Issue
Article
Vol.28 No.4, October 1996
Article
Issue
Issue |
Article |
Vol.28 No.4, October 1996 |
Article |
Issue |
For each of the last five years, there has been a workshop on HCI Education at the annual CHI conference. What makes these workshops so interesting isn't just the variety of people it brings together or issues discussed, it's the way the workshops have changed over the years. Just as HCI has evolved as a discipline, the topics of these and other workshops have also evolved. These changes are one indication of how much we have learned and what we have left to understand.
One of the earliest workshops on HCI education, held in 1985, proposed developing HCI courses (Mantei, 1985). In 1988, a two and one-half year project began that culminated in the now famous lime green report - the ACM SIGCHI Curricula for Human-Computer Interaction (Hewett et al., 1992, available at http://www.acm.org/sigchi/cdg/). Following the release of this report, a series of workshops on HCI education began at the annual CHI conferences. The first, held at CHI'92, discussed HCI education in general (Gasen & Aiken, 1993). More recently, the 1994 workshop focused on helping HCI educators solve specific problems (Løgren et al., 1994). Participants brought examples of successful exercises as well as problems they faced in teaching HCI courses. By the end of the workshop, most participants had practical solutions to their problems. At CHI'95, the focus shifted to begin looking at possible collaborations between industry and academia (Cohen et al., 1995). This year, at CHI'96, the workshop continued to seek input from both academia and industry. The focus was on evaluating the results of HCI courses and programs in an attempt to determine if they really deliver what is needed. This is a difficult process, so hopefully, the CHI'96 workshop will lead to additional activities focused on evaluating HCI courses and programs.
The official results of this most recent workshop appear elsewhere in this issue of the Bulletin. While the official results are important, the unofficial results may have a more dramatic impact on the future of HCI education. Just as the official topics have evolved, the questions attendees discuss during breaks and after the workshops end have also changed. In just five years, we have gone from general discussions on HCI education, to identifying issues HCI professionals should understand, to discussing how to teach our HCI courses, to questioning the outcomes of existing courses and programs.
We used to ask ourselves what topics students should study and how to improve the position of HCI within higher education (Gasen, 1994). In reality, this often came down to designing the content of a course or two and fitting them into an existing curriculum. Things have changed. Introductory HCI courses are becoming more common and new courses, concentrations, and degrees are popping up all the time. As a result, some new questions are being asked. Does everyone teach their introductory courses in the same way? If not, what are the differences, why do they exist, and how well do the different approaches work? What material is covered in the various HCI concentrations and degrees?
Unfortunately, these changes are not universal and many faculty continue to struggle to introduce HCI into their existing curriculum. Clearly, we must continue to think about the fundamental questions of what to teach and how to position HCI within higher education while also considering these new questions. Can those of us that have successfully introduced courses, concentrations, and degrees provide advice, guidance, and resources for those interested in expanding their own HCI programs?
The remainder of this article serves two purposes. First, it provides a brief snapshot of where HCI education is today. Second, it begins to address the questions of providing guidance and resources, evaluating outcomes, and helping HCI education continue to spread.
After CHI'94, Jean Gasen wrote about HCI education being "on the outside looking in." HCI courses didn't exist and certainly were not required in the majority of behavioral and computer science programs (Gasen, 1994). Today, HCI education is bigger than ever. Many faculty report having hundreds of students signing up for HCI courses each year.
To provide a better understanding of where HCI education is today, a request for information was distributed over the internet. Instead of asking about individual classes, this request asked for information about degrees that focus on HCI and HCI concentrations within traditional degrees. The responses to this request, and the HCI Education Survey maintained by Jean Gasen and Gary Perlman (available at: ftp://archive.cis.ohio-state.edu/pub/hci/Education/), were used as a foundation for what follows.
As of August 1995, the HCI Education Survey listed 174 courses offered in 77 different "programs". While some universities continue to offer just a single course, more are beginning to offer a collection of courses that discuss various aspects of HCI.
Over a dozen departments report having "HCI concentrations" within traditional programs such as Cognitive Science, Computer Science, Electrical Engineering, Information Systems, Psychology, and Sociology. Concentrations exist at the undergraduate level, but are more commonly offered to masters students. Clearly, HCI is becoming more widely accepted.
Surprisingly, more institutions reported offering degrees than concentrations. Although some existed prior to 1994, many have been created in just the last 2 or 3 years. The names include everything from a BS in HCI to an MA in Human Factors and Applied Cognition, but looking at the curriculum makes it clear that the theme is the same: Human-Computer Interaction. Perhaps as HCI becomes more widely accepted, the names of these degrees will become more uniform. Some of these degrees include:
We certainly have made progress! Unfortunately, based on the number of faculty that still report that they have difficulty introducing HCI courses and degrees, HCI still appears to be "on the outside" at many institutions.
