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.
10 Team Leadership and Crew Coordination Effective leadership and communications among crew members are nec- essary to ensure optimal team performance during a workload transition. It is necessary for the tank commander to direct the actions of the crew mem- bers to form an effective team and to coordinate the actions of his tank with those of other tanks and combat units. The tank environment is a classic example of a small work group embedded in a multigroup setting- a situa- tion demanding effective leadership and crew coordination. The critical issue is that crew performance must be considered as a team endeavor. While this statement seems to be belaboring the obvious, training in many, if not most, team activities focuses on individual tasks, and evaluation concentrates on individual performance and error. For ex- ample, operating teams are composed of surgeons, anesthesiologists, and supporting nurses for each. Medical training, however, concentrates on the specialty, with the implicit assumption that the subgroups will coordinate their activities effectively when they come together to deal with a patient. A growing body of evidence suggests that, when teams face sudden transi- tions from routine procedures to a medical emergency, coordination can break down and conflicts can occur (Howard et al., in press). Similarly, although flying a modern jet transport, patrol aircraft, or bomber is patently a team endeavor, the formal evaluation and certification of pilots have his- torically concentrated on individual proficiency. What is characteristically lacking is attention to processes by which teams as a unit accomplish their tasks. Within this context, the key to success in effecting a transition from a 229
230 WORKLOAD TRANSITION standby or vigilance status to an action (combat) posture is in keeping each crew member alert, fully apprised of the operational situation, and prepared for coordinated action. Maintaining a cohesive, efficient team is made more difficult in the tank environment by the fact that communications are degraded by the physical layout and habitability of the vehicle. As noted earlier, face-to-face interaction is generally precluded, with intracrew com- munications taking place over interphones and interunit communications by radio. Several channels of communications operate simultaneously, creat- ing a need to monitor and screen communications for relevant information. Chemical or nuclear alerts require MOPP gear, which further increases dif- ficulties in communications. Crews transitioning to combat may have spent several days in cramped, poorly ventilated, readiness conditions in or near their tank. It is possible that when the sudden need arises for effective, coordinated communications and action, crew members will be in less than peak physical and psychological readiness. Clearly, the success or failure of tank crews (and teams in general) in workload transition depends to a large extent on the flow of communication among the team members, how this flow is affected by stress, and how these effects, in turn, are moderated by personality qualities of the team leader and by the organizational structure. Somewhat relevant here is an extensive literature on group decision making and problem solving (see Davis, 1992, or Hastie, 1986, for a good synthesis). The vast majority of this research has been involved with fairly abstract tasks. Nevertheless, certain conclusions from this work are worth summarizing: (1) Generally, performance of a group solving a problem or reaching a decision falls somewhere between the average of the individual competen- cies of group members and the performance of the most competent mem- bers. Confidence in decision quality is generally higher than individual competence. (2) Where performance ranges between these limits depends on a num- ber of other factors: performance appears to be a function of (a) how clearly demonstrable is the correct solution, (b) the length of time the orga- nization remains constituted (improves), (c) the nature of factions that form within the group, (d) the size of the group (improves up to five members), and (e) the persuasiveness and status of more competent (improves) or less competent (degrades) members. (3) There is little evidence that different structural techniques, such as Delphi or brainstorming, systematically improve decision performance (Hastie, 19864. (4) There is some evidence that face-to-face settings are beneficial. It also appears that computer-mediated communications allow less dominant
TEAM LEADERSHIP AND CREW COORDINATION 231 group members to provide more input. Group decisions reached in this computer-mediated manner tend to be more extreme but not necessarily of higher quality (Kiesler and Sproull, 19924. As noted above, most of the studies have employed fairly abstract tasks (some of the work on jury verdicts is an exception), using naive subjects, in nonstressed conditions, and groups that were constituted specifically for the purposes of the experiment. These characteristics somewhat limit the generalizability of these conclusions to team performance in crisis, the fo- cus of this report. To address the latter domain, we turn instead to one environment the aircraft cockpit in which a substantial amount of rel- evant research has been conducted. Hence, the cockpit environment is the primary focus of this chapter. AVIATION RESEARCH FINDINGS ON LEADERSHIP AND CREW COORDINATION On preliminary examination, many issues seem common to tank crews and flight crews. One, of course, is long periods of passive vigilance, which may be supplanted by a high-workload period with high needs for effective, coordinated behavior. A second is a high volume of (often de- graded) multichannel communications and the need to comprehend and act on critical information embedded in extraneous transmissions. Also com- mon to many flight as well as tank operations are high levels of fatigue that may impair individual and group function. Finally, the psychological sense of danger and the possibility of fatality following an inflight emergency is certainly common with the tank crew facing battle. The physical environment of the tank is also clearly less benign and provides less opportunity for face-to-face interaction than does the aircraft. It is also possible that the decision-making and information processing tasks incumbent in tank warfare are less dependent on open communications among crew members, making the issues cited above less serious in this environ- ment. It is possible that the military chain of command and organizational structure may override individual and group idiosyncrasies and result in effective information transfer, even under highly stressful situations (or, conversely, may impede effective interaction). Despite these caveats, examination of crew performance issues in other military environments, such as shipboard combat information centers, and demanding civilian settings, such as nuclear power plant control rooms and hospital emergency rooms, leads to the tentative conclusion that team effec- tiveness depends heavily on effective resource management; that is, person- nel within the team share information effectively and are appropriately co- ordinated in their monitoring and task performance responsibilities. In fact,
