The development of a tutoring system to assist students develop answering techniques

Abstract

End-user computing is a compulsory course that all first year technikon students who are not registered for a computer-related diploma. However, most students ( especially those from disadvantaged educational backgrounds) experience much difficulty in grasping theoretical computing concepts. For this reason we are currently developing CAI tools to assist students overcome learning difficulties. One of the problems experienced by students is interpreting and answering questions in tests and examinations. Consequently, we are currently developing an CAI application, namely INTELLIQUEST, to provide students with practice with respect to answering questions. End-user computing tests are usually comprised of three types of questions: • multiple choice • explanation of terms • short questions This paper looks at the technical aspects of developing such a system. In order to implement a multiple choice tutor traditional procedural programming techniques are sufficient. However, in order to mark responses to the second two types of questions these techniques are inadequate. An examination of different artificial intelligence constructs and methods for this purpose is presented in the paper. In order to reduce the variety of responses to a particular question lecturers are required to set questions and solutions to questions using templates in a structured editor. Students are also required to use the editor to respond to questions. Finally, the paper discusses a pilot evaluation of the system to determine how accurate.the system is in marking responses and advising students.