A vigilance experiment was performed on the characteristics of visual monitoring behavior in complex tasks like those found in modern semi-automatic systems. The activationist hypothesis, which contends that human alertness is a function of stimulation level, served as framework for the experiment. Under investigation were sources of environmental and response-produced stimulation that might he related to hitman alertness. A. simulated semi-automatic air defense surveillance task was used. Environmental stimulation was manipulated by requiring six or thirty-six visual stimulus sources to be monitored for a 3-hr observation period. Response-produced stimulation was a function of response complexity. No vigilance decrement was found for per cent of signals correctly detected. Response latency declined significantly, but slightly, for groups that had simple response conditions but not for groups with complex response requirements. Results were discussed in terms of issues in operationally defining sources of stimulation for the activationist hypothesis, and the cautions that must be observed in generalising from the simple tasks of most vigilance experiments to the complex tasks of semi-automatic systems.
Observers detected many more of a fixed number of signals when these were among stimuli presented at 5 per minute than when these were among stimuli presented at 30 or 60 per minute. The effect, which is associated with either the signal probability or the nonsignal stimulus density, is analyzed with conventional measures and with measures from the theory of signal detectability (TSD). The TSD measures were used to define several possible modes of observing, and the model of vigilance based on decisions about observing could then be related to decision processes in detection performance as considered by TSD. If a single measure of the probability of alert observing is required, the best one is the percentage of detections of the readily detectable signal of the vigilance task. However, the TSD analysis suggested various different “mixes” of modes of observing for the subgroups in this experiment, and these mixes could be specified with the the help of heuristic models relating performance measures to the probability of observing.
This study relates the perceived complexity of 20 random forms to their physical factor structure. Ten principal axes, accounting for 94 per cent of the total variance of 24 physical measures, were rotated using the Varimax criterion. Factor scores for each form were correlated with the complexity ratings of the forms by each of 11 Ss. A sing Ie factor accounted for most of the variance in the complexity ratings. This factor was best described by four physical measures: the number of turns in the form, the length of the perimeter, the perimeter squared to area ratio, and the variance of the internal ang Ies of the form.
Six psychological dimensions were recovered from similarity judgments of 20 random forms using Kruskal The judgment of similarity among members of a set of objects requires an observer to make simultaneous use of all the perceptual dimensions that he deems relevant. Thus, data on judged similarities contain information on the dimensions of perception. A variety of techniques for recovering the nature and the number of these perceptual dimensions from similarity data has been developed, beginning with the work of Richardson (1938) and the theorems of Young and Householder (1938). Torgerson (1958) reviews the scaling theory underlying the techniques in use up until the time of his book. Since Torgerson's book, important breakthroughs have been accomplished in the analysis of similarities by Shepard (1962) and Kruskal (1964).The work of Kruskal d, to be described later, is of major concern in the present study. Kruskal's method is used here to recover the psychological dimensions underlying the judged similarities among members of a set of random shapes. These psychological dimensions are then related through the use of canonical correlation analysis to a set of physical dimensions that describe the same set of shapes. The result of this analysis is a physical description of the intersection of a set of psychological measures of these shapes with a set of physical measures of the same shapes.The following sections on the general nature of the similarity scaling model, the non-metric scaling technique, and canonical correlation are designed to introduce the reader informally to concepts necessary for the understanding of this study. So as not to break the continuity between this Introduction to the concepts and their use in the following sections, the discussion of literature pertinent to the substantive results of this study has been deferred to the Discussion section. There the results of relevant studies are stated and compared with the results of the present study. The ModelThe basic model on which this study is based is quite an old one. It is the same general model as that described by Torgerson (1958) for analyzing similarity data, and the same model forms the basis for the methods of Shepard (1962) and Kruskal (1964).In this model the percept of a stimulus object is represented as a point located in a multidimensional Cartesian coordinate system, whose reference axes are (unknown) dimensions of perception. The points representing two very similar percepts are close to each other in this psychological space, and very dissimilar percepts are represented by points that are far apart. The interpoint distances, which represent similarities, have usually been treated as Euclidean distances: That is, the distance between any two points is equal to the squared differences of their projections on each reference axis of the space summed over all axes. A notable exception to this practice is, Attneave's "city block" distance model for form perception (Attneave, 1950). In this model the "distance" between any two points...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.