232 WORKLOAD TRANSITION a root cause of the majority of accidents involving aircraft with multiperson crews has proved to be breakdowns in leadership and coordination among crew members resulting in flawed decision making and improper actions (e.g., Cooper et al., 1979~. Analytically we address the role of coordination or resource manage- ment in transition teams in this chapter by first describing crew resource management training. We then consider the influences on effective voice communications. Finally, we focus on the broader implications of research on coordinated activities within the team, addressing factors related to orga- nizational climate, leadership, and personality (e.g., Foushee, 1984; Foushee and Helmreich, 1988; Helmreich et al., in press; Helmreich and Foushee, in press). Many of the valid conclusions in these domains come directly from research in aviation; this domain therefore represents the focus of much of this chapter. CREW RESOURCE MANAGEMENT TRAINING The aviation community has responded to evidence that failures in team coordination have been implicated in a majority of commercial jet transport accidents by initiating formal training in communications and group coordi- nation, known generically as crew resource management (CRM) training. Early courses grew out of traditional management development training, but recent programs have evolved to deal with very specific behaviors and encompass full mission simulation (Line Oriented Flight Training or LOFT) designed to require high complex decision making and group coordination under conditions with a high level of time pressure (Butler, in press; Helmreich and Foushee, in press). This integrated approach to training is augmented by videotaping the simulator session and an intensive debriefing that allows crews to observe their own behavior (Butler, in press). Validation research in commercial aviation has shown highly signifi- cant changes in crew member attitudes and, more critically, positive changes in behavior in operational settings following training (Helmreich and Wilhelm, 1991~. Critical elements that determine the success of CRM training in- clude not only courses that address concrete behaviors rather than abstract, psychological constructs, but also strong organizational support for con- cepts taught and recurrent training accompanied by continuing feedback and reinforcement for the practice of effective teamwork (Chidester, in press; Helmreich and Foushee, in press). The success of CRM training in aviation has resulted in the adaptation of the approach to other settings in which effective teamwork is essential for mission success. These include operating room teams using simulated patients and videotaped interactions (Howard et al., in press), aircraft main- tenance groups (Taggart, 1990), and nuclear power plant control room teams.
TEAM LEADERSHIP AND CREW COORDINATION 233 The Federal Aviation Administration is also developing plans to adapt CRM training for air traffic controllers. Given the conceptual similarities, it would appear that the outcomes of tailoring these training approaches to the tank environment would have a high probability of success, not only in improving transitions from monitor- ing to action, but also in improving overall team effectiveness. This ap- proach would seem to fit naturally into the types of simulations accom- plished using SIMNET. VOICE COMMUNICATIONS Breakdowns in communications between and within aircrews have been documented as a major source of human error and potential disaster (Hawkins, 1987; Nagel, 1988~. The most salient example is the decision made by a KLM 747 pilot at the Tenerife Airport in the Canary Islands to proceed with a takeoff, despite the presence of a second jumbo jet still on the runway (Hawkins, 19871. The resulting collision led to over 500 fatalities. Three features of this accident are worthy of analysis because of their similarity to characteristics of the tank crew environments: (1) the auditory quality of the message was poor" a degradation that may be caused by high ambient noise or by electronic "clipping" of radio messages; (2) the communications were not face-to-face, thereby eliminating many of the nonverbal cues that have been documented to improve communications, particularly with de- graded speech (Chapanis et al., 1972; Kryter, 1972~; and (3) the environ- ment was stressful (although not involving the stress of combat). The flight crew was on the final leg of a long and fatiguing international flight, weather conditions (and visibility) were rapidly deteriorating, and there was obvious time stress to proceed with the decision action (takeoff) as soon as possible. These potentially degrading characteristics of communications may be balanced by a list of other factors that can foster good communications. Restricting vocabulary and standardization reduces the possibility of con- fusing messages, and introducing redundancy, by repeating key messages or key words or echoing auditory with visual displays, ensures that ambiguous or unexpected messages are not interpreted incorrectly. Many short mes- sages can be better understood by providing a redundant verbal context, like "your fuel is low," rather than "fuel low." The likelihood of misinterpreta- tion (and the resulting need for redundancy) is particularly high whenever a message involves conveying information that is unexpected, or negating a statement (e.g., do not proceed) as was the case with the KLM flight. In general, affirmative communications are better understood and more reli- ably received than negative messages (Wickens, 1992~. Using acknowledg- ments and read-backs can trap errors before they are executed (Helmreich and Foushee, 1988~.
