sufficient but not necessary and that students producing this pattern had mastery of another sufficient prerequisite not represented in the model. This limitation can be overcome by a representation, used both in the task models of Heines and O'Shea (1985) and in Pask's entailment structures (Pask, KaUikourdis and Scott, 1975), which would indicate that any one of a specified set of subobjectives can serve as a prerequisite. In other words, hierarchies augmented by such a representation are AND/OR graphs, allowing for both conjunction and disjunction of prerequisites where standard Gagn6 hierarchies allow only conjunction.…”