Table of Contents

• Safety Cultures and the Importance of Human Factors, by: John K. Lauber, National Transportation Safety Board
• FAA Sponsored Research Planning Conference Report, by: Elaina Edens, PhD, Federal Aviation Administration
• Successful Communications for Maintenance, by: James C. Taylor & Michelle M. Robertson, University of Southern California
• Cockpit Crises & Decision Making: Implications for Pilot Training, by: Dr. Maureen Pettitt, Western Michigan University
• On the Design of Flight Deck Procedures, by: Asaf Degani & Earl Wiener

by:
John K. Lauber
National Transportation Safety Board
Washington, D.C.

The art and science of human factors has undergone significant transformation since its inception during the early years of World War II During the period when the war grew to encompass the globe, nations were faced with the necessity for a massive buildup of fighting forces. In particular, aviation technology had evolved to the point where aircraft of all kinds--transports, bombers, and fighters--were to play a crucial role in the conduct of both offensive and defensive warfare. But technology alone was not sufficient: these new, highly complex machines, manufactured by the hundreds of thousands, were only effective in carrying out their intended mission when maintained and operated by skilled mechanics, pilots, navigators, bombardiers, and a myriad of other technical support personnel. The necessity to select and train millions of people was the immediate driving force that led to the emergence of human factors--the application of scientific practices and principles to the study of complex, person-machine systems--a half century ago.

Although selection and training were primary issues at the time, it was quickly learned that this, too, was not sufficient to ensure an adequate relationship between complex systems and the people who operate then-L Selection and training focus upon the central question of how to best adapt people to machines. Although such adaptation is an absolute necessity, it was soon recognized that it was also necessary to attend to the converse issue: how to best adapt the machines to the abilities and capabilities of the people who must operate them? Ergonomics--the science of work--is that aspect of human factors that focuses upon such issues. These factors range from the simple and obvious such as the reachability of controls and displays to the more arcane and complex, including such concepts as "control-display compatibility" which recognizes that humans tend to perform best when there is some rational "mapping of control and display layout reflecting the cognitive, characteristics of the human operator.

In the years since WW II, these fundamental human factors domains have continued to evolve, and each of them--selection, training, and ergonomics--is an important component of modem human factors theory and practice. But we have also learned that even the successful application of these fundamentals does not ensure optimal person-machine system design and function. It has become increasingly clear that there are other important determinants of human performance in complex systems. Foremost among these is the fact that one must address not only upon the performance of individuals but also the performance of collections of individuals, or teams to optimize safety and efficiency. Although some early work was done on team performance in the early postwar period, it was not until roughly two decades ago that the formal emergence of such team-oriented approaches to the study and design of aviation systems occurred. Now applied with varying degrees of success on a nearly-global basis, crew resource management concepts at their very heart recognize that the safe and efficient operation of the modem aviation system depends upon well-designed (from a human as well as an aerodynamic point of view) aircraft operated by capable, well-trained individuals operating as highly integrated, well-coordinated teams. Fundamental individual skills and characteristics, such as leadership, decision-making, and communications are central to the successful achievement of such team performance, and are the elements from which present-day approaches to the training and evaluation of CRM in the aviation world (and more recently, in other domains as well) arc built.

There are many indications, both empirical and anecdotal, that the adoption of such team-centered approaches is paying off in terms of improved efficiency and system safety. But, good as they may be, even these are not sufficient; safety experts now recognize that there are even higher level determinants of human performance in the aviation system, and that on occasion, these too can cause 'dents just as inadequate individual and team performance does. The fundamental notion emerging here is that of the safety culture--the basic characteristics, shared values, and beliefs of an organization. These organizational, or cultural, determinants of human performance are increasingly recognized as powerful factors in the operation of modem, complex systems.

Several years ago, safety engineers at Boeing conducted a study that was stimulated by accident records showing that some airlines are clearly safer than others. In fact, the majority of hull losses and fatalities in commercial civil aviation result from accidents happening to a relatively small number of airlines. Based on extensive surveys and audits of both "good" and "bad" airlines--that is, those with safety records clearly much better, or worse, than average--the Boeing engineers drew several interesting conclusions. For example, it was noted that the "good" airlines invariably utilized CRM training (along with a closely-related simulation technique called Line Oriented Flight Training, or LOFT), thus confirming the importance and utility of these relatively new training approaches to aviation system safety. But in addition, they also noted that another factor seemed to be present in the "good" airlines; this they called, for want of a better term, "management commitment to safety." Airline safety, they recognized, begins at the very top of the corporation--indeed in the Board Room--and filters down to the very bottom thus infusing every player, no matter what role that individual plays in the functioning of the company, with a personal responsibility for safety. The successful establishment of such a safety culture is the greatest safeguard against accidents an organization can possess.

