Generalized Teaching Machine Decision Structure with Application to Speed Reading

Abstract
A relatively new type of automated instruction called the computer-directed teaching machine is discussed in this thesis. Typical present-day teaching machines either give every student the same instruction material or choose what material the student receives on the basis of his answer to the last question. The computer-directed machine chooses instruction material by making a statistical evaluation of the student's total behavior in comparison with other students' total behaviors. This machine's statistics are actually changed as new students take the course. Such a teaching machine can perform very much like a human tutor who adjusts his presentation to fit the individual student's capabilities and who improves his teaching technique with each student. In this paper a technique is suggested for comparing teaching machines. The machine's tutorial functions would be fitted to a very general model of the tutorial teaching cycle. This allows the various automated instruction devices to be discussed in terms of a common model. An application of the computer-directed machine was made to a speed reading course. Preliminary experiments with this course indicate that the computer-directed machine can perform like a human tutor.

This publication has 0 references indexed in Scilit: