Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
6 Vigilance and Target Detection As noted in Chapter 1, a large portion of the responsibility of armored vehicle teams, and indeed the teams of many analogous systems, prior to transition is simply to monitor the environment for events that might signal the need for the team to mobilize into action. This task of monitoring for infrequent signals is one for which humans are not well suited, particularly after periods of sleep disruption. In this chapter we examine in detail the research on human performance in monitoring, vigilance, and target detec- tion. HISTORICAL BACKGROUND World War II Vigilance or sustained attention refers to the ability of observers to maintain their focus of attention and to remain alert to stimuli for prolonged periods of time (Davies and Parasuraman, 1982; Warm, 1984a). Systematic study of sustained attention began during World War II. It was stimulated by surprising fallibility in the performance of British airborne radar observ- ers while on patrol for enemy submarines. These individuals were often required to maintain continuous observation of their radar scopes during long flights over the Bay of Biscay watching for telltale "blips" that sig- naled the presence of German U-boats in the sea below. Despite their extensive training and obvious motivation to perform well, the observers failed with increasing frequency to notice the critical signals displayed by 139
140 WORKLOAD TRANSITION their equipment as time on watch progressed. As a result, the submarines passed undetected and were free to prey on Allied shipping. In response to a request from the Royal Air Force to study the problem, Mackworth (1948, 1961) initiated a series of ingenious experiments that formed the first systematic effort to bring the study of vigilance into the laboratory. Using a simulated radar display known as the "Clock Test," Mackworth was able to chart the course of performance over time and to confirm the suspicions generated in the field that the quality of sustained attention is fragile: it wanes over time. He found that his subjects became progressively more inefficient at detecting signals as the watch continued, and that this inefficiency did not take long to develop. In general, the accuracy of signal detections declined about 10 percent after only 30 min- utes of watch, and then showed a more gradual decline during the remainder of the 2-hour session. The progressive decline in performance over time noted in Mackworth's pioneering experiments has been confirmed in a large number of subsequent investigations. It has been labeled the "decrement function" or the "vigi- lance decrement" and is the most ubiquitous finding in vigilance studies. Many investigations using a broad assortment of vigilance tasks indicate that the decline in performance is complete from 20 to 35 minutes after the initiation of the vigil and at least half of the final loss is completed within the first 15 minutes (Teichner, 1974~. Under especially demanding circum- stances, the decrement can even appear within the very first few minutes of watch (Jerison, 1963; Nuechterlein et al., 1983~. As Dember and Warm (1979) have noted, the most striking aspect of this finding is that it seems to result merely from the necessity of looking or listening for a relatively infrequent signal over a continuous period of time. Implications of Automation for Vigilance The study of vigilance has generated considerable interest among hu- man factors and ergonomics specialists, who are concerned not only with the decrement function, but also with the factors that affect the overall level of performance. Comprehensive reviews of the extensive experimental lit- erature on vigilant behavior can be found in Craig (1985a, 1985b, 1991), Davies (1985), Davies and Parasuraman (1982), Dember and Warm (1979), Parasuraman (1986), Warm (1977, 1984b), Warm and Berch (1985) and Warm and Parasuraman ~ 1987~. The importance of vigilance research for human factors concerns is based in part on the fact that the surveillance problems encountered during World War II are still with us in one form or another. The generation of automatic control and computing systems for the acquisition, storage, and processing of information has altered the role of the human operator from
VIGILANCE AND TARGET DETECTION 141 that of active controller to a more executive function, a development that Sheridan (1970) has characterized as a shift from active to supervisory control. Vigilance is a crucial aspect of the reliability of human performance in a wide variety of activities including industrial quality control, robotic manufacturing, air traffic control, nuclear power plant operations, long dis- tance driving, and the monitoring of life signs in medical settings (Warm, 1984a). As Parasuraman (1986) has noted, target detection in today's highly automated systems may be executed by instruments and controls, but the same problems of vigilance occur when the systems malfunction or unusual events appear. In some cases, the vigilance functions demanded of the operator can be overwhelming (Parasuraman, 1987; Wiener, 1984, 1985, 1987~. Implications for Armor Personnel Modern tank warfare is still another situation in which vigilant behav- ior is critical. Indeed, the maintenance of a proper level of alertness is the price of survival in combat. As described in Armor Field Manual 17-12-1, Tank Combat Table (U.S. Department of the Army, undated), future battle- fields are expected to require tank crews to move and engage rapidly under conditions in which our forces are intermingled with those of the enemy. Survival will depend on the crew's ability to detect and locate opposing forces rapidly in order to maintain the advantage of a first strike. Target acquisition, however, can be a difficult task for all crew members, espe- cially the gunner. Toward that end, they must search both ground and sky, looking and listening for target signatures (critical signals) involving acous- tic and visual patterns (engine noise, aircraft noise, broken vegetation, weapon smoke, glare from airplane canopies or wings), using just their eyes and ears or with the assistance of field glasses, telescopes, and thermal imagery. All of this must be accomplished under extreme conditions of heat, noise, vibration, and danger. In order to provide some insight into factors that might affect the vigilant behavior of armor crews and ways to counter threats to their ability to remain alert, this chapter describes a variety of task and environmental factors that have been discovered to play an important role in the quality of vigilance performance. PSYCHOPHYSICAL DETERMINANTS A Functional Equation The Components of Vigilance Performance efficiency in vigilance tasks is closely tied to the nature of the stimuli that demand attention. Consequently, the study of vigilance,
42 WORKLOAD TRANSITION like that of other perceptual phenomena, has profited from the precise deter- mination of the stimulus conditions that influence performance. In summa- rizing these conditions, it is helpful to follow the framework of an empiri- cally determined functional equation developed by Jerison (1959b) and modified by Warm and Berch (1985), which takes the following form: P=f(M,S, U,B,C) According to this relation, performance (P) is a function of the sensory modality of signals (M), the salience of signals (S), stimulus uncertainty (U), the characteristics of the background of nonsignal events in which critical signals for detection are embedded (B), and task complexity (C). Signal Modality Vigilance experiments have used auditory, visual, and cutaneous stimu- lation. However, the sensory modality of signals is not a matter of indiffer- ence when the quality of sustained attention is concerned. Auditory tasks tend to be associated with a higher level of overall efficiency and with greater stability over time than their cutaneous and visual analogs (Davies and Parasuraman, 1982; Warm and Jerison, 1984~. In addition, intersensory correlations are often low or nonsignificant (Hatfield and Loeb, 1968~. Fortunately, techniques are available that can enhance visual perfor- mance and improve intersensory correlations in vigilance. The level of visual vigilance can be elevated to that characteristic of audition by closely coupling subjects to the display so that they cannot avoid stimulation by looking away (Hatfield and Loeb, 1968), and audiovisual correlations can be increased by equating the types of discriminations required in the two modalities (Galinsky et al., 1990; Hatfield and Loeb, 1968; Parasuraman and Davies, 1977~. Performance with redundant displays in which analo- gous signals are presented simultaneously to the auditory and visual chan- nels exceeds that with single-mode displays. The dual-mode superiority effect has been shown to stem from the integrative action of the sensory systems and not from a fortuitous combination of their independent activi- ties (Craig et al., 1976; Doll and Hanna, 1989~. It represents a potentially important technique for the general enhancement of monitoring efficiency. Signal Salience A common finding in studies of target detection under alerted condi- tions is that the frequency of detection is positively related to stimulus amplitude and duration. These factors are also important under conditions of sustained attention. Investigations by Adams (1956), Guralnick (1972), and Loeb and Binford (1963) have demonstrated that the overall quality of sustained attention can be enhanced and performance rendered more stable