One implication of the concept of safety culture is that organizations, like individuals, can cause accidents. Increasingly, aviation accident investigative bodies, including the National Transportation Safety Board (NTSB), are formally including organizational components of accidents in their findings and safety recommendations. In some notable instances (for example, the investigation of the tragic capsizing of the Herald of Free Enterprise, a British ferry leaving Zeebrugge harbor, the investigation of a British Rail accident at Clapham Junction, near London, and the takeoff icing accident involving an Air Ontario F28 at Dryden, Ontario) the official reports are exemplary in their discussions of the roles not only of specific operational personnel, but also of high level management in the causation of the accident. At the NTSB, several examples can also be found illustrating the role of management and organizations in the genesis of accidents. For instance, the ground collision between a SkyWest Metroliner and a USAir B737 at Los Angeles several years ago happened "because" an air traffic controller forgot that she had cleared the SkyWest airplane onto the runway to hold for takeoff a couple of minutes before the approaching B737 called for landing clearance. However, the Safety Board's determination of probable cause cited failures on the part of FAA management to implement appropriate, redundant, air traffic control procedures at Los Angeles tower as causal. Numerous other accident reports illustrating such organizational causes for accidents have been issued by the Board and by other agencies in the past several years.

Although transportation system safety in general, and aviation safety in particular, have been the primary drivers for much of the "new" approach to human factors discussed above, it is clear that the utility of such concepts is not limited to the design, operation, and maintenance of vehicles, aeronautical or otherwise. Recent work has shown the applicability of many of these notions to such diverse areas as nuclear power plant design and operation, industrial processes, medicine, and, importantly for this audience, space operations. The fundarmntals--good ergonomic design of work stations and tasks, the selection of individuals with appropriate attributes, and the training of essential skills and knowledgc--are critically important in the field of nuclear operations as in aviation, as the landmark Three Mile Island accident amply demonstrated. But that accident, and others, also illustrate other critical aspects of human performance in any complex system, including the central importance of effective team work and the fundamental role of organizational culture in the determination of the level of risk posed by those systems.

In the classic report on the Herald of Free Enterprise disaster, it is stated, "All concerned in management ... must be regarded as sharing responsibility for the failure of management. From top to bottom the body corporate was infected with the disease of sloppiness" (emphasis added). Can there be a more eloquent statement of the role of organizational factors in the causation of system accidents?

by:
Elaina Edens, PhD
Federal Aviation Administration

The FAA Research and Development Service and Flight Standards sponsored a CRM research planning conference on August 16 and 17 in Newport Beach California. The meeting brought together researchers, air carrier operational representatives, and ALPA representatives to define future CRM research objectives. The participants examined past research accomplishments and the current state of CRM knowledge and implementation. A research agenda was established by the air carriers and ALPA. The prime topics center on automation, CRM assessment strategies and instructor/evaluator selection and training. A complete transcript and conference report will be available approximately November 15. The report will be sent to major air carrier training departments.

For additional copies E-mall your request to: EEdens@mail.hq.FAA.gov after November 15, 1994.

by:
James C. Taylor & Michelle M. Robertson
Institute of Safety & Systems Management
University of Southern California

If 1980's CRM was cockpit resource management, the 1990's CRM is crew resource management. In this decade, management concerns have expanded to include not only the cockpit or flight deck, but the passenger cabin, the hangar, the ramp, and the ticket counter as well. This paper deals with achieving and sustaining effective crew resource management in hangar and line maintenance.

The Three Pillars of Change.

Successful organizational change can be defined as long lasting positive effects on end results, which are diffused throughout a company. Wherever it is applied, CRM is becoming a fundamental source of organizational change and improvement in aviation. To be successful, organizational change (including CRM) requires three elements to be present:

1. unequivocal top management support and vision of the purpose for the change,
2. a well conceived and relevant intervention, and
3. timely appropriate feedback, through a broad range of measurement and evaluation activities.

The necessity of these three aspects in successful cockpit resource management training is well known and has been documented in the ICAO Digest #2: Flight Crew Training (1989).

The Challenge in Maintenance

Today's aircraft mechanics, inspectors and foremen face an extremely high-tech maintenance task. These individuals are successfully learning to work with fly-by-wire and by-light, the "glass cockpit BITE and other computer-based tools, as well as with composite structures, new repair procedures. They are also learning to work with increasing legal obligations and responsibilities. Maintenance people however, are delayed in their rapid deployment of these new lessons because they have not mastered how to successfully communicate. Unlike cabin personnel and gate agents, who self-select and are selected because of their skills and ability to interact with others, mechanics select their occupation because they want to work with tools and with large and complex machinery. But, the complexity of today's civil aviation industry demands that mechanics (as well as all other aviation occupations) learn to communicate effectively among themselves as well as with other groups. Just as CRM, can help maintenance employees change and improve, so the three pillars of change similarly apply to sustaining and diffusing maintenance CRM.