234 WORKLOAD TRANSITION Within many multicrew teams, the issue of confusion of the message source is also a very real danger. A large research base in experimental psychology has documented the likelihood of confusion of the source of verbal and nonverbal auditory messages that may be heard simultaneously or in close temporal proximity (Hirst, 1986; Treisman, 1984; see Wickens, 1984, for a summary). Within the tank, this confusion could result between messages intended for any of the four crew members. For the tank com- mander, this confusion might well be enhanced by the need to monitor both intratank and intertank (i.e., company and battalian level) communications. Research data indicate that confusions of this sort will be lessened (but not necessarily eliminated) if different channels can be made physically dis- tinct by different voice characteristics, modulation, or apparent spatial lo- cation (Treisman, 1964) or by some other clearly discriminable (perhaps visual) cue identifying whom is speaking what. Flight Deck Communications Communication patterns among crew members have proven to be the Rosetta stone for understanding the nature of effective and ineffective flight crew performance. Because the flight deck is a constrained environment with well-defined crew roles and well-defined tasks, variations in the qual- ity, quantity, and nature of communications are relatively easy to isolate and quantify. Cockpit voice recorders provide a tangible record of communica- tions in accidents in which crew performance is implicated, and high-fidel- ity simulators allow experimental manipulation of critical factors and tests of hypotheses regarding determinants of performance. Foushee and Manos (1981) pioneered a methodology of coding the interpersonal interactions of flight crews from audio tapes of cockpit com- munications. Their seminal work grew out of an experimental simulation conducted in a Boeing 747 simulator (Ruffle-Smith, 1979~. This study consisted of a simulated mission flown by volunteer line crews from an airline's B747 fleet. During the two-segment flight there were several equipment malfunctions requiring the diversion of a scheduled trans-Atlantic crossing under deteriorating weather conditions. Foushee and Manos found that crews with a higher frequency of operational communications and informa- tion exchanges committed fewer operational errors and had a more even distribution of workload during critical phases of flight. Subsequent work has resulted in the development of coding schemata that allow precise delineation of the group processes involved in normal and emergency situations. These have been applied both to experimental simu- lations and to the cockpit voice recorder tapes and transcripts from aircraft accidents and incidents (Kanki and Foushee, 1989; Kanki et al., 1989~. These schemata allow specification of the content of information exchange,
TEAM LEADERSHIP AND CREW COORDINATION 235 sequences of communications among crew members, and shifts in commu- nication patterns during periods of high stress and/or workload. Work in progress by Predmore (1991) further categorizes communications in terms of action decision sequences (ADS), interactions dealing with particular tasks or decisions that face crews in normal and abnormal circumstances. Using data from accidents and from an experimental simulation flown by three-person crews in a NASA B727 simulator, he is exploring the patterns that emerge when crews have to deal with multiple ADSs requiring immedi- ate action. The data from all of these investigations indicate that there is great variability in communications among crews faced with the same stressful flight scenario and that effective crews exchange information and utilize available resources better. Systematic Observational Studies of Crew Performance A methodology has also been developed for the collection of reliable data on flight crew performance in both normal line operations and full mission training in flight simulators (LOFT: Line Oriented Flight Training; Butler, in press). In this methodology, expert raters observe crews and make real-time assessments of crew behavior and performance. Data are collected on a rating form, the NASA/UT Line/LOS (Line Operational Simulation) Checklist (Helmreich et al., 1990a, 1990b, 1991) that elicits Likert-scaled ratings of elements of communications practices that have been identified as critical determinants of team performance. Also recorded are comments on special circumstances and unusual reactions. In addition to global rat- ings of technical proficiency and crew effectiveness, nine human factors components are assessed: (1) conduct and quality of pre-event briefings; (2) effectiveness and openness of communications and processes of deci- sion making; (3) inquiry/assertion/advocacy the willingness of crew mem- bers to question proposed actions and decisions and to propose alternatives; (4) crew self-critique of decisions and actions; (5) leadership/followership; (6) interpersonal relations and group climate; (7) preparation/planning/vigi- lance; (8) workload distribution and avoidance of distractions; and (9) con- flict resolution (when occurring). Training in the use of the form centers on the concept of specifically defined behaviors (called behavioral markers) that represent effective enact- ment of the components of each of the nine elements. Data from more than 10,000 actual or simulated flight segments have been collected in several airlines and military units. Of particular concern is isolating the factors that trigger superior and poor overall performance. Comparable data from two airlines show similarities, but also organizational differences, in the deter- minants of performance. In one organization, the most frequent causes of poor ratings were (1) poor critique, (2) poor inquiry/assertion/advocacy,
236 WORKLOAD TRANSITION (3) ineffective conflict resolution, and (4) poor management of distractions. In a second airline, poor performance was associated with (1) poor inquiry/ assertion/advocacy; (2) poor management of distractions; (3) inadequate briefings; and (4) ineffective conflict resolution. In the first airline, supe- rior performance ratings were triggered by excellence in (1) preparation/ planningivigilance; (2) establishing a positive group climate; (3) briefings; and (4j conflict resolution. In the second airline, the most frequent causes were (1) group climate; (2) briefings; (3) conflict resolution; and (4) leader- ship/followership. Flight operations, particularly long flights involving extended periods of cruise at high altitude, are characterized by periods of low activity and system monitoring followed by transition to high workloads during terminal approach and landing. Inflight mechanical emergencies and changing weather conditions also lead to abrupt workload transitions. The factors cited above, as they determine overall crew performance, also determine readiness for workload transitions and the effectiveness of such transitions. Leader Behavior A common element in the observational data on crew performance and in analyses of the causes of accidents is the actions of the designated leader. The captain of an aircraft bears ultimate responsibility for the management of the flight deck, the provision of briefings, planning, management of workload, resolution of conflicts, and the