A survey of the literature related to man-computer interaction reveals the many aspects of this problem, which appears to be in the crossroads among such diverse fields as computer languages, computer systems operational characteristics, control theory, decision theory, information theory, applied psychology, computer display and interface engineering, etc. In this paper we have chosen to present the on-line interaction from an information and decision point of view. After a brief discussion of classes of on-line situations and tasks, we propose an information-processing model of the case in which a human operator is engaged on-line in the solution of a problem like debugging a program, testing a model in a scientific application, or performing a library search. In this model the human operator is considered to seek to maximize overall cost. This cost is obtained by adding the operational cost of both man and computer to a remnant terminal cost originated by the remaining uncertainty. This analysis, performed for each of a set of possible alternatives for action, may lead the man to select and execute one of them, to terminate the process, or to reevaluate the possible alternatives and/or hypotheses in a search for new ones. Some practical applications in terms of response time and other characteristics of a computer utility are discussed, as well as some theoretical implications from an informational point of view. Copyright © 1969 by The Institute of Electrical and Electronics Engineers, Inc.
