Hollnagel, E. (2003). (Ed.), Handbook of cognitive task design. Mahwah, NJ: Erlbaum.

Erik Hollnagel (CSELAB,Department of Computer and Information Science, University of Linköping, Sweden)

Abstract. The design of artefacts, and in particular artefacts involving computing technology, is usually focused on how the artefact should be used. The aim of Cognitive Task Design (CTD) is to go beyond that by emphasising the need of considering not only how an artefact is used, but also how the use of the artefact changes the way we think about it and work with it. This is similar to the envisioned world problem, i.e., the paradox that the artefacts we design change the very assumptions on which they were designed. The ambition is not to make CTD a new discipline or methodology, but rather to offer a unified perspective on existing models, theories, and methods that can be instrumental in developing improved systems. In this context cognition is not defined as a psychological process unique to humans, but as a characteristic of system performance, namely the ability to maintain control. The focus of CTD is therefore the joint cognitive system, rather than the individual user. CTD has the same roots as Cognitive Task Analysis, but the focus is on macro-cognition rather than micro-cognition, i.e., the requisite variety of the joint system, rather than the knowledge, thought processes, and goal structures of the humans in the system.

2 - HIERARCHICAL TASK ANALYSIS

John Annett (Department of Psychology, University of Warwick , CoventryCV4 7AL, UK)

Abstract. This chapter begins by placing Hierarchical Task Analysis in its historical context. The method is presented as a general analytic strategy for providing solutions to initially specified performance problems. The unit of analysis is an operation specified by a goal and activated by an Input, attained by an Action, and terminated by Feedback. The method is based on systematic decomposition of goals and sub-goals, operations and sub-operations to any desired level of detail until the source of performance failure, physical or cognitive, is identified and a solution can be hypothesised. A seven step procedure is described and illustrated. A variety of uses and adaptations of the method are outlined, including cognitive task design, hazard and operability assessment, training needs in contexts such as power generation, air traffic control and military command and control tasks. Finally Hierarchical Task Analysis is realistically evaluated as requiring both time and skill to yield useful results.

3 - DISCOVERING HOW DISTRIBUTED COGNITIVE SYSTEMS WORK

David D. Woods (Cognitive Systems Engineering Laboratory, Institute for Ergonomics,The Ohio State University, USA)

Abstract. In cognitive work there is a mutual adaptation of agents’ strategies, affordances of artefacts, and demands of the work setting. Regularities in cognitive work concern dynamic interactions across of these sets of factors. Since cognitive work systems are not decomposable into independent basic elements, different empirical tactics are necessary, though core values of observation, discovery, and establishing warrant remain fundamental guides.

Functional analysis is a mode of research that can cope with the unique difficulties of studying and designing cognitive work in context. Functional analysis is a process that coordinates multiple techniques in order to unpack complex wholes to find the structure and function of the parts within the whole. Central to the empirical techniques orchestrated is observation. Three families of techniques emerge which vary in how they shape the conditions of observation: natural history or in situ observation, staged world or observation of performance in simulated situations as models of what is important in situ, and spartan lab settings where observation occurs in experimenter created artificial tasks. The chapter discusses how to coordinate these techniques in a discovery process that reveals the mutual adaptation of strategies, affordances, and demands and predicts how these dynamic processes will play out in response to change.

4 - COURSE-OF-ACTION ANALYSIS & COURSE-OF-ACTION CENTRED DESIGN

Jacques Theureau (CNRS / UTC,Compiègne, France)

Abstract. This chapter presents a theoretical and methodological framework that integrates (1) an approach to the cognitive task analysis, coined as the “course-of-action analysis”, that considers the cognitive tasks as embodied, situated, indissolubly individual and collective, cultured and experienced, (2) an approach to cognitive task design, the “course-of-action centred design”, which concerns the situation (spatial, informational, technical, organisational) as a whole, the training and, more generally, the culture of the operators, thanks to (3) a paradigm of human cognition stemming from theoretic biology, the “enaction paradigm”, and various philosophical and scientific contributions that go well beyond analytical philosophy, cognitive psychology, and computer science.