climate of the group. Recogniz- ing that successful performance is a group endeavor requiring the coordi- nated activities of all members, the leader remains the most important single component. Two patterns of leader behavior have been isolated as causal elements in many accidents and sources of poor performance in experimen- tal simulations (see, e.g., Chidester and Foushee, 1988~. One pattern is characterized by autocratic behaviors that inhibit communication from sub- ordinates and result in a hostile group climate, leading subordinates to with- hold critical information, even when it is needed to avoid catastrophe. An- other reflects a leadership vacuum and failure to coordinate and guide the actions of group members. An important study by Ginnett (1987) examined leader behaviors surrounding initial crew briefings as they relate to crew performance. Ginnett observed airline crews from initial briefings through a three-day scheduled trip. The crews he followed were classified into two groups by independent expert observers. One group had captains who were rated as outstanding in terms of observed leadership and crew coordination. The second had captains who were rated as deficient on these dimensions. Ginnett found that he could predict the overall performance of the crews from behaviors manifested during the initial briefing. Effective captains consistently established a positive group climate and bases for open com
TEAM LEADERSHIP AND CREW COORDINATION 237 munications during the briefing. They also expanded the definition of the crew to include cabin personnel and stressed role definitions. Leaders of ineffective crews showed a variety of behaviors but failed to establish the crew as an integrated team. Leader Personality Attempts to isolate personality traits associated with effective leader- ship have resulted in a large literature with many conflicting findings (Hol- lander, 1985~. Part of the difficulty in isolating global factors stems from the fact that requirements for leadership are not common to all groups and are contingent on the group's structure and the nature of task to be accom- plished (Fiedler, 1964~. Efforts to use personality traits to select pilots (who ultimately become leaders of their flight crews) have similarly had mixed outcomes (e.g., Dolgin and Gibb, 1988~. Problems with the predictive validity of personality mea- sures in aviation may stem from the nature of the performance criterion employed. The overwhelming majority of the research on pilot selection utilized either performance in initial training or success or failure in com- pleting training as the criterion measure. Research by Helmreich et al. (1986b) found personality to be a poor predictor of job performance imme- diately after completion of training but to have significant relationships with objective performance measures after six and eight months on the job. This phenomenon has been labeled "the honeymoon effect," and it is inter- preted as reflecting the fact that initial motivation to obtain a desired posi- tion may initially overcome the influence of personality on performance. With the passage of time and familiarity with the work role, however, the underlying relationships between personality and performance emerge. This implies that, if the criterion is the operational performance of experienced pilots, personality may be a valid predictor of leader and crew effective- ness. Recent research supports this view. A set of traits reflecting positive and negative manifestation of two broad, orthogonal dimensions has been used in research with flight crews (Spence and Helmreich, 1978; Helmreich and Spence, 1978~. The first dimension consists of traits associated with achievement motivation and instrumental goals (with the negative compo- nent reflecting an autocratic, dictatorial orientation). The second consists of traits defining expressivity and interpersonal sensitivity (with the nega- tive components reflecting either subservience and passivity or verbal ag- gression). To reduce the personality battery to a smaller set of categorical factors and to reflect the distribution of component variables in the research popu- lation, cluster analytic techniques were employed to isolate frequently oc- curring constellations of traits (Chidester et al., 1991~. In a validation of
238 WORKLOAD TRANSITION these personality constellations and their relationships with flight crew per- formance, Chidester and Foushee (1988) conducted a two-day experimental simulation in a Boeing 727 simulator using line airmen assigned to this aircraft. Three experimental groups were formed based on the personality cluster assignment of the captain. The first consisted of crews led by a captain with high levels of positive, instrumental traits and high levels of positive, expressive traits (a group labeled as having the "right stuffy. The second group had captains from a cluster characterized by high scores on negative instrumental traits (e.g., domineering, autocratic) and low levels of positive expressivity (labeled as having the "wrong stuff''). The third group of captains were low on both positive instrumental and positive ex- pressive dimensions (labeled by default as having "no stuff". Each crew flew a two-day simulation involving five flight segments with two having high workloads, adverse weather, and mechanical prob- lems. Significant differences in performance (based on expert ratings and objective measures of errors) were found as a function of the leader's per- sonality. Crews led by "right stuff" captains performed best across all flight segments; those led by "no stuff,' leaders performed worst under all condi- tions. Crews led by "wrong stuff'' captains performed badly on initial segments but showed much improvement in performance by the second-day segment involving high workload and mechanical abnormalities. A theo- retical explanation for the latter finding is that crew members became able, over time, to cope with the domineering leader who had strong achievement motivation although he was lacking in interpersonal skills. No interper- sonal strategies appear to have allowed crews to overcome the problems caused by captains lacking in both achievement motivation and interper- sonal skills. These personality dimensions have also been shown to relate signifi- cantly to the acceptance of training in crew coordination concepts as mea- sured by changes in attitudes regarding appropriate leadership and interper- sonal communications (Chidester et al., 1991; Helmreich and Wilhelm, 19893. Those high on both instrumental and expressive dimensions ("right stuff") show the strongest positive, attitude change. Those in the other two clusters showed much less change, and there was evidence that training actually had a regression effect for those low on both dimensions ("no stuffy. A theo- retical explanation for the latter finding is that those lacking in attributes whose importance is stressed in leadership and crew coordination training are threatened and respond defensively by changing their attitudes away from the direction advocated. It is obvious that the personalities of junior crew members also play a role in determining the overall effectiveness of work groups. The multiplic- ity of combinations possible even in very small groups has had limited empirical investigations and the research base remains sparse.