The Three Pillars of Change for CRM in Maintenance

1. Management support.

Successful change requires unequivocal top management support, or making CRM part of the culture. What happens when a dedicated, inspired Senior Technical Operations Vice President creates a CRM program for all 2,200 of his management and staff support personnel? A real-life case has shown that if that executive dedicates himself to that vision long enough, if he is persistent in its visible sponsorship, and if he is clear in his conviction that scientific evaluation of the program will improve its acceptance and continued development as well as validate his vision, then results actually happen. His maintenance managers begin to seriously value the open, assertive communication, safe work habits, and problem-solving methods that the program espouses. Those managers go on to believe (and report in surveys and interviews) that this program unlike most others they have experienced, really works, and they hope it will stay long enough "to make a difference."

2. Quality intervention.

Successful change requires a well conceived and relevant intervention. In this case the maintenance CRM program thus created was well planned and efficiently executed. It was based on the best of the lessons learned from flight-deck CRM training. The training was managed, administered and conducted by Technical Operations line managers who (not incidentally) also had good training skills. The training content and its illustrations were relevant to the work the participants do. In addition the training design required trainees to engage in active student-centered learning. As a result, they cheered the training as unparalleled in appropriate content, timely in the delivery of its message, and useful in its ability to be practically applied. As increasing numbers of people experienced the training and its effects became manifest, most of these participants recommended recurrent training, and many went on to endorse the wider use of the training throughout the company, including similar training for their mechanics and inspectors as well.

3. Feedback

Successful change requires timely, appropriate feedback -- through a broad range of measurement and evaluation -- for continual reinforcement and refinement. In this case, FAA and NASA funding was available to facilitate and enable novel and extensive data collection before and during the period of CRM maintenance training. Such support permitted appropriate scientific analysis of those data to test the effects of the training on maintenance effectiveness and on air worthiness and safety. The data collection, analysis, and reporting, were done with strict confidentiality by a neutral third party, ensuring enthusiastic and uncensored response to surveys and interviews. Given that those measures of CRM-relevant attitudes and behaviors were reliable and were related to measures of effective maintenance performance, important and interesting questions about the practical impact and further improvement of maintenance CRM training could be answered.

Experience with this maintenance system demonstrates that the three pillars of change act together to assure successful use of CRM training.

Some Case Specifics.

The training design. The program was the personal creation of the Vice President (VP) of Technical Operations who intended it for his departments' directors, managers, supervisors, assistant supervisors, engineers, analysts, coordinators, his other staff specialists, and eventually his inspectors. Following the company's model of flight deck CRM, this training was designed to improve safety and efficiency by identifying organizational norms, using assertive behaviors, understanding individual leadership, managing stress, improving problem solving/decision making, and enhancing interpersonal behavior. Technical Operations personnel conducted the training with assistance from professional training consultants. Each session began with a reference to the VP's direct interest and confidence in the program Training included a combination of short lectures, discussions in large and small groups, individual study, and role-playing with others. That training was delivered to virtually all 2,200 maintenance management and staff personnel (hereafter called "participants") over a period of two and one-half years. In order to accomplish this, the two day training sessions (the location rotated among three cities) were repeated weekly for groups of about 20 participants each.

The survey measurement. A customized version of the well-tested University of Texas CRM attitude questionnaire (CMAQ) became the core measure for participants. The survey was used to test the direct training effects of participants immediately preceding and following the training itself, as well as follow-ups at two, six, and twelve months later. The company provided the follow-up questionnaires to each participant, at the appropriate interval following their training, together with a letter from the VP expressing his continued support for the program and the importance of the survey for its evaluation. All completed questionnaires were mailed by participants directly to the University of Southern California, for data entry and processing. These "pre," "post" and "follow-up surveys" were used to obtain short-term, direct feedback and suggestions to the trainers for improving their program as well as gathering data for longer term evaluation of the attitude changes among the participants themselves.

Changes in attitudes showed more than a "honeymoon effect." The baseline attitudes of maintenance managers measured weeks before the program began were shown to be stable indicators of attitudes measured again for participants immediately prior to their training. Sizable and statistically significant improvements in attitudes about "communication and coordination," about the "value of managing stress," and about "delegating authority" were noted immediately following the training. In the months following training, participants' responses to the follow-up surveys showed that these attitudes remained stable and high. In addition, participants' attitudes about the value of assertiveness improved dramatically and significantly two months after their training, and remained high thereafter.

