Preface to Sparks of Innovation in Human-Computer Interaction

The occasion for this book is the 10th Anniversary of the Human-Computer Interaction Laboratory (HCIL) at the University of Maryland. I have selected two dozen key papers from more than a hundred to represent the work of many participants. My section introductions tell how we do what we do, including some of our failures and background stories that are not appropriate for journal papers. Many papers were trimmed to emphasize the cogent points. They weave together the threads of our work into a unified fabric that reveals the patterns of development. It was difficult to choose the best papers; these exemplify different research methodologies and show the maturation of human-computer interaction research. This book is a tribute to the faculty, staff, visitors, and students who have shared in a decade of work.

Sparks of innovation are the ideas and processes that stimulated these creative and devoted participants. This metaphor emerged as I told potential supporters that our lab felt like a group around a blazing campfire: telling stories, sharing ideas, and enjoying each other's company. Some people (academic and industrial researchers, and commercial developers) enjoyed watching our fire from a distance (we are happy to send technical reports), while others preferred to come and join us for a story or two (we enjoy doing demonstrations and having long-term visitors participate in projects), but we do need some people to throw a log on the fire occasionally (collaborate and support our efforts financially). I thought this told our story well, but an old friend, Fred Hansen, now a researcher affiliated with Carnegie-Mellon University, commented that our lab had something more. He said that we understood the art of fire-making, and that the largest benefit to our visitors and supporters was learning how to spark fires within their own organizations.

We are proud of our role in the emergence of human-computer interaction, a new inter-discipline devoted to researching how people use computers and to improving user interface designs. As in any new venture, we question and are challenged about what we are doing, but there is a growing confidence that comes when there are ten scientific journals in an area that had only one just a decade ago. The confidence stems also from the numerous well-attended and invigorating conferences that are held around the world almost every month, and the stream of national policy documents, corporate business plans, and academic departments declaring their focus on human-computer interaction.

I have gained much by reading earlier work and by frequent contact with colleagues. By learning about our experiences, I hope other academic researchers and advanced commercial developers can speed their efforts and make their own fires.

Acknowledgements

This book was created by the diligent and intense efforts of Lian Arrow, Teresa Casey, Ruth Golembiewski, and Ara Kotchian, who massaged various forms of electronic documents and figures into a consistent and appealing style. Individual authors contributed by writing new sections, redoing figures to put everything in electronic form, revising and trimming published papers, and by commenting on drafts. Robert Jacob, Larry Koved, Charles Kreitzberg, Janis Morariu, and Bob Singers were especially helpful with their constructive and supportive guidance. The staff at Ablex Publishers worked rapidly in making arrangements to publish this book.

Ben Shneiderman
College Park, Maryland
April 1993

Overview:
fuel for a new discipline

Increasingly, researchers and designers are conducting experiments on the profound effects that design improvements can have on users: reduced learning times, faster performance on tasks, lower rate of errors, higher subjective satisfaction, and better human retention over time. Theories, taxonomies, and models at differing levels of abstraction are competing for attention. Empirical research has produced breakthroughs in the design of menu selection, form fill-in, pointing devices, and direct manipulation interactions.

Knowledgeable managers are recognizing that excellent user interfaces produce dramatic marketing advantages because they can greatly increase productivity, substantially reduce fatigue and errors, and enable users to be more creative in solving problems. When the user interface is well-designed, users should not only be performing well, but should also experience a sense of accomplishment and a positive regard for the designer of the interface. Usability testing, guidelines documents, and user interface management software tools (UIMS) are the three pillars of successful user interface development. Repeated testing in a usability lab with small numbers (3-12) of typical users performing typical tasks has proven to be very successful in inspiring improved designs and finding flaws. Hundreds of labs have been created in development organizations and a society of usability professionals has sprouted. Guidelines documents are successful in promoting consistency, defining organizational identity, and speeding development. Of course methods for enforcement, enhancement, and exemption must be part of the process. UIMSs dramatically speed development and allow easy modification, thereby supporting the pursuit of quality.