VIGILANCE AND TARGET DETECTION 143 over time by increasing the amplitude of the signal-to-noise ratio of critical signals. In addition, Corcoran and his coworkers (Corcoran et al., 1977) have shown that it is also possible to reverse the typical time course of performance by "turning up the gain" on the sensory channel being moni- tored. Using acoustic pulses, they found that an abrupt increment in the amplitude of the stimuli midway through the vigil bolstered the frequency of signal detections for the remainder of the session. In addition to ampli- tude increments, signals can also be rendered more salient by increasing their duration. Signals of brief duration are more likely to be missed than those that remain visible for longer periods of time. Indeed, the rate of gain in detection efficiency is a negatively accelerated increasing function of duration up to a limit of about four seconds (Warm et al., 1970~. Stimulus Uncertainty Alluisi (1966) has pointed out that, like lookouts in old-time sailing ships, subjects in vigilance experiments can be faced with considerable uncertainty regarding the signals they are to detect they may not know when such signals will appear (temporal uncertainty) or where they will appear (spatial uncertainty). The same can be said for armor personnel in combat. Both types of uncertainty degrade performance efficiency. One means of manipulating the subject's temporal uncertainty is through variations in the density or the number of critical signals presented during a vigilance session. The more frequently such signals occur within a fixed time period, the greater the a priori signal probability and the lower will be the subject's average uncertainty as to when critical signals will occur. The accuracy of signal detections varies directly as a function of signal density (Warm and Jerison, 1984~. Increases in signal density also increase the speed of detection in a regular manner. Using an information theory analy- sis to measure the density-determined temporal uncertainty in the appear- ance of critical signals, Alluisi and his coworkers (Smith et al., 1966; Warm and Alluisi, 1971) found that response times to detections increased as a linear function of uncertainty. An important aspect of the effects of signal probability in vigilance performance is that these effects perseverate. Several studies have demon- strated that subjects trained initially under conditions of high signal prob- ability do better during subsequent testing than do those initially exposed to a low probability, regardless of the probability condition in effect (high or low) during the testing phase of the experiment (Colquhoun and Baddeley, 1964, 1967; Griffin et al., 1986; Krulewitz and Warm, 1977; McFarland and Halcomb, 1970~. Temporal uncertainty in vigilance experiments has also been manipu- lated through variations in the intervals of time between signals or the
44 WORKLOAD TRANSITION intersignal intervals. These intervals can be made quite regular and easily predictable or irregular and unpredictable. The speed and accuracy with which signal detections occur is greater in the context of regular compared with irregular signal conditions (Adams and Boulter, 1964; Warm et al., 1974~. Spatial uncertainty has been introduced into vigilance experiments by varying the probability that signals will appear in different locations of a monitored display or by using an unpredictable sequence of display loca- tions. Under such conditions, performance efficiency is lowered and sub- jects tend to bias their attention toward those portions of the display in which the likelihood of signal appearance is highest (Adams and Boulter, 1964; Joshi et al., 1985; Milosevic, 1974; Nicely and Miller, 19574. The Background Event Context Vigilance experiments frequently employ dynamic displays in which critical signals for detection are embedded within a matrix of recurrent background events. For example, subjects may be asked to detect occa- sional brighter flashes of light in a background of dimmer flashes or the presence of a stronger pulse of acoustic stimulation in a background of less intense pulses. Although the background events may be neutral in the sense that they require no overt response, they are far from neutral in their influ- ence on the quality of sustained attention. The frequency of background events, or the background event rate, is a very important determinant of performance efficiency. Both the speed and the accuracy of signal detec- tions vary inversely with event rate, and the vigilance decrement tends to be more pronounced in the context of a fast compared with a slow event rate (Jerison and Pickett, 1964; Lanzetta et al., 1987; Parasuraman, 1979; Parasuraman and Davies, 1976~. From the earlier discussion of signal density, one might be tempted to conclude that the effects of event rate are artifactual. If critical signal density is held constant, as is usually the case when event rate is varied, increases in event rate reduce the conditional probability that a stimulus event is a critical signal. Thus, event rate may simply be another example of signal uncertainty. This is not the case, however. The quality of sus- tained attention decreases with increased event rate even when signal den- sity is adjusted so that the conditional probability is equated within event rates (Loeb and Binford, 1968; Parasuraman, 1979; Taub and Osborne, 1968~. In addition to its own influence on performance, the background event rate also modifies that of other stimulus parameters. The effects associated with signal amplitude and signal regularity as well as the perseverative effects of signal density are dependent on the event rate context in which critical signals must be detected (Krulewitz and Warm, 1977; Metzger et al.,
VIGILANCE AND TARGET DETECTION 145 1974; Moore and Gross, 19739. Moreover, the speed of probe detections for a secondary task also depends on the event rate in force in a primary vigi- lance task (Bowers, 1983; Parasuraman, 1985~. Taken together, findings such as these have led several investigators to conclude that background event rate is probably the prepotent psychophysical factor in vigilance per- formance (Parasuraman et al., 19871. For the most part, neutral events in vigilance experiments have been presented in temporally regular intervals, such as 1 event every 12 seconds at an event rate of 5 per minute or 1 event every 2 seconds at a faster rate of 30 per minute. Under such conditions, subjects can predict when an event that needs to be inspected for the possibility that it is a critical signal will appear. Accordingly, they can take task-contingent timeouts and rest be- tween the appearance of events. Recent studies by Scerbo and his col- leagues have denied subjects such rest intervals by using temporally irregu- lar or asynchronous event schedules so that subjects could not be certain when an event that could be a signal would appear and had to observe the vigilance display continuously. As might be anticipated, event asynchrony degrades performance efficiency compared with a synchronous schedule of background events (Scerbo et al., 1987a, 1987b). The effects of both event rate and event asynchrony are striking; they demonstrate that signal detec- tion in vigilance experiments is determined as much by what transpires in the interim between signals as by the characteristics of the signals them- selves. Stimulus Complexity Vigilance studies typically make use of relatively simple perceptual discriminations involving the detection of discrete changes in the intensity, extensity, duration, or movement of stimuli on a single display (see Hancock, 1984, for a description of the kinds of displays used). In operational set- tings, however, more complex discriminations are often involved and ob- servers may be required to cope with multiple signal sources. A number of attempts have been made to explore the consequences of increased task complexity for the quality of vigilant behavior, but the results are not clear cut. In a well known study, Jerison (1963) asked observers to monitor three displays simultaneously and found that the vigilance decrement was greatly enhanced under such conditions. Rather than appearing after the first 5 or 10 minutes of watch, he found that the decrement could be observed from the very first signal onward. In contrast, however, are several experiments by Adams and his associates (Adams and Humes, 1963; Adams et al., 1963; Montague et al., 1965) that employed very complex vigilance tasks with multiple stimulus sources (from 6 to 36) under conditions in which any one
146 WORKLOAD TRANSITION source could present a signal at any moment in time. The vigilance decre- ment was either absent or minimal in these experiments even though the vigil lasted for several hours. However, the absolute level of vigilant per- formance (the number of signals missed) was far from perfect. Adams's findings have been confirmed recently by Warm and his collaborators who varied complexity by increasing the cognitive demand placed on observers in a vigilance task (Dember et al., 1984; Lysaght et al., 1984; Warm et al., 1984~. But Loeb and his coworkers have reported that the beneficial effects of increased cognitive demand are limited- increasing demand beyond a rather low optimal level degrades performance and restores the decrement function (Loeb et al., 1987~. The problem of stimulus complexity is perplexing. Depending on the approach employed, it is possible to amplify or eliminate the vigilance decrement through modifications of complexity. Clearly, a resolution of these disparities will be necessary in order to develop a complete functional equation for the psychophysics of vigilance. Some insight into ways in which this conundrum might be unraveled comes from a study by Fisk and Schneider (1981), who approached the issue of complexity in terms of auto- matic and controlled processing theory (see Schneider and Shiffrin, 1977~. According to this view, automatic processes are fast, fairly effortless, skilled behaviors while controlled processes refer to relatively slow, effortful, ca- pacity-limited processes. In a carefully conducted experiment in which automatic processing was developed over several hundred trials, Fisk and Schneider found that the vigilance decrement was restricted to controlled processing tasks; automatic tasks were performed in a stable manner throughout the vigilance session. Evidently, proper training might be a key in aiding subjects to cope with demanding vigilance tasks. Sensing and Decision Making The Theory of Signal Detection At first glance, a subject's success in correctly detecting a signal (a predesignated stimulus event) in vigilance or other detection situations may seem to be a rather direct measure of the individual's perceptual ability in that situation. However, if one examines the process of signal detection more carefully, it becomes evident that perceptual reports are not that simple (see Natsoulas, 1967~; the affirmation of signal presence or absence does not depend solely on perceptual sensitivity. Such a report is also contingent on nonperceptual factors that include the subject's detection goals, expecta- tions about the nature of the stimuli, and the anticipated consequences of correct and incorrect responses. Taken together, these factors compose the subject's response criterion or willingness to emit a detection response.
