DEMAND TRANSITIONS IN VIGILANCE: EFFECTS ON ... - CiteSeerX

2 downloads 0 Views 216KB Size Report
The effects of such demand transitions on vigilance performance have been explored in .... In the low salience condition, the contrast between the letter stimuli.
Helton, W.S., Shaw, T.H., Warm, J.S., Matthews, G., Dember, W.N., & Hancock, P.A. (2004). Demand transitions in vigilance: Effects on performance efficiency and stress. In: D.A. Vincenzi., M. Mouloua., and P.A. Hancock, (Eds.). Human performance, situation awareness and automation: Current research and trends. (pp. 258- 262), Erlbaum: Mahwah, NJ..

DEMAND TRANSITIONS IN VIGILANCE: EFFECTS ON PERFORMANCE EFFICIENCY AND STRESS William S. Helton Wilmington College Tyler H. Shaw, Joel S. Warm, Gerald Matthews William N. Dember University of Cincinnati Peter A. Hancock University of Central Florida

ABSTRACT The National Research Council (Huey & Wickens, 1993) has identified transitions in task demand as an important dimension of human operator response. The effects of such demand transitions on vigilance performance have been explored in only two published studies (Krulewitz, Warm, & Wohl, 1975; Gluckman, Warm, Dember, & Rosa, 1993). The results of the first study suggested that the effects of transitions in task demand are characterized by simple psychophysical contrast. However, the latter study failed to confirm these results. The present study was designed to further explore whether demand transitions can be adequately described by a psychophysical contrast model. With that goal in mind, observers were shifted from a high salience (low demand) to a low salience (high demand) condition and vice-versa. Results failed to confirm the psychophysical contrast model. However, they did show that transitions in task demand have important implications for task-induced stress. Key Words: Vigilance, Demand Transition, Stress INTRODUCTION Laboratory studies of vigilance or sustained attention traditionally maintain a constant information-processing load throughout the experimental session (Davies & Parasuraman, 1982). However, the vigilance tasks encountered in many operational settings, such as air-traffic control, process control, and medical monitoring, can contain abrupt changes in the demands placed upon operators during a duty cycle. Consequently, the National Research Council (Huey & Wickens, 1993) has identified transitions in task demand as an important dimension for study in vigilance research. Only two published investigations have focused upon this issue. One of these made use of changes in background event rate or the rate of repetition of stimulus events that need to be scanned in order to detect signals. Performance efficiency typically varies inversely with event rate (Warm & Jerison, 1984). Krulewitz, Warm, and Wohl (1975) found that observers shifted abruptly from a low to a high event rate during a vigil did more poorly than non-shifted controls on the high event rate. Those shifted in the high to low direction exceeded the performance of non-shifted controls on the low event rate. An outcome of this sort suggests that the effects of transitions in task demand are characterized by simple psychophysical contrast. However, a subsequent study by Gluckman, Warm, Dember, and Rosa (1993), using two different vigilance tasks to shift observers from single-task to dual-task monitoring and vice versa, produced performance changes that were devoid of contrast effects. In that study, the post-shift performances of the shifted groups equaled that of their non-shifted controls. Gluckman et al. (1993) suggested that single-task and dual-task conditions in their study might have been perceived as so qualitatively different as to preclude any kind of direct contrast between them. Thus, contrast effects in demand transitions may be limited to changes in a single common dimension. In addition to event rate, performance efficiency in vigilance varies directly with signal salience (Warm & Jerison, 1984). Accordingly, this

Helton, W.S., Shaw, T.H., Warm, J.S., Matthews, G., Dember, W.N., & Hancock, P.A. (2004). Demand transitions in vigilance: Effects on performance efficiency and stress. In: D.A. Vincenzi., M. Mouloua., and P.A. Hancock, (Eds.). Human performance, situation awareness and automation: Current research and trends. (pp. 258- 262), Erlbaum: Mahwah, NJ..

study was designed to determine whether contrast effects would be obtained when signal salience was shifted in high-to-low and low-to-high directions. To date, studies of demand transitions have only examined effects on performance efficiency. However, vigilance tasks are stress-inducing, and the stress of sustained attention is closely tied to psychophysical demands (Galinsky, Rosa, Warm, & Dember, 1993; Hancock & Warm, 1989; Temple et al., 2000; Warm, 1993). Hence, a second goal for this study was to examine the effect of transitions in task demand on task-induced stress. Toward that end, the study made use of the Dundee Stress State Questionnaire (DSSQ; Matthews et al., 1999; 2002), an instrument that provides factor-analytically differentiated scales of task engagement, distress, and worry. Prior research has shown a vigilance signature in regard to the DSSQ – observers feel less task engaged and more distressed after a vigil than prior to its start (Grier, et al, 2003; Helton, Dember, Warm, & Matthews, 2000; Matthews, Joyner, Gilliland, Huggins, & Falconer, 1999; Matthews et al., 2002; Temple et al., 2000). The present study examined the effects of transitions in task demand on this pattern of stress response. METHOD Twenty observers (10 males and 10 females) were assigned at random to each of four conditions resulting from the factorial combination of signal salience (high and low salience signals) and switching (switch and no-switch). All observers participated in a 12-min vigil divided into six continuous 2-min periods. They inspected the repetitive presentation on a VDT of light gray capital letters consisting of an “O,” a “D,” and a “backwards D.” The letters were exposed for 40 msec at a rate of 57.5 events/min against a visual mask consisting of unfilled circles on a white background. Critical signals for detection (p = 0.20/period of watch) were the appearance of the letter “O.” Observers signified their detection of critical signals by pressing the key on a response pad. Prior to the main vigil, all participants were given a 2-min. period of practice to familiarize themselves with the vigilance task. In the high salience condition, the contrast between the letter stimuli and the background was 59 percent, as indexed by the Michaelson contrast ratio ([maximum luminance - minimum luminance / maximum luminance + minimum luminance] x 100; Coren, Ward, & Enns, 1999). In the low salience condition, the contrast between the letter stimuli and the background was 45 percent. Switch participants performed for 6 min at one salience level and then for 6 min at the other. This abbreviated vigilance task has been found to duplicate the general effects of signal salience and task-induced stress noted with more traditional long-duration tasks (Temple et al., 2000). The DSSQ was administered in two sessions: a pre-vigil questionnaire completed prior to the practice period and a post-vigil questionnaire completed after the vigil. RESULTS Signal Detections. Mean percentages of correct detections in all experimental conditions are presented in Figure 1. Separate split-plot analyses of variance (ANOVA) based on arcsin transformations of the detection scores were conducted on the pre-switch and post-switch segments of the vigil. The analysis of the pre-switch data revealed that signals were detected significantly more often in the high salience (M = 98.5%) than in the low salience condition (M = 95.5%), F (1, 76) = 10.37, p