Andrew Sears, Mary Czerwinski, Laurie P. Dringus, and Barbara Bernal Thomas
For the last five years, there has been a workshop at CHI that focused on HCI education. This year, the goal was to begin evaluating the outcome of existing HCI courses and programs. The idea was to attempt to identify skills and knowledge that recent graduates must be taught by industry when working on HCI projects. Once the skills and knowledge that are missing are identified, we can begin to determine where they should be taught. Are these things that academia should be teaching? Or are they better left to industry? Are they fundamental issues that everyone should learn or are they specific to one organization?
The objective of the CHI 96 workshop was to bring together a mixture of industrial and academic representatives to discuss these issues. There were a total of 14 participants (including four organizers) for this workshop. The participants included industry representatives from Microsoft, Rank Xerox Parc, Bellcore, Intel, and Interval Research, and academicians from a variety of U.S. and international institutions. Some participants from industry also had connections to academia. The workshop was conducted over one and a half days.
The following people participated:
- Daniel Boyarski, Design Department, Carnegie Mellon University
- Martin Colbert, Ergonomics & HCI Unit, University College London
- Mary Czerwinski, Microsoft Corporation (organizer)
- Laurie Dringus, School of Computer and Information Sciences, Nova Southeastern University (organizer)
- Judy Hammond, School of Computing, University of Technology, Sydney
- Julie Jacko, Department of Engineering, Trenton State College
- Alistair Kilgour, Department of Computing and Electrical Engineering, Heriot-Watt University
- Robert Kraut, Human-Computer Interaction Institute, Carnegie Mellon University
- William Newman, Rank Xerox Research Centre
- Christian P. Leo Rohrer, School of Education, Stanford University
- Tony Salvador, Ethnography & Applications Research, Intel Corporation
- Andrew Sears, School of Computer Science, Telecommunications and Information Systems, DePaul University (organizer)
- Barbara Bernal Thomas, Computer Science Department, Southern College of Technology (organizer)
- Nick Wilde, Department of Computer Science, University of Montana
The workshop began with general introductions. Next, the results of previous workshops were summarized including those from CHI 93, CHI 94, and CHI 95. Materials from these workshops, as well as the NSF/ARPA report "New Direction in HCI Education, Research and Practice", (Strong, 1995) were distributed. Summarizing these previous activities highlighted their contributions and helped prevent participants from reinventing the wheel.
Three official goals were established for the workshop:
- Identify HCI skills and knowledge that industrial representatives must teach recent graduates
- Identify and discuss techniques academia can use to convey HCI issues to students to prepare them for industry work
- Develop an action plan for collaboration between industry and academia
The participants broke into two groups to discuss each of the goals and then gathered together to discuss the results. The final activity was to develop a list of specific suggestions that may help resolve some of the issues raised and improve the status of HCI education.
The two groups took different approaches to dealing with this problem. One generated a single list of issues companies often expect recent graduates to have and issues companies often have to teach recent graduates. The items listed ranged from general personality traits and skills to knowledge of specific techniques and procedures. Some items seem to be included several times in slightly different ways. Apparently, these issues were considered quite important.
- How to objectively observe users while they work
- Understanding that software development must be seen in a business context
- Formal methods for design specification (GOMS, claims analysis, etc.), their payoffs and limitations
- Informal methods, their payoffs and limitations including how they fit into the design cycle and into the different industries
- Teamwork
- Communication skills
- Cost-benefits analysis (via case studies)
- How to contextualize the HCI skills, the impact of technology on organizations, and how our tools and techniques are affected by the contexts in which they are applied
- Flexibility and creativity
- A love of learning
- How to identify problems, as well as solve them
- An appreciation for different points of view
- Basic visualization skills, low-fidelity prototyping skills. Currently, the tendency is to use technology and this often gets in the way
- Prototyping tools
- Storyboarding, flowcharting
- What are the factors that come from being human? What we know about humans, their information processing capabilities, and how these factors impact designs
- Project management (timetable, milestones, etc.)
- Information design
- Visual information design
The second group generated three lists. The first contained issues based in computer science that graduates from other disciplines should understand. The second focused on issues often associated with the behavioral sciences and the third focused on issues often associated with graphic or industrial design. The goal was to develop lists of issues that faculty in the various disciplines could focus on when teaching HCI courses.
Several specific issues were included in this list. Initially, programming and data structures were suggested for inclusion. After a discussion with all participants in the workshop, programming and data structures were replaced with a more general statement. While being able to program could be useful, the conclusion was that graphic designers and psychologists did not need to understand the details of C, stacks, and queues in order to be effective members of a development team. Instead, these individuals need to understand the basic terminology, how computers work, what can be done easily by computers, what is hard, and what is impossible. In addition, understanding the basics of testing software, project management, and working on teams would be useful.
Data collection and analysis skills topped the list from the behavioral sciences. This referred to both qualitative and quantitative data. Fundamental knowledge was also considered important. Specific topics that were listed included mental models, cognitive modeling, basic human attributes, and basic organization behavioral issues. Students should learn not only fundamental facts about how people think, organize things, solve problems, etc., but should also learn about group dynamics. Teamwork and the ability to read and interpret published reports were also included in this list.
This list was short, but the items included were considered critical. The ability to sketch an idea was the first item included. It was stressed that this did not mean the ability to create a beautiful drawing, but simply the ability to sketch ideas so others could understand them. Storyboarding, flowcharting, and general information design were also listed.
Once an initial list of topics had been identified, the next step was to identify specific tools and techniques faculty could use to convey these ideas to students. While the initial goal was to develop a list of tools and techniques that had been proven effective, it was soon decided that a more general list was necessary. Adequate solutions had not been found for many of the topics students were not currently learning, so this collection soon became a wish-list. Some items in this list have been tested in isolated situations, but should generalize. Others are simply ideas that appear promising.