VIGILANCE AND TARGET DETECTION 147 Failure to detect a signal can result from a lack of perceptual sensitivity for the signal or from a conservative criterion that leads the subject to withhold the detection response. The theory of signal detection is concerned with these aspects of per- formance. It is a major theory for the general study of perception and has been used widely in investigations of sustained attention. A detailed tuto- rial on the nature of this approach is beyond the scope of this chapter. Appropriate descriptions can be found in Dember and Warm (1979), Green and Swets (1974), and McNicol (1972~. For special sections on the applica- tion of signal detection theory to vigilance, see Davies and Parasuraman ( 1982) and Warm and Jerison ( 1984~. Applications to Vigilance In essence, signal detection theory makes use of the frequency of cor- rect detections (hits) and the frequency of false detections (false alarms), which also decline over time in vigilance studies, to derive two independent measures of performance: an index of perceptual sensitivity (d'), and an index of criterion in responding Gil. The former describes the keenness of the observer's senses: the ability to discriminate signals from nonsignals. The latter describes the observer's tendency to say "yes, I see a signal," thus possibly detecting more signals but also generating more false alarms. These measures have provided an important insight into the nature of the vigilance decrement. A substantial number of investigations have indi- cated that the decrement often does not involve a decline in alertness during a vigil (drop in d'). Instead, it reflects a shift to a more conservative response criterion (rise in p), perhaps because subjects develop more ratio- nal expectations of the actual signal probability in force with experience in the experiment (Davies and Parasuraman, 1982; Swets, 1977; Warm and Jerison, 1984; Williges, 1969~. True perceptual decrements seem to be restricted to extremely demanding situations with low levels of signal discriminability, which may be brought about by low signal-to-noise ratios, fast event rates, and memory demanding tasks (Parasuraman et al., 19873. Task Taxonomy A major feature of vigilance tasks is their diversity. As noted earlier, such tasks can be presented in different sensory modalities and utilize a variety of psychophysical dimensions to define critical signals for detec- tion. Davies and Parasuraman (1982) have argued, however, that there is an essential order to this diversity that can be made clear by appealing to a general view of attention known as resource theory. According to this position, the quality of performance in many situations is directly related to
148 WORKLOAD TRANSITION the amount of mental resources or capacity invested in the task at hand (Gopher and Kimchi, 1989; Kantowitz, 1985; see also Chapters 3 and 49. With that notion in mind, Davies and Parasuraman have suggested a classi- fication system (taxonomy) in which vigilance tasks can be characterized as successive or simultaneous. The former are absolute judgment tasks in which subjects need to compare current input with a standard retained in working memory in order to separate signals and noise. Simultaneous tasks are comparative judgment tasks in which all of the information needed to distinguish signals from nonsignals is present in the stimuli themselves, and there is little involvement of recent memory for the signal feature. Davies and Parasuraman maintain that because of the memory imperative, succes- sive tasks are more resource demanding than their simultaneous cohorts A number of studies have supported this point of view. In so doing, they have adopted the strategy of comparing the effects of factors known to degrade the quality of sustained attention by increasing the subject's information processing load on the performance of successive and simultaneous tasks. If the former are more resource demanding than the latter, then any factor that places an additional drain on the subject's information processing re- sources should have a more pronounced effect on vigilance efficiency when presented within the context of a successive compared with a simultaneous task. Along these lines, the degrading effects of increments in event rate, event asynchrony, and the spatial uncertainty of signals have all been found to be more notable with successive than with simultaneous tasks (Josh) et al., 1985, Lanzetta et al., 1987; Scerbo et al., 1987a). ENVIRONMENTAL STRESS Environment and Task In addition to psychophysical factors, vigilance performance is influ- enced considerably by conditions in the ambient environment. These condi- tions are often teed environmental stressors. The concept of stress, which is discussed more fully in Chapter 4, is difficult to define, but for the purpose of this chapter, stress is considered to be any threat to the physical or psychological well-being of the organism (Wingate, 1972~. The discus- sion centers on the effects on vigilance of four basic stressors that are particularly likely to be a part of the combat environment of tanks: tem- perature, noise, vibration, and sleep loss. Temperature As in the case of the psychophysics of vigilance, Mackworth's seminal studies (1948, 1961) also provided the earliest data regarding the effects of
VIGILANCE AND TARGET DETECTION 149 heat on performance efficiency. Using a range of effective temperatures from 21 to 36 degrees centigrade, he found an inverted U-shaped function between performance efficiency and the temperature of the testing chamber. Performance was maximal at about 26 degrees centigrade and was poorer at temperatures on either side of that point. Later studies have demonstrated that while performance may improve when subjects are first exposed to moderate levels of heat (Kerslake and Poulton, 1965), the quality of vigilant behavior is impaired with continued exposure to temperatures above 32 degrees centigrade. In these conditions the probability of signal detection declines, as does d', the sensitivity (Benor and Shvartz, 1971; Mackie and O'Hanlon, 1977; Pepler, 1953; Poulton and Edwards, 1974; Poulton et al., 1974). The effects of heat stress on vigilance performance have also been studied more directly in terms of body temperature itself. Bell and his colleagues have found that performance suffers with rising body tempera- ture (Bell et al., 1964), while other studies have reported that performance efficiency improves with elevations in body temperature (Colquhoun and Goldman, 1972; Wilkinson et al., 1964~. In an effort to account for this disparity, Hancock (1984) has noted that the subjects' body temperatures were rising during the vigilance session in the Bell et al. (1964) experiment, while in the other studies, subjects were established in a static hyperthermic state during the session. Hancock (1984, 1986) has suggested that the key to understanding the effects of thermal stress lies in the action of the stres- sor on deep body temperature. Vigilance performance breaks down with perturbations in core temperature, remains unaffected when there is no variation in that temperature, and can be facilitated when subjects are established in a static hyperthermic state. Only a few studies have examined the effects of cold on vigilance performance. Nevertheless, they suggest that, like exces- sive heat, cold also impairs the quality of sustained attention when deep body temperature is perturbed (Hancock, 1984, 1986~. Noise As described by Loeb (1986), noise, or unwanted sound, is the most ubiquitous pollutant in our industrialized society and a major environmental stressor. Accordingly, its effects have been studied extensively in a variety of situations, including those requiring sustained attention (Jones' 1983, 1984~. The effects of noise on vigilance performance are intricate and depend on many factors. Two of the more important are the type of noise involved, intermittent or continuous, and task complexity. Studies with intermittent noise have led to a plethora of conflicting results. In different circumstances, such noise can either enhance or de- grade vigilance performance or have no effect at all (Hancock, 1984; Koelega
150 WORKLOAD TRANSITION and Brinkman, 1986; Poulton, 1977~. Conditions in which noise is loud and continuous, such as might be expected in armored vehicles during combat, have led to more consistent results. Subjects appear to be relatively im- mune to the adverse effects of continuous noise when required to monitor only a single stimulus source (Blackwell and Belt, 1971, Davies and Hockey, 1966; Jerison, 1959a; Poulton and Edwards, 1974~. They are far less im- mune, however, when attention must be directed to several stimulus sources. For example, Broadbent (1954) and Jerison (1959a) have reported that loud continuous noise degrades the speed and accuracy of signal detections in multidisplay monitoring tasks. The general effects of continuous noise on the quality of vigilant behavior can be summarized as follows: performance is degraded by loud noise (above 9OdB SPL) when the information process- ing or resource demands of the task are high while it remains unaffected by loud noise when the processing demands are low. In addition, performance under low demand conditions can be facilitated by low-level (approximately 64dB SPL) continuous noise that is variegated in character (Hancock and Warm, 1989; Loeb, 1986~. In careful reviews of the literature on noise and vigilance, Jones (1983, 1984) has noted that continuous noise leads to two interesting changes in the manner in which monitors process information. One of these is a ten- dency for noise to increase their confidence in the adequacy of a detection decision. For example, given a choice of using one of three response cat- egories when interrogating stimulus events for the presence of signals, "Sure Yes," "Unsure," and "Sure NO," subjects exposed to noise increased the use of the two extreme categories at the expense of the intermediate one (Broadbent and Gregory, 19651. A similar result has also been reported by Hartley and Shirley (19774. As Jones (1984) has noted, it would appear that the range over which sensory information is considered becomes narrowed in noise, and in terms of signal detection theory, the distance in decision-space sepa- rating risky and cautious response criteria is lessened in noise. lIence, in settings in which risky responding is encouraged, such as combat situations wherein detecting the enemy first is vital for success, noise could reduce the likelihood of early target acquisition. The second noise-related information processing change is in the man- ner in which monitors interrogate the vigilance display itself. Hockey (1970) has reported that noise induces subjects to follow the tendencies toward attentional narrowing (Easterbrook, 1959), discussed in Chapter 4, and shift their inspection away from peripheral, low-priority events, toward more central high-priority events. Jones (1983) has argued that this effect is essentially strategic in character; resources are invested in some sources of information rather than others. Such a noise-induced tradeoff has poten- tially important combat implications. It could serve to enhance target ac
VIGILANCE AND TARGET DETECTION 151 quisition where it is likely that the enemy may be located on the battlefield, but at the same time, render a tank crew more vulnerable to threats from less likely locations. Vibration Vibration is a common experience among travelers regardless of whether their vehicles move about in the air, on the land, or in the sea. It is also encountered frequently in the workplace with machinery of all sorts. Ac- cordingly, vibration stress has been a serious human factors concern (Goether, 19711. Surprisingly, however, few studies have focused on the relation between vibration and vigilance performance. The data that are available indicate that, except for situations in which the amplitude and frequency (above 10-15 Hz) of vibration is strong enough to blur vision (see Hancock, 1984; Poulton, 1977), vibration seems to have little impact on the quality of sustained attention. Studies by Schohan et al. (1965) using visual discrimi- nations and Weisz et al. (1965) using both visual and auditory discrimina- tions failed to find any significant effects for vibration on performance efficiency. Indeed, two experiments using vibration at SHz, the resonant frequency of the human body to vertical vibration (Poulton, 1977), found that such stimulation enhanced the frequency of signal detections (Shoenberger, 1967; Wilkinson and Gray, 19741. Poulton (1977) has suggested that this effect may be based on an increase in arousal level, but his explanation is not shared by all investigators (Hancock, 1984~. All things considered, it would appear that, unless quite severe, vibration is not likely to be a serious environmental threat to the alertness of rotorcraft or armor crews. Sleep Loss As described in Chapter 5, a substantial body of data is available to indicate that sleep loss reduces efficiency on a variety of tasks (Anch et al., 1988; Broadbent, 1963; Colquhoun, 1972; Johnson and Naitoh, 1974~. As described here, sleep deprivation, in particular, has been found to have a strong impact on the ability of monitors to remain alert. For example, Bergstrom et al. (1973), using a simulated radar task, found that the per- centage of target detections dropped consistently from 100 percent to only 70 percent as the amount of time subjects went without sleep increased from 6 to 66 hours. Several other studies have shown that, after one night of total sleep deprivation, the overall speed and accuracy of signal detec- tions declined as did perceptual sensitivity, and that the vigilance decrement can be exacerbated. In air crews, vigilance tasks are found to be the first affected by fatigue. Moreover, the amount of impairment resulting from
