A theoretical framework for designing interfaces for complex humanmachine systems is proposed. The framework, called ecological interface design (EID), is based on the skills, rules, knowledge taxonomy of cognitive control. The basic goal of EID is twofold: first, not to force processing to a higher level than the demands of the task require, and second, to support each of the three levels of cognitive control. Thus, an EID interface should not contribute to the difficulty of the task, and at the same time, it should support the entire range of activities that operators will be faced with. Three prescriptive design principles are suggested to achieve this objective, each directed at supporting a particular level of cognitive control. In this paper, the theoretical foundations of the framework are laid out. Particular attention is paid to presenting a coherent deductive argument justifying the principles of EID. In addition, three sources of converging support for the framework are presented. First, a review of the relevant psychological and cognitive engineering literature reveals that there is a considerable amount of research that is consistent with the principles of EID. Second, an examination of other approaches to interface design indicates that EID has a unique and significant contribution to make. Third, the results of an initial empirical evaluation also provide some preliminary support for the EID framework. The paper ends by outlining some issues for future research.

Previous research has shown a significant correlation between domain expertise and memory recall performance after a very brief exposure time. Despite the large number of such studies, several findings in the literature have no satisfactory theoretical explanation. A novel theory based on an ecological approach is proposed to explain these results. This constraint attunement hypothesis provides a framework for identifying and representing the various levels of goal-relevant constraint in a domain. The theory predicts that there will be a memory expertise advantage in cases in which experts are attuned to the goal-relevant constraints in the material to be recalled and that the more constraint available, the greater the expertise advantage can be. The theory explains a number of diverse empirical findings in the literature in a coherent, unique, and parsimonious fashion and suggests a number of promising issues for future research.

Ecological interface design (EID) is a theoretical framework for designing human-computer interfaces for complex sociotechnical systems. Its primary aim is to support knowledge workers in adapting to change and novelty. This literature review shows that in situations requiring problem solving, EID improves performance when compared with current design approaches in industry. EID has been applied to industry-scale problems in a broad variety of application domains (e.g., process control, aviation, computer network management, software engineering, medicine, command and control, and information retrieval) and has consistently led to the identification of new information requirements. An experimental evaluation of EID using a full-fidelity simulator with professional workers has yet to be conducted, although some are planned. Several significant challenges remain as obstacles to the confident use of EID in industry. Promising paths for addressing these outstanding issues are identified. Actual or potential applications of this research include improving the safety and productivity of complex sociotechnical systems.

Recently, a new class of artifacts has appeared in our environment: complex, high technology work domains. An important characteristic of such systems is that their goal-relevant properties cannot be directly observed by the unaided eye. As a result, interface design is a ubiquitous problem in the design of these work environments. Nevertheless, the problem is one which has yet to be addressed in an adequate manner. An analogy to human perceptual mechanisms suggests that a smart instrument approach to interface design is needed to supplant the rote instrument (single-sensor-single-indicator) approach that has dominated to this point. Ecological Interface Design (EID) is a theoretical framework in the smart instrument vein that postulates a set of general, prescriptive principles for design. The goal of EID is twofold: first, to reveal the affordances of the work domain through the interface in such a way as to take advantage of the powerful capabilities of perception and action; and second, to provide the appropriate computer support for the comparatively more laborious process of problem solving. An example of the application of the EID framework is presented in the context of a thermal-hydraulic system. The various steps in the design process are illustrated, showing how the abstract principles of EID can be applied in a prescriptive manner to develop a concrete design product. An important outcome of this discussion is the novel application of Rasmussen's means-end hierarchy to structure the affordances of an ecosystem. The means-end hierarchy is a generic framework for describing goal-oriented systems with many degrees of freedom, and therefore represents a useful addition to the conceptual framework of ecological psychology. All sorts of instruments have been devised for mediating apprehension. Some optical instruments merely enhance the information that vision is ready to pick up; others ... require some inference; still others ... demand a complex chain of inferences. Some measuring instruments are closer to perception than others. (Gibson, 1979, p. 260) The process operator lives in a complex world. The information presented to him is a code for the physical, dynamical process in the interior of the plant. He is able to ... operate on the physical meaning of the symbols by rational deductive reasoning. During frequent routine tasks, however, .... he may be able to improvise rapidly ... if he is allowed to break down ... the information patterns into familiar generic units. This is only possible if he can control the process directly-the display system therefore should be 'transparent' and the physical process should be directly 'touchable' on the control desk. (Rasmussen, 1974, p. 11

No abstract

A fundamental challenge in studying cognitive systems in context is how to move from the specific work setting studied to a more general understanding of distributed cognitive work and how to support it. We present a series of cognitive field studies that illustrate one response to this challenge. Our focus was on how nuclear power plant (NPP) operators monitor plant state during normal operating conditions. We studied operators at two NPPs with different control room interfaces. We identified strong consistencies with respect to factors that made monitoring difficult and the strategies that operators have developed to facilitate monitoring. We found that what makes monitoring difficult is not the need to identify subtle abnormal indications against a quiescent background, but rather the need to identify and pursue relevant findings against a noisy background. Operators devised proactive strategies to make important information more salient or reduce meaningless change, create new information, and off-load some cognitive processing onto the interface. These findings emphasize the active problem-solving nature of monitoring, and highlight the use of strategies for knowledge-driven monitoring and the proactive adaptation of the interface to support monitoring. Potential applications of this research include control room design for process control and alarm systems and user interfaces for complex systems.

Individual differences among users of a hierarchical file system were investigated. Results indicate that psychometric tests of vocabulary and spatial visualization are the best predictors of task performance, accounting for 45% of the variance in the data. The spatial predictor was found to be the most influential. This was dramatically illustrated by the fact that, on the average, subjects with low spatial ability took twice as long to perform the task as those with high spatial ability. Surprisingly, experience alone does not predict task performance. A comparison of the frequency of command usage between subjects with high and low spatial ability revealed that those with low spatial ability were getting lost in the hierarchical file structure. These data have implications for redesigning the software interface so as to accommodate people with low spatial ability.

Adverse drug events are the single leading threat to patient safety. Human factors engineering has been repeatedly proposed, but largely untested, as the key to improving patient safety. The value of this approach was investigated in the context of a commercially available patient-controlled analgesia device that has been linked with several alleged patient injuries and deaths. Several reports have stated that errors in programming drug concentration were made during these adverse drug events. A simulation of the commercially available interface was compared experimentally with a simulated prototype of a new interface designed according to a human factors process. Professional nurses, averaging over 5 years of clinical experience with the commercially available interface and only minimal experience with the new interface, programmed both interfaces. The new interface eliminated drug concentration errors, whereas the simulated commercially available interface did not. Also, the new interface led to significantly fewer total errors and faster performance. These findings may have broad implications for the design, regulation, and procurement of biomedical devices, products, or systems that improve patient safety in clinical settings.