Individual and Team Decision Making Under Stress: Theoretical Underpinnings
Total Page:16
File Type:pdf, Size:1020Kb
2 Individual and Team Decision Making Under Stress: Theoretical Underpinnings Janis A. Cannon-Bowers and Eduardo Salas As previously discussed, the Tactical Decision Making Under Stress (TADMUS) program was designed to (a)define the decision problem facing Navy tactical teams, (b) develop measures of tactical decision making (TDM) performance, (c) collect empirical data to document the impact of stress on TDM, and (d) develop and test principles for decision support, information display, training system design, and simulation that would mitigate these stress effects. Given the complexity of the environment (see Collyer & Malecki, chap. 1, this volume), the project team was presented with a formidable challenge: how to weed through the universe of things that we could do to improve performance and select only those interven- tions that would yield the most significant gains. We were motivated both practically and scientifically to answer this question. On the practical side, it was clear that a program afforded the resources and visibility of TADMUS needed to provide some demonstrable results fairly quickly. More important, we were committed from the outset to develop and test interventions that would actually be implemented in operational settings -we did not have the time or money to suffer many failures. From the scientific side, we were determined to make best use of what we considered to be a tremendous opportunity-the mandate and resources to study a difficult, real-world problem in depth. This was an opportunity that does not present itself very often. Given this dilemma, we fell back to an old adage that traces back to Kurt Lewin: “There’s nothing more practical than a good theory” (Marrow, 1969). Our reasoning was that if we adopted a few strong theoretical po- sitions and generated promising hypotheses from these, our chances of success would be greatly enhanced. Therefore, several emerging areas of theory and research were exploited as a means to drive empirical work. The purpose of this chapter is to review briefly these theoretical perspec- tives. To do this, we first provide a description of the boundaries of the problem we were dealing with and offer definitions of key concepts. Next, we discuss the way we approached the problem of selecting theories that we believed could drive the specification of training interventions. Follow- 17 18 CANNON-BOWERS AND SALAS ing this, we describe the major theoretical perspectives we adopted for the program. Finally, we summarize the major hypotheses associated with these theories and describe briefly the training interventions that were developed from them. Boundaries of the Problem To select theoretical perspectives appropriate for the problem with which we were dealing, it was necessary to establish the boundaries of the prob- lem and delineate important definitions. In the following sections, we pro- vide a description of the decision environment and task, the nature of stressors characteristic of the environment, and major definitions. Defining the Decision Environment Demands on the human decision maker in military tactical environments are becoming more complicated with advances in technology and changes in the world order. Modern combat scenarios are often characterized by rapidly evolving and changing conditions, severe time compression, and high degrees of ambiguity and uncertainty. In addition, such situations often present the decision maker with an overwhelming amount of data and require the coordinated performance of a team of operators who must gather, process, integrate, communicate, and act on these data in support of a tactical decision. A variety of other stressors (both physical and psy- chological) also exist in the operational setting, not the least of which is the catastrophic costs of making an error, which mitigate against effective individual and group performance. Coupled with the fact that the modern military scenario is likely to be complex and multinational, these factors have provided unprecedented demands on the human decision maker. An example may help to illustrate this contention (see Johnston, Poirer, & Smith-Jentsch, chap. 3, this volume, for details). One of the tasks facing a team of operators in a Navy combat information center (CIC) is to defend the ship against hostile aircraft. This task is accomplished by a hierarchically structured team of operators/decision makers, with final de- cision authority retained by a single individual. Team members perform a variety of functions in support of the final decision: They operate sensor consoles to detect aircraft, integrate information collected regarding the aircraft’s intent, make decisions about how and when to seek additional information, and make decisions about how and when to transmit perti- nent situation assessment information. Once information is passed to a final decision maker, it must be considered against the situational con- straints (e.g., rules of engagement) and potential consequences before a decision can be reached or action taken. To function in such a situation, team members must understand the system at several levels. First, they must understand the dynamics and control of the equipment (both hardware and software) with which they INDMDUAL AND TEAM DECISION MAKING 19 are interacting to extract information. Second, they must understand the demands of the task and how to accomplish them (e.g., the significance of information, types of information required, strategies to combine infor- mation, necessary procedures, etc.). They must also understand how var- ious facets of the environment affect the task and task demands (e.g., when workload increases as a function of air traffic in the area, or when radar reception is affected by weather conditions). Third, they must un- derstand sound decision-making process so that they optimize the use of available information and avoid errors. Fourth, they must understand their role in the task, that is, what their particular contribution is, how they must interact with other team members, who requires particular classes of information, and so forth. Finally, to perform optimally, tactical team members must be familiar with the knowledge, skills, attitudes, preferences, and other task-relevant attributes of their teammates. This is because expectations for the behav- ior of their teammates will vary as a function of the individuals who com- prise the team. When working with a particularly competent teammate, for example, a team member may alter his or her behavior so that it is consistent with how he or she thinks that teammate will perform. Defining Stressors in the Environment There are many definitions of stress in the literature (e.g., Hogan & Ho- gan, 1982; Ivancevich & Matteson, 1980; Janis & Mann, 1977). For our purposes, we adopted the definition of stress offered by Salas, Driskell, and Hughes (1996) as “a process by which certain environmental demands . evoke an appraisal process in which perceived demand exceeds re- sources and results in undesirable physiological, psychological, behavioral, or social outcomes” (p. 6). Practically, our initial investigations involved extensive interviewing and observing of actual Navy crews. As a function of this effort, we found that the following characteristics, which can be defined as stressors, all appear to be present in the operational environ- ment: 0 multiple information sources 0 incomplete, conflicting information 0 rapidly changing, evolving scenarios 0 requirement for team coordination 0 adverse physical conditions 0 performance pressure 0 time pressure 0 high workhnformation load 0 auditory overloadhnterference 0 threat A second benefit of these interviews and observations was to deter- mine which stressors are likely to have an impact on decision-making per- 20 CANNON-BOWERS AND SALAS formance. Of these, it was necessary to select a subset of stressors that could be induced reliably and safely in experiments. It was decided that the following stressors were candidates for manipulation in experimental studies. Task-related stressors workloadhime pressure 0 uncertainty/ambiguity 0 auditory overload Ambient stressors auditory interference performance pressure 0 fatigudsustained operations Once potential stressors for manipulation were identified, a survey of the literature was conducted to determine the most common ways these stressors have been manipulated in past research. Manipulation tech- niques were assessed on the basis of reliability with which stress could be created, the face validity of the procedure, ethical concerns, safety, and feasibility. On the basis of this review and interaction with fleet personnel, the following initial stress manipulations were established. Workloadtime pressure. Workload and time pressure were manipu- lated using the tactical task (problem) scenario. Specifically, the number of tasks an operator performed simultaneously was increased as a manip- ulation of workload. Time pressure was defined as the time allowed to process a radar contact, and manipulated by changing the distance at which a potentially hostile contact was inserted into the scenario (ie., the closer the contact was to own ship on initial detection, the less time the team had to prosecute it). Uncertaintylambiguity. Uncertainty was described as the amount of information available regarding a contact’s intent. Under high uncertainty conditions less information was available, as, for example, when an iden- tification mode was not available on a contact. Auditory overload.