Participants were optimistic about the value of the training. immediately following training, participants reported that the experience would result in "moderate" to "large" changes in their behavior on the job. This result exceeds similar data from pilots, who typically report that CRM training will result in only "slight" to "moderate" changes in their behavior at work- In the months following their training the participants in the maintenance program were also asked to describe how they expected to use their training on the job, as well as how they actually did use it. These self-reports of intended and actual use included both passive and active behaviors. Over the months following training, intended and reported passive behaviors (such as "better use of active listening") were replaced by active uses such as "dealing better with others" and "being more assertive." This shift from passive to active use of the CRM training coincides with (and is explained by) the improved attitude toward assertiveness found in these participants' CMAQ results two months following training.

Improvements in maintenance performance occurred after the training began. A large number of robust and appropriate indicators of maintenance performance were collected monthly for a substantial period prior to and following the onset of the CRM training. Fourteen measures of Technical Operations performance were selected by the CRM trainers and graded as to their suitability for assessing the training. Trainers rated several measures as best, because they combined the following: they measured safety or dependability, they reflected the efforts of people by work unit, they measured changes resulting from human behavior, and they were largely independent of the effects of other performance measures. Post-training performance improved following the onset of training. Safety (measured by both ground damage incidents and occupational injuries) had been declining in the year prior to training, but both measures improved after the training began. Dependability (on-time performance) continued to improve post-training.

Improved attitudes about delegation, stress reduction, goal-sharing, and assertiveness are related to improved performance. By all accounts, the program is having the desired results. Participants' improved end-of-training attitudes about the value of managing stress and about the value of communicating and coordinating, and their reports of sharing work goals, are strongly and significantly associated with their work groups' subsequently improved scores for on-time departures (dependability). Participant attitudes about the value of assertiveness (which increased two months after their training) and of delegating authority are also significantly related to improved subsequent work group safety performance (both ground damage and occupational injury), as well as to improved dependability.

Maintaining and Diffusing Change -- Dancing with the Bear.

Can an effective program like this one be assured of continued success? Can it be stopped, and if so under what circumstances? The answers lie in all three fundamental elements of change. To keep up with the state of the art in CRM training, the program needs continual improvement. It can be improved in a variety of ways. For instance all mechanics and other wage- grade personnel can be trained in CRM. New material can be developed and provided as recurrent training to the participants from the original program If the program does not evolve it is likely to become a one-shot "project," the positive effects of which will be arrested.

If the program loses its management support, a gradual reversal in effects is likely. if that support is lost suddenly, the reversal is accelerated and measurable decay is virtually guaranteed.

The cessation of evaluation is the least important of the three pillars, and the absence of negative results during reversal may even prolong the illusion of continued progress. Without feedback, however, even the most effective program will gradually drift in the direction of presumed strengths which in reality can become the source of complacency and arrested development, if not decay and reversal. The effects of frequent third party measurement on program visibility and evidence of its continuance is an unmeasurable benefit at relatively small cost. The opportunity third-party measurement provides for participant (indirect) dialogue with trainers and top management also cannot easily be replaced.

In addition to the three pillars of change, we cannot escape the effects of the context of change -- the external environment. For example, changes in the market, customers, or competition can cause management to react with policies that seem to work at cross purposes with the involvement and open communication of CRM. Such ambiguity or conflicting policies require that familiar programs such as CRM are supported with even greater clarity. These programs deserve that clarity of focus, especially during environmental turmoil, if they are to survive and succeed.

What permanent effects do an organization's changes have on its members? Nothing is totally neutral or completely reversible. An effective program such as the one described here cannot survive without support, content, and feedback, but its unintended effects are not fragile either. The abandonment of such a program as this CRM in maintenance will result in discouragement if not a cynical despair. The resulting loss of employee motivation alone can have a considerable effect on maintenance performance. As the story goes, one cannot stop dancing with a bear merely because one wants to, but only when the bear is satisfied.

Research was conducted under a cooperative agreement between NASA Ames Research Center and the University of Southern California. Funding was provided by NASA, the Office of Life and Microgravity Sciences and Applications, and with contract funds from FAA Office of Aviation Medicine.

by:
Dr. Maureen Pettitt
Western Michigan University

Abstract

This paper presents the theoretical foundations and a description of a research study designed to examine pi 'lots' attitudes about cockpit crises and the processes used to make decisions in crises. The findings suggest that a "high crisis perception/low urgency/low rigidity" pattern may be an optimal approach to crisis decision making. In other words, the decision maker recognizes the situation as a crisis and is motivated to act, but the low sense of urgency encourages flexibility with respect to roles, responsibility, participation, and procedures. Suggestions for both ground training and simulator instruction am offered which expand situational awareness to include the concepts of crisis and available decision time.

