Latent Degradation under Stress

Although stress does not always result in any obvious reduction in performance, this should not be taken to mean that there is no threat to task goals. There is now considerable evidence of knock-on effects of performance protection to secondary aspects of behavior, related to both performance and costs; I have referred to these as latent degradation.

By reducing the safe working margins of the adaptive control process, these may threaten the integrity of performance - for example, strategies that work only if there are no new problems to deal with. Four kinds of latent degradation may be identified: two performance indices (secondary task decrements and strategy changes) and two cost indices (psychophysiological activation and fatigue aftereffects).

Secondary Task Decrements. Decrements in the secondary aspects of performance are commonly observed in studies of the effects of high workload, providing an indirect measure of increasing load on primary tasks. Such effects have been studied less systematically in assessing threats from environmental stressors, although they are, in fact, also common.

One of the best-documented forms of secondary task decrement under stress is the narrowing of attention found in spatially complex tasks. For example, although a central tracking task may be carried out effectively under noise, the detection of signals in the visual peripheral may be impaired.

Similar attentional narrowing has been found under both laboratory and field conditions and for a wide range of environmental conditions (noise, heat, anxiety associated with deep sea diving, and threat of shock). This type of decrement may be related to strategy changes because they depend on a shift of priority between task elements.

Strategy Changes. Strategy changes are (usually adaptive) changes in the way that tasks are carried out under stress. One obvious way is to minimize the disruption to primary activities by reducing the time spent on secondary tasks. However, there may also be more subtle changes, involving a shift to less resource-intensive modes of task control, reducing dependency on effort-demanding processes such as WM, which is known to be impaired under stress conditions.

Despite their obvious diagnostic value, such effects have not been well studied, partly because of the complex task environments necessary to analyze strategy changes. It has been known for some time that, under periods of difficulty or stress, industrial process operators may shift from an attention-demanding open-loop control (in which whole sequences of actions are guided mainly by the operator’s internal mental model) to a simpler closed-loop strategy (in which actions are carried out one at a time, paying more attention to feedback). This may slow the process or fail to make optimal use of available options, but it minimizes the likelihood of serious errors.

A good example of this is a well-known study of French air-traffic controllers, carried out in the 1970s. The controllers adopted a simplified method of dealing with aircraft contacts when they exceeded a comfortable number, but under very high workload they switched from individual plane-by-plane routing instructions to a fixed procedure for all contacts. By minimizing the demands for planning and aircraft management, they reduced the need to involve the vulnerable WM system.

The strategy change is adaptive in that secondary goals such as airport schedules and passenger comfort are compromised in the service of the primary goal of safety. A second example is the work of Schonpflug’s group in Berlin, which used a simulated office environment to examine decision making in stock control. Under normal conditions, participants typically held background information (on prices, stock levels, etc.) in WM while making a sequence of decisions. However, under time pressure or loud noise, they tended to check lists containing such information before making each decision.

Reducing the load on WM helped people to keep decision errors to a minimum, although at the expense of increased time costs. Again, the change is adaptive because accuracy matters more than speed in such work. However, in situations in which speed is also important, the hidden loss of efficiency represents a genuine stress-induced impairment.

Psychophysiological Activation. One of the most reliable costs of the use of increased effort to protect performance is the observation of increased levels of activation. This is particularly true of the physiological systems involved in emergency reactions (e.g., sympathetic and musculoskeletal responses, and responses of the neuroendocrine stress systems).

These effects are typically accompanied by changes in subjective reports of emotional and mood states reflecting the affective response to emergency and sustained coping effort. These may be thought of as the unwanted side-effects of the compensatory behavior that helps to maintain primary performance under threat from environmental conditions.

The effect is illustrated in an early study of sleep deprivation, in which decrements in arithmetic computation following a night without sleep were smaller for participants found to have increased muscle tension (interpreted as evidence of greater effort to combat sleepiness and maintain orientation toward the task).

This performance-cost trade-off is seen more clearly in several studies of noise effects using more meaningful psychophysiological measures. Noise has been shown to increase heart rate, blood pressure, adrenaline, and subjective effort in tasks in which performance decrement is forestalled. In the Swedish study referred to earlier, two different patterns of arithmetic performance and costs were observed in different experiments. In one, performance was impaired by noise, with no change in levels of adrenaline or effort.

In the other, performance was maintained, but adrenaline and effort levels were both greater. Unfortunately, because of the difficulty of obtaining psychophysiological measures under such circumstances, there are few studies within real work contexts, although another Swedish study found that the absence of decrement in work output during an intense period of organizational change was, again, accompanied by a compensatory increase in adrenaline and cognitive effort.

Such effects illustrate the role of compensatory regulation in the protection of performance and may be seen as a trade-off between the protection of the primary performance goal and the level of mental effort that has to be invested in the task. They also indicate that the regulation of effort is at least partially under the control of the individual rather than being an automatic feature of task or environmental conditions.

Fatigue Aftereffects of Stress. A final form of latent degradation is one that appears only after set tasks have been completed, in terms of decrements on new (and less critical) tasks. Such aftereffects have also been studied very little, and then normally within a workload-fatigue paradigm. However, they are equally appropriate as a response to the sustained effort required to maintain effective levels of work under stressful environmental conditions.

Given its long-recognized importance, work fatigue has been studied extensively since the early days of psychology, although it has proved surprisingly difficult to demonstrate carry-over effects of this kind. Even intensive research programs carried out by the U.S. army failed to find any marked fatigue effects of periods of up to 60 h of continuous work. Holding and others have showed that there are methodological difficulties in the analysis of this apparently straightforward problem.

As with the compensatory response to stressors, participants in such experiments appear able to work harder (make more effort) for brief periods to respond to the challenge of any new test, effectively compensating for any reduction in capacity. However, when tired people are provided with alternative ways of carrying out the postwork test they are more likely to choose one requiring low effort, even though it entails more risk of error.

Similar results of high workload and stressful jobs have been found for driving examiners, bus drivers, and junior doctors. This approach to fatigue reveals it to be a state in which there is a shift toward preferring activities requiring less effort or use of WM. So far, there has been little direct research on this form of decrement with laboratory stressors, although similar effects have been found for noise and high workload.

The link between stress and fatigue is a very strong one. It is likely that actively managing stress in order to protect performance leads directly to fatigue, so that recovery is necessary before we can function even when the stressor is no longer present. Recent work carried out in the Netherlands suggests that fatigue impairs the effectiveness of the executive control system that maintains the activation of tasks in working memory, triggering both the withdrawal of effort and compensatory changes in information processing strategy.

At present, we have no direct evidence of the brain processes involved in this stressor - control/ effort - fatigue chain, but it is currently being addressed in a number of laboratories. Clearly, a better understanding of the physiological basis of the control of stress during task performance will help us manage work and other tasks more effectively, as well as informing our approach to the design and management of the working environment.