Jodi Heintz Obradovich

SPECIAL SECTION: Users as Designers: How People Cope with Poor HCI Design in Computer-Based Medical Devices

In this paper we examine how users interact with a computer-based infusion device adapted for terbutaline infusion to treat preterm labor in women experiencing high-risk pregnancies. This study examines (a) the human-computer interaction (HCI) deficiencies in the device as related to this context of use, (b) how the device characteristics increase the potential for error, and (c) the tailoring strategies developed by users to insulate themselves from failure. Interviews with nurses and tests of the behavior of the infusion device in different conditions identified several classic HCI deficiencies: complex and arbitrary sequences of operation, mode errors caused by poor differentiation of multiple operating modes intended for different contexts, ambiguous alarms, getting lost in multiple displays, and poor feedback on device state and behavior.

Successful Use of an Expert System to Teach Diagnostic Reasoning for Antibody Identification

A previously reported study indicated that, when used by an instructor as a tool to assist with tutoring in a class laboratory setting, use of the Transfusion Medicine Tutor (TMT) resulted in improvements in antibody identification performance of 87-93% (p<.001). Based on input from teachers requesting that TMT be designed for use without the presence of an instructor, a new study on the use of TMT without instructor assistance found that performance improved by 64-66% (p<.001). Finally, based on the results of these two studies, TMT was mailed to 7 sites for beta-testing. In exchange for a free copy of the kit, the instructors (and their students) were asked to fill out questionnaires. Results of these questionnaires are summarized.

Cooperative Problem-Solving Challenges for the Movement of Aircraft on the Ground

In this paper, we discuss issues surrounding aircraft surface movement that were uncovered through a series of ethnographic investigations, including (a) observations of surface movement operations, (b) structured interviews with airline ramp control and dispatch personnel, and (c) critical incident reports. These results are part of the “problem identification” stage of a study aimed at the design of cognitive tools to improve the safety and efficiency of aircraft surface movement. This study is focusing on such issues from an airline operations control perspective (i.e., ramp control and dispatch). We have identified areas of opportunity for aiding airline staff and controllers in this information-intensive cognitive work, which will lead to identifying the nature of cognitive tools and procedures that could improve prioritization, planning, and coordination during surface movement

The antibody identification assistant (AIDA), an example of a cooperative computer support system

Performance when using a critiquing expert system was compared to performance with no decision support for two groups of medical technologists solving antibody identification cases. The treatment group had a significantly lower misdiagnosis rate than the control group across the four cases tested (p<0.000005). This is one of the few studies conducted to evaluate critiquing as a form of decision support for practitioners solving real-world tasks. In particular, there was a trend for improved performance even on a case for which the computers knowledge was not fully competent. This is in contrast to the usual problems with people not being able to recover from faulty reasoning exhibited by a brittle, partially automated decision support system. Users of critiquing systems are doing the task themselves and given feedback in the context of what they are doing. Thus, the computer can monitor for errors in the humans reasoning, and the human has a basis for judging the computers reasoning, resulting in cooperative problem-solving between the two decision makers.

Users as Designers: How People Cope with Poor HCI Design in Computer-Based Medical Devices

In this paper, we examine how users interact with a computer-based infusion device adapted for terbutaline infusion to treat preterm labor in women experiencing high-risk pregnancies. This study examines: (1) the HCI deficiencies in the device as related to this context of use, (2) how the device characteristics increase the potential for error, and (3) the tailoring strategies developed by users to insulate themselves from failure. Interviews with nurses and bench tests of the behavior of the infusion device in different conditions identified several HCI deficiencies: complex and arbitrary sequences of operation, mode errors due to poor differentiation of multiple operating modes intended for different contexts, ambiguous alarms, getting lost in multiple displays, and poor feedback on device state and behavior.

Design Concepts for Distributed Work Systems: An Empirical Investigation Into Distributed Teams in Complex Domains

Coordination and collaboration across distributed teams during planning and execution activities present many challenges to organizations and operations, including military command and control. With the increase in spatially and temporally distributed work teams, it is essential to achieve an understanding of how distributed work can be effectively supported. In this paper, we discuss issues surrounding distributed work systems that were gained through a series of investigations, including (a) field studies, (b) structured interviews, (c) critical incident reports, (d) participation in concept experiments, and (e) literature reviews. Characteristics held in common by the domains explored include the following: high cognitive complexity, distribution of the work (and associated responsibilities) among many people, distribution of data and knowledge among many people, uncertainty about how scenarios will actually play out, and geographic and temporal distribution of the participants. The results of these investigations presented in the form of case studies, and design concepts to support distributed work systems are suggested.

C-SLANT: an asynchronous communications tool to support distributed work in the National Airspace System

In recent years there have been a number of efforts under the Collaborative Decision Making (CDM) program that have demonstrated the benefits of disseminating both knowledge and goals between organizations in the National Airspace System (NAS), and of collaboratively building decision support tools to assist in the interpretation of the large interrelated dataset produced. The paper describes the development of the Collaborative SLide ANnotation Tool (C-SLANT), an asynchronous communications tool that was developed to support discussion of data from the CDM product Post Operations Evaluation Tool. It is suggested that slide show based, multi-modal, asynchronous, messaging systems appear potentially valuable in this distributed work domain for problem solving during post operations analysis.

Cooperative problem-solving in the air traffic management system

A focus group was held to explore different perspectives on the interactions of operational staff at the airlines and the FAA regarding their real time interactions prior to and during flights. To elicit these different perspectives, four scenarios were prepared based on actual events. The resultant discussion raises issues about information exchange, communication, training and coordination.

The Transfusion Medicine Tutor: Using Expert Systems Technology to Teach Domain-Specific Problem-Solving Skills

This study provides data regarding the effectiveness of the expert system-based Transfusion Medicine Tutor (TMT) when used by medical technology students to learn an important problem-solving task, the identification of alloantibodies in a patients blood for the purpose of finding compatible blood for transfusion. The results show that the students who were taught by an instructor using TMT to provide the instructional environment went from 0% correct on a pre-test case to 87%–93% correct on post-tests (N=15). This compares with an improvement rate of 20% by a control group (N=15) who used a passive version of the system with the tutoring functions turned off. The results also demonstrate the importance of relying on objective performance data rather than questionnaire data to evaluate systems, as there was no difference in the subjective responses of the students to these two different versions of the system.

Information Management to Support Distributed Decision-Making in the National Airspace System

For Summer 2000, the Federal Aviation Administration introduced a new collaborative procedure for reducing departure delays during weather events. This initiative involved the used of predefined Coded Departure Routes (CDRs). Goals of this procedure included: Providing airline operations centers and FAA air traffic facilities with a process for working collaboratively to develop earlier plans for dealing with predicted constraints. Giving airport control towers greater flexibility to respond to the often rapidly changing picture regarding available airspace during weather events. Keeping airline dispatchers in the loop through the early identification of alternative departure routes to expedite departures from an airport. An analysis of the use of CDRs was completed, based on interviews with staff at six airlines and four enroute air traffic control Centers, on analyses of System Command Center advisories and on analyses comparing filed with flown routes. Based on this analysis, recommendations are made for the design of improved tools for information dissemination and decision support.