Each year, the attendees at the HCI education workshop demonstrate the diversity of our experiences. We teach HCI in a variety of departments, to a variety of students, using a variety of materials, in both quarter and semester systems. We use a variety of approaches to teach everything from single courses to entire degrees. The workshop also attracts faculty trying to develop more comprehensive HCI programs. Some are interested in developing concentrations or degrees while others simply want to teach a single HCI course. For many of these faculty, attending these workshops is part of their search for answers. Among other things, those interested in developing concentrations and degrees wonder about what topics should be required and which should be optional. Faculty finding themselves in charge of their department's one HCI course are frequently looking for answers to more fundamental questions: what material to teach, what techniques work best, and what resources, exercises, and assignments others have found effective. Still other faculty are at institutions where HCI courses do not exist. These faculty often ask about fitting HCI into existing courses. What common courses can be used as a means of teaching the fundamentals of HCI? What resources, exercises, and assignments have worked in these courses?
Perhaps it is time to organize the knowledge and experiences of those who have already solved these problems. Different approaches are used in different situations, but the reasons are not always obvious. Documenting our experiences, both successes and failures, may help guide others facing the same problems. For example, the most common situation is still the single HCI course. Several distinct approaches have been used for this type of course. Why do the differences exist, when do the different approaches work, and what do students really learn? If there are several approaches to an individual course, just imagine the differences between degrees and concentrations. What topics are required? What electives are available? How much time is spent on theory vs. projects? What practical issues must be dealt with when introducing a new degree or concentration? The answers exist, but are hard to find.
Jean Gasen included a list of useful resources in a recent issue of the Bulletin (Gasen, 1996). These include web pages full of pointers HCI resources as well as published documents. The published reports, including the ACM SIGCHI Curricula for Human-Computer Interaction and the NSF/ARPA report on New Directions in HCI Education, Research and Practice are available in paper and electronically (Hewett et al., 1992, available at http://www.acm.org/sigchi/cdg/ and Strong et al., 1994, available at http://www.sei.cmu.edu/arpa/hci). While all of these are useful, something is missing.
Gary Perlman describes the process of designing, implementing, and evaluating user interfaces in his SEI Module on User Interface Development (available at: ftp://archive.cis.ohio-state.edu/pub/hci/SEI, 1989). This described both what to teach and how to teach it. More recently, Saul Greenberg provided an excellent example by putting materials for one of his courses, right down to his overheads, on the web (Greenberg, S., 1996, course material available at: http://www.cpsc.ucalgary.ca/projects/grouplab/481/481.html). This site documents an entire course, but many different approaches are used to teach this same type of course. Just imagine if we had the same resources available for a dozen different introductory courses. Maybe we could distill the dozen different courses into two or three basic approaches. We could document the approaches including when, where, and why they have been used. The resources, including syllabi, reading lists, homework assignments, project ideas, and classroom exercises could also be documented. Individuals developing new courses could select one of these approaches as a starting point and then incorporate whatever resources they found most beneficial.
Imagine a single, more complete set of resources, that represents the diverse approaches used to teach HCI in a variety of situations. This would include descriptions for individual courses, concentrations, and entire degrees. Syllabi, exams, exercises, homework assignments, project ideas, and perhaps even outlines for lectures would be included. Now we could begin to provide answers to question such as: How do I fit HCI into an existing software engineering course? What can I teach in an introductory HCI course in a quarter system? What material can be conveyed successfully in a four course concentration? What courses and topics should be required as part of a baccalaureate degree in HCI? What should be required as part of a masters degree?
As HCI continues to expand, we must begin to seriously examine the results of our efforts. HCI is a young discipline that continues to evolve, so we must continue to evaluate our curriculum. What core knowledge can we provide our students that will allow them to adapt as the field evolves? What topics can be left as electives? Are students that take our courses and earn our degrees learning what they need to know to be successful? How can we really evaluate the success of our programs?
In addition to providing the necessary knowledge, skills, and experiences, we must also think about the attitudes of students after they complete our courses. HCI is no longer just a single course that computer science or psychology students take. Some universities are offering degrees dedicated to educating students about HCI as a discipline.