- Films and videos that capture real design projects
- A web site to hold HCI educational resources
- Case studies that highlight successes and failures
- Using examples of success and failure to contextualize design principles and theories
- Design projects for students
- Tours of industry labs
- Funded research
- Guest lecturers from industry to emphasize specific parts of the process or techniques
- Industry representatives visiting universities for longer periods of time to share their experiences
- Student visits to industry to work on a design project
- Student internships
- Short courses for people in industry
- Distance collaborations between universities and between universities and industry
- Guest lecturers from other disciplines (like theater, law, etc.)
This has been a recurring goal for several workshops. Some representatives from both academia and industry recognize the importance of getting industry involved in educating students about HCI. The goal for participants of this workshop was to explore some existing collaborations and to try to find methods of increasing this collaboration. At least four different models were discussed. Each had advantages and disadvantages.
- The first model was characterized by both the Apple Design Project and the Interval Design Project. Workshop participants that had been part of these projects in the past characterized them as using a hands-off, coaching, approach. This was contrasted with a more hands-on style that could result in teaching students the skills they really need to learn. While these "Design Projects" are an attractive idea, the opportunities are far too limited for all of the students that could benefit from this type of experience.
- A more hands-on approach was also discussed. This involved identifying companies that are willing to fund projects that have students work on products that are important to the company, but are not on a critical path. These projects tend to last three years, but the bulk of the work is done during the summers. Students that have completed a one-semester design course implement the product during a summer. Since the product is a real system, students are divided into teams that focus on one part of the process (software implementation, documentation, etc.). The result is a series of prototypes developed by different students. Ideally, the industry representatives visit the university several times during the project to help guide the development process. This type of project involves several faculty and student assistants and up to 30 student workers. Resources, deadlines, deliverables, and funding are all negotiated. Funding is often in the range of $150,000/year. Once again, this type of project provides a valuable experience, but the opportunities are limited. Finding an adequate source of industrial partners that are willing to fund projects at this level does not seem feasible for most universities.
- The third model involves smaller groups of students working on an internal university project for up to fourteen months. Students work for a small company developing a real, but small application. This involves determining system requirements, conducting design reviews, building the application, and writing documentation. The projects must be small and well-defined. Students benefit by developing a real system. Since these are small organizations, they benefit by getting a system implemented for free. An alternative is to develop systems for non-profit organizations. While non-profits are readily available, students end up working with outdated systems. This model seemed more feasible to most participants, but would certainly involve a serious commitment by the faculty.
- The final model involved industry representative presenting design problems to students. Students then present their responses to the industry representatives. The companies treat these presentations like job interviews and hire the student they feel had the best response. Students then spend six months redesigning the system for the company. One problem that can occur is a lack of supervision. Students typically work on their own, so teamwork skills are not developed as part of this type of project. Once again, identifying adequate industrial partners could be a problem for many universities.
All four models involved presentations to one extent or another. Some included recruiting faculty from other departments to listen to and critique dry runs of presentations. Everyone wished they had more time to spend working on writing and presentation skills. Many participants stressed the importance of giving realistic presentations, including the duration. In general, presentations were about 15 minutes. Once again, all four models appear to involve a significant commitment from the faculty if they are to be successful.
In addition to the results listed above, the final part of the workshop was spent highlighting some of the more important activities we identified and thinking about future workshops on HCI education. Four activities dominated the discussions.
- Projects appear to be the single most important idea. Ideally, these projects should be realistic industry projects. There should be a commitment by the industrial partner to provide feedback and to participate as much as possible. Projects can be short- or long-term, but should all try to include interdisciplinary teams. Students should be required to give presentations at key milestones, work to deadlines, manage resources, and write final reports. The projects can involve developing a new system or redesigning existing systems.
- The skills identified in response to the first goal appear to be a minimal requirement for certain disciplines to foster better communication and teamwork on interdisciplinary teams.
- Case studies appear to be a valuable, but scarce resource. Unfortunately, developing a complete case study appears to be a time consuming task and many organizations are reluctant to release all of the details. Important parts of the process, including failures, are seldom shared.
- Finally, we discussed sharing these results and educational resources in general. This article is a beginning, but it cannot convey all of the experiences. Additional methods of communicating these results, and the results of other activities focused on HCI education were discussed and plans are under way for several additional activities and events in the next year. So keep your eyes open and think about participating at a workshop next year.
- Strong G. W., Gasen, J. B, Hewett, T., Hix, D., Morris, J., Muller, M. J., and Novick, D. G. (1994).
- A Report on New Directions in Human-Computer Interaction Education, Research and Practice. Washington DC: Sponsored by National Science Foundation and Advanced Research Projects Agency. (Available at: http://www.sei.cmu.edu/arpa/hci)
Andrew Sears, Ph.D., is an Assistant Professor in the School of Computer Science at DePaul University, Chicago, Illinois. He is the Education Editor for the SIGCHI Bulletin, and can be contacted at
sears@cs.depaul.edu.
Mary Czerwinski, Ph.D., currently works at Microsoft as a Usability Research Manager in the Interactive Media Division. She can be contacted at
marycz@microsoft.com.
Laurie P. Dringus, Ph.D. is an Associate Professor in the School of Computer and Information Sciences at Nova Southeastern University, Fort Lauderdale, Florida. She can be contacted at
laurie@scis.acast.nova.edu.
Barbara Bernal Thomas, is a Professor of Computer Science at Southern College of Technology, Marietta, Georgia. She can be contacted at
bthomas@sct.edu.