52 WORKLOAD TRANSITION sleep loss is linked to circadian rhythms. It is greater at night than during the day and reaches its maximum during the early morning hours (Davies and Parasuraman, 1982; Horne et al., 1983~. Williams et al. (1965) have pointed out that, rather than resulting in a system shutdown, the loss of adequate sleep appears to produce brief peri- ods of inefficiency or lapses in what is otherwise normal functioning. These lapses are considered to stem from intervals of microsleep, which become common with fatigue (Hockey, 1986~. It is encouraging to know that the effects of total sleep deprivation are not necessarily long-lasting, a point that might be of importance when considering the recovery of armor crews from sleep stress (Anch et al., 19881. In this regard, Rosa et al. (1983) have reported that subjects who suffered performance impairments after 40 or 64 hours of sleep deprivation returned to a baseline level of efficiency after 4 or 8 hours of recovery sleep, respectively. Recently, studies have shown the benefits of brief naps in air crew performance (Rosekind et al., 19911. Rather than prolonged sleep deprivation, battle scenarios are more likely to include partial deprivation or limitations of normal sleep length. Studies by Hartley (1974), Webb and Agnew (1974), and Wilkinson and his col- leagues (Wilkinson, 1968; Wilkinson et al., 1966) have shown that reduced sleep regimens impaired vigilance performance in comparison to control sleep schedules. Along these lines, it is worth noting that acute shifts in the sleep-wakefulness cycle can also affect vigilance performance. Taub and Berger (1973) have reported that signal detectability is impaired in subjects permitted the usual eight hours of rest when their sleep schedules were shifted forward or backward by several hours from their normal bedtimes. A similar result has been reported by Seidel et al. (1984~. In addition, Taub and Berger (1969) have described what they have called the "Rip van Winkle effect," in which auditory signal detections were significantly lower after 11 hours of sleep than after the typical 8 hours. As Davies and Parasuraman (1982) point out, it would seem that any alteration in the normal sleep schedule can impair the quality of sustained attention. Task-Induced Stress in Vigilance Physiological Indices Historically, vigilance tasks have been viewed as tedious but benign situations that place little demand on those engaged in them (Dember and Warm, 1979; Parasuraman, 1984~. Evidence is accumulating, however, to indicate that vigilance tasks are not benign. Instead, such tasks can be quite demanding and induce considerable stress in subjects. Consequently, armor crew members who must perform tasks requiring vigilance functions during combat may be stressed not only by environmental factors, but also by the
VIGILANCE AND TARGET DETECTION 153 tasks they need to perform. This aspect of sustained attention tasks led Hancock and Warm (1989) to suggest that, in order to understand the role of stress in human performance, it is necessary to revise traditional views of stress as an independent environmentally and/or socially determined agent that acts on performance and recognize that tasks themselves can be a sig- nificant source of stress. The stress response of a monitor during a vigilance session can be assessed by measuring the levels of circulating catecholamines (epinephrine and norepinephrine) and corticosteroids released into the bloodstream (Parasuraman, 1984; Wesnes and Warburton, 1983~. This method was used by Frankenhaeuser et al. (1971) in comparing subjects' responses with a supposedly overstimulating sensorimotor task, in which they responded to multisensory stimuli with appropriate button-pressing, lever-pulling, and pedal- pushing activities and an understimulating vigilance task, involving the de- tection of increments in the intensity of visual targets, during a three-hour experimental session. As expected, the complex sensorimotor task pro- duced elevated catecholamine levels indicative of stress. However, both epinephrine and norepinephrine levels were also elevated during the vigi- lance task, indicating that the latter also induced effort and stress. Similar results for vigilance tasks have been reported by Frankenhaeuser and Patkai (1964) and by Lundberg and Frankenhaeuser (1979~. More recently, Hovanitz et al. (1989) measured the vigilance-induced stress response by means of a frontal electromyogram and found a significant increase in muscle tension over the course of a one-hour session. Moreover, they reported that the task brought on tension headaches in sensitive subjects. Mood Measures Still another way to examine the stress induced by sustained attention tasks is through the use of self-reports of mood after participating in a vigil. The initial investigation along this line was conducted by Thackray et al. (1977), who asked monitors to rate their levels of boredom, monotony, irritation, attentiveness, fatigue, and strain at the beginning and at the end of a one-hour vigilance session. The task consisted of a simulated air traffic control radar system that displayed targets flying along specified routes at different speeds. Critical signals for detection consisted of departures from altitude as reflected in alphanumeric readouts on the radar screen. The subjects rated themselves as feeling significantly less attentive and more bored, strained, irritated, and fatigued after the vigil than before its start. Subsequent studies using the Thackray scales have reported similar mood changes following participation in a watchkeeping session (Hovanitz et al., 1989; Lundberg et al., 1980; Thiemann et al., 1989~. In addition, Warm and his coworkers (Dittmar, 1989; Macomber, 1967; Thiemann et al., 1989;
154 WORKLOAD TRANSITION Warm et al., 1991) have reported that subjects feel more sleepy and fatigued after a vigil than at its start, as measured by the Stanford Sleepiness Scale (Hoddes et al., 1973) and a scale of symptoms of fatigue developed by Yoshitake (19781. Indeed, in one study, self-reports of fatigue increased by 120 percent over the course of a 40-minute vigil (Thiemann et al., 1989~. In addition to these investigations, negative mood changes during the course of vigilance performance have also been reported in two additional experiments that had more direct operational bearing. In one of these, Warm et al. (1989) examined the effects of extratask demands and long hours of work on the performance of simultaneous and successive vigilance tasks in a simulated work environment. For 3 consecutive 12-hour days, subjects engaged in 4 1-hour vigilance tasks interspersed with work at a heavy-load or light-load data entry task. They reported becoming more drowsy, strained, and fatigued and they experienced more somatic com- plaints over the workday and the workweek. These mood changes were maximal with the successive task and the heavy auxiliary workload, sug- gesting that in order to maintain performance standards in the successive task, subjects expended more processing resources, which led to a greater cost in fatigue and strain. An interesting report by Johnson and McMenemy (1989) on mood changes among U.S. soldiers during sentry duty is also consistent with the idea that vigilance tasks are inherently stressful and provides striking evidence for the ecological validity of the findings of the other investigations just de- scribed. During three hours of simulated sentry duty, soldiers assumed a standing foxhole position and monitored a realistic target scene for pop-up targets silhouettes of enemy troops. On detecting targets, the soldiers pressed a telegraph key and fired a Weaponeer rifle simulator. Target detec- tion speed deteriorated with time on duty, a result that could have rather disastrous consequences in combat if the enemy's marksmanship is good, and the sentinels' predominant mood shifted during the assignment from one of vigor at the outset to fatigue at the end. Johnson and McMenemy suggest that, in order to optimize sentry duty performance, it should be limited to one hour or less. Extrapolating to armor crews, sentry duty is often required in the so- called quiet time prior to combat when the crews are preparing defensive positions or waiting to begin offensive action themselves. Clearly, what was once thought to be a relatively nonstressful activity may place a much greater degree of strain on sentinels in the field than previously believed. Perceived Workload Additional evidence for the stress of sustained attention tasks comes from measurements of their perceived mental workload. This concept re
VIGILANCE AND TARGET DETECTION 155 fers to the information processing load or resource demands that are associ- ated with a task (O'Donnell and Eggemeier, 1986~. Recently, several inves- tigations of sustained attention used a subjective scale of mental workload known as the NASA Task Load Index, or the TLX. The instrument is a multidimensional scale that provides a global measure of workload and identifies specific components of workload along three demand dimensions (mental, physical, and temporal) imposed on the observer by the task and three dimensions (effort, frustration, and performance) related to the inter- action of the observer and the task (Hart and Staveland, 19881. These studies showed that the cost of mental operations in vigilance is indeed high. For example, Gluckman et al. (1988b) reported uniformly high workload responses to both simultaneous and successive sustained attention tasks, and that overall workload levels varied inversely with the psychophysical salience of the signals; the more difficult it was to see the signals, the higher the workload. This result was confirmed by Galinsky et al. (1989), who found in addition that the powerful effect of event rate in vigilance was reflected directly in the perception of mental workload. The fact that workload scores in vigilance can be brought under psychophysical control enhances the validity of these subjective ratings (see Natsoulas, 1967~. Vigilance studies by Deaton and Parasuraman (1988), Decker et al. (1991), and Dittmar (1989) also found overall workload scores at the upper end of the TLX scale. These values are generally higher than those ob- tained for tasks such as memory search, choice reaction time, mental arith- metic, and grammatical reasoning (Dittmar, 1989~. In all of these experi- ments, the major factors that contributed to overall workload were mental demand and frustration, a result that provides further indication of the stress of sustained attention. Moreover, it appears that stress effects may be so pervasive that they are more difficult to alleviate than performance effects, at least by noninvasive means. More specifically, Warm et al. (1991) found that providing subjects with brief whiffs of air containing the scent of mugger or peppermint can enhance the probability of signal detections in vigilance compared with plain air control conditions. Although overall performance was affected by the olfactory stimulation, neither fragrance made the task seem less demanding in terms of TLX workload ratings or reduced the feelings of stress obtained from self-report mood measures. OPERATIONAL RELEVANCE Workload Transition Current Army doctrine calls for crews to remain with their tanks at the periphery of a battle zone for up to 72 hours prior to combat. During this time, there is relatively little to do. With the order to engage the enemy,