Academic research in human-computer interaction combines the experimental methods and intellectual framework of cognitive psychology with powerful tools from computer science. HCI benefits from related fields such as education


where computers are increasingly used in programs ranging from elementary school through professional skills development. The theory and measurement techniques of educational psychology are applicable to studying the learning process in novice computer users. Business system design and management decision making are endeavors which are being increasingly shaped by the nature of the computer facilities. Library and information services are also dramatically influenced by the availability of computer-based systems.

At the Human-Computer Interaction Laboratory, our goal is to do research on the theory and design of interactive systems that enable users to perform tasks rapidly, learn skills easily, and communicate in an atmosphere of competence, satisfaction, and confidence.

We want to replace arguments about ìuser friendly systemsî with a more scientific approach. We emphasize controlled experiments which yield more objective and reliable results, but also find informal usability studies are helpful in understanding design problems. We specify user communities carefully and identify tasks as thoroughly as possible. Then we turn to measurable criteria such as:

These criteria can be established before implementation and measured during an acceptance test. These criteria also serve as dependent variables in experimental research. Our orientation is towards direct manipulation designs that empower users, rather than anthropomorphic agents (e.g. talking bank machines or deceptive teaching machines) and artificially intelligent expert systems (designed to replace rather than support human users). We blend basic research themes and theories with practical projects that can inspire commercial developers.

A typical project team has a senior researcher and one to two graduate students, with help from other faculty, students or staff as needed. There is usually some software development (we try to limit the effort by using powerful tools), which generates comments to spur refinement, and then an empirical evaluation with typical users as subjects. I am devoted to empirical evaluations because they overcome the fog of wishful thinking and produce the clarity that leads to further innovation. Successful experiments can be difficult to conduct, especially in novel domains. We often must re-run experiments till we develop the appropriate training, controls, and tasks that lead to significant results and useful insights.

Administrative organization

The Human-Computer Interaction Laboratory (HCIL) brings together faculty at the University of Maryland who share an interest in these topics. The main participants are the Department of Computer Science, the Department of Psychology, and the College of Library and Information Services, with contributions from the School of Education, College of Business, Computer Science Center, and other units. I've especially appreciated the active and enduring efforts of my respected colleagues Prof. Kent Norman (Department of Psychology), Prof. Gary Marchionini (College of Library and Information Services), and Research Scientist Catherine Plaisant (Center for Automation Research). Valuable contributions over the years have been made by Charles Grantham, Yoram Kochavy, and Richard Chimera. We have conducted joint research projects, interdisciplinary seminars, an annual symposium and open house with presentation of research, cooperation in graduate programs, coordination in developing experimental facilities, work with state and federal agencies, assistance to Maryland corporations, and collaboration with corporate sponsors.

Since May 1983, the Human-Computer Interaction Laboratory has been one of the constituent laboratories in the Center for Automation Research under the direction of Prof. Azriel Rosenfeld. The Center for Automation Research provides assistance in securing supported projects and in administration. This interdisciplinary Center also includes the world-famous Computer Vision Laboratory and a Robotics Laboratory.

We publish in leading journals and present results at major conferences. Our work has led to commercial products such as our hypertext system,Hyperties, which is now disseminated and expanded by Cognetics Corporation, Princeton Jct., NJ, and home automation systems created by Custom Command Systems, College Park, MD. In addition, the University's Office of Technology Liaison licenses several of our software systems and our Questionnaire for User Interface Satisfaction.

We receive external support from the projects that we pursue for corporations (such as Apple, AT&T, General Electric, IBM, Johnson Controls, NCR Corporation, and Sun Microsystems) and government agencies (such as NASA, National Science Foundation, Library of Congress, National Library of Medicine, and the National Center for Health Statistics). The State of Maryland has supported us in working with two large companies (GE Information Services and Hughes Network Systems) and two small companies (Corabi Telemetrics and Custom Command Systems) under the Maryland Industrial Partnerships (MIPS) program. Several Japanese companies (such as NEC, Sony, Panasonic, and Toshiba) have provided support and sent their employees to work on our projects during year-long visits.

We've appreciated other support from University of Maryland units such as the Institute for Systems Research, the Departments of Computer Science and Psychology, the College of Library and Information Services, the Institute for Advanced Computer Studies, and the Engineering Research Center.