5 - COGNITIVE WORK INVESTIGATION AND DESIGN IN PRACTICE: THE INFLUENCE OF SOCIAL CONTEXT AND SOCIAL WORK ARTEFACTS

John R Wilson, Sarah Crawford and Sarah Nichols (Institute for Occupational Ergonomics, School of Mechanical , Materials, Manufacturing Engineering and Management, University of Nottingham, Nottingham, NG7 2RD, UK)

Abstract. This chapter is concerned with cognitive work investigations, which are necessary as a foundation for cognitive task design. It draws on two programmes of research, into the work of planners and schedulers in manufacturing industry and into the work of signallers and controllers in the railway network. In particular, the chapter stresses the fundamental influence of context on the performance and understanding of cognitive work, and stresses the need for focus on social factors and social work artefacts, not just as part of this context but also as vital work elements in their own right. As a consequence, the case is made for a high quality field study approach and the need to develop new methods of measurement and assessment, for use in complex interacting and distributed sociotechnical systems.

6 - Group Task Analysis and Design of Computer Supported Cooperative Work

Holger Luczak, Manfred Mühlfelder and Ludger Schmidt ( Institute of Industrial Engineering and Ergonomics of the University of Technology in Aachen , Germany.)

Abstract. Pushed by the connectivity of the internet, the emerging power of the world wide web, and the prevailing nature of “virtual” organizations, i.e. locally distributed teams collaborating by using networked computer systems (groupware), many new forms of computer supported cooperative work (CSCW) are becoming increasingly common. The communities of science and practice dealing with analysis and design of CSCW are truly multi-disciplinary and combine knowledge from computer science, software engineering, human factors, psychology, sociology, ethnography, business administration and training. Despite this quantitative and qualitative growth, many groupware systems still have serious usability problems. At worst, cooperating by using groupware can be awkward and frustrating compared to face-to-face collaboration. Still, research on CSCW has only sketchy knowledge of how people really collaborate, and translating what people actually do into effective designs is difficult. As well, the effort involved in assessing groupware prototypes and systems is onerous. There are neither any universal evaluation techniques nor any generally approved design criteria existing. The methods of cognitive task analysis can contribute to improve both groupware design, e.g. by giving insight in demand, structure and process of collaborative task execution, and task design for locally distributed teams, e.g. by analysing the effects of groupware on work structure and process, allocation of responsibility and resources, and task bound communication. This paragraph joins the domains of task analysis and CSCW by explaining cognitive issues in collaboration and applying them in a collaborative project-planning scenario.

7 - Cognitive Workload and Adaptive Systems

James L. Alty (IMPACT Research Group, Dept. of Computer Science, Loughborough University,Loughborough, Leicestershire, UK)

Abstract. The development of adaptive systems is suggested as a possible way forward for decreasing cognitive workload particularly in the control of large dynamic systems. The nature of adaptation is discussed and two types of adaptation are identified – short-term and long-term. In the former case the system returns to equilibrium after a disturbance. In the latter case the disturbance becomes the new equilibrium point. Early attempts at developing adaptive systems are reviewed and some construction principles identified including the identification of adaptive triggers and adaptive functions. Software agent technology is identified as an appropriate technology for developing adaptive systems and a research project using these techniques – AMEBICA – is described in some detail. The AMEBICA project has developed an adaptive interface for a large dynamic system. The implementation is briefly discussed and the experiences gained in using it in an Electrical Network exemplar are presented together with the lessons learned.

8 - Adaptive Automation: Sharing and Trading of Control

Toshiyuki Inagaki (Institute of Information Sciences and Electronics, University of Tsukuba, Japan)

Abstract. Function allocation is the design decisions to determine which functions are to be performed by human and which by machines to achieve the required system goals, and is closely related to the issues of automation. Some of the traditional strategies of function allocation include (1) assigning each function to the most capable agent (either human or machine), (2) allocating to machine every function that can be automated, and (3) finding an allocation scheme that ensures economical efficiency. However, such “who does what” decisions are not always appropriate from human factors viewpoints. This chapter will clarify why “who does what and when” considerations are necessary, and will explain the concept of adaptive automation in which the control of functions shift between human and machines dynamically depending on environmental factors, operator workload and performance. Who decides when the control of function must be shifted? That is one of the most crucial issues in the adaptive automation. Letting the computer be in authority may conflict with the principle of human-centred automation claiming that, “the human must be maintained as the final authority over the automation.” Qualitative discussions cannot solve the authority problem. This chapter proves the need for quantitative investigations with mathematical models, simulations, and experiments, for a better understanding of the authority issue.