156 WORKLOAD TRANSITION however, the crews will suddenly be thrust into a period of high workload and stress. What are the implications for the efficiency of performance at this juncture of a prior period of relative inactivity? The question of workload transition on the battlefield is the central issue of this volume. Unfortu- nately, while there is a considerable literature on vigilance, very little infor- mation is available on the transition problem with regard to the quality of sustained attention. Nevertheless, the information that is available suggests that vigilance efficiency could be threatened considerably by a shift in the workload level. Assuming that the modern battlefield will feature an intermingling of friendly and enemy forces, it would appear that one consequence of the onset of combat would be an increased event rate with respect to potential targets. A similar situation is found in the nuclear power plant following an abnormality. Scores of lights and warnings may be illuminated, but only a small percentage of these may contain useful diagnostic information. A study by Krulewitz et al. (1975) addressed this issue in the labora- tory. These investigators examined the effects of abrupt shifts in event rate in a low to high as well as in a high to low direction and found that such shifts had strong effects on monitoring efficiency. The frequency of signal detections for subjects shifted from a low to a high event rate was consider- ably poorer than that for control subjects maintained on the high event rate throughout the vigilance session. Conversely, a shift in the opposite direc- tion resulted in enhanced performance on the part of the shifted subjects. A similar result has been reported by Wiener (1977) in a study in which the event rate was not altered within a session, but instead between sessions. Wiener's finding suggests that the event rate shift effect is most likely a contrast phenomenon in which prior experience determines the perceived magnitude of stimulus events (see Corso, 1967~. Consequently, the high resource demands of a fast event rate in combat may be exacerbated by the lower demands of the lower event rate prior to the onset of combat. Still another aspect of the post-transition phase compared with the prior passive waiting period is that team members will probably be confronted with more dual-task requirements. For example, the air crew may need to fly a crippled airliner while simultaneously engaging in demanding fault diagnosis and problem solving, and the tank commander may find it neces- sary to watch for both hostile aircraft and ground forces. A recent study by Gluckman et al. (1988a) demonstrated that efficiency is degraded when two vigilance tasks must be handled simultaneously and that this effect is stron- gest when resource demands are increased by the presence of a successive task in the dual-task ensemble. Further work by Gluckman on dual-task vigilance performance has in- dicated that, as in the case of event rate, changes in workload wrought by shifts in dual-task requirements during a vigil can also have a deleterious
VIGILANCE AND TARGET DETECTION 157 effect on performance efficiency. This effect, however, is quite different from that which results from alterations in event rate. More specifically, Gluckman (1990) has reported that subjects shifted from single-task to dual- task monitoring (low to high workload) during a vigil did as well on the dual-task assignment as controls who were maintained on that assignment throughout the vigil. By contrast, subjects shifted from dual-task to single- task monitoring (high to low workload) performed more poorly immedi- ately after the shift compared with controls who were maintained on the single task continuously. A result of this sort suggests that the cost in resource expenditure incurred by a high workload during combat may perseverate and have a negative impact on subsequent tasks requiring sustained atten- tion (e.g., sentry duty) that armor personnel may have to perform after combat. When compared with the effects of shifts in event rate described by Krulewitz et al. (1975) and Wiener (1977), Gluckman's (1990) findings also suggest that the effects of workload transitions on vigilance perfor- mance are complex-they seem to depend on the precise manner in which such changes are brought about. This interpretation is supported in a recent report by See (1992), which indicated that shifts in workload brought about by changes in signal salience produced still a different result from the Krulewitz et al. (1975) and Wiener (1977) experiments. In See's investigation, sub- jects shifted from high-salience to low-salience signals performed in a man- ner equivalent to their continuous low-salience controls, while subjects shifted in the opposite direction performed in a manner similar to their continuous high-salience controls. Given that the quality of vigilant behavior may be affected adversely by increments in workload, can anything be done to minimize the potential problem of workload transition and to enhance the general vigilance level of tank crews? Some suggestions toward these ends are offered below. Remediation Engineering Solutions In a cogent analysis of human factors principles in the control of vigi- lance, Craig (1984) suggested a number of general approaches for enhanc- ing the quality of sustained attention in operational settings. Two of these, reductions in signal uncertainty and the moderation of environmental sources of stress, might be followed by focusing on enhanced engineering design. In Craig's analysis, the phrase signal uncertainty has broad meaning. It refers to anything that might aid monitors in detecting small, transient, or otherwise inconspicuous targets. In view of what is known with regard to the importance of signal amplitude and duration in vigilance in the tank environment, devices for aiding target acquisition, such as thermal imaging
158 WORKLOAD TRANSITION or computer-assisted detection systems, might be built to amplify the inten- sity of such critical signals as hot spots and weapon flashes and increase their dwell time of such signals on the crews' target acquisition displays. Computer aids might also be developed to provide crews with information as to the most probable times of appearance and spatial locations of targets and thus reduce the effects of both temporal and spatial uncertainty on vigilance efficiency. As noted earlier, the sensory modality of a signal plays an important role in the quality of vigilant behavior, and sensory redundant displays aid in detection efficiency. Accordingly, thought might be given to ways of taking advantage of this effect by designing a form of radar-sonar system for tanks that would make use of analogous audio-visual signals, as has been successfully demonstrated for submarine sonar monitors (Lewandowski and Kobus, 1989~. Such a system might increase overall efficiency in target acquisition. It might also reduce the powerful effects of event rate, since these effects can be eliminated in the laboratory when subjects are perm~it- ted to search for targets by alternating their inspection of analogous audi- tory and visual displays (Galinsky et al., 19901. Throughout this chapter emphasis has been placed on the importance of resource utilization as a way of understanding vigilance performance. It is clear that successive tasks drain more processing capacity than simulta- neous tasks. Hence, designers should strive whenever possible to provide simultaneous displays for target acquisition functions. This might be achieved by the use of instruments that furnish crews with immediate reference stan- dards against which events can be compared instead of forcing them to rely on short-term memory to separate signals and noise. In addition to these psychophysically oriented solutions, Craig (1984) has suggested that still another means by which the observer's uncertainty about the signal to be detected might be reduced and the likelihood of target acquisition enhanced is to employ multiple monitors under the assumption that failures to detect a target are likely to be uncorrelated; thus, what is missed by one monitor might be detected by another. In this way, detection probability by the team of monitors may be greater than that of only one monitor. Team monitoring is an attractive possibility for enhancing target acqui- sition in armor crews, since Army doctrine calls for overlapping fields of inspection on the battlefield (Armor Field Manual 17-12-1, Tank Combat Tables, no date). Unfortunately, the experimental evidence with regard to the effectiveness of team monitoring in vigilance is too equivocal at present to permit a strong endorsement of this practice. While some experiments have demonstrated an improvement in signal detectability for teams of monitors compared with lone monitors (Baker et al., 1962; Konz and Osman, 1977, Mico, 1965; Morgan and Alluisi, 1965; Wiener, 1965), other studies have
VIGILANCE AND TARGET DETECTION 159 reported no advantage for monitoring teams (Bergum and Lehr, 1962~. In at least one case, performance for multioperator teams was poorer than that of single monitors working alone (Ware et al., 1964~. The effects of team monitoring are complicated by the fact that social interactions between monitors plays a role in determining the effectiveness of the procedure. For example, a study by Klinger (1969) showed that the performance of a monitor is enhanced by the presence of a coactor, but only when the coactor served as a potential evaluator by having access to the quality of the monitor's perfor mance. It is worth noting that the lack of consistency in team monitoring stud- ies was first pointed out by Davies and Tune (1969) several years ago. Little work has been done since that time to clarify the issue. Moreover, except for one study involving signal density (Bergum and Lehr, 1962), none of the available experiments has explored the influence of important psychophysical variables on the effectiveness of team monitoring. Given its potential applicability to target acquisition in armored combat, the team concept warrants further investigation. Unlike a factory environment or that of the air crew in fixed-wing aircraft, both of whom can be under the control of the intrinsic interests and the wishes of management, battlefield conditions are more capricious, and the battlefield is a more difficult arena in which to modify sources of envi- ronmental stress. Noise and vibration are likely to be unavoidable, and complete noise abatement is probably not advisable, since noise in battle can be informative as well as distracting. In contrast, the elimination of temperature extremes of the sort that can degrade vigilance efficiency in tanks or helicopters would appear to be worthwhile. This can be accom- plished by building appropriate cooling and heating systems into such ve- hicles and/or by providing protective clothing that can insulate crew mem- bers from climatic extremes. These steps seem to be especially necessary if tank crew members are to remain buttoned up inside their vehicles for several hours, as in chemical warfare scenarios. Training A third approach to the control of vigilance recommended by Craig (1984) and also by Parasuraman (1986) is training. Considerable evidence exists to indicate that training with target cueing or with knowledge of results can have a beneficial effect on performance in tasks requiring sus- tained attention (Decker et al., 1991; Dittmar et al., 1985; Warm, 1977; Warm and Jerison, 1984; Wiener, 1984~. With regard to their vigilance functions, crews of any team confronting vigilance problems may profit from instructions designed to optimize response strategies. More specifi- cally, it may help crew members develop more effective procedures for
160 WORKLOAD TRANSITION coping with the vigilance challenges that confront them in combat situa- tions to advise them of such effects as the tendency of monitors to: (1) transfer signal probability estimates from one situation to another, (2) ig- nore low-probability areas under spatial uncertainty, (3) become more con- servative in responding over time, (4) ignore peripheral stimuli in noise, and (5) become more certain (perhaps unduly so) of signal and nonsignal decisions in noise. This training of operators to understand the biases and tendencies in their own performance has an analogy in decision making discussed in Chapter 8, wherein such debiasing procedures have proven successful. The most impressive testimonial for the potential benefits of training comes from Fisk and Schneider's (1981) finding that the decrement func- tion can be eliminated when vigilance tasks become automatic. Fisk and Scerbo (1987) provide several suggestions for how training should be car- ried out to maximize automatic processing development. Appropriate strat- egies for this type of training are discussed in Chapter 11. Fisk and Scerbo (1987) have also pointed out that training for automatic processing can make vigilance performance more resistant to the effects of environmental stress and increased mental workload, findings that add to the potential importance of a good training regimen and that fit well with Wilkinson's (1969) claim that familiarity with a task is a primary way to insulate perfor- mance against the effects of stress. As for stress itself, evidence is also available that indicates that training can have an important role in helping individuals to develop appropriate coping strategies (Hockey, 1986; see also Chapter 4~. REFERENCES Adams, J.A. 1956 Vigilance in the detection of low-intensity visual stimuli. Journal of Experimental Psychology 52:204-208. Adams, J.A., and L.R. Boulter 1964 Spatial and temporal uncertainty as determinants of vigilance performance. Jour- nal of Experimental Psychology 64:127-131. Adams, J.A., and J.M. Humes 1963 Monitoring of complex visual displays: Training for vigilance. Human Factors 5: 147-153. Adams, J.A., J.M Humes, and N.A. Sieveking 1963 Monitoring of complex visual displays. Effects of repeated sessions and heavy visual load on human vigilance. Human Factors 5:385-389. A]luisi, E.A. 1966 Attention and vigilance as mechanisms of response. Pp. 201-213 in E.A. Bilodeau, ea., Acquisition of Skill. New York: Academic Press. Anch, A.M., C.P. Brownian, M.M. Mitler, and J.K. Walsh 1988 Sleep: A Scientific Perspective. Englewood Cliffs, New Jersey: Prentice-Hall.
