Experimental Investigation of User Performance in Time-Shared Computing Systems: Retrospect, Prospect, and the Public Interest.

摘要

This study was conducted to survey the field of user studies in time-sharing, and to develop a conceptual framework for cooperative, long-range applied research in this area--ultimately to serve the public interest in the development of the computer utility. The introduction traces the historical roots of user problems and develops the need for experimental studies of user performance in time-sharing systems. The literature review reveals a large and growing experimental lag between the extension of information services and verified knowledge of user performance. A conceptual framework for user studies in time-sharing is constructed following three basic steps. The first defines this field of inquiry. The definition essentially portrays this area as experimentally derived techniques and findings comprising the shared and verified experiences of the user community. The second step builds an evolutionary systems framework for user studies, encompassing the design, development and operation of user systems, and relating time-shared user systems to other types of computer-aided systems. The last is a classification of user problems into four broad areas--methodological, normative, behavioral, and social effectiveness. Numerous problems, hypotheses and recommendations for experimental investigation of user performance are made for each of these four categories. The study concludes with a plea for interdisciplinary applied research to meet the imminent challenge of the public information utility.