Starting with the concept of function allocation, this chapter describes how the concept of adaptive automation was invented. The concept of levels of automation is used to explain interactions between humans and machines. Sharing and trading are distinguished to clarify the types of human-automation collaboration. Algorithms for implementing adaptive automation are categorized into three groups, and comparisons are made among them. Benefits and costs of adaptive automation, in relation to decision authority, trust-related issues, and human-interface design are discussed with some examples.

9 - A Search for Meaning: A Case Study of the Approach-to-Landing

John M. Flach, Paul F. Jacques, Darby L. Patrick (Department of Psychology, Wright State University, Dayton, OH, USA)

Matthijs Amelink, M. M. (Rene) van Paassen and Max Mulder (Aerospace Engineering, DelftUniversity of Technology, The Netherlands)

10 - Requisite Imagination: The Fine Art of Anticipating What Might Go Wrong

Anthony J. Adamski and Ron Westrum (Eastern MichiganUniversity)

Abstract. Good design for cognitive tasks is critical in today’s technological workplace. Yet the engineering record shows that often designers fail to plan realistically for what might go wrong. Either the design fails to meet the designer's purpose or it holds hidden traps for the users. We call the ability to foresee these potential traps “requisite imagination.” While many designers learn to avoid such pitfalls through experience or intuition, we propose a systematic way to consider them. We argue that by the use of a sound conceptual model, designers can foresee side effects and avoid unintended consequences. This chapter presents a systematic way to practice requisite imagination.

Part II: Methods

11 - Cognitive Factors in the Analysis, Design and Assessment of Command and Control Systems

Jonathan Smalley (Airspace Management Systems Department, QinetiQ, Malvern Technology Centre, St Andrews Road, Malvern, Worcestershire, England)

12 - Basic Cognitive Principles Applied to the Design of Advanced Supervisory Systems for Process Control

Bernard Riera and Serge Debernard (LAMIH – UMR CNRS 8530 – Department of Human-Machine Systems, University of Valenciennes - Le Mont Houy,59313Valenciennes cedex 9 – France)

Abstract. Supervision of highly automated processes is an activity in which human being, despite the increasing automation of the last years, is still present. Consequently, the design of efficient supervisory systems requires a different approach of automation integrating cognitive task design, and being more systemic than analytic. With this approach, cognitive task design can lead to original solutions in matter of automation based around the concept of human-machine cooperation. In this chapter, it is explained how to support cooperation between a human operator and a technical system in the case of supervisory control. Cognitive models and three generic forms of cooperation (augmentative, integrative, debative) are the foundations to develop a common frame of reference supporting human-machine cooperation. To illustrate the chapter, two design applications of advanced human adapted supervisory system are proposed. The first application concerns air traffic control, a process with a low level of automation. The second application deals with a process with a high level of automation: a nuclear fuel reprocessing system. In both cases, cognitive task design has led to original solutions in matter of automation and has been evaluated by the mean of experiments performed with experienced operators.

13 - Cognitive Task Load Analysis: Allocating Tasks and Designing Support

Mark A. Neerincx (TNO Human Factors, P.O. Box 23, 3769 ZG, Soesterberg, The Netherlands)

Abstract. We present a method for Cognitive Task Analysis that guides the early stages of software development, aiming at an optimal cognitive load for operators of process control systems. The method is based on a practical theory of cognitive task load and support. In addition to the classical measure percentage time occupied, this theory distinguishes two load factors that affect cognitive task performance and mental effort: the level of information processing and the number of task-set switches. Recent experiments provided empirical support for the theory, showing effects of each load factor on performance and mental effort. The model can be used to establish task (re-)allocations and to design cognitive support.

This chapter provides an overview of the method’s foundation and two example applications. The first example is an analysis of the cognitive task load for a future naval ship control centre that identified overload risks for envisioned operator activities. Cognitive load was defined in terms of task demands in such a way that recommendations could be formulated for improving and refining the task allocation and user-interface support. The second example consists of the design and evaluation of a prototype user interface providing support functions for handling high-demand situations: an information handler, a rule provider, a diagnosis guide and a task scheduler. Corresponding to the theory, these functions prove to be effective, in particular when cognitive task load is high. The user interface is currently being implemented at a bridge of an icebreaker.

The two examples comprise an integrated approach on task allocation enhancement and design of cognitive support. The theory and method are being further developed in an iterative cognitive engineering framework to refine the load and support model, improve the empirical foundation and extend the examples of good practices.