TEAM LEADERSHIP AND CREW COORDINATION Automation, Leadership, and Crew Coordination 239 The introduction of more automated flight management systems has created major changes in flight operations. This has resulted in a reduction in crew complement from three to two, even in widebody aircraft involved in long transoceanic flight. Empirical data are just beginning to accumulate regarding the impact of automation on leadership and crew interaction (Wiener, 1988, in press). One finding is clear: automation, especially with a re- duced crew complement, results in a redistribution of workload and may lead to higher levels of workload, especially when reprogramming of flight management computers is required during critical phases of flight such as airport approaches (see also Chapter 3~. On a theoretical level, automation of the flight deck may shift the balance of authority away from the captain. In many instances, the copilot may be assigned primary responsibility for programming the flight management system. This can have the unintended effect of shifting information control and, hence, de facto control of the flight to the more junior crew member. Another outcome of automation is a tendency, implicated as causal factor in several accidents involving automated flight systems, to allow the computer to retain control even in situations in which reversion to manual flight is indicated by evidence of computer malfunction or excessive repro- gramming requirements. In some ways this phenomenon, which has been labeled "automation complacency," is another form of erosion of leader- ship. Systematic studies of crew behavior in automated aircraft are just be- ginning. A recent NASA study contrasted the performance of crews flying the same scenario during an experimental simulation flown either in a con- ventional DC-9 or an automated MD-88 (Wiener et al., 1991~. The results provide some justification for concern over the impact of high levels of automation. Although there were generally few differences in performance of the highly automated (MD-88) and less automated (DC-9) crews in their effectiveness in dealing with simulated inflight crisis situations, those dif- ferences that were observed tended to favor the DC-9 crews. It is clear that prior to finalizing the design of automated systems either for aircraft or for tanks, simulations that examine the dynamics of crew interaction and per- for~ance should be conducted and the results of such investigations should guide hardware development and crew training practices. Organizational Cultures and Subcultures Another factor that has been implicated in crew coordination is the culture of the organizations or subunit within organizations in which crews operate (Hackman, 1987~. Embedded in the larger study of crew perfor
240 WORKLOAD TRANSITION mance have been surveys designed to measure crew members' attitudes regarding optimal crew management (Cockpit Management Attitudes Ques- tionnaire (CMAQ): Helmreich, 1984; Gregorich et al., 1990~. Data have now been collected from more than 20,000 crew members in military and civilian organizations in the United States and Europe. The data indicate that, even in a highly regulated environment, organizations and units within them (for example, particular aircraft fleets or bases) show highly signifi- cant differences in attitudes indicating the presence of unique norms regard- ing appropriate behavior that may or may not have formal sanctions. These attitudes have operational significance and have been validated as predic- tors of crew performance in line operations (Helmreich et al., 1986a). In addition, systematic observational research confirms significant differences in crew performance between aircraft fleets within particular organizations (Helmreich and Wilhelm, 1989; Helmreich and Foushee, in press; Clothier, 1991~. Ethnographic research into organizational cultures and their develop- ment and manifestations in crew behavior is under way under a cooperative agreement between J. Richard Hackman (of Harvard University) and NASA. One ultimate outcome of this research should be a better understanding of how to change such norms in order to optimize crew performance. The issues involved are clearly relevant to the tank corps of the Army as evi- denced by a review of culture and military performance by Tamir and Kunda (1987). ENGINEERING MODELS OF COORDINATION One promising avenue of research, addressing organizational factors in a nonaviation team performance environment, has been the program of re- search on distributed decision making carried out by Kleinman and his colleagues at the University of Connecticut (Kleinman et al., 1992; Miao et al., in press). Their approach departs from much of the research on group processes by adopting an engineering-oriented quantitative modeling per- spective. Examining teams of operators who may be in different physical locations (i.e., not in face-to-face contact) and must solve a common prob- lem with different amounts of information available, they have formed a number of tentative conclusions with direct relevance to the current issues. For example, they find that human operators within a team tend to overvalue their own information and tasks relative to others (Kleinman et al., 19921; they observe the particular value of a coordinating leader in times of stress induced by information overload (Miao et al., in press); they document substantial differences in team performance induced by differing perceptions of team goals that may be held by members at different vertical