VIGILANCE AND TARGET DETECTION Baker, R.A., I.W. Ware, and R.R. Sipowitz 161 1962 Signal detection by multiple monitors. Psychological Record 12:133-137. Bell, C.R., K.A. Provins, and R.W. Hiorns 1964 Visual and auditory vigilance during exposure to hot and humid conditions. Ergo- nomics 7:279-288. Benor, D., and E. Shvartz 1971 Effect on body cooling on vigilance in hot environments. Aerospace Medicine 42:727-730. Bergstrom, B., M. Gillsberg, and P. Arnberg 1973 Effects of sleep loss and stress on radar watching. Journal of Applied Psychology 58:158-162. Bergum, B.O., and D.J. Lehr 1962 Vigilance performance as a function of paired monitoring. Journal of Applied Psychology 46:341-343. Blackwell, PA., and J.A. Belt 1971 Effect of differential levels of ambient noise on vigilance performance. Percep- tual and Motor Skills 32:734. Bowers, J.C. 1983 Stimulus Homogeneity and the Event Rate Function in Sustained Attention. Doc- toral dissertation, University of Cincinnati, Cincinnati, Ohio. Broadbent, D.E. 1954 Some effects of noise on visual performance. Quarterly Journal of Experimental Psychology 6:1-5. 1963 Differences and interactions between stresses. Quarterly Journal of Experimental Psychology 15:205-211. Broadbent, D.E., and M. Gregory 1965 Effects of noise and of signal rate upon vigilance analyzed by means of decision theory. Human Factors 7:155-162. Colquhoun, W.P., ed. 1972 Aspects of Human Efficiency: Diurnal Rhythm and Loss of Sleep. London, UK: English Universities Press. Colquhoun, W.P., and A.D. Baddeley 1964 Role of pretest expectancy in vigilance decrement. Journal of Experimental Psy- chology 68:156-160. Influence of signal probability during pretraining on vigilance decrement. Journal of Experimental Psychology 73 (1):153-155. Colquhoun, W.P., and R.F. Goldman 1972 Vigilance under induced hyperthermia. Ergonomics 15:621-632. Corcoran, D.W.J., J. Mullin, M.T. Rainey, and G. Frith 1977 The effects of raised signal and noise amplitude during the course of vigilance tasks. Pp. 645-664 in R.R. Mackie, ea., Vigilance: Theory, Operational Perfor- mance and Physiolog~cal Correlates. New York: Plenum. 1967 Corso, J.F. 1967 The Experimental Psychology of Sensory Behav~or. New York: Holt, Rinehart and Winston. Craig, A. 1984 Human engineering: The control of vigilance. Pp. 247-291 in J.S. Warm, ea., Sustained Attention in Human Performance. Chichester, UK: Wiley. 1985a Field studies of human inspection: The application of vigilance research. Pp. 133-144 in S. Folkard and T.H. Monk, eds., Hours of Work: Temporal Factors in Work-Schedul~ng. Chichester, UK: Wiley.
162 WORKLOAD TRANSITION 1985b Vigilance: Theories and laboratory studies. Pp. 107-121 in S. Folkard and T.H. Monk, eds., Hours of Work: Temporal Factors in Work-Scheduling. Chichester, UK: Wiley. 1991 Vigilance and monitoring for multiple signals. Pp. 153-172 in D. L. Damos, ea., Multiple Task Performance. London, UK: Taylor and Francis. Craig, A., W.P. Colquhoun, and D.W.J. Corcoran 1976 Combining evidence presented simultaneously to the eye and the ear: A compari- son of some predictive models. Perception and Psychophysics 19:473-484. Davies, D.R. 1985 Individual and group differences in sustained attention. Pp. 123-132 in S. Folkard and T.M. Monk, eds., Hours of Work: Temporal Factors in Work-Scheduling. Chichester, UK: Wiley. Davies, D.R., and G.R.J. Hockey 1966 The effects of noise and doubling the signal frequency on individual differences in visual vigilance performance. British Journal of Psychology 57:381-389. Davies D.R., and R. Parasuraman 1982 The Psychology of Vigilance. London, UK: Academic Press. Davies, D.R., and G.S. Tune 1969 Human Vigilance Performance. New York: American Elsevier. Deaton, J.E., and R. Parasuraman 1988 Effects of task demands and age on vigilance and subjective workload. Pp. 1458- 1462 in Proceedings of the Human Factors Society 32nd Annual Meeting. Santa Monica, California: Human Factors Society. Decker, A.B., J.S. Warm, W.N. Dember, and P.A. Hancock 1991 Effects of feedback on perceived workload in vigilance performance. Pp. 1491- 1494 in Proceedings of the Human Factors Society 35th Annual Meeting. Santa Monica, California: Human Factors Society. Dember, W.N., and J.S. Warm 1979 Psychology of Perception, Second Edition. New York: Holt, Rinehart and Winston. Dember, W.N., J.S. Warm, J.C. Bowers, and T. Lanzetta 1984 Intrinsic motivation and the vigilance decrement. Pp. 21-26 in A. Mital, ea., Trends in Ergonomics/Human Factors, I. Amsterdam: Elsevier (North-Holland). Dittmar, M.L. 1989 Sex Differences and Stress In Vigilance Performance. Doctoral dissertation, Uni- versity of Cincinnati, Cincinnati, Ohio. Dittmar, M.L., J.S. Warm, and W.N. Dember 1985 Effects of knowledge of results on performance in successive and simultaneous vigilance tasks: A signal detection analysis. Pp. 195-202 in R.E. Eberts and C.G. Eberts, eds., Trends In Ergonomics/Human Factors II. Amsterdam: Elsevier (North-Holland). Doll, T.J., and T.E. Hanna 1989 Enhanced detection with bimodal sonar displays. Human Factors 31:539-550. Easterbrook, J.A. 1959 The effect of emotion on cue utilization and the organization of behavior. Psycho- log~cal Review 66:183-201. Fisk, A.D., and M.W. Scerbo 1987 Automatic and control processing approach to interpreting vigilance performance: A review and reevaluation. Human Factors 29:653-660. Fisk, A.D., and W. Schneider 1981 Control and automatic processing during tasks requiring sustained attention: A new approach to vigilance. Human Factors 23:737-750.
VIGILANCE AND TARGET DETECTION 163 Frankenhaeuser, M., B. Nordheden, A.L. Myrsten, and B. Post 1971 Psychophysiological reactions to understimulation and overstimulation. Acta Psychologica 35:298-308. Frankenhaeuser, M., and P. Patkai 1964 Catecholamine excretion and performance under stress. Perceptual and Motor Skills 19:13-14. Galinsky, T.L., W.N. Dember, and J.S. Warm 1989 Effects of Event Rate on Subjective Workload in Vigilance Performance. Paper presented at the meeting of the Southern Society for Philosophy and Psychology, Chicago, Illinois. Galinsky, T.L., J.S. Warm, W.N. Dember, E.M. Weller, and M.W. Scerbo 1990 Sensory alternation and vigilance performance: The role of pathway inhibition. Human Factors 32:717-728. Gluckman, J.P. 1990 Changing Task Demands in Sustained Attention: Effects on Performance and Perceived Workload. Doctoral dissertation, University of Cincinnati, Cincinnati, Ohio. Gluckman, J.P., W.N. Dember, and J.S. Warm 1988a Capacity demand in dual-task monitoring of simultaneous and successive vigi- lance tasks. Pp. 1463-1465 in Proceedings of the Human Factors Society 32nd Annual Meeting. Santa Monica, California: Human Factors Society. Gluckman, J.P., J.S. Warm, W.N. Dember, J.A. Thiemann, and P.A. Hancock 1988b Subjective Workload in Simultaneous and Successive Vigilance Tasks. Paper presented at the meeting of the Psychonomic Society, Chicago, Illinois. Goether, W.F. 1971 Vibration and human performance. Human Factors 13:203-216. Gopher, D., and R. Kimchi 1989 Engineering psychology. Annual Review of Psychology 40:431-455. Green, D.M., and J.A. Swets 1974 Signal Detection Theory and Psychophysics. New York: Krieger. Griffin, J.A., W.N. Dember, and J.S. Warm 1986 Effects of depression on expectancy in sustained attention. Motivation and Emo- tion 10:195-205. Guralnick, M.J. 1972 Observing responses and decision processes in vigilance. Journal of Experimental Psychology 93:239-244. Hancock, P.A. 1984 Environmental stressors. Pp. 103-142 in J.S. Warm, ea., Sustained Attention in Human Performance. Chichester, UK: Wiley. 1986 Sustained attention under thermal stress. Psychological Bulletin 99:263-281. Hancock, P.A., and J.S. Warm 1989 A dynamic model of stress and sustained attention. Human Factors 31:519-537. Hart, S.G., and L.E. Staveland 1988 Development of NASA-TLX (Task Load Index): Results of empirical and theo- retical research. Pp. 139-183 in P.A. Hancock and N. Meshkati, eds., Human Mental Workload. Amsterdam: North-Holland. Hartley, L.R. 1974 A comparison of continuous and distributed sleep schedules. Quarterly Journal of Experimental Psychology 26:8-14. Hartley, L.R., and E. Shirley 1977 Sleep loss, noise and decisions. Ergonomics 20:481-482.