14 - Cognitive Task modelling for system design and evaluation in non-routine task domains

Nikolaos K. Tselios and Nikolaos M. Avouris (University of Patras, ECE Department, HCI Group, GR-265 00 Rio Patras, Greece)

Abstract. This chapter concerns task-based design and evaluation of computer environments to support non-routine tasks, like exploratory learning and problem solving. Non-routine tasks are characterised by exponential growth of expected task structures. Ways of tackling the complexity of this task structure is the objective of the described approach. The proposed development life-cycle involves a model-driven requirements specification, design and implementation phase, followed by a field study evaluation phase. During the latter phase, evaluation of the effectiveness of the system in supporting the expected typical tasks is measured and new requirements are derived that modify the original design. This spiral design process is supported by appropriate tools. Such tools typically allow expression of the designer expectations, concerning user cognitive tasks. Subsequently, during field studies, the tools support tracking of user interaction and inferring cognitive task models, which are matched against the original designer’s assumption on the user cognitive models.

In this chapter a survey of application of cognitive modelling techniques in the design of systems involving non-routine tasks is attempted first. Subsequently, an outline of a proposed design and evaluation framework and the related tools are described. Existing cognitive task modelling techniques like Hierarchical Task Analysis have been adapted, in the frame of the reported framework, in order to be used effectively for expression of both the designer model and the user task model concerning observed interaction. Examples of application of the developed techniques and tools in the design and evaluation of open problem solving environments are also described.

15 - Design of Systems in Settings with Remote Access to Cognitive Performance

Oliver Sträter (Human Factors & Manpower Unit, Eurocontrol, Belgium)

Heiner Bubb ( Institute of Ergonomics, University of Technology Munich, Germany)

Abstract. In some fields of cognitive task design the user, or user group, cannot be involved directly in the design process. Typically this is the case in incident investigations and assessment of human reliability. They require a prediction of cognitive processes from a distance and the system-design relies on remote access to cognitive performance. These settings are called remote settings in this chapter.

The chapter describes a method on how to proceed in such situations. It will show how information about cognitive demands, possible human behaviour and contextual factors can be obtained, and how it can be used to improve the design of technical systems. Examples are provided from incident analysis in nuclear power plants, from design of alarm systems for steer-by-wire systems in automobiles, and from human reliability assessment in regulatory tasks for nuclear industry. Finally, the chapter will reflect the method presented here in respect to possible implications of cognitive modelling.

16 - Applied Cognitive Work Analysis: A Pragmatic Methodology for Designing Revolutionary Cognitive Affordances

William C. Elm, Scott S. Potter and James W. Gualtieri (Aegis Research Corporation; Cognitive SystemsEngineering Center; Pittsburgh, PAUSA)

Emilie M. Roth (Roth Cognitive Engineering; Brookline, MAUSA)

17 - Decision-Centred Design: Leveraging Cognitive Task Analysis in Design

Robert J. B. Hutton, Thomas E. Miller, Ph.D. and Marvin L. Thordsen (Klein Associates Inc., 1750 Commerce Center Blvd. North, Fairborn, OH 45324-6362)

Abstract. Decision-Centred Design is a design approach that focuses on understanding and supporting cognitive tasks. Research in the areas of naturalistic decision-making and expertise has provided the cognitive models and analytic methods that enable this design approach. DCD provides a means for communication and understanding between designers and the individuals for whom interventions are being designed. It targets the critical and often challenging cognitive tasks confronting these individuals, including individual tasks where cognitive performance may break down, and at the team level where collaboration and coordination are required. The approach consists of task analysis, design, and evaluation.

The analysis portion is made possible by continuing advances in analytic methods that enable subject-matter experts to articulate complex cognitive activities, the analysis of task performance, and the environments in which performance occurs. Analysis is also grounded by existing and developing models of applied cognition that guide observations and interviews (such as the Recognition-Primed Decision model, Advanced Team Decision Making model, and evolving models of macrocognition). The output of the analysis is an understanding of cognitive requirements illustrated by critical incidents, and an understanding of the strengths and weaknesses of existing strategies and supports for these cognitive tasks.

The design aspect of the approach creates designs for decision-centred information technology, organizations, task design, and training. Design interventions are focused on reducing the cognitive challenges, improving error management, and supporting the acquisition of skilled performance, raising novice performance to the level of experts. The design approach also emphasizes the collaboration of designers, analysts, and subject-matter experts through design prototypes and mock-ups using challenging scenarios that provide the context for performance of these cognitive tasks.