TEAM LEADERSHIP AND CREW COORDINATION 241 levels of the hierarchy; and they note the particularly debilitating effects of uncertainty on team performance. These effects are complex, and the conclusions, being recent, await further replication. The approach, however, represents a much-needed di- rection of research, modeling performance limits of mission-oriented teams. Such research, with its controlled, laboratory-based approach, can be used to provide convergence with the more observational techniques employed by Helmreich and his colleagues in the air crew domain. CREW PERFORMANCE RESEARCH The preceding discussion reveals that a series of related actions to as- sess the need for formal training in leadership and crew coordination are recommended. Additional areas for action include assessment of the social psychological impact of automation and reduced crew complement and in- vestigation of the role of personality factors as determinants of crew perfor- mance. First, an evaluation of the extent and nature of leadership and crew coordination problems in tank crews is needed. A number of methodologi- cal approaches can be used to obtain a valid representation of leadership and crew coordination issues in current battle tanks. These include: (a) An examination of archival operational records from a social psy- chological perspective to determine the extent of leadership, communica- tions, and crew coordination problems. Especially important is the isolation of instances of extremely effective and ineffective crew performance, simi- lar to that which has been undertaken with aircrews (see Ruffle-Smith, 1979). (b) A survey of crew members to determine normative attitudes regard- ing leadership, crew coordination, and personal capabilities under stressful . . cone ltlons. (c) An assessment of communications patterns and their operational implications for tank operations using the research paradigms employed in aviation. A useful research strategy would be to obtain recordings of crew communications (intra- and interunit) from a sample of tanks in both field operations and simulations. Using adaptations of coding schemata em- ployed in flight crew research, it should be possible to analyze communica- tions to evaluate leadership behaviors, gaps in communication, failures in information transfer, and situation analysis and awareness and decision pro- cesses and to relate these to indices of unit performance. Comparative analyses of communications protocols from differing organizations (e.g., European-based versus continental United States-based units) should pro- vide some information on the possible existence of different organizational cultures. In particular, protocols should be analyzed to determine the exist
242 WORKLOAD TRANSITION ence of performance, communications, and crew coordination problems during workload transition situations. (d) Experimental simulations investigating leadership and crew coordi- nation. An investigation of the effects of tank automation and complement reduction on crew communications and performance is required. In addi- tion to more traditional human factors analyses of crew task performance in more automated units, it is vital to determine how crew interactions change as a function of complement reduction. It is becoming increasingly clear that this area was not sufficiently investigated with the introduction of so- phisticated automation on the flight deck. The most viable approach would be to conduct full mission simulations (including interunit communications) focusing on the crew interactions, vigilance, decision making, and task per- formance, using as subjects crew members with experience levels represen- tative of the tank force. The simulations should be of sufficient duration to allow assessment of behavior during workload transition periods. Refine- ment of coding schemata for analysis of intracrew and interunit communi- cations should allow evaluation of the impact of hardware and staffing changes on group processes and performance. It is important to note that the fourth crew member in current opera- tional tanks may fill a number of critical roles affecting operational effec- tiveness outside combat (e.g., watch standing, maintenance, etc.~. In as- sessing the operational significance of complement changes, it is critical to examine all the operational implications of such changes. In particular, the additional crew member could be more beneficial in workload transition situations. Second, an evaluation of the utility of formal training in leadership and crew coordination should be conducted. Should the existence of significant problems in crew coordination and communications be confirmed through this research paradigm, the next logical step would be the adaptation of existing training approaches in crew coordination to the tank environment and experimental investigation of whether such training influences commu- nications processes and, most critically, crew performance criteria. (It should be noted that the Army is implementing this type of training for rotorcraft crews at the present time.) The training should focus on the particular behavioral skills required for effective team functioning and should provide clear exemplars of positive and negative leadership and interactions. Tests of the efficacy of such training should include opportunities for crew members receiving formal instruction, to practice the concepts learned, and to receive feedback on their behavior. A conceptual analog to LOFT in aviation is needed with operational scenarios designed to require effective communications and coordination for mission accomplishment. While cur- rent simulation training may serve some of these functions, experiences