Introduction

Crisis situations in the cockpit occur in a relatively complex, often ambiguous, sometimes hostile environment. However, the high achievement and task orientations of crew members seem to produce optimal responses in most situations that would be described as crises. Extensive technical training and highly structured emergency procedures provide pilots a means for diagnosing and responding to critical situations. Nevertheless, the history of commercial aviation is marred with many accounts of mismanaged crises.

The aviation community has long realized that the effective performance of cockpit crews is essential to aviation system safety. Early research in the area of flight crew performance focused primarily on skills acquisition and retention, perceptual requirements, and physical stress. Much less attention was given to the psychosocial aspects of the cockpit environment. Air transport accident analyses and related research during the past decade have, however, produced convincing evidence that pilot training and evaluation systems must address the crucial dimension of crew interaction and decision making in the cockpit (Foushee 1984; Ruffell Smith 1979; Sears 1986; Wheale 1984).

In recent years many airlines have initiated training programs to encourage effective cockpit resources management (CRM). Although CRM programs vary, they are essentially designed to educate pilots (and, more recently, cabin and ground crews) about the importance of interpersonal relations, communication skills, synergistic activity, and participatory decision making to safe flight operations training quite in contrast to the traditional focus on the technical means of accomplishing the goals of flight operations rather than the process.

Subsequent, evaluative research has supported the notion that CRM training can improve cockpit performance and, further, suggests that performance related attitudes are significant predictors of crew coordination in line operations (Helmreich, Foushee, Benson, and Russini 1986). However, research also indicates that crewmembers lack awareness of the deleterious effects of stress and have unrealistic attitudes about their personal vulnerability to stress (Helmreich 1984). Some researchers caution that during a crisis situation the crew is likely to revert to prior well-learned behaviors rather than the concepts espoused by CRM (Hackman 1987b).

It is expected, however, that over a period of time CRM training-- if an accepted and customary component of initial and recurrent training programs will encourage behaviors which lead to more effective coordination and decision making in critical or crisis situations (Hackman 1987b, Helmreich 1984; Helmreich and Wilhelm 1987).

Crisis and Group Decision Making

A crisis is triggered by incremental or radical changes in the environment. A crisis threatens the goals of the decision unit and involves risks with the potential for substantial losses. There is a relatively short time in which to make a decision before loss occurs. The crisis is further exacerbated by the lack of a contingency plan and of decision-relevant information. Not surprisingly, the psychological and physiological manifestations of stress accompany the crisis situation. In sum, the crisis situation is characterized by threat, uncertainty, limited information, time pressure and tension (Billings, Milburn and Schaalman 1980; Hermann 1969; Milburn 1972; Miller 1963; Staw, Sanderlands and Dutton 1981).

The psychological and physiological responses to crisis as well as individual differences in personality predispositions and perceptions affect the processes employed by the group, and, ultimately, decision outcomes. The evidence suggests that individual cognitive flexibility is restricted in a crisis. Decision makers may overemphasize similarities between the past and the present (Milburn 1972), gravitate toward simple decision rules (Janis and Mann 1977; Nagel 1988), rely on well-learned, dominant behavior (Zajonc 1965), suffer from a foreshortened time perspective (Milburn 1972) and be indiscriminate in the search for information or, conversely, ignore relevant information (Janis and Mann 1977).

In addition, the strategies utilized by the crew and the control structures within which decisions are made are determined to varying degrees by leadership--the leader's abilities, style, influence, and orientations toward the task and relational aspects of group performance. The evidence suggests that, in a crisis, leaders often tend to restrict information, participation and shared responsibility. These strategies are designed to reduce error, losses, and uncertainty and to encourage quick and decisive action. However, such strategies often result in information overload, omission, or distortion; role overload; and reducing the number of decision participants (Billings, Milburn and Schaalman 1980; Dutton 1986; Smart and Vertinsky 1977; Staw, Sanderlands and Dutton 1981).

Alternatively, decision making is likely to be more successful in cockpit crises if (1) group norms encourage shared information and responsibility for decision making (Foushee 1984; Foushee and Helmreich 1988), (2) the crew has confidence that a satisfactory solution exists and believes that there is sufficient time available to search for and evaluate alternative courses of action (Janis and Mann 1977)-, (3) the crew evaluates and utilizes available resources to develop alternatives, strategies or new resources, rather than relying upon existing strategies and resources to resolve the situation (Hackman 1987a); and (4) individuals are trained for cognitive flexibility under adverse conditions (Dutton 198 1).