The iterative design approach requires continuous evaluation, testing, and feedback. The decision-centred design approach provides the input for cognitively challenging test and evaluation scenarios as well as context sensitive measures of performance. Subject-matter experts’ feedback is elicited and focused on the cognitive aspects of task performance, and the data are fed back into the design process.

This chapter provides an overview of the decision-centred design approach, its critical phases, methodologies, and theoretical underpinnings. It then provides two context-rich case studies where the approach has been successfully applied. Decision-centred design explicitly addresses the need for task and technology design required for the increasing cognitive complexity of modern work environments.

18 - Requirement for a Cognitive Framework for Operation in Advanced Aerospace Technologies

Kevin M. Corker (Department of Computer Information and Systems Engineering, San Jose State University, San Jose, CA.)

Abstract. The international aviation community is advocating goals for the first quarter of the new millennium that compel a radical revamping of the practice of aircraft and air traffic management. This vision finds expression in both the European and United Statescommissions and offices as well as in corporate statements of its intent to invest in requirements analysis of an international air traffic management coordinated via a satellite-based information exchange processes and reduced constraint in control of aircraft movement. Other advancements envisioned include an increased use of the airspace to support access to space with single-stage to launch vehicles and a distribution of space access ports through out the United Statesand around the world. The visions assert new modes of operation and technological requirements. Essentially without exception these technologies fundamentally change the process of the work of air traffic and air space management. The “advancements” include a redistribution of information and control among the humans and the automation systems in airspace operations that alters decision modes, execution modes and optimisation processes among all participants in the aerospace transportation process. I feel that the changes in the work of air transportation operations requires an approach to analysis that includes concern for the change in the cognitive processes that supports the work in context.

19 - Groupware Task Analysis

Martijn van Welie and Gerrit C. van der Veer (Department of Software Engineering, Vrije Universiteit, The Netherlands)

20 - KOMPASS: A Method for Complementary System Design

Toni Wäfler, Gudela Grote, Anna Windischer and Cornelia Ryser (Institute of Work Psychology, Swiss Federal Institute of Technology (ETH), Zurich, Switzerland)

Abstract. In order to take into account the complex interplay of human, technology, and organisation the KOMPASS method provides criteria for task design on three levels. On the level of the human-machine function allocation, those aspects are considered, that are directly influenced by process automation. On the level of the individual work task the whole task to be performed by an individual is taken into account. On the level of the work system tasks of sociotechnical units (e.g. work teams, departments) are focused. The design criteria on all three levels are deduced from the key notion to provide opportunities for process control as well as to promote respective human motivation and competencies. Furthermore KOMPASS also defines guidelines for a participatory approach to system design.

The chapter has three parts. In the first part the theoretical background of KOMPASS is outlined. Next, the KOMPASS criteria are described. The final section presents the guidelines for the support of the design process.

21 - Relating the automation of functions in multi-agent control systems to a system engineering representation

Michael D. Harrison, Philip D. Johnson and Peter C. Wright (Department of Computer Science, University of York, Heslington, York, YO10 5DD. UK.)

Abstract. Engineers of safety critical systems are beginning to recognise that human issues are critical to their safe automation, and that appropriate techniques for taking account of the people in the system should be integrated into the design process. This chapter gives a brief introduction to a two-step decision procedure that can be used to help decide how to automate an interactive system. The procedure is intended for use early in the development of systems in larger scale collaborative settings with the aim of improving their safety and performance. Two issues are particularly important. The first concerns the appropriate choice of automation so that the tasks designed for the different roles satisfy criteria that have significance from a cognitive perspective. The second is to understand the mapping from concepts of function allocation to notations that are meaningful and usable by system engineers. The method has received preliminary evaluation in aviation and naval contexts.

22 - Product Design to Support User Abstractions

Alan F. Blackwell and Rachel L. Hewson ( Computer Laboratory, University of Cambridge, UK)

Thomas R. G. Green (Department of Computing, University of Leeds, UK)

Abstract. Both software products and new consumer products mask substantial abstractions behind apparently simple button-pressing interfaces. However the simplicity of the interface belies the complexity of the device. We analyse a range of new product types, including a survey of retail consumer devices and a series of interviews with the designers of simple home automation devices. We find that there is little support for usable access to abstraction facilities. A series of interview with office workers confirms that they are able to create abstractions in their working environment, but are inhibited from doing so when using software products. We propose an alternative analysis of product abstractions, where the presentation of the abstraction is considered as a notational system. A case study of an well established and sophisticated abstract notational system provide a test case for this analysis technique.