TEAM LEADERSHIP AND CREW COORDINATION 243 drawn from the incorporation of LOFT into flight training indicate that, to be effective, the simulation scenarios must be designed to exercise the par- ticular skills stressed in the formal training programs. Those charged with administering the simulation training should be given special training in evaluating and debriefing crew coordination and leadership. Utilizing video or audio tapes of crew interaction for both research and crew feedback can greatly enhance the impact of the experience. Finally, the role of personality as a determinant of tank commander and crew member performance should be investigated. Recent evidence from civil and military aviation suggests the importance of personality, especially that of the leader, as a determinant of crew performance. It is recommended that the operational significance of personality be investigated, first at the level of the tank commander. Assessment of personality constellations of a sample of tank commanders using validated measures in a nojeopardy situ- ation should be followed by exploration of the relationships between these factors and leadership behaviors and crew interaction patterns. This evalua- tion could be conducted both in field operations and experimental simula- tions. Should it be determined that personality accounts for a significant component of performance variance, the utility of several alternative strate- gies can be evaluated. One strategy involves designing procedures to select in for optimal characteristics. This strategy works well when there is a substantial pool of qualified candidates. The alternative strategy, more fea- sible when there is a limited pool of candidates, is to screen out those with personality constellations associated with poor leadership and ineffective crew performance. Logical follow-on research would investigate the importance of the per- sonalities of junior crew members as determinants of the crew's perfor- mance. This leads into investigation of the importance of the mix of per- sonalities comprising effective and less effective crews. As noted, investigations in this area are much more exploratory as the research base is limited. REFERENCES Butler, R.E. In press Line Oriented Flight Training: Full mission simulation as Crew Resource Man- agement training. In E.L. Wiener, B.G. Kanki, and R.L. Helmreich, eds., Cockpit Resource Management. San Diego, California: Academic Press. Chapanis, A., R.B. Ochsman, R.N. Parrish, and G.D. Weeks 1972 Studies in interactive communications I. Human Factors 14:487-509. Chidester, T.R. In press Critical issues for CRM training and research. In E.L. Wiener, B.G. Kanki, and R.L.Helmreich,eds.,Cockpit Resource Management. San Diego,California: Academic Press. Chidester, T.R., and H.C. Foushee 1988 Leader personality and crew effectiveness: Factors influencing performance in
244 WORKLOAD TRANSITION full-mission air transport simulation. In Proceedings of the 66th Meeting of the Aerospace Medical Panel on Human Stress Situations in Aerospace Operations. The Hague, Netherlands: Advisory Group for Aerospace Research and Develop- ment. Chidester, T.R., R.L. Helmreich, S.E. Gregorich, and C. Gels 1991 Pilot personality and crew coordination: Implications for training and selection. International Journal of Aviation Psychology 1:23-42. Clothier, C. 1991 Behavioral interactions across various aircraft types: Results of systematic obser- vations of line operations and simulations. In R.S. Jensen, ea., Proceedings of the Sixth International Symposium on Aviation Psychology. Columbus, Ohio: Ohio State University. Cooper, G.E., M.D. White, and J.K. Lauber 1979 Resource Management on the Flight Deck. NASA Conference Publication No. 2120. Moffett Field, California: NASA-Ames Research Center. Davis, J.H., ed. 1992 Special issue on group decision making and problem solving. Organizational Behavior and Human Decision Processes 52. Dolgin, D.L., and G.D. Gibb 1988 A Review of Personality Measurement in Aircrew Selection. DJAMRL Monograph no. 36. Pensacola, Florida: Naval Aerospace Medical Research Laboratory. Fiedler, F.E. 1964 A contingency model of leadership effectiveness. In L. Berkowitz, ea., Advances in Experimental Social Psychology, Volume 1. New York: Academic Press. Foushee, H.C., and R.L. Helmreich 1988 Group interactions and nightcrew performance. Pp. 189-227 in E.L Wiener and D. Nagel, eds., Human Factors in Modem Aviation. New York: Academic Press. Foushee, H.C., and K.L. Manos 1981 Information transfer within the cockpit: Problems in intracockpit communica- tions. In C.E. Billings and E.S. Cheaney, eds., Information Transfer Problem in the Aviation System. NASA Technical Paper No. 1875. Moffett Field, California: NASA-Ames Research Center. Ginnett, R.O. 1987 First Encounters of the Close Kind: The First Meetings of Airline Flight Crews. Ph.D. dissertation, Yale University. Gregorich, S., R.L. Helmreich, and J.A. Wilhelm 1990 The structure of cockpit management attitudes. Journal of Applied Psychology 75(6) :682-690. Hackman, J.R. 1987 Group level issues in the design and training of cockpit crews. In H.W. Orlady and H.C. Foushee, eds., Cockpit Resource Management Training: Proceedings of the NASA/MAC Workshop. Report No. CP-2455. Moffett Field, California: NASA Ames Research Center. Hastie, R. 1986 Review essay: Experimental evidence on group accuracy. Pp. 129-164 in B. Grofman and G. Owen, eds., Information Pooling and Group Decision Making. Greenwich, Connecticut: JAI Press. Hawkins, F.H. 1987 Human Factors in Flight. Brookfield, Vermont: Gower Press. Helmreich, R.L. 1984 Cockpit management attitudes. Human Factors 26:583-589.