Description of the Research

The purpose of the study was to examine pilots' perceptions about three constructs central to decision making in cockpit crisis situations--the perception of crisis, sense of urgency, and response rigidity. The survey materials included a crisis scenario, a two-part, nineteen-item questionnaire and a background information sheet. The scenario and questionnaire items had been pretested during personal interviews conducted with twenty-four Denver-based line pilots employed by a major airline.

Subjects

The subjects of the study were Los Angeles-based line pilots from three major airlines. Six hundred and fifty-seven surveys were distributed. One hundred and eighty-five usable surveys were returned and used in the analysis. The subjects represented a relatively broad cross-section of the pilot population (i.e., 46% were captains, 33.5% were first officers, and 20% were second officers; 38% were not yet forty years old, the other 62 were; 67% had no formal CRM training while the other 33% had).

The concept of crisis was measured in two ways. The perception of crisis was determined by asking pilots if they believed that this crew was in crisis situation (Question 1, Part 1). They were asked to respond on a Likert-type scale numbered 1 (strongly agree) through 7 (strongly disagree). In Part II of the questionnaire pilots were asked to rate five crisis characteristics of the scenario on a Likert-type scale numbered 1 (low) to 9 (high). The second measure of crisis is a combination of the mean ratings of four crisis characteristics--level of threat to the safety of the flight, level of situational uncertainty, availability of decision-relevant information, and the level of tension.

The perception of urgency--the perceived time available in which to search for and evaluate alternative courses of action--was measured by combining the responses to three questions in Part I in addition to the rating of the fifth crisis characteristic, level of time pressure, in Part II of the questionnaire.

Response rigidity is characterized by the restriction of participation, limiting the search for and evaluation of viable alternatives, and adherence to/reliance on authority and procedures. Nine questions in Part I of the questionnaire were used to determine response rigidity.

It was hypothesized that (1) the perception of crisis would have a positive correlation to the ratings of the crisis characteristics, (2) the higher the rating of the situation as a crisis, the higher the response rigidity, and (3) a high sense of urgency would result in high rigidity.

Results of the Study

Ninety percent of the pilots surveyed indicated, at some level of agreement, that the crew in the scenario was in a crisis situation. This crisis perception positively correlated with their ratings of the level of threat, uncertainty, information availability and tension--characteristics commonly attributed to crisis situations.

The hypothesis that a higher perception of crisis would result in higher response rigidity was not supported by the data. The results indicated that a higher perception of crisis resulted in a lower rigidity score (r=.18). Further, pilots had, overall, a lower rigidity score than expected.

It was also expected that a higher sense of urgency would result in higher response rigidity. Previous studies of decision making suggest that a high sense of urgency (the belief that there is little time to search for and evaluate alternative courses of action) may evoke dysfunctional decision making behavior characterized by the restriction of information, authority, and participation. The hypothesis that high urgency would result in high response rigidity was supported by the data. The higher the sense of urgency the higher the response rigidity (r=.25).

The findings indicate that the perception of crisis and response rigidity are negatively related (high crisis/low rigidity, low crisis/high rigidity) and the sense of urgency and response rigidity are positively related (low urgency/low rigidity, high urgency/high rigidity). Rigidity scores were not significantly different as a result of a high or low perception of crisis, however, pilots with a low sense of urgency had significantly lower rigidity scores than pilots with a high sense of urgency. These results suggest the interpretation that sense of urgency--the time component of crisis--drives decision making behavior, more so than the perception of crisis.

While the results of this study were not conclusive, they strongly suggest the possibility that the high crisis perception/low urgency/low rigidity pattern may be an optimal approach to crisis decision making. First, the decision maker recognizes the situation as a crisis. As a result of this awareness the decision maker experiences mild but "helpful" stress and is, consequently, motivated to act on the situation. The low sense of urgency--the belief that there is sufficient time to search for and evaluate alternative courses of action--encourages flexibility with respect to roles, responsibility, participation, and procedures.

Conversely, pilots exhibiting the high crisis/high urgency/high rigidity pattern may be behaviorally similar to people who suffer from the high, debilitating stress which ultimately inhibits performance. They may resemble the hypervigilant decision maker described by Janis and Mann (1977) whose high arousal state results in impaired cognitive functioning and narrowed time perspectives. The study similarly suggests (although less conclusively) that a high, or at least moderate, perception of crisis is antecedent to the low urgency/low rigidity response pattern. The present research makes a stronger case for low sense of urgency as antecedent to flexible, participatory decision making.

Implications for Pilot Training

The research lends additional support to the findings that cockpit resource management training improves attitudes toward crew coordination and decision making (e.g., Helmreich 1989). Pilots who had attended formal CRM training programs exhibited a significantly lower sense of urgency and significantly lower response rigidity than pilots who had no formal CRM training.