Part III: Field Studies

23 - Mapping the Design Space for Socio-Cognitive Task Design

Alistair Sutcliffe (Centre for HCI Design, Department of Computation, UMIST, P.O. Box 88, ManchesterM60 1QD, UK)

Abstract. This chapter describes a framework for cognitive task design that focuses on social and communication issues. It is intended to complement the more traditional task oriented views. The chapter starts by describing a process model for cognitive task design then introduces cognitive and social pathologies: heuristics for analysing potential causes of failure in socio-technical systems. This is followed by an introduction to generic tasks as reusable models for specifying human action, to which requirements for computer system task support can be attached, as well as the pathologies to warn the designer about potential failure points. A general model of discourse is described which is also associated with pathologies and generic requirements to guide the designer towards reusable knowledge at appropriate stages in the communication process. The discourse model is supplemented with patterns for specific types of communication such as Command and Control, proposals and liaison, etc. The next section proposes models for analysing power and trust in human-human and inter-organisational relationships. Once again models are associated with pathologies and countermeasures. Use of the models is illustrated by analysis of Command and Control tasks in the combat sub-system of a Navy frigate. A scenario-based approach is taken to walk through the models using operation scenarios of a missile attack. The design implications for cognitive task support and communication are assessed, then two alternative designs for intelligent computerised radar and standard radar are compared using the generic models for analysing communication, perceptions of power and trust. The chapter concludes by reviewing the lessons learned for functional allocation, and future work in providing more effective decision support for operators in socio-technical systems.

24 - The Design of POGO Story World

Antonio Rizzo and Patrizia Marti(University of Siena, Via dei Termini 6, 53100 Siena – Italy)

Françoise Decortis (Université de Liège, 5, Bd. du Rectorat, 4000 Liège - Belgique)

Claudio Moderini (Domus Academy, Via Savona 97, 20144 Milano – Italy)

Job Rutgers (Philips Design, Witte Dam 1, Eindhoven - Holland)

Abstract. In this paper we describe the design process that led us to the development of an innovative narrative environment for children at school and home, the POGO story world. The design process was iterative and co-evolutionary among four design components: user study, concept design, content design and technology design. Grounded on a shared vision this design process allowed us to use principles and concepts of the Cultural Psychology theoretical approach to drive the design concepts and development of the tools of POGO world. The process we describe show how Cognitive Task Design can envision and guide technological development and innovation. To this aim, we firstly present the pedagogical principles derived from the work of Lev Vygotsky and Jerome Bruner that became the conceptual basis for assessing the results of the design process. Subsequently we present an analysis of the pedagogical activity involving narratives carried out in two European schools and the development of the design concepts that were embodied in mock-up and working prototypes. Finally, we briefly report the testing at school of the mock-ups/prototypes, and conclude with the implication of the described design process for extending Cognitive Design to more formal working context.

25 - Approaches to transportation safety: Methods and case studies applying to track maintenance train operations

Kenji Itoh (Department of Industrial Engineering and Management. Tokyo Institute of Technology, Japan)

Masaki Seki (Central Japan Railway Company, Nagoya, Japan)

Henning Boje Andersen (Systems Analysis Department, Risø National Laboratory, Roskilde , Denmark)

Abstract. The objective of this chapter is to describe several methods for addressing transportation safety with illustration of case studies that have been applied to the operations and organisation of track maintenance trains for a Japanese high-speed railway. This chapter discusses the following six methods: (1) experiments, (2) observation of operator tasks, (3) PRA (probabilistic risk assessment) and HRA (human reliability analysis), (4) cognitive simulation, (5) accident/incident analysis, and (6) questionnaire-based surveys. For each method, the authors offer a general description of its purpose, main characteristics, benefits, limitations and the steps involved in its application. Then, three case studies are described of track maintenance train operations and their associated organisations. Throughout the review of the case studies, we seek to bring out implications from each of the methods to the design of cognitive tasks including safety management issues in organisations.