TEAM LEADERSHIP AND CREW COORDINATION 245 Helmreich, R.L, T.R. Chidester, H.C. Foushee, S.E. Gregorich, and J.A. Wilhelm In press How effective is cockpit resource management training? Issues in evaluating the impact of programs to enhance crew coordination. In T.R. Chidester, ea., Emerg- ing Issues for the Second Decade of Line Oriented Flight Training: Proceedings of a Workshop Held at the Pan American International Flight Academy. Helmreich, R.L., and H.C. Foushee 1988 Flightdeck communications. In E. Wiener and D. Nagel, eds., Human Factors in Aviation. Orlando, Florida: Academic Press. In press Why Crew Resource Management: The history and status of human factors train- ing programs in aviation. In E.L. Wiener, B.G. Kanki, and R.L. Helmreich, eds., Cockpit Resource Management. New York: Academic Press. Helmreich, R.L., H.C. Foushee, R. Benson, and R. Russini 1986a Cockpit management attitudes: Exploring the attitude-perfo~mance linkage. Avia- tion, Space and Environmental Medicine 57:1198-1200. Helmreich, R.L., L.L. Sawin, and A.L. Carsrud 1986b The honeymoon effect in job performance: Delayed predictive power of achieve- ment motivation. Journal of Applied Psychology 7 1 :1 085- 1 088. Helmreich, R.L., and J.T. Spence 1978 The work and family orientation questionnaire: An objective instrument to assess components of achievement motivation and attitudes toward family and career. JSAS Catalog of Selected Documents in Psychology (MS 1677, Vol. 8). Austin, Texas: University of Texas. Helmreich, R.L., and J.A. Wilhelm 1989 When training boomerangs: Negative outcomes from Cockpit Resource Manage ment programs. Proceedings of the Fifth Aviation Psychology Symposium. Co lumbus, Ohio: Ohio State University. 1991 Outcomes of crew resource management training. International Journal of Avia tion Psychology 1:287-300. Helmreich, R.L., J.A. Wilhelm, S.E. Gregorich, and T.R. Chidester 1990a Preliminary results from the evaluation of Cockpit Resource Management Train- ing: Performance ratings of flightcrews. Aviation, Space, and Environmental Medicine 6 1 :576-579. Helmreich, R.L., J.A. Wilhelm, J.E. Kello, and W.R. Taggart 1990b Reinforcing and Evaluating Crew Resource Management: Chock AirmanlLOFT Instructor Reference Manual. Technical Manual 90-1. Austin, Texas: NASA/ University of Texas. Helmreich, R.L., J.A. Wilhelm, J.E. Kello, W.R. Taggart, and R.E. Butler 1991 Reinforcing and Evaluating Crew Resource Management: EvaluatorlLOS lnstruc tor Reference Manual. Technical Manual No. 90-2. Austin, Texas: NASA/Uni versity of Texas. Hirst, W. 1986 Aspects of divided and selected attention. In J. LeDoux and W. Hirst, eds., Mind and Brain. New York: Cambridge University Press. Hollander, E. 1985 Leadership and power. In G. Lindzey and E. Aronson, eds., Handbook of Social Psychology, Vol~cme 2. New York: Random House. Howard, S.K., D.M. Gaba, K.J. Fish, G. Yang, and F.H. Samquist In press Anesthesia crisis resource management training: Teaching anesthesiologists to handle critical events. Aviation, Space, and Enrironmental Medicine. Kanki, B.G., and H.C. Foushee 1989 Communication as a group process mediator of aircrew performance. Aviation, Space, and Environmental Medicine 60:402-410.
246 WORKLOAD TRANSITION Kanki, B.G., S. Lozito, and H.C. Foushee 1989 Communications indexes of crew coordination. Aviation, Space, and Environmen- tal Medicine 60:56-60. Kiesler, S., and L. Sproull 1992 Group decision making and communications technology. Organizational Behav- ior and Human Decision Processes 52:96-123. Kleinman, D.L., P.B. Luh, K.R. Pattipati, and D. Serfaty 1992 Mathematical models of team distributed decision making. In R.W. Swezey and E. Salas, eds., Teams, Their Training and Performance. New York: Ablex. Kryter, K.D. 1972 Speech communications. In H.P. Van Cott and R.G. Kinkade, eds., Human Engi- neering Guide to Systems Design. Washington, DC: U.S. Government Printing Office. Miao, X., P.B. Luh, and D.L. Kleinman In press A normative-descriptive approach to hierarchical team resource allocation. IEEE Transactions on Systems, Man, and Cybernetics 22(3). Nagel, D. 1988 Human error in aviation operations. In E. Wiener and D. Nagel, eds., Human Factors in Aviation. Orlando, Florida: Academic Press. Predmore, S. 1991 Communications and multi-task processing on the flightdeck. In B. Kanki, ea., Approaches to Studying Group Productivity and Group Processes Using High Fidelity Flight Simulators. Technical Memorandum. Moffett Field, California: NASA-Ames Research Center. Ruffle-Smith, H.P. 1979 A Simulator Study of the Interaction of Pilot Workload with Errors, Vigilance, and Decisions. NASA Technical Memorandum No. 78482; A-7354. Moffett Field, California: NASA-Ames Research Center. Spence, J.T., and R.L. Helmreich 1978 Masculinity and femininity as personality dimensions. Society for the Advance- ment of Social Psychology Newsletter 4:2-3. Taggart, W.R. 1990 Introducing CRM into maintenance training. Proceedings of the Third Interna- tional Symposium on Human Factors in Aircraft Maintenance and Inspection. Washington, DC: Federal Aviation Administration. Tamir, B., and G. Kunda 1987 Culture and military performance. Background paper for D. Druckman and J.A. Swets, eds., Enhancing Human Performance. Issues, Theories, and Techniques. Washington, DC: National Academy Press. Treisman, M. 1984 A theory of criterion setting: An alternative to the attention band and response ratio hypotheses in magnitude estimation and cross-modality matching. Journal of Experimental Psychology 1 13(3):443-463. 1964 Verbal cues, language, and meaning in selective attention. American Journal of Psychology 77 :206-219. Wickens, C.D. 1984 Processing resources in attention. Pp. 63-102 in R. Parasuraman and D.R. Davies, ea., Varieties of Attention. San Diego, California: Academic Press. 1992 Engineering Psychology and Human Pe7 formance. New York: Harper Collins. Wiener, E.L. 1988 Cockpit automation. Pp. 189-227 in E.L. Wiener and D. Nagel, eds., Human Factors in Aviation. New York: Academic Press.
TEAM LEADERSHIP AND CREW COORDINATION 247 In press Human factors of the high technology cockpit. In Proceedings of the ICAO Hu- man Factors Seminar, 1990. Leningrad, USSR. Wiener, E.L., T.R. Chidester, B.G. Kanki, E.A. Palmer, R.E. Curry, and S.E. Gregovich 1991 The Impact of Cockpit Automation on Crew Coordination and Communication: I. Overview, Loft Evaluations, Error Severity, and Questionnaire Data. Contractor Report no. 177587. Moffett Field, California: NASA.