The results of this study also provide some possibilities for expanding or modifying content in pilot training programs in industry and those used in colleges and universities. While there are implications for several concepts central to most CRM training programs--decision making, communication, stress management (U.S. Federal Aviation Administration 1989) the concept of situational awareness is most relevant to the present discussion.

It has been argued that an accurate assessment of crisis is a necessary first step toward crisis resolution (Billings, Milburn and Schaalman 1980). In aviation, this notion is embodied in the concept of situational awareness. Situational awareness is generally considered to be the accurate perception of the factors and conditions that affect the aircraft. In other words, a pilot who is situationally aware has made an accurate assessment of reality. However, the concept of situational awareness as currently conceived focuses primarily on information processing and communication (e.g., Nagel 1988) with little reference to temporal structure or awareness.

The results of this study suggest that it would be beneficial to expand the concept of situational awareness to include, (a) the concept of crisis and, (b) the accurate assessment of decision time available in critical situations (as opposed to relying on an internally-perceived time frame).

Pilots should be aware that in a crisis time perspectives are constricted and short-term goals and consequences tend to be overemphasized.

One means of accomplishing this in a ground training session is the use of the critical incident technique. Flannagan (1954) used this technique successfully in the 1940s to collect observations of flight crew behavior (incidents) which significantly contributed to a specified outcome (were critical). The critical incident technique can be used to sensitize pilots to the importance of an accurate perception of time to situational awareness and informed decision making. For example, the instructor might ask pilots for a factual, detailed account of the behavior of the flight crew in a situation which could be considered--by them or by someone else--a crisis. The resulting list of behaviors could then be categorized along behavioral dimensions normally associated with CRM but with special emphasis on temporal perceptions and errors. The criteria used to establish the situation as a crisis could be similarly examineed.

An expanded concept of situational awareness is applicable to simulator training as well as ground training. Debriefing sessions could include evaluation of how the crew assessed the decision time available, the accuracy of that assessment, and how time perspectives affected the processes used to resolve the crisis. This approach is not limited to the full-motion, high-fidelity simulators utilized by major airlines. Such training might be accomplished just as effectively in low-fidelity simulators or using interactive video workstations and video recordings of behavior (e.g., Foushee and Helmreich 1988).

The roles, the skills, and the training required of 'lots are changing. The successful pilot of the next century will advance based not only on technical skills and total experience but on his or her management skills as well. Cockpit resources management is designed, as the name implies, to improve the management of resources in the cockpit. These resources are commonly categorized as people, information, and equipment. The present paper suggests that the concept of time should be added to this list of interrelated resources, especially in critical or crisis situations. Further research is also needed to clarify how patterns of information exchange and processing might encourage the accurate perception of crisis and assessment of decision time. The findings could contribute a great deal to understanding crew performance in crisis situations and provide valuable information relevant to CRM and other pilot training programs.

by:
Asaf Degani & Earl Wiener

Summary by:
Lou Nemeth

The following list of recommendations come from Degani & Wiener's NASA Contractor report 177642. The report is rich with information report on the need for an institutional approach in the design and development of flight deck procedures.

The report begins by referring to other studies that conclude that 33 percent of crew caused factors in 93 hull loss accidents were for "Pilot deviation from basic operational procedures." Why would a professional choose to deviate from published procedures?

Degani and Wiener offer these reasons among others:

• Individualism
• Complacency
• Frustration

Degani and Wiener offer that these underlying causes should be expected and respected. In some cases the adopted practices may be a better alternative.

What is needed is an institutional method that reduces the differences (delta) between management expectations (procedures) and crew actions (practices).

Guidelines to the establishment of an institutional method for the development of flight deck procedures (see graphic this page) are offered by Degani and Wiener in their report and offered to our readers here.