164 WORKLOAD TRANSITION Hatfield, J.L., and M. Loeb 1968 Sense mode and coupling in a vigilance task. Perception and Psychophysics 4:29- 36. Hockey, G.R.J. 1970 Effect of loud noise on attentional selectivity. Quarterly Journal of Experimental Psychology 22:28-36. 1986 Changes in operator efficiency as a function of environmental stress, fatigue, and circadian rhythms. Pp. 44.1-44.49 in K.R. Doff, L. Kaufman, and J.P. Thomas, eds., Handbook of Human Perception and Performance: Volume II, Cognitive Processes and Performance. New York: Wiley. Hoddes, E., V. Zarcone, H. Smythe, R. Phillips, and W.C. Dement 1973 Quantification of sleepiness: A new approach. Psychophysiology 10(4):431-436. Home, J.A., N.R. Anderson, and R.T. Wilkinson 1983 Effects of sleep deprivation on signal detection measures of vigilance: Implica- tions for sleep function. Sleep 6:347-358. Hovanitz, C.A., K. Chin, and J.S. Warm 1989 Complexities in life stress-dysfunction relationships: A case in point-tension headache. Journal of Behavioral Medicine 12: 55-75. Jerison, H.J. 1959a Effects of noise on human performance. Journal of Applied Psychology 43:96 101. 1959b Experiments on Vigilance: The Empirical Model for Human Vigilance. WADC Report No. 58-526. Wright-Patterson Air Force Base, Ohio: Aero-Medical Labo ratory, Wright Air-Development Center. 1963 On the decrement function in human vigilance. Pp. 199-212 in D.N. Buckner and J.J. McGrath, eds., Vigilance: A Symposium. New York: McGraw-Hill. Jerison, H.J., and R.M. Pickett 1964 Vigilance: The importance of the elicited observing rate. Science 143:970-971. Johnson, L.C., and P. Naitoh 1974 The Operational Consequences of Sleep Deprivation and Sleep Deficit. Report No. AGARD-OGRAPH-193. Paris, France: Advisory Group for Aerospace Re- search and Development. Johnson, R.F., and D.J. McMenemy 1989 Target detection, rifle marksmanship and mood during three hours of sentry duty. Pp. 1414-1418 in Proceedings of the Human Factors Society 33rd Annual Meet- ing. Santa Monica, California: Human Factors Society. Jones, D.M. 1983 Noise. Pp. 61-95 in R. Hockey, ea., Stress and Fatigue in Human Performance. Chichester, UK: Wiley. 1984 Performance effects. Pp. 155-184 in D.M. Jones and A.J. Chapman, eds., Noise and Society. Chichester, UK: Wiley. Joshi, A., W.N. Dember, J.S. Warm, and M.W. Scerbo 1985 Capacity Demands In Sustained Attention. Paper presented at the meeting of the Psychonomic Society, Boston, Massachusetts. Kantowitz, B.H. 1985 Channels and stages in human information processing: A limited analysis of theory and methodology. Jourrlal of Mathematical Psychology 29:135-174. Kerslake, D.M., and E.C. Poulton 1965 Initial stimulating effect of warmth upon perceptual efficiency. Transient per- petual efficiency variation on temperature increase, noting arousal level. Aero- space Medicine 36:29-32.
VIGILANCE AND TARGET DETECTION Klinger, E. 165 1969 Feedback effects and social facilitation of vigilance performance: Mere coaction versus potential evaluation. Psychonomic Society 14:161-162. Koelega, H.S., and J.A. Brinkman 1986 Noise and vigilance: An evaluative review. Human Factors 28:465-482. Konz, S., and K. Osman 1977 Team efficiencies on a paced visual inspection task. Journal of Human Ergology 6:111-119. Krulewitz, J.E., and J.S. Warm 1977 The event rate context in vigilance: Relation to signal probability and expectancy. Bulletin of the Psychonomic Society 10(5):429-432. Krulewitz, J.E., J.S. Warm, and T.H. Wohl 1975 Effects of shifts in the rate of repetitive stimulation on sustained attention. Per- ception and Psychophysics 18(4):245-249. Lanzetta, T.M., W.N. Dember, J.S. Warm, and D.B. Berch 1987 Effects of task type and stimulus heterogeneity on the event rate function in sus- tained attention. Human Factors 29:625-633. Lewandowski, L.J., and D.A. Kobus 1989 Bimodal information processing in sonar performance. Human Performance 2:73- 84. Loeb, M. 1986 Noise and Human Efficiency. Chichester, UK: Wiley. Loeb, M., and J.R. Binford 1963 Some factors influencing the effective auditory intensive difference limen. Jour- nal of the Acoustical Society of America 35:884-891. 1968 Variations in performance on auditory and visual monitoring tasks as a function of signal and stimulus frequencies. Perception and Psychophysics 4:361-367. Loeb, M., T.K. Noonan, D.W. Ash, and D.H. Holding 1987 Limitations of the cognitive vigilance increment. Human Factors 29:661-674. Lundberg, U., and M. Frankenhaeuser 1979 Pituitary-Adrenal and Sympathetic-Adrenal Correlates of Distress and Effort. Re- port No. 548. Department of Psychology, University of Stockholm, Sweden. Lundberg, P.K., J.S. Warm, and W. Seeman 1980 Sustained Attention and the Type A Individual: Attentive, Aroused and Able. Paper presented at the meeting of the Midwestern Psychological Association, Chi- cago, Illinois. Lysaght, R.J., J.S. Warm, W.N. Dember, and M. Loeb 1984 Effects of noise and information-processing demand on vigilance performance in men and women. Pp. 27-32 in A. Mital, ea., Trends in ErgonomicslHuman Fac- tors I. Amsterdam: Elsevier (North-Holland). Mackie, R.R., and J.F. O'Hanlon 1977 A study of the combined effects of extended driving and heat stress on driver arousal and performance. Pp. 537-558 in R.R. Mackie, ea., Vigilance: Theory, Operational Performance, and Physiological Correlates. New York: Plenum. Mackworth, N.H. 1948 The breakdown of vigilance during prolonged visual search. Quarterly Journal of Experimental Psychology 1:6-21. Research on the measurement of human performance. Pp. 174-331 in H.W. Sinaiko, ea., Selected Papers on Human Factors in the Design and Use of Control Systems. New York: Dover (Reprinted from Medical Research Council Special Report Series 268, London, UK: HM Stationery Office, 1950). 1961
166 WORKLOAD TRANSITION Macomber, R.M. 1967 Effects of Noise on Stress and Performance in Sustained Attention Tasks. Unpub- lished Senior Thesis, University of Cincinnati, Cincinnati, Ohio. McFarland, B.P., and C.G. Halcomb 1970 Expectancy and stimulus generalization in vigilance. Perceptual and Motor Skills 30:147-151. McNicol, D. 1972 A Primer of Signal Detection Theory. London, UK: Allen and Unwin. Metzger, K.R., J.S. Warm, and R.J. Senter 1974 Effects of background event rate and critical signal amplitude on vigilance perfor mance. Perceptual and Motor Skills 38:1175-1181. Mico, H.C. 1965 On the most efficient use of two observers in vigilance tasks. Pp. 449-453 in Proceedings of the Second Annual Congress OF Ergonomics. Dortmund, London, UK: Taylor and Francis. Milosevic, S. 1974 Effect of time and space uncertainty on a vigilance task. Perception and Psychophysics 15:33 1-334. Montague, W.E., C.E. Weber, and J.A. Adams 1965 The effects of signal and response complexity on eighteen hours of visual monitor- ing. Human Factors 7: 163-172. Moore, S.F., and S.J. Gross 1973 Influence of critical signal regularity, stimulus event matrix and cognitive style on vigilance performance. Journal of Experimental Psychology 99:137-139. Morgan, B.B., and E.A. Alluisi 1965 On the inferred independence of paired watchkeepers. Psychonomic Science 2:161- 162. Natsoulas, T. 1967 What are perceptual reports about? Psychological Bulletin 67:249-272. Nicely, P.E., and G.A. Miller 1957 Some effects of unequal spatial distribution on the detectability of radar targets. Journal of Experimental Psychology 53:195-198. Nuechterlein, K.H., R. Parasuraman, and Q. Jiang 1983 Visual sustained attention: Image degradation produces rapid sensitivity decre- ment over time. Science 220:327-329. O'Donnell, R.D., and F.T. Eggemeier 1986 Workload assessment methodology. Pp. 42.1-42.49 in K.R. Boff, L. Kaufman, and J.P. Thomas, eds., Handbook of Perception and Human Performance: Volume II, Cognitive Processes and Performance. New York: Wiley. Parasuraman, R. 1979 Memory load and event rate control sensitivity decrements in sustained attention. Science 205:924-927. 1984 The psychobiology of sustained attention. Pp. 61-101 in J.S. Warm, ea., Sustained Attention in Human Performance. Chichester, UK: Wiley. 1985 Sustained attention: A multifactorial approach. Pp. 493-511 in M.I. Posner and O.S. Marlin, eds., Attention and Performance I. Hillsdale, New Jersey: Erlbaum. 1986 Vigilance, monitoring and search. Pp. 43.1-43.39 in K.R. Boff, L. Kaufman and J.P. Thomas, eds., Handbook of Perception and Human Performance: Volume II, Cognitive Processes and Performance. New York: Wiley. Human computer monitoring. Human Factors 29:695-706. 1987
VIGILANCE AND TARGET DETECTION 167 Parasuraman, R., and D.R. Davies 1976 Decision theory analysis of response latencies in vigilance. Journal of Experimen tal Psychology: Human Perception and Performance 2:578-590. 1977 A taxonomic analysis of vigilance performance. Pp. 559-574 in R.R. Mackie, ea., Vigilance: Theory, Operational Performance and Physiological Correlates. New York: Plenum. Parasuraman, R., J.S. Warm, and W.N. Dember 1987 Vigilance: Taxonomy and Utility. Pp. 11-32 in L.S. Mark, J.S. Warm, and R.L. Huston, eds., Ergonomics and Human Factors: Recent [Research. New York: Springer-Verlag. Pepler, R.D. 1953 The Effect of Climatic Factors on the Performance of Skilled Tasks by Young European Men Living in the Tropics. 4. A Task of Prolonged Visual Vigilance. Medical Research Council Applied Psychology Unit Report No. 156/53, London, UK. Poulton, E.C. 1977 Arousing stresses increase vigilance. Pp. 423-459 in R.R. Mackie, ea., Vigilance: Theory, Operational Performance and Physiological Correlates. New York: Ple num. Poulton, E.C., and R.S. Edwards 1974 Interactions and range effects in experiments on pairs of stresses: Mild heat and low frequency noise. Journal of Experimental Psychology 104:621-628. Poulton, E.C., R.S. Edwards, and W.P. Colquhoun 1974 The interaction of the loss of a night's sleep with mild heat: Task variables. Ergonomics 17:59-73. Rosa, R.R., M.H. Bonnet, and J.S. Warm 1983 Recovery of performance during sleep following sleep deprivation. Psychophysiol- ogy 20:152-159. Rosekind, M.R., P.H. Gander, and D.F. Dinges 1991 Alertness management in flight operations: Strategic napping. Pp. 1-12 in Aero- space Technology Conference and Exposition. Long Beach, California. Scerbo, M.W., J.W. Warm, V.S. Doettling, R. Parasuraman, and A.D. Fisk 1987a Event asynchrony and task demands in sustained attention. Pp. 33-39 in L.S. Mark, J.S. Warm, and R.L. Huston, eds., Ergonomics and Human Factors: Recent Research. New York: Springer-Verlag. Scerbo, M.W., J.S. Warm, and A.D. Fisk 1987b Event asynchrony and signal regularity in sustained attention. Current Psychologi cal Research and Reviews 5:335-343. Schneider, W., and R.M. Shiffrin 1977 Control and automatic human information processing: I. Detection, search, and attention. Psychological Review 84: 1 -66S. Schohan, B., H.E. Rawson, and S.M. Soliday 1965 Pilot and observer performance in simulated low-altitude high-speed flight. Hu man Factors 7:257-265. See, J.E. 1992 Effects of Transitions in Signal Salience on Vigilance Performance. Masters the sis, University of Cincinnati. Seidel, W.F., T. Roth, T. Roehrs, F. Zorick, and W.C. Dement 1984 Treatment of a 12-hour shift of sleep schedule with Benzodiazepines. Science 224: 1262-1264.
