Advances in Human Performance and Cognitive Engineering Research: Volume 1

The purpose of this chapter is to contribute to ongoing discussions defining the future of cognitive engineering research by examining a part of its past. The intellectual history of one particular line of research, that of the Electronics Department at Risø National Laboratory, is reviewed. A number of influential studies, conducted between 1962 and 1979, are described. Among these are operational experience acquired from the introduction of a prototype digital console in a nuclear research reactor, two field studies of professional operators conducting representative tasks in representative settings (electronic trouble-shooting and conventional power plant control), and analyses of over 645 human error reports in the nuclear and aviation industries. Examples of the influence that the Risø work has had on basic and applied problems are reviewed. Also, some of the themes characterizing the Risø research in cognitive engineering are identified. These themes help define what cognitive engineering is, and what it might be concerned with in the future.

The ALS is a theoretically driven approach to the design of integrated-embedded training systems that is highly flexible and offers ease of implementation. It operates by exerting leverage on foci of the self-regulation system, which recent research has demonstrated to be central to learning and performance for difficult, complex, and dynamic tasks. The training strategy incorporated in the ALS constructs instructional interventions by combining specific training components that affect different aspects of the SRS. By designing synergistic combinations, instructional interventions can be tailored to the developmental progress of trainees and can enhance learning, performance, and adaptability.Our research will target those training components that offer the greatest practical and theoretical potential for improving complex skill acquisition, and the enhancement of adaptive capabilities. By building on existing principles of training design (e.g. mastery goals, sequencing), and examining promising new ideas (e.g. information, interpretation) that are likely to be key capabilities of the next generation of advanced technology systems, the research is expected to yield new principles of training design uniquely suited for the design of integrated-embedded training systems.

Research on human cognition in complex tasks, such as interacting with advanced technology, requires the development and validation of new methods. This paper describes PRONET, a method for summarizing, representing, and analyzing event sequences. The first section outlines how the PRONET method can be applied to any sequence of events, with lessons learned from previous applications of the method. The second section presents demonstrations of the application of PRONET. In the first demonstration — a computer-based simulation of operant training - the PRONET analysis and representation clearly shows the change in the behavior of the simulation produced by changes in reinforcement contingencies, but also shows interesting aspects of behavior that were not affected. In the second demonstration — involving transfer of word processing skill — network-related and performance measures showed the expected pattern of positive transfer. In addition, the network of the far transfer participant suggested that she used task knowledge to search for conditions that would permit the correct action. Two previously-published examples showed the usefulness of the PRONET method in characterizing a hybrid event sequence consisting of environmental conditions and behavioral actions and a sequence of events from a team.

Cognitive Task Analysis (CTA) attempts to explain the mental processes involved in performing a task. These processes include the knowledge, skills and strategies that are needed to accomplish the task functions. The criteria for success in a CTA study are: making a useful discovery about the cognitive skills being studied; being able to communicate the discovery to the users (i.e. those who will need to use the CTA for design); and having a meaningful impact on the eventual design.Currently, a wide variety of CTA methods are being used. As we learn how to define the cognitive demands presented by a task/situation, we hope we will be able to map CTA methods onto these demands, so that we can more efficiently select and apply the appropriate methods. This should result in more efficient studies, and greater user satisfaction. It should also help move the field of CTA into becoming more of a technology.

This chapter reviews the ability of the emerging human performance modeling technologies to support the design and operation of complex systems. The ability of existing technologies to meet current application needs is analyzed, and the results are then used to assess the areas where additional research and development is most needed. Following a brief history of human performance modeling, a taxonomy of models and modeling techniques is established, as a framework for remaining discussion. The human performance modeling technology base is separately analyzed for its ability to support system design processing and to support system operation. The system design process analysis considers the various roles that human performance models may play during that process, ranging from generating design concepts to affording simulation-based range of roles, from training to performance support to automation. These analyses demonstrate that human modeling technology has reached a sufficient state of maturity and has become a proven contributor of the complex systems engineering process. Challenges for further high-payoff research are also presented in five categories: cognition, knowledge management, team and organizational structure and processes, predictive models of training, and human-centered systems engineering.