Pioneering Work Helped Integrate Computers Into Our Lives

Successful interaction between a human and a computer doesn't happen when a computer does what a programmer wants, but when a computer has been programmed to do what a user wants.

To help his Virginia Tech computer science students "get it," John Carroll has them watch people try to use the students' software. Such approaches have resulted in increasingly user-friendly computers. For his contributions to the study of computer-human interaction (CHI), the Association for Computing Machinery's CHI special-interest group will present Carroll with the CHI Lifetime Achievement Award at their annual conference in Fort Lauderdale, Florida, on April 10.

The award recognizes "a lifetime of innovation and leadership including cumulative contributions to the field, influence on the work of others, and development of new research directions." Carroll is the fifth recipient of the award.

Carroll, a professor of computer science, psychology, and education, and founding director of the Center for Human Computer Interaction at Virginia Tech, said he "fell into" the field of CHI. When he graduated from Columbia in 1976 with a Ph.D. in psychology, he wanted to stay around until his girlfriend graduated--so he took a post-doctoral position at the IBM Thomas Watson Research Center and stayed 18 years.

Carroll's post-doc assignment was to study the problems of professional programmers and software designers. "The opportunity to study problem solving in this complex, open-ended domain was eye-opening and refreshing," he said. "We had to create new methods and discover research goals."

As the Watson group grew from three to 30 people, Carroll became a pioneer in the transition from programming based on technology constraints to software development based on user experiences and activity. It was a paradigm change. Programming was transformed from "functional specification"--listing functions and then writing code to achieve the functions--to "scenario based design."

"It was the end of the mainframe era, although no one knew it," Carroll said. "In the mainframe world, full-time technicians were on hand to run the machines. Now personal computers are in our homes, our cars, as well as our work places. It has to make sense. It has to be transparent. An important representation became a narrative description of use. A use scenario could be created even before anything was built. It puts the emphasis on human interactions with technology."

Carroll, who was the founding manager of IBM's User Interface Institute, remembers his scenario-based epiphany: "We were talking about scenarios--how people use PCs--and I said, 'What if scenarios were used as the functional specification?' This technical person I was talking to was aghast. Her reaction and my realization that scenarios were really scary to programmers motivated me to be more aggressive."

He developed the concepts of usability specification and user-interface metaphor, and originated the minimalist model of information design. His work on scenario-based design has had impacts beyond CHI in design studies, requirements engineering, human factors, and home-oriented informatics.

The next strategy for making computers extensions of human creativity was "participatory design."

"If you had told me in the 1970s that retirees and children should have direct input in software design, I would have laughed," Carroll said. By the time he joined Virginia Tech's faculty in 1994, participatory design had reached the United States. "It came out of Scandinavia. I heard about it at an international conference at the end of the 1980s. When I came to Virginia Tech, I wanted to change my technical direction. The Blacksburg Electronic Village (BEV) was a major initiative to put the town that surrounded the university online. I decided to work on community computing and computing in the public schools. That would absolutely require participatory design."

"Many of the programs and software packages developed for teachers have pretty much been failures," Carroll noted. "My thought was, if teachers are directly involved, we might have more success."

By the early 1990s most participatory design projects lasted six months, at most. Carroll undertook a long-term study, working with five Montgomery County teachers for six years. "I was able to look at the way participatory design develops and the consequences," he said. "You have to worry about trust, respect, and bridging communities. Participation is not a simple matter of inviting people to participate. You can't have respect just because you want it."

With the teachers, the nature of the relationship changed profoundly through time. "At first, the teachers totally deferred to us. We were the people with the grant, so they thought we must know what we are doing. It was good they respected us, but bad that they didn't think they had much to give. We needed people willing to share their expertise--but they didn't think they had any," said Carroll. "Many people who think they are doing participatory design are actually engaging in cosmetic tyranny."

A breakthrough came in Carroll's school project when he video-taped a ninth grade classroom and organized a joint discussion of the tape. "It was clear to the teachers that they knew more than we did--we didn't know what was going on in that classroom."

Carroll's participatory design work continues through the Center for Human Computer Interaction. His research on networking tools for collaborative learning activities is supported by the National Science Foundation, the Office of Naval Research, and the Hitachi Foundation.

Carroll has written more than 250 technical papers and 13 books, including Making use: Scenario-based design of human-computer interactions (MIT Press, 2000), Human-Computer Interaction in the New Millennium (Addison-Wesley, 2001), Usability engineering: Scenario-Based Development of Human-Computer Interaction (Morgan-Kaufmann, 2002, with M.B. Rosson), and HCI Models, theories, and frameworks: Toward a multidisciplinary science (Morgan-Kaufmann, 2003).

He has presented invited lectures worldwide on interactive system design, user-interface design, requirements engineering, and home-oriented informatics. He serves on nine editorial boards for journals and handbooks. In 1994 he won the Rigo Career Achievement Award from ACM SIGDOC for his work on the minimalist information design model and in 2002 he was elected to the ACM CHI Academy, a small group of pioneers whose technical work enabled key advances in personal computing.

Carroll earned bachelor's degrees in mathematics and in information sciences from Lehigh University in 1972 before pursuing his graduate degrees at Columbia.