168 WORKLOAD TRANSITION Sheridan, T. 1970 On how often the supervisor should sample. IEEE Transactions on System Sci- ence and Cybernetics SSC-6:140-145. Shoenberger, R.W. 1967 Effects of vibration on complex psychomotor performance. Aerospace Medicine 12: 1265-1269. Smith, R.P., J.S. Warm, and E.A. Alluisi 1966 Effects of temporal uncertainty on watchkeeping performance. Perception and Psychophysics 1:293-299. Swets, J.A. 1977 Signal detection theory applied to vigilance. Pp. 705-718 in R.R. Mackie, ea., Vigilance: Theory, Operational Performance and Physiological Correlates. New York: Plenum. Taub, H.A., and F.H. Osborne 1968 Effects of signal and stimulus rates on vigilance performance. Journal of Applied Psychology 52:133- 138. Taub, J.M., and R.J. Berger 1969 Extended sleep and performance: The Rip Van Winkle effect. Psychonomic Science 16:204-205. 1973 Performance and mood following variations in the length and timing of sleep. Psychophysiology 10:S59-570. Teichner, W.H. 1974 The detection of a simple visual signal as a function of time of watch. Human Factors 16(4):339-353. Thackray, R.I., J.P. Bailey, and R.M. Touchstone 1977 Physiological, subjective and performance correlates of reported boredom and mo- notony while performing a simulated radar control task. Pp. 203-215 in R.R. Mackie, ea., Vigilance: Theory, Operational Performance, and Physiological Cor- relates. New York: Plenum. Thiemann, Job., J.S. Warm, W.N. Dember, and E.B. Smith 1989 Effects of Caffeine on Vigilance Performance and Task-Induced Stress. Paper presented at the meeting of the Southern Society for Philosophy and Psychology, New Orleans, Louisiana. U.S. Department of the Army No date FM 17-12-1 Tank Combat Tables. Fort Knox, Kentucky: U.S. Army Armor School. Ware, J.S., R.A. Baker, and E. Drucker 1964 Sustained vigilance II. Signal detection for two-man teams during a 24-hour watch. Journal of Engineering Psychology 3:104-110. Warm, J.S. 1977 Psychological processes in sustained attention. Pp. 623-644 in R.R. Mackie, ea., Vigilance: Theory, Operational Performance and Physiological Correlates. New York: Plenum. 1984a An introduction to vigilance. Pp. 1-14 in J.S. Warm, ea., Sustained Attention in Human Performance. Chichester, UK: Wiley Warm, J.S., ed. 1984b Sustained Attention in Human Performance. Chichester, UK: Wiley. Warm, J.S., and E.A. Alluisi 1971 Influence of temporal uncertainty and sensory modality of signals on watchkeeping performance. Journal of Experimental Psychology 87:303-308.
VIGILANCE AND TARGET DETECTION 169 Warm, J.S., and D.B. Berch 1985 Sustained attention in the mentally retarded: The vigilance paradigm. Pp. 1-41 in N.R. Ellis and N.W. Bray, eds., International Review of Research in Mental Retar- dation fVolume 13). Orlando, Florida: Academic Press. Warm, J.S., W.N. Dember, and R. Parasuraman 1991 Effects of olfactory stimulation on performance and stress in a visual sustained attention task. Journal of the Society of Cosmetic Chemists 42:199-210. Warm, J.S., B.D. Eppe, and R.P. Ferguson 1974 Effects of knowledge of results and signal regularity on vigilance performance. Bulletin of the Psychonomic Society 4:272-274. Warm, J.S., S.R. Howe, H.D. Fishbein, W.N. Dember, and R.L. Sprague 1984 Cognitive demand and the vigilance decrement. Pp. 15-20 in A. Mital, ea., Trends in Ergonomics/Human Factors I. Amsterdam: Elsevier (North-Holland). Warm, J.S., and H.J. Jerison 1984 The psychophysics of vigilance. Pp. 15-60 in J.S. Warm, ea., Sustained Attention in Human Performance. Chichester, UK: Wiley. Warm, J.S., M. Loeb, and E.A. Alluisi 1970 Variations in watchkeeping performance as a function of the rate and duration of visual signals. Perception and Psychophysics 7:97-99. Warm, J.S., and R. Parasuraman, eds. 1987 Vigilance: Basic and applied. Human Factors 29:623-740. Warm, J.S., R.R. Rosa, and M.J. Colligan 1989 Effects of auxiliary load on vigilance performance in a simulated work environ- ment. Pp. 1419-1421 in Proceedings of the Human Factors Society 33rd Annual Meeting. Santa Monica, California: Human Factors Society. Webb, W.B., and H.W. Agnew 1974 The effects of chronic limitation of sleep length. Psychophysiology 11:265-274. Weisz, A.A., C. Goddard, and R.W. Allen 1965 Human Performance Under Random and Sinusoidal Vibration. Aerospace Medi- cal Research Laboratories Report No. AMRL-TR-65-209. Wright-Patterson Air Force Base, Ohio. Wesnes, K., and D.M. Warburton 1983 Stress and drugs. Pp. 203-243 in R. Hockey, ea., Stress and Fatigue In Human Performance. Chichester, UK: Wiley. Wiener, E.L. 1965 The performance of multi-man monitoring teams. Human Factors 6:179-184. 1977 Stimulus presentation rate in vigilance. Human Factors 19:301-303. 1984 Vigilance and Inspection. Pp. 207-246 in J.S. Warm, ea., Sustained Attention in Human Performance. Chichester, UK: Wiley. 1985 1987 Beyond the sterile cockpit. Human Factors 27:75-90. Application of vigilance research: Rare, medium or well done? Human Factors 29:725-736. Wilkinson, R.T. 1968 Sleep deprivation: Performance tests for partial and selective sleep deprivation. Pp. 28-43 in L.A. Abt and B.F. Reiss, eds., Progress in Clinical Psychology, VoIume 7. New York: Grune and Stratton. 1969 Some factors influencing the effect of environmental stressors upon performance. Psychological Bulletin 72:260-272. Wilkinson, R.T., R.S. Edwards, and E. Haines 1966 Performance following a night of reduced sleep. Psychonomic Science 4:471-472.
170 WORKLOAD TRANSITION Wilkinson, R.T., R.H. Fox, R. Goldsmith, I.F.G. Hampton, and H.E. Lewis 1964 Psychological and physiological responses to raised body temperature. Journal of Applied Physiology 19:287-291. Wilkinson, R.T., and R. Gray 1974 Effects of duration of vertical vibration beyond the proposed ISO "fatigue-de- creased proficiency time," on the performance of various tasks. Proceedings of the AGARD Conference on Vibration and Combined Stresses In Advanced Systems. NO. 145. OS1O7 Norway: AGARD/NATO. Williams, ILL., O.F. Kearny, and A. Lubin 1965 Signal uncertainty and sleep loss. Journal of Experimental Psychology 69:401- 407. Williges, R.C. 1969 Within-session criterion changes compared to an ideal observer criterion in a vi- sual monitoring task. Journal of Experimental Psychology 81:61-66. Wingate, P. 1972 The Penguin Medical Encyclopaedia. Harmondsworth, Middlesex, UK: Penguin Books. Yoshitake, H. 1978 Three characteristic patterns of subjective fatigue symptoms. Ergonomics 21:231- 233.