26 - Lessons From A Focus On Artefacts And Implicit Theories: Case Studies In Analysis And Design

Ann M. Bisantz (University at Buffalo, The State University of New York, Amherst, NY 14020)

Jennifer J. Ockerman (Georgia Tech Research Institute, Georgia Institute of Technology, Atlanta, Georgia 30332)

Abstract. A claim in human-computer interaction is that technological artefacts embody implicit theories about people including how they will use the artefacts. More broadly, artefacts embody theories of the work setting in which the artefact will be implemented. By identifying these embedded theories, and examining the use of a built artefact in a real life setting,one can assess the adequacy of those theories in describing how humans will interact with the technology and how work in the setting is conducted. This chapter describes such an investigation of the use of decision support technology in a quick service restaurant environment. Analysis indicated that the decision support technology was not used as intended, in part because the theories embodied in the technology did not represent conditions in the restaurant. Specifically, the system embodied theories regarding human decision-making, the use of decision aids, and the way work was organized which ran counter to those descriptive of the way these activities were actually conducted. This contrast illustrates differences between competing theories of decision making, decision support, and work organization: the study serves as a falsifying case to refute those theories that are highly prescriptive in nature. The chapter describes the work environment, methods of data collection, and then provides evidence from the field observations which support accounts of human decision making, and work organization that emphasize the flexible, distributed, and continuous nature of those activities. Additionally, the analysis provides information regarding the types of decision aids that are more appropriate for intentional work environments. Finally, the role of similar embedded theories in informing design are explored through a second case study in which computerized technologies were being introduced into a different work domain.

27 - Case Studies: Applied Cognitive Work Analysis in the Design of Innovative Decision Support

Scott S. Potter, James W. Gualtieri and William C. Elm (Aegis Research Corporation, Cognitive Systems Engineering Center, Pittsburgh, PAUSA)

28 - An Analysis of Procedure Following as Concerned Work

Peter Wright (HCI Group, Department of Computer Science, University of York, UK)

John McCarthy (Department of Applied Psychology, University College Cork, Ireland)

Abstract. This chapter is concerned with the design and use of operating procedures. Operating procedures feature strongly in safety critical work domains where it is argued they reduce the likelihood of human error. Procedures can be viewed as the result of cognitive task design. Information processing analyses of procedure following have shown how the design of procedures can be improved. In this chapter we explore an alternative to the prevalent information processing approach to take a more critical view of the proceduralisation of the work place. Our approach takes as its starting point ideas from the field of literary studies and narrative rather than information processing psychology. The analysis focuses on autobiographical data. Unlike a more traditional cognitive analysis it highlights the concerns facing operators who have to make procedures work in practice, the paradoxes and dilemmas they face in doing this, and the amount of intelligent effort that goes into making sense of procedures. Our analysis leads us to conclude that the gap between procedures and practice is filled by the creative work of the operator. Such creative work is based on a history of experiences both inside and outside the cockpit. Currently there are very few ways in which this experience can be given a legitimate voice in the process of procedure design.

29 - Representations for Distributed Planning

Valerie L. Shalin (Department of Psychology, Wright State University, Dayton, OH, USA 45435)

Pamela M. McCraw (Mission Operations Directorate, Johnson Space Center , NASA, Houston, TX 77058)

30 - Pilot’s understanding process when coping with “automation surprises”

Marielle Plat (Université ParisVIII,Laboratoire cognition et activités finalisées,2 rue de la liberté, 93526 Saint Denis France)

31 - Team design

Jan Maarten Schraagen & Peter Rasker (TNO Human Factors,Soesterberg, The Netherlands)

Abstract. Team design is about achieving a coordinated effort through the structuring of tasks, authority, and work flow. Just as with the design of an artefact or a technological system, team design affects the cognitive tasks of the members in the team. In this chapter, we will provide an overview of criteria that are important to consider in team design. These criteria are derived from six large, multi-year research programs in the area of team design. Based on these criteria, we will propose recommendations for team design, and show how particular choices for a certain design affect the cognitive tasks of the team members. Seemingly disparate results may be reconciled by taking into account whether team members have to deal with unanticipated disturbances in the environment or not. Team structures that are better suited to deal with unanticipated disturbances are characterized by four R's: redundancy, robustness, review, and re-allocation. The four R's are likely to be all mediated by the single cognitive construct of 'shared mental models'. These in turn are facilitated by certain antecedents, such as continuity of personnel, cross training, self-managed teams, consensual leadership, and organizational slack.