1. A feedback loop from line pilots to flight management and procedure designers should be established. This feedback loop should be a formal process, as opposed to an informal process. It must be maintained as a non-punitive, reactive system, with mandatory feedback from management to the initiating line pilot about the progress of his report and/or suggestion.
2. When designing procedures for automated cockpits, the designer should recognize that many tasks that involve the use of automation are too complex and interactive to allow a stringent set of SOPs to be mandated.
3. It is essential that management develop a philosophy of its operations. This is particularly important for operating automated cockpits.
4. When introducing new technology into the cockpit, the procedure designer should reevaluate all of the existing procedures and policies in light of the new technology and support the new technology via new procedures.
5. Management, through the feedback loop and the line check airman program should be watchful of techniques that arc used on the line. Techniques that conform to procedures and policies should not be interfered with. Techniques that have a potential for policy and procedure deviation should be addressed through the normal quality assurance processes. Techniques that yield better and safer ways of doing a task may be considered for SOP.
6. Care must be taken that not only the principal participants of a system (e.g., flight crews in this case), but also others that are affected (e.g., controllers, ground crews, cabin attendants) be involved and informed in the design and modifications of a system procedure.
7. Procedures must be tailored to the particularities of the type of operation. Ignoring these particularities can foster low compliance with procedures on the line.
8. The procedures designer must be mindful of the limitations and capabilities of the device he or she is designing a procedure for. Devices that are well designed for the human user require minimal procedurization. Less robust devices will require more thought on the part of the designer, and will probably require more complex and lengthy procedures.
9. Management must guarantee that any procedure is compatible with the engineering of the aircraft or any subsystem Care must be taken when there are subtle differences between aircraft (especially if these differences are invisible or difficult to detect). Incompatibility can be resolved by re-engineering.
10. Airframe manufacturers and component suppliers (such as avionics forms) must be attuned to general airline procedures. Knowledge of such procedures may influence ergonomic considerations.
11. The entire documentation supplied to the cockpit (and elsewhere) should be regarded as a system, and designed accordingly as a system, not a collection of independent documents. A clear and logical (from the user's view) structure for this system and a criterion for the location o different procedures is important. An effective index in each manual would go a long way toward aiding pilots in finding materials they seek, especially when it is an unfamiliar, obscure, or seldom accessed procedure.
12. Paperwork should be designed carefully to be compatible with the device for which it is intended. Particular care should be exercised in preparing materials for computer based systems. It may be necessary to provide differently formatted documents for different cockpit configurations.
13. Procedure design includes intra-cockpit communication. The expected communication should be specified, trained, and subject to standardization like any other procedure.
14. In managing automated cockpits, briefing becomes an critical crew coordination tool not so much to reduce variance, but rather to reduce the level of ambiguity of other agents (e.g., PNF or F/O) by increasing expectations. The more one allows for technique, the more one has to stress briefing.
15. If the same procedure can yield significantly different outcomes, then the procedure must be modified in order to eliminate its embedded ambiguity. In brief, a procedure should lead to a totally predictable outcome.
16. Particular attention should be paid in order to safeguard information transfer during critical and high workload phases of flight. Callouts should be economical, unambiguous, and should convey only the information needed by the other crew member(s). They should not distract the crew member from his primary task(s). Finally, we urge frequent review of callout procedures: as other procedures change, callouts should be reexamined.
17. Procedure designers should always bear in mind the contribution which any procedure makes to total workload of the crew at any given time. They should be especially sensitive to procedures that may require crew attention in times of high workload, and should strive to "manage" workload by moving tasks that are not time-critical to periods of low workload.
18. The designer of flightdeck documentation should search for situations where procedures are tightly coupled, and exploit the opportunity to decouple them.
19. Frequent procedure and checklist changes lead flight crews to conclude that the system is unstable. This may diminish the importance they attribute to new and modified procedure. Therefore, management should minimize frequent procedures or checklist changes. It is probably better to bunch them together and make larger, less frequent "bundles" of changes if the items are not time-critical.
20. The SOP documentation should not only explain the mechanics of the procedure, but also state the logic behind it. A detailed account of the operational logic, system constraints, and the link to the "Four P" model should be part of the documentation.
21. While benefits of cross fleet standardization are quite obvious, there are certain situations where this type of standardization is just inappropriate. It may lead to suboptimal procedures by superimposing procedures that are suitable for one type of cockpit operation on another.
22. We recommend a three-way approach for a cross-fleet standardization. (1) Development of a cross-fleet philosophy, (2) creating a crossfleet standardization forum, and (3) obtaining input for procedural design from personnel that design, certify, teach, use, and check procedures.
23. The flow of any procedure through design, training, checking, implementation, and finally feedback, must be supported by the organizational structure. When a new procedure, or a modified procedure is established, it should be closely monitored (by standardization and check airmen, and LOFT instructors) for compliance.

For a full copy of this report (highly recommended) contact Dr. Earl Wiener 305.284.6595.

The material contained in The CRM Advocate back issues is the property of the contributing editors. No duplication of any kind is authorized without the express written permission of the editor. All rights reserved. For training and information purposes only. The intent of the editors is shared information, through controlled distribution to the benefit of the safety of flight.

[ Home ] [ Up ] [ 93.1 ] [ 94.1 ] [ 94.2 ] [ 94.3 ] [ 94.4 ] [ 95.1 ] [ 95.2 ] [ 95.3 ] [ 95.4 ] [ 96.1 ] [ 96.2 ]