Even with the growing number of concentrations and degrees, we must recognize reality: most students will continue to experience HCI as a single class. These students are unlikely to take additional HCI classes and many will find themselves in situations where they have the most experience designing interfaces of anyone on the development team. Since these students may never take a second HCI class, this first class should provide as much knowledge and experience as possible. Students should certainly understand the process of developing effective user interfaces. They should understand what makes a design effective and may have experienced some of the tools and techniques that are frequently used. Ideally, they should be able to develop better interfaces than they could before taking the class.
In addition to providing knowledge and experience, this first class will shape the students' feelings about HCI as a discipline. So another important question is: How should these students feel about their knowledge and skills? What should their first thoughts be when they are the person with the most experience designing user interfaces?
A single course in a traditional discipline such as computer science, psychology, or graphic design hardly qualifies an individual to practice in that field. People do write programs without taking any courses, but we believe that the quality of their programs increases after taking a course or two. Sure, some people write commercial quality software after taking one course, but this is rare. More often, it takes numerous courses and years of practice before reaching this point. Similarly, some people design user interfaces without any formal training. As a matter of fact, this used to be standard practice. Fortunately, this is becoming less common and more students are taking the one HCI course their departments offer. Some students may be ready to design commercial quality user interfaces after taking this one introductory course, but this is also rare. Designing commercial quality user interfaces typically requires understanding more than one course can teach. Students completing a single HCI course should not only gain knowledge, skills, and experience, but should also gain an appreciation for what they do not know. They should realize that they have only seen the tip of the iceberg. This is why HCI concentrations and degrees exist.
So, how should recent graduates respond when asked to design an interface? They may have more formal training than anyone else on the development team. They may even have more experience designing interfaces. Should they jump at the opportunity to design the interface or should they try another approach? If they understand how much more there is to know, maybe they would push to include an HCI specialist on the development team. Then, only if this cannot be accomplished, the student would apply what they have learned and design the best interface they can.
This is not to say that students completing a single course have not learned anything. Hopefully they will be able to design better interfaces than they could before taking the course. They would understand the process involved and could insist on understanding who their users are and what the users' goals are. Hopefully they will understand the basics of presenting information effectively, guiding users through their tasks, and the importance of iteration and testing. At the same time, they should realize just how much they do not understand. If we convince computer science or psychology students that they have mastered the knowledge and skills necessary to design user interfaces after a single HCI course, we may be doing more harm than good.
Finally, as we develop, gather, and evaluate these resources, we must also spread the word. The idea of gathering resources and making them easily accessible as discussed earlier is a starting point. Putting resources on the web increases accessibility, but does not solve the problem. Access to resources is useful, but talking with people who faced and solved the same problems is even more useful. Publishing reports and writing articles for publications such as the Bulletin reminds everyone of the importance of HCI education, but doesn't compare with face-to-face meetings. Holding workshops at events such as CHI helps, but does not reach everyone that may benefit. Even if individuals can attend CHI and can arrive early, they must choose between dozens of tutorials and workshops.
Many faculty have the desire to teach HCI, but do not have the resources necessary to attend major conferences. Providing information on the web or in widely distributed publications helps, but cannot compare with face-to-face conversations. Face-to-face question and answer sessions may be the most effective way of spreading HCI to new institutions. Perhaps it is time for a new event. Can an HCI Education workshop be held at a place and time that is more accessible to individuals with limited funding? Is funding available to cover travel expenses for those without the resources?
As HCI evolves, courses and degrees get created, and demand for practitioners continues to increase, we must begin to take HCI education even more seriously. We must think about the content of courses, concentrations, and degrees. We must make resources easier to find, evaluate the outcomes of our programs, and spread the word to even more academic institutions.
Plans have been put into motion to begin addressing some of these questions, but resources and energy are needed. If you have resources to share or time to dedicate to these issues, please let me know. If you are trying to expand the HCI offerings at your institution and have questions without answers, let me know. I'll do my best to provide answers or to put you in touch with someone else who might be able to. Finally, if you have issues you would like to see discussed in the Bulletin, please contact me. Contact information can be found inside of the front cover of the Bulletin.
Andrew Sears is the Education Chair for SIGCHI. He is an Assistant Professor in the School of Computer Science at DePaul University, Chicago, Illinois. He received his Ph.D. from the University of Maryland -- College Park in Computer Science in 1993. Andrew was involved in developing DePaul's HCI program, including both a BS in HCI and an MS in HCI. Andrew teaches graduate and undergraduate HCI courses. He participated in the workshop on teaching HCI at CHI 94 and was an organizer for the CHI 95 and CHI 96 workshops on HCI education. He can be contacted at sears@cs.depaul.edu.
Issue |
Article |
Vol.28 No.4, October 1996 |